added
stringdate 2002-09-30 06:55:26
2023-01-01 16:16:21
| created
stringdate 1-01-01 00:00:00
2023-03-01 00:00:00
| id
stringlengths 4
9
| source
stringclasses 2
values | text
stringlengths 324
264k
| version
stringclasses 1
value |
|---|---|---|---|---|---|
2022-11-21T14:30:03.142Z
|
2022-11-21T00:00:00.000Z
|
253709738
|
s2orc/train
|
The impact of the COVID-19 pandemic on palliative care practice: A survey of clinical oncologists
Background Palliative care is an essential intervention to improve the quality of life for patients with cancer, whereas the ongoing COVID-19 pandemic poses a challenge to supportive and palliative care providers. This survey aims to explore the current status of palliative care practice for cancer and the influence of COVID-19, from the perspective of oncologists. Methods The semi-structure electronic questionnaire was designed. Mixed-mode surveys including electronic questionnaires, face-to-face interactions, and telephone interviews were adopted according to the willingness of respondents. Face-to-face and telephone interviews were based on same questions in the online questionnaire. Participants working in cancer-related departments with frontline palliative care experience during the COVID-19 outbreak were included. Surveys covered experiences and perspectives regarding the impact of COVID-19 on clinical work, personal lives, and palliative care practice. Suggestions on coping strategies were further proposed and qualitatively analyzed. Results Thirty-seven oncologists participated in this study from September 2021 to January 2022. The majority of them believed COVID-19 significantly and negatively affected their clinical work routines (75.7%), personal daily lives (67.6%), and palliative care practice (64.9%). Most specialists considered that currently the palliative care system remained underdeveloped (73.0%), and other factors besides COVID-19 were associated with this situation (78.4%). Seventeen participants further made suggestions on how to promote palliative care during COVID-19, and three themes emerged through the qualitative analysis: (1) Remote or online service (88.2%); (2) Publicity, education, or shared decision-making for patients (29.4%); (3) Guidelines, training, or programs for care providers (23.6%). Conclusion Oncologists consider that COVID-19 has an adverse impact on their palliative care practice and daily routine. In addition to COVID-19, other factors affecting palliative care should not be neglected. Corresponding measures are warranted to encourage palliative care practice during COVID-19.
Background: Palliative care is an essential intervention to improve the quality of life for patients with cancer, whereas the ongoing COVID-pandemic poses a challenge to supportive and palliative care providers. This survey aims to explore the current status of palliative care practice for cancer and the influence of COVID-, from the perspective of oncologists.
Methods:
The semi-structure electronic questionnaire was designed. Mixed-mode surveys including electronic questionnaires, face-to-face interactions, and telephone interviews were adopted according to the willingness of respondents. Face-to-face and telephone interviews were based on same questions in the online questionnaire. Participants working in cancer-related departments with frontline palliative care experience during the COVID-outbreak were included. Surveys covered experiences and perspectives regarding the impact of COVID-on clinical work, personal lives, and palliative care practice. Suggestions on coping strategies were further proposed and qualitatively analyzed.
Results: Thirty-seven oncologists participated in this study from September to January . The majority of them believed COVID-significantly and negatively a ected their clinical work routines ( . %), personal daily lives ( . %), and palliative care practice ( . %). Most specialists considered that currently the palliative care system remained underdeveloped ( . %), and other factors besides COVID-were associated with this situation ( . %). Seventeen participants further made suggestions on how to promote palliative care during COVID-, and three themes emerged through the qualitative analysis: ( ) Remote or online service ( . %); ( ) Publicity, education, or shared decision-making for patients ( . %); ( ) Guidelines, training, or programs for care providers ( . %).
Conclusion:
Oncologists consider that COVID-has an adverse impact on their palliative care practice and daily routine. In addition to
Introduction
In December 2019, coronavirus diseases 2019 , caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), emerged in China. Subsequently, COVID-19 has rapidly become a severe pandemic and significantly impacted various clinical practices (1,2). Although the global mortality rate estimated by the World Health Organization (WHO) was 3.4%, mortality and morbidity rates tend to be higher among older people and cancer patients (1,3). The outbreak of the COVID-19 pandemic results in a global shortage of healthcare resources, presumably including supportive and palliative care resources and applications, especially for the large number of patients with cancer (4). At present, cancer remains the leading cause of worldwide medical burden and brings tremendous physical and mental stress on patients and their families (5). Palliative care is an essential component of the cancer comprehensive treatment, aiming to alleviate the suffering and improve the quality of life (6). Given that COVID-19 is expected to surpass our capacity to provide supportive and palliative care to all patients, which poses a unique challenge to healthcare teams of rationing care during pandemic when resources are scarce (4), in-depth investigations on the specific impact of the ongoing COVID-19 pandemic on the clinical practice of palliative care for cancer patients as well as effective coping strategies are necessary.
The two-sides of practice indicates that one side is the recipient of care and the other side is for the care providers. Thus, both care practitioners and patients play an important role in the practice of palliative care. With early studies suggesting that cancer patients are particularly susceptible to COVID-19, the current pandemic is forcing oncology professionals to explore and practice more (3,7,8). Meanwhile, not only vulnerable patients affected by COVID-19 should be concerned, but also palliative care practitioners. The experiences, perspectives and thoughts of care providers during COVID-19 are of great importance and value, even though sometimes we may mainly focus on patients and neglect that doctors are passive sufferers of the COVID-19 pandemic as well. In fact, the health care workers struggling to cope with the current situation face more stress and anxiety, due to the heavier medical burden under COVID-19 and their overwork (9). The fact that the medical staff come so close to the disease puts their mental health at a higher risk than the general population (10). Recently, an increasing prevalence of mental health symptoms has been reported among physicians who had direct contact with infected patients (11,12). With the explosive growth of the number of diagnosed COVID-19 cases, the stress, anxiety, depression, and feelings of negativity became more and more common in Chinese medical workers (13,14).
With the increasing difficulty to provide palliative care during the COVID-19 pandemic, the wide emphasis on the experiences and viewpoints of palliative care providers, as well as the urgent need for useful coping strategies to better tackle the influence of COVID-19 on palliative care practice, this survey was conducted in order to shed light on these issues. We not only explored the impact of the COVID-19 pandemic on the clinical practice of supportive and palliative care for patients with cancer, from the perspective of oncologists, but also proposed some useful countermeasures to promote and encourage the clinical practice of palliative care during COVID-19.
Study design
We designed a semi-structured electronic questionnaire to elicit the perspectives of clinical oncologists through both quantitative and open-ended qualitative questions. Mixed-mode surveys including online electronic questionnaires, face-to-face interactions, and telephone interviews were adopted based on the willingness of respondents. If they agreed to receive faceto-face or telephone surveys, they would be individually asked the same questions in the online electronic questionnaire in person, and the information was collected. After obtaining permission from respondents in face-to-face and telephone surveys, all interviews were digitally recorded and transcribed verbatim to assure accuracy. If participants chose to answer the electronic questionnaire, they would complete a selfadministered anonymous web-based questionnaire in both Chinese and English. The Independent Ethics Committee of National Cancer Center approved this research. Informed consent was obtained from all individual participants included in the study. They can access the online Participant Information Consent Form via a secure web link and complete it using mobile phones or computers. After the completion of survey, .
/fpubh. . participants could receive an e-card gift and a thank-you note via email, if they were willing to provide their email addresses and some other personal information only for this purpose.
Participants and study setting
The study was conducted from September 2021 to January 2022 in China. In the approximately two-year pandemic background, this survey was primarily in the setting where palliative care was provided for cancer patients during the COVID-19 period, including medical centers, hospitals, nursing homes, palliative care institutions, community healthcare centers, etc.
Eligible participants were those over 18 years of age currently working in cancer-related departments, such as the department of medical oncology, radiation oncology and surgical oncology, etc, in medical establishments or other sites that provide palliative care practice. Among them, oncologists who had clinical front-line working experiences during the COVID-19 pandemic (from December 2019 to the date participating in the survey), possessed and were able to use online electronic mobile devices autonomously, and proficiently mastered Chinese or English language, were finally included.
Survey content
The questionnaire survey or interview outline covered: (1) Demographic characteristics: Age, gender, country, educational attainment, workplace and currently working department, supportive and palliative care training experience, the primary place of palliative care practice.
(2) Subjective perceptions regarding the influence of the COVID-19 pandemic: Do you think the COVID-19 pandemic has a significant impact on your clinical work in oncology? Yes, mainly negative impacts. / Yes, mainly positive impacts. / No. / Other. Do you think the COVID-19 pandemic has a significant impact on your personal life or daily routine? Yes, mainly negative impacts. / Yes, mainly positive impacts. / No. / Other. Do you think the COVID-19 pandemic has a significant impact on clinical practice of palliative care for cancer patients? Yes, mainly negative impacts. / Yes, mainly positive impacts. / No. / Other.
(3) Subjective perspectives regarding the status quo of palliative care: Do you agree that the current supportive and palliative care system in your working environment is adequate or fully developed? Yes. / No. / I am not sure. / Other. Do you agree that other factors, except for the influence of COVID-19, are associated with the current status of palliative care practice system? Yes. / No. / I am not sure. / Other.
(4) Suggestions and advice: Do you have any suggestions to improve the clinical practice of palliative care during the COVID-19 pandemic? If yes, please give your precious and specific advice.
Data collection and analysis
Data was collected from the electronic questionnaire surveys and interviews on oncologists in China from September 2021 to January 2022. Demographics and subjective perspectives of respondents were quantitatively summarized mainly using descriptive statistics. Personal suggestions of free-text narrative responses were qualitatively analyzed through inductive thematic analyses. All data were translated into English before analysis. All face-to-face and telephone interview surveys were digitally recorded and transcribed verbatim by two researchers (W.Y. and X.M.) together to assure maximum accuracy. Two investigators (W.Y. and D.W.), without previous knowledge of the participants and not involved in the distribution of questionnaires, independently collected and analyzed the questionnaire content, and they further compared and verified their research results. Any discrepancies between two researchers (W.Y. and D.W.), especially in summarizing countermeasures proposed by participants, were solved by consulting senior investigators (X.Z. and N.B.).
Results
Thirty-seven eligible clinical oncologists participated in this study. Among them, 32 (86.5%) were surveyed by online questionnaires, 4 (10.8%) by face-to-face interviews, and 1 (2.7%) by a telephone interview. The baseline demographic characteristics were presented in Table 1. All 37 participants were from China, including 29 (78.4%) males. The median age was 40 (22-56). A large proportion of them had doctoral degrees (67.6%) and worked in urban areas (81.1%). One-third (32.4%) participants were from the department of radiation oncology, 24.3% from medical oncology, and 16.2% from surgical oncology. Meanwhile, clinicians were more likely to practice palliative care in medical centers or hospitals (70.3%) than in the community or elsewhere (29.7%) in China. However, only 10.8% of them had obtained the accredited professional training certification in palliative medicine, and most were with non-accredited training experience (56.8%).
In terms of subjective perceptions and opinions with respect to the impact of COVID-19, the majority of participants agreed that the COVID-19 pandemic had a significantly negative effect on their clinical work in the cancer field (75.7%, Figure 1A), as well as their daily routines or personal lives (67.6%, Figure 1B). In addition, as many as of 64.9% specialists considered that the clinical practice of palliative care for cancer patients had been significantly and negatively affected ( Figure 1C). Moreover, 2 (5.4%) oncologists believed that the current supportive and .
/fpubh. . palliative care system was fully developed in China, while 73.0% of them deemed that it remained underdeveloped. It was quite common for them to agree that many other factors besides COVID-19 were associated with this present status (78.4%), but 2.7% of participants disagreed with that. Furthermore, a total of 17 specialized physicians proposed their suggestions on how to tackle the adverse influence of COVID-19 on palliative care practice. The qualitative analysis resulted in the following three themes: (1) Remote or online service (88.2%); (2) Publicity, education, or shared decision-making for patients (29.4%); (3) Guidelines, training, or programs for care providers (23.6%). We reported some suggestions made by participants (P) in Box 1.
Discussion
It is known that the integration of team-based, timely and targeted supportive and palliative care into standard oncology care for all patients with cancer is of great significance (15). The characteristics of palliative care mainly lie in the team-based care, allowing the interdisciplinary members to address comprehensively the multi-dimensional care needs of patients and their caregivers; the timely intervention, becoming preventative care to minimize crises at the end-of-life; and the targeted treatment, referring to the identification of the patient most likely to benefit from a specialized palliative treatment (15). Nevertheless, palliative care services are under-resourced at the best of times (16). To date, providing effective palliative care has become more and more difficult for specialists, as worldwide health systems become strained under the ongoing COVID-19 pandemic (16). Therefore, we have conducted this survey, from the professional perspective of medical workers, to explore the influence of COVID-19 on palliative care practice, and further put forward some interventions to deal with the status quo. Overall, this report not only identified specific aspects that have been negatively affected by COVID-19, but also underscored the need for useful coping strategies.
Although it is universally accepted that palliative care should be adopted by specialists in all oncology settings to benefit cancer patients and their families (17). The lack of integrating supportive care into comprehensive cancer treatment has become strikingly evident in the current context of the COVID-19 pandemic (18). In the present study, we found the majority of oncologists agreed that COVID-19 adversely impacted their routine clinical work (75.7%) and palliative care practice (64.9%) to a very large extent. Perhaps it is because this pandemic has created more uncertainty and disrupted the way that we practice medicine, including palliative and supportive cancer for cancer patients. Through ongoing international conversations pertaining to COVID-19, palliative care practitioners are asking whether we should attach more importance to patients with cancer, who are often the most vulnerable (19,20). We consider that because of the underlying suppression of immune system and poor general condition heightening the risk for susceptibility to COVID-19 and relative complications, cancer BOX | Quotes per theme.
"Theme 1. Offering remote or online service" Offering more online guidance on palliative care for cancer patients and their families "Before the COVID-19 epidemic, there were more cancer patients from other provinces in our hospital, but due to the restrictions of the epidemic, the access and follow-up of these patients were very limited. Adding more online consultations may help them" #P25 Promoting remote multidisciplinary cooperation "Multidisciplinary cooperation members cannot meet on-site due to epidemic restrictions, but the mechanism of online MDT (multidisciplinary treatment) is not mature enough and needs to be improved urgently" #P34 "Invite the professional palliative medicine team to join the multidisciplinary consultation" #P9 Strengthening home-based palliative care ". . . palliative care at home for cancer patients should be given a higher priority, and access to remote medical guidance is less clear for physicians as well as patients" #P20 "The trend of shifting palliative care settings from hospital to home has been accelerated by COVID-19" #P37 Using more modern technology in palliative care ". . . mobile phones and WeChat could become important tools to provide remote palliative care guidance during this COVID-19 period. It would be better to have some platforms like WeChat public account for telemedicine that do not involve doctors' personal privacy" #P32 Theme 2. Increasing the publicity, education, or shared decision-making for patients Encouraging shared decision-making "The situation of patients who need palliative care is often complex, especially end-of-life patients, and it could be more helpful and effective if patients and doctors share the decision-making process" #P35 Increasing the publicity and education of palliative medicine "Increase palliative care education, positive publicity, and concept shaping, the actual potential demand in China is huge" #P19 Theme 3. More guidelines, training, or programs for care providers Standardizing palliative medicine training for medical workers "Palliative care providers working in different sites, different regions, and different fields will need more specialized training" #P11 "Most care providers actually do not specialize in this. They might lack relevant professional experience. More professional training in palliative medicine might be helpful" # P36 Enhancing personalized palliative care programs "During COVID-19, government support may be more helpful, such as more palliative care programs and financial investments" #P5 ". . . promoting projects about personalized palliative care practices could be of great benefit, given that each doctor's situation (under the COVID-19 pandemic) is different" #P37 Updating more information on COVID-19 in palliative care guidelines "The prevention and control of the COVID-19 in China is becoming more and more regular. Information and guidelines on palliative care also should be updated accordingly" #P35 patients ought to be paid more attention and supports (21)(22)(23)(24). Another concern is that palliative care practice to address certain circumstances of cancer patients remains inadequate and immature, such as in the particular case of the severe COVID-19 pandemic (25). Despite the fact that individuals with cancer on active palliative healthcare are more likely to require frequent hospital visits or meeting with professionals, the isolation of interpersonal contact and restrictions on patient access to hospitals, in order to reduce the risk of spreading SARS-CoV-2, bring great difficulties to the practice and promotion of palliative care (26). Hence, palliative care in oncology should be an explicit part of international response plans for COVID-19, especially considering the high morbidity and mortality from COVID-19 in patients with cancer (27). More importantly, 67.6% oncologists in our study considered that their personal lives and daily routines also had been significantly and negatively affected by the COVID-19 outbreak. In a study from Italy during COVID-19, frontline .
/fpubh. . health care workers were reported severe posttraumatic stress symptoms, which could seriously affect their lives and careers (28). Previous studies also showed that medical workers tended to worry a lot about possibly infecting their families and thus usually were highly stressed even after coming back home (29,30). Besides, numerous medical professionals around the world were sent to quarantine after contacting and fighting against COVID-19, which might cause a significant impact on the mental health and daily life of medical staff (31,32). While patients are often the focus of attention during COVID-19, we believe that medical workers should be given full supports, appropriate comforts, and positive encouragements, as well, which may be one of the potential ways to motivate palliative care practitioners. Furthermore, when it comes to the current status of the palliative care system in China, most specialists deemed that it remained underdeveloped. They also agreed that many other factors besides COVID-19 were associated with it, presumably on account of the unbalanced medical resources, conflicts of traditional values, and reluctance from patients (33)(34)(35). Thus, the development of palliative care still needs further efforts, not only in the current context of the pandemic but in the future. Despite challenges experienced during the pandemic, the global oncology community has responded with an unprecedented level of investigation and collaboration (36). This research also proposed some viable coping strategies, including promoting online palliative care guidance and home-based supportive care. Especially at present, the healthcare place is constantly shifting from the clinic to the home, where people can be treated via telehealth services, digital consultations, and intelligent devices (25,37). These modern technologies are of great help and may encourage palliative care practice, even at the self-quarantine time during the COVID-19 outbreak. Similarly, based on digital equipment, remote multidisciplinary consultation and modern technology were also underlined by oncologists in this study. Emerging hi-technology will significantly contribute to palliative therapy if adopted properly and integrated into comprehensive care plan. However, novel technologies could augment traditional health strategies but cannot entirely replace them. As a result, shared decision-making, standardized palliative care training for medical workers, and personalized palliative medicine programs, etc., were also emphasized in this study.
Palliative care providers and hospice sectors play an essential role in the response to COVID-19 (38). Oncologists are known to provide supportive care with professional decisions, psychological counseling, and complex symptom management, especially for patients with advanced cancer (39). Providing such care is particularly challenging but also tremendously meaningful, given that humanitarian palliative caregivers with sufficient preparedness and capacity to cope with the current high-stress conditions of the COVID-19 pandemic may further improve the quality of life and optimize overall survival for cancer patients (35). Moreover, our study also highlighted that the issues and needs of palliative care practitioners should raise public concern and be further addressed.
There are several limitations. First, the sample size was moderate, but we have adopted some design strategies to improve response rates, including small financial incentives, mixed-mode survey, and brief questionnaire (40). In the current context of rapid spread of COVID-19, the opportunity to gather in-depth information was limited due to the extreme pressure on medical service system. Nonetheless, the results from our study supported a cross-sectional survey with larger sample to identify more examples of innovate practice in palliative care in the future. It also would be helpful to incorporate more diverse viewpoints from other palliative care participants in future research, such as nurses, as nurses .
are also involved in palliative care and COVID-19 to a large extent. Second, using self-reported results may bias the conclusion, although this study focused primarily on the subjective feelings of the oncologist community. Finally, the dynamic state of pandemic and different medical backgrounds among countries may limit the generalizability of the results for other settings. In the future, larger-scale studies involving more countries and regions to examine the impact of COVID-19 on palliative care under different epidemic prevention policies are warranted.
Conclusion
The COVID-19 pandemic has a significant adverse impact on palliative care practice, daily clinical routine, and personal lives, from the perspective of oncologists. Most of them consider the current palliative care system underdeveloped in China, and other factors besides COVID-19 may be associated with this situation. The corresponding measures should be taken to improve the clinical practice of palliative care during COVID-19, such as incorporating more online guidance and remote assistive technology in palliative care, encouraging home-based and personalized palliative care treatment according to the condition of patients, and promoting up-to-date information and practical training for palliative care practitioners.
Data availability statement
The datasets presented in this article are not readily available because questionnaire data and interview transcripts are available upon reasonable request.
Requests to access the datasets should be directed to YW, [email protected].
Ethics statement
This study was approved by the Ethics Committee of National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College. Informed consent was obtained from all individual participants included in this survey.
Author contributions
YW, DW, and XM acquired, analyzed, and interpreted all data. YW and YH drafted and revised the manuscript. XZ and NB reviewed and edited this manuscript critically. All authors contributed to the concept and design of this work and approved the final version to be published.
|
v2
|
2022-11-22T06:17:22.756Z
|
2022-11-21T00:00:00.000Z
|
253732946
|
s2ag/train
|
Peritoneal metastatic mixed adenoneuroendocrine carcinoma treated with cytoreduction surgery and hyperthermic intraperitoneal chemotherapy: a case report.
A 61-year-old man presented with abdominal distension without any symptoms. On colonoscopy and computed tomography findings, it was clinically diagnosed as peritoneal metastasis of sigmoid colon cancer, and diagnostic laparoscopy was performed. Only the peritoneum was partially resected, and the pathology was signet ring cell carcinoma with predominantly local mucinous carcinoma component. However, the patient complained of persistent symptoms and, despite the progress of chemotherapy, the peritoneal dissemination worsened, and additional cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (HIPEC) was performed. Mixed adenoneuroendocrine carcinomas (MANECs) were reported in the appendix with perforated visceral peritoneum. After additional chemotherapy, the patient was discharged. Patients with advanced MANEC with peritoneal spreading may benefit from aggressive treatment by cytoreduction surgery with HIPEC, followed by intravenous chemotherapy.
|
v2
|
2022-11-22T06:17:25.311Z
|
2022-11-21T00:00:00.000Z
|
253733566
|
s2ag/train
|
Custom designed and 3D-printed titanium pelvic implants for acetabular reconstruction after tumour resection.
BACKGROUND
Reconstructive procedure following resection of large pelvic tumours around the hip joint remains a complex challenge.
METHODS
This study presents a retrospective case series of patients presenting with benign or malignant pelvic tumour for which an internal hemipelvectomy including the hip joint and subsequent reconstruction with a custom designed 3-dimensional printed titanium pelvic implant (3DPPI) has been performed between August 2013 and January 2018.
RESULTS
15 consecutive patients with a median age of 33.9 years (IQR 26.4-72.2) and a median BMI of 20.7 kg/m2 (IQR 19.0-33.3) were reviewed after median follow-up of 33.8 months (IQR 24.0-78.1). The majority of patients presented with a malignant tumour as their principal diagnosis (n = 13, 86.7%). The median surgical time was 5.5 hours (IQR 4.5-8.5) and median peri-operative blood loss was 5000 ml (IQR 2000-10000). The median MSTS score at follow-up was 63.3% (IQR 51.7-86.7%). The median NRS in rest was 0.0 (IQR 0.0-5.0), the median NRS during activity was 2.0 (IQR 0.5-7.0) and the median HOOS-PS was 76.6% (IQR 67.9-91.0). 4 patients had implant-specific complications (n = 4, 26.6%); 1 hip dislocation (Henderson type 1a), 3 structural complications (type 3a), 1 deep infection (type 4a) and 1 local tumour recurrence (type 5b). At follow-up, 4 out of 15 implants were classified as a failure, resulting in an implant survival rate of 73.3%.
CONCLUSIONS
Acceptable peri-operative outcomes, functional results, complication rates and short-term implant survival can be achieved in a cohort of complex patients undergoing 3DPPI reconstruction after hemipelvectomy including the acetabulum.
|
v2
|
2022-11-22T06:17:25.355Z
|
2022-11-21T00:00:00.000Z
|
253734085
|
s2ag/train
|
Synthesis, X-ray structure and anticancer activity evaluation of a binuclear La(III) complex with anthranilic acid.
A binuclear La(III) complex {[La2(HA)4(H2O)4(C2H5OH)2Cl2]Cl4 (C1)} with 2-aminobenzoic acid (HA) was prepared from the ligand and heptahydrated lanthanum chloride. The complex was characterised by X-ray crystallography that revealed anti-prismatic geometry around both of the lanthanum. In the complex, the four 2-aminobenzoic acid ligands are zwitter ionic and the two lanthanum(III) ions net charge is only counterbalanced by chloride ions. The complex cytotoxicity was determined against human breast (MDA-MB-231), prostate (PC-3) and bladder (T-24) cancer cells. This complex afforded cytotoxicity towards the T-24 bladder cancer cells with an IC50 value of 383.5 µg/mL (319 µM). In contrary, activities by the lanthanum complex with IC50 values of 1124 µg/mL (934 µM) and 739 µg/mL (614 µM) were, respectively, shown against the MDA-MB-231 and PC-3 cancer cells. This means the complex is more cytotoxic against the T-24 cells, despite that its activity is less compared with activities shown by classical drugs.
|
v2
|
2022-11-23T06:17:33.250Z
|
2022-11-21T00:00:00.000Z
|
253759981
|
s2orc/train
|
Xeno-Estrogenic Pesticides and the Risk of Related Human Cancers
In recent decades, “environmental xenobiotic-mediated endocrine disruption”, especially by xeno-estrogens, has gained a lot of interest from toxicologists and environmental researchers. These estrogen-mimicking chemicals are known to cause various human disorders. Pesticides are the most heavily used harmful xenobiotic chemicals around the world. The estrogen-mimicking potential of the most widely used organochlorine pesticides is well established. However, their effect is not as clearly understood among the plethora of effects these persistent xenobiotics are known to pose on our physiological system. Estrogens are one of the principal risk modifiers of various disorders, including cancer, not only in women but in men as well. Despite the ban on these xenobiotics in some parts of the world, humans are still at apparent risk of exposure to these harmful chemicals as they are still widely persistent and likely to stay in our environment for a long time owing to their high chemical stability. The present work intends to understand how these harmful chemicals may affect the risk of the development of estrogen-mediated human cancer.
Introduction
Xenobiotics (xeno: foreign; biotic: life form) are chemicals of synthetic origin that are foreign to living systems. These chemicals are not part of the normal metabolic activity of living organisms and therefore may interfere with the functioning of the physiological system. These unwanted chemicals inside the body are known to have harmful effects on functioning and are responsible for diseases in humans. One such detrimental effect caused by some of these chemicals is xeno-estrogenicity [1].
'Xeno-estrogenicity' (xeno: foreign; estrogenicity: resembling natural estrogens) denotes the property of foreign chemicals being able to produce responses in physiological systems comparable to estrogens produced by the body by different mechanisms. Xenoestrogens are a group of chemicals that disrupt the endocrine system by mimicking the activity of natural estrogens or producing a similar response in the body by other mechanisms. Majorly known persistent organochlorine pesticides (OCPs) and other classes of chemicals present in the environment are known to bind estrogen receptors and produce a physiological response equivalent to the hormone in various organs and tissues [2]. Therefore, these chemicals can produce adverse effects such as increasing the risk of estrogen-mediated cancers [3]. These chemicals can easily enter the human body due to their physiochemical characteristics. They not only widely persist in our environment, including water, soil, and food, but also exhibit high volatility at environmental temperatures [4]. Moreover, a much greater risk is rendered by such chemicals due to their biomagnification through to the top
Carcinogenic Effects of Organochlorine Pesticides
Pesticides are the most well-known carcinogenic chemicals to humans. Studies from both humans and animals have unambiguously confirmed the development of pesticide exposure-mediated human cancer [20]. All known OCPs possessing xeno-estrogenic potential are listed in the 'first dirty dozen' in the Stockholm Convention as they were widely reported to cause various adverse health outcomes in humans around the world [21]. Human studies support the association of various OCPs (aldrin, dieldrin, endosulfan, HCH, DDT, 2,4,5-trichlorophenoxyacetic acid, phenoxy acid herbicides, and methoxychlor). The exposure (both occupational and non-occupational) has been strongly linked to increased incidence of non-Hodgkin's lymphoma [22]; multiple myeloma [23]; soft tissue sarcomas [24]; lung cancers [25]; and cancers of the pancreas, stomach, liver, and kidney as well as urinary and gall bladder cancer [26][27][28][29][30].
The International Agency for Research on Cancer (IARC) has listed OCPs as potent carcinogens in animal studies [31,32]. Cancer development also depends on genetic susceptibility as well. Xenobiotic metabolic enzymes are metabolic susceptibility enzymes whose metabolization leads to the elimination of a diverse range of xenobiotics, including OCPs. These predominantly hepatic enzymes are responsible for xenobiotic detoxification in two phases: 'phase I' enzymes (functionalization reactions), mainly cytochrome P450, and 'phase II' enzymes (conjugation reactions), including glutathione S transferases (GSTs). Cytochrome P450 (abbreviated as CYP P450 and infrequently as CYP450) is a superfamily of monooxygenase enzymes that catalyze metabolism by oxidation of their substrates. On the other hand, GSTs, important phase II enzymes of the xenobiotic metabolizing enzyme family, eliminate phase I CYP450 enzymes' catalyzed intermediates, leading to their excretion out of the body [33,34].
Individual susceptibility to environmental chemical exposure leading to diseases, including cancer, is modified by genetic susceptibility, and polymorphism in such gene families is a major risk. Genetic variations in phase I and/or phase II genes may alter the activity of the corresponding enzyme responsible for the bioactivation of xenobiotics; individuals with genetically impaired xenobiotic elimination functions will be at higher risk of disease susceptibility over their lifetime. Identification of the genes responsible for the elimination of xenobiotics and their variability is key for the identification of individuals at higher risk [12,[33][34][35].
Evidence of Xeno-Estrogenicity
Since the advent of modern agriculture, pesticide use has become a primary means to enhance crop productivity. Amongst all the pesticides, OCPs became the first choice of farmers around the world after their introduction, attributed to their low cost and versatile effectiveness against a wide range of agricultural pests. Table 1 presents the current usage status, commercial name, and other relevant details related to pesticides. Organochlorine pesticides were banned in developed countries after their listing in the Stockholm Convention on Persistent Organic Pollutants (an international agreement aimed at the banning of harmful environmental pollutants), which is still used for the development and control of vector-borne diseases such as malaria. The overuse of OCPs has been rising alarmingly, and it is raising serious threats to both human health and the environment [36]. Every year, billions of pounds of OCPs are used for boosting agriculture productivity to fulfill the dietary needs of the world population. It is not surprising that high levels of these harmful chemicals are also detected in water and soil [37,38]. The hormone-like activity of these man-made chemicals was realized long after they were released into our environment. Krishnan et al. (1993) accidentally found unknown chemicals to be estrogenic because they disrupted experiments conducted in laboratories while studying the effects of natural estrogens. Later experiments confirmed that the disruption of the experiments was due to the plastic tubes used in the experiments possessing estrogenic activity [42]. In humans, indirect positive evidence of "pesticide exposure-related estrogenicity" was reported among farm workers. In 1949, workers occupationally exposed to dichlorodiphenyltrichloroethane (DDT) were found to have significantly lowered sperm counts and loss of fertility [43].
These chemicals are known to stay in the environment for a long period of time, which could be from a few years to several decades in different media; for instance, the most well-known pesticide, DDT, has a half-life of up to 30 years in soil and 3 to 6 years in the human body (up to 150 years in water) [44]. Therefore, long-term exposure may lead to considerable accumulation in adipose tissue due to the lipophilic nature of agricultural chemicals [45]. High levels of pesticides, including OCPs, have been reported in different samples from all over the world, including blood, adipose tissue, placental tissue, umbilical cord, colostrum, and even cord blood, etc. [6][7][8]. In a recent study, groundwater polluted with pesticides was found to increase susceptibility to estrogen-mediated cancers, in spite of the ban on these chemicals for decades [46]. Likewise, alarmingly high levels of these chemicals have also been reported from other regions of the world in recent studies [47][48][49]. Additionally, a Brazilian study from the year 2022 linked pesticide residue with an increased risk of cancer in one province [50].
These xenobiotics contaminate water resources and animals living in pesticide-contami nated water, leading to their accumulation in the animals' bodies. The quantity of OCPs in their bodies can be several folds higher than that in the surrounding water. If these animals are consumed by other organisms, the OCPs will enter and accumulate in those animals' bodies (bioaccumulation) [51]. Studies on the OCP levels in edible fishes found alarmingly high levels of these compounds [52][53][54], suggesting that these compounds are bioaccumulating in our bodies at an alarming rate, mainly due to the biomagnification of pesticides through seafood. These compounds are still found in our food and increasing the risk of human diseases despite the ban.
Confirmation of the Pesticide's Xeno-Estrogenicity
Data from the observation of occupational exposure were not sufficient to understand the adverse effects of these OCPs. Nevertheless, animal-and cell-line-based studies unambiguously confirmed the xeno-estrogenic potential of heavily used and well-known pesticides [3,55,56].
Animal-based experiments first confirmed the biological response to DDT comparable to that for estrogens, such as increased uterine weight and vaginal epithelial cornification in lab animals, known as uterotrophic assay [57]. Later, uterotrophic assays of aldrin, dieldrin, endosulfan, and HCH confirmed that all these compounds were xeno-estrogenic in nature [55,56]. Furthermore, a cell-line-based study found evidence of the pesticides modulating estrogen-dependent genes as natural estrogens do [58].
The highest isoform of dichlorodiphenyltrichloroethane, the most well-known pesticide in the world, p,p -DDT (more than 85% of w/w), was found to be estrogenic in animal and cell lines experiments [59]. Additionally, experiments proved that DDT metabolite binds to the estrogen receptor (ER), further consolidating its xeno-estrogenic activity [60].
Dieldrin was reported to cause various forms of cancer in at least seven strains of mice upon oral administration [61]. In one study, dieldrin was found to decrease the binding of natural estrogens to its receptor in female rats' uterine tissue extracts. Additionally, in another animal-based study, the dieldrin treatment was found to competitively inhibit the binding of 17β-estradiol (the most potent form of estrogen) to the receptor in the rat uterus, indicating the strong similarities between these two compounds [62]. Dieldrin treatment at a concentration of 10 µM in MCF-7 cells produced a significant increase in proliferation. The proliferation induced due to dieldrin at this concentration was 54.89% that of estradiol, indicating that it is a strong xeno-estrogen [63]. Both isoforms of the pesticide endosulfan, αand β-endosulfan, were reported to possess xeno-estrogenic potential [64]. Furthermore, atrophy was noticed in different testicular tissues of male rats fed meal containing 10 ppm concentration of endosulfan [65]. In another study, it was noted that endosulfan-induced atrophy relates to biochemical changes in testicular tissue that translate into a significant fall in sperm counts in the testicular epididymis region and reduced spermatid counts in intratesticular tissue [66]. Endosulfan-induced testicular atrophy has also been reported to induce infertility in males due to a dramatic decrease in sperm count after exposure [67].
Chemicals produced by humans and nature possess estrogenic activity. However, the molecular details or structure rendering a chemical/compound its xeno-estrogenic activity are not well known. However, the majority of environmental xeno-estrogens share a common structural motif either of phenol or structurally similar to phenol as observed in all the OCPs [68].
The understanding of mechanisms of endocrine disruptors exerted by such chemicals is growing with time. Scientists proposed that by following these mechanisms, these environmental chemicals (including OCPs) are reported to lead to endocrine disruption (including xeno-estrogenicity) [69]. Figure 1 presents the different mechanisms of action of these compounds:
1.
Mimicking the effect of endogenous steroidal hormones (androgens and estrogens).
3.
Altering the synthesis and metabolism of endogenous steroidal hormones.
4.
Modifying hormone receptor expression in different tissues.
Chemicals produced by humans and nature possess estrogenic activity. However, the molecular details or structure rendering a chemical/compound its xeno-estrogenic activity are not well known. However, the majority of environmental xeno-estrogens share a common structural motif either of phenol or structurally similar to phenol as observed in all the OCPs [68].
The understanding of mechanisms of endocrine disruptors exerted by such chemicals is growing with time. Scientists proposed that by following these mechanisms, these environmental chemicals (including OCPs) are reported to lead to endocrine disruption (including xeno-estrogenicity) [69]. Figure 1 presents the different mechanisms of action of these compounds: 1. Mimicking the effect of endogenous steroidal hormones (androgens and estrogens).
Association of Xeno-Estrogenic Pesticides with Endocrine-Related Cancer
Through human studies, the role of estrogens in female as well as male cancers has been well established. Breast and endometrial cancers in females are primarily estrogenmediated [70]. In males, declining testosterone with age and rising estrogens are known to elevate prostate cancer risk with age [71].
Therefore, the risk of estrogen-mediated cancer has been analyzed in relation to exposure to OCPs by researchers. Risk of breast, endometrium, and ovary cancers in females and risk of prostate and testis cancers in males are estrogen-dependent to some extent [71]. Studies indicate a strong relation between exposure to estrogens as the principal risk factor in the development of these estrogen-mediated cancers [15,16,18,19,32,38,57,60,62,70].
Association of Xeno-Estrogenic Pesticides with Endocrine-Related Cancer
Through human studies, the role of estrogens in female as well as male cancers has been well established. Breast and endometrial cancers in females are primarily estrogenmediated [70]. In males, declining testosterone with age and rising estrogens are known to elevate prostate cancer risk with age [71].
Therefore, the risk of estrogen-mediated cancer has been analyzed in relation to exposure to OCPs by researchers. Risk of breast, endometrium, and ovary cancers in females and risk of prostate and testis cancers in males are estrogen-dependent to some extent [71]. Studies indicate a strong relation between exposure to estrogens as the principal risk factor in the development of these estrogen-mediated cancers [15,16,18,19,32,38,57,60,62,70].
Xeno-Estrogenic Pesticides and Female Cancer
The role of these xeno-estrogenic pesticides in elevated breast cancer risk has been most extensively studied compared to any other cancer in humans [11,[72][73][74][75][76][77][78][79][80][81][82][83]. Over the past few decades, the incidence and mortality of breast cancer have increased worldwide by more than 33% [72,73]. Numerous studies have linked the contaminants we humans are unavoidably exposed to as the prime cause of higher breast cancer risk. Sufficient evidence points to the fact that breast cancer is strongly related to exposure to such contaminants, including DDT and other pesticides that act as estrogenic stimulants [40]. Significantly higher levels of pesticides such as DDT along with its persistent metabolites (especially DDE and DDD), heptachlor, dieldrin, and hexacyclohexane were detected in samples of patients with carcinoma of the breast, irrespective of any other factor when compared to controls [74].
Incidence of breast carcinoma is the world's second highest, and it has the fifth highest mortality rate, while 36.8% of all incidences directly link to lifestyle and/or environmental factors in females over the age of 30 years. A significant association was found between the breast cancer epidemic and environmental contaminants exposure-especially DDT [11,75]. In one of the most significant studies linking OCPs with breast cancer, the decline in mortality rates observed in a decade (1976 to 1986) related significantly to decreases in specific OCPs (DDT and lindane) in colostrum [38]. Studies also suggest that removal of xeno-estrogens can prevent the incidence of and mortality due to breast cancer in females [76]. In prospective studies, exposure to OCPs was measured over a long time before diagnosis, and a statistically significant link was reported between breast cancer risk and DDT exposure. Similarly, two studies consistently linked the pesticide DDT with higher breast cancer risk when significant exposure occurred before teenage years (below the age of 14 years) [77,78]. These observations were further consolidated by a cohort study where it was found that above a certain cut-off concentration of dieldrin (>57.6 ng/g), morbidity and mortality amongst breast cancer patients significantly increased [78]. Dieldrin weakly induced both ER expression and cell proliferation in MCF-7 cell lines in a study, indicating dieldrin to be weakly estrogenic in females [82].
It is not clear how the estrogenic mechanism increases breast cancer risk. In a toxicoproteomic study amongst breast cancer patients, these OCPs were found to downregulate the expression of ER (a common event in a large number of breast cancer cases) by disrupting the relevant pathways [83]. In vitro studies also confirmed estrogen deregulation and increases in the concentration of cellular metabolites that activate a number of oncogenes [84,85]. In a recent study from China, OCPs were not only associated with breast cancer risk but were also found to elevate oxidative stress biomarkers in both serum ad urine [86].
Studies analyzing the role of xeno-estrogenic pesticides in endometrial cancer risk are few and rather inconclusive. Previous studies did not find any correlation between xeno-estrogenic OCP levels and endometrial cancer risk [87,88]. More human and animal experiment-based studies are needed to establish a link between these contaminants and endometrial cancer risk.
Xeno-Estrogenic Pesticides and Male Cancer
The role of xeno-estrogens in modifying the risk of male cancer and other related disorders is heavily debated and not well understood. In utero and early postnatal estrogen exposure is a significant contributory factor for testicular cancer risk in young men [89,90]. Studies from our lab have found a high level of these pesticides in mothers' milk and umbilical cord, indicating continuous transfer of these pesticides during the early development of children [6][7][8].
Endocrine-disrupting chemicals were found to contribute to increased testicular cancer risk in some studies [91][92][93]. Later, this hypothesis was further consolidated when it was found that exposure to different xeno-estrogens in utero increases testicular cancer risk by fourfold in men (between the ages of 16 and 59) [64]. In an occupational study, xenoestrogenic OCPs were found to increase testicular cancer risk among youngsters born to Norwegian farmers from 1952 to 1991, reporting a higher-than-expected testicular cancer incidence compared to controls [94]. In a nested case-control study, it was found that mothers of patients with testicular cancer had significantly higher DDT metabolites in their body during the lactational period [95]. A pooled study found 1.29 times higher testicular dysfunction risk, the main cause of testicular cancer, amongst the population living in regions having higher occupational and non-occupational exposure to OCPs compared to those in regions with limited or lower chances of exposure [96].
Due to the hormonal basis of prostate development, researchers have valid reasons to understand the potential relationship of xeno-estrogenic pesticides with prostate cancer risk [97]. The human prostate gland is the organ most affected by malignant neoplasm in elderly men. With the increase in life expectancy, the incidence of cancer has increased, affecting almost 90% of males over 80 years of age. The basis for this high incidence is not clear despite decades of exhaustive research [98]. In contrast to testicular cancer, direct connections between xeno-estrogenic pesticides and prostate cancer have not been established [90,99]. Animal-and cell-line-based studies are a better indicator of the effect of xeno-estrogenic pesticides on the functioning and transformation of prostate cancer. One study reported that endocrine-disrupting chemicals can transform and/or reprogram prostate stem cells and potentially elevate prostate cancer risk in experimental animals [100].
In an animal-based study, it was found that feeding regular small doses of xeno-estrogens to pregnant females led to a significantly increased prostate weight in adulthood in their pups; increased prostate weight is one of the hallmarks of prostatic disorders, including both prostate cancer and benign prostatic hyperplasia (BPH) [101]. The most compelling proof for a link between xeno-estrogenic OCP exposure and prostate cancer comes from studies from our laboratory. We found that the levels of some of these pesticides were significantly higher amongst patients with higher-stage and more aggressive prostate cancer, indicating the development of an aggressive form of cancer. In addition, these pesticides influence prostate weight as well amongst BPH patients [102,103]. Occupational pesticide exposure is regarded as an established factor elevating prostate cancer risk [104][105][106]. Furthermore, numerous epidemiological studies reported a positive relation between (non-occupational) pesticide exposure and the risk of prostate cancer [106][107][108]. In relatively recent Asian studies, DDT and endosulfan were found to increase prostate cancer risk. This hints that in spite of the partial ban on these two compounds, they may still be increasing the risk [109,110]. However, more studies are needed to understand if there is a relation before drawing any conclusion.
Conclusions
The association of pesticides, including OCPs, with cancer risk is well established. However, some of the mechanisms, such as xeno-estrogenicity, that increase the risk of carcinogenic progression are not clearly understood. Accumulating evidence is consolidating the role of xeno-estrogenic OCPs in the risk of not only cancers but also a plethora of other human disorders. These chemicals, despite their apparent adverse health effects, are still used in some countries. Although the risk of some cancers such as breast cancer is clearly modified with exposure to the discussed chemicals, the link with the risk of other estrogen-mediated cancers, especially in males, is not clearly understood and needs to be studied more. Figure 2 represents the risk of different cancers reported to be associated with exposure to xeno-estrogenic pesticides. In recent decades, breast cancer in females and prostate cancer in males have become the top causes of morbidity and mortality, and both are estrogen-mediated cancers. More studies are needed to find out how much this increased incidence can be attributed to such harmful environmental factors. To the best of our knowledge, this is the only study that focuses on all-gender estrogen-mediated cancer risk modification by xeno-estrogenic OCPs. These chemicals must be completely phased out and replaced with less toxic and affordable alternatives that have negligible adverse health effects on mammalian systems. phased out and replaced with less toxic and affordable alternatives that have negligible adverse health effects on mammalian systems.
Data Availability Statement:
The data used to support the findings of the study are included within the article.
Conflicts of Interest:
The authors declare no conflict of interest.
Data Availability Statement:
The data used to support the findings of the study are included within the article.
Conflicts of Interest:
The authors declare no conflict of interest.
|
v2
|
2022-11-23T06:17:33.455Z
|
2022-11-21T00:00:00.000Z
|
253760489
|
s2orc/train
|
Real-World Experience of Monitoring Practice of Endocrinopathies Associated with the Use of Novel Targeted Therapies among Patients with Solid Tumors
Background: Cancer treatments have gradually evolved into targeted molecular therapies characterized by a unique mechanism of action instead of non-specific cytotoxic chemotherapies. However, they have unique safety concerns. For instance, endocrinopathies, which are defined as unfavorable metabolic alterations including thyroid disorders, hyperglycemia, dyslipidemia, and adrenal insufficiency necessitate additional monitoring. The aim of this study was to assess the prevalence of monitoring errors and develop strategies for monitoring cancer patients who receive targeted therapies. Method: A retrospective chart review was used to assess the prevalence of monitoring errors of endocrinopathies among cancer patients who received targeted therapies over one year. All of the adult cancer patients diagnosed with a solid tumor who received targeted therapies were included. The primary outcome was to determine the prevalence of monitoring errors of endocrinopathies. The secondary outcomes were to assess the incidences of endocrinopathies and referral practice to endocrinology services. Results: A total of 128 adult patients with solid tumors were involved. The primary outcome revealed a total of 148 monitoring errors of endocrinopathies. Monitoring errors of the lipid profile and thyroid functions were the most common error types in 94% and 92.6% of the patients treated with novel targeted therapies, respectively. Subsequently, 57% of the monitoring errors in the blood glucose measures were identified. Targeted therapies caused 63 events of endocrinopathies, hyperglycemia in 32% of the patients, thyroid disorders in 15.6% of them and dyslipidemia in 1.5% of the patients. Conclusion: Our study showed a high prevalence of monitoring errors among the cancer patients who received targeted therapies which led to endocrinopathies. It emphasizes the importance of adhering to monitoring strategies and following up on the appropriate referral process.
Introduction
In the last decade, the robust development of cancer treatments has been achieved, with the management of cancer moving more towards the use of targeted molecular therapies instead of non-specific cytotoxic chemotherapies [1]. The targeted therapies have changed the paradigm of cancer treatment as these are characterized by a unique mechanism of action which is considered to be highly specific for the key cellular biological pathways implicated in the cancer process [2]. Notwithstanding their effectiveness, these new, targeted molecular therapies exhibit certain side effects such as endocrinopathies which have a significant impact on the patients' quality of life in the long run [1].
Endocrinopathies are among the most common side effects associated with the use of immune checkpoint inhibitors and targeted therapies (refer to Appendix A) [3]. Endocrinopathies are defined as unfavorable metabolic alterations which include hyperthyroidism, hypothyroidism, hyperglycemia, hypertriglyceridemia, hypercholesterolemia, hypogonadism and adrenal insufficiency. The incidence of endocrinopathies with the use of immune checkpoint-blocking antibodies and targeted therapies has been difficult to accurately state due to the varied methods of assessment, diagnosis and monitoring in different clinical trials [4]. The most commonly reported endocrinopathies are hypothyroidism, hyperthyroidism, hyperglycemia, dyslipidemia and hypophysitis [5][6][7][8][9]. As a result of numerous endocrinopathies being reported in real-life practice, the European Society for Medical Oncology (ESMO), the American Society of Clinical Oncology (ASCO) and the National Comprehensive Cancer Network (NCCN) have established clinical practice guidelines for the diagnosis and management of immunotherapy-related toxicities [10][11][12]. On the other hand, targeted therapies which include small molecule tyrosine-kinase inhibitors (TKIs), the mammalian target of rapamycin (mTOR) inhibitor, immunomodulatory agents and other monoclonal antibodies have had no established guidelines for clinical practice in the diagnosis, monitoring and management of their related toxicities apart from the recommendations that are mentioned in their prescribing information. As many studies have shown, novel targeted therapies have durable clinical benefits which highlight the importance of establishing safe standards in preventing endocrinopathies that are associated with the use of novel targeted therapies [3][4][5][6][7][8][9].
Hence, novel targeted therapies require additional monitoring and an appropriate clinical review [13]. The failure to review a prescribed targeted therapy is defined as a monitoring error. The American Society of Health System Pharmacists (ASHP) defined a monitoring error as a failure to review a prescribed regimen for the appropriateness and detection of problems, or the failure to use appropriate clinical or laboratory data for an adequate assessment of the patient's response to prescribed therapy [14].
Di Lorenzo et al. published a systematic analysis of the abstracts issued by ASCO and ESMO between 2005 and 2010. Their aim was to describe the side effects associated with targeted therapy that are used to treat metastatic renal cell carcinoma (mRCC) and its management. Furthermore, they investigated the incidence and grading of the toxicities associated with certain targeted therapies which are used in the management of mRCC, and they concluded that targeted therapies caused severe toxicities, thereby requiring external specialist consultation and treatment modification [15]. Another review article was published in 2013 by Grace K et al. with the aim of providing a brief overview of the toxicities associated with novel agents and relevant implications for the management of their side effects in cancer patients. They found that targeted therapy agents rose to unanticipated toxicities despite their efficacy profile due to previously unknown mechanisms and/or the multiplicity of the affected off-target proteins [16]. Moreover, many studies have evaluated the incidence and prevalence of targeted therapy-related toxicity, and they have recommended effective strategies to prevent these safety hazards [17,18].
The growing evidence of a significant number of incidences of endocrinopathies associated with the use of novel targeted agents means that healthcare providers ought to be aware of these side effects and know how to monitor them periodically.
At our institution, multiple targeted therapies are used for a wide variety of oncology indications. These medications can cause many types of endocrinopathies that need close monitoring and regular follow-ups in cancer patients. Unfortunately, there is no standard practice for the safe monitoring of these medications to ensure patient safety, especially since this medications usually given in an outpatient setting. The aim of our study was to assess the prevalence of monitoring errors of endocrinopathies and then, develop strategies for safe practices in the monitoring of patients who are undergoing targeted therapies.
Study Design
This retrospective cohort study evaluated the prevalence of monitoring errors of endocrinopathies in patients receiving novel targeted therapies who were included in our hospital formulary for the treatment of adult patients with solid tumors at the National Guard of Health Affairs-King Abdulaziz Medical City-in the western region (KAMC-WR), Saudi Arabia, during the period from 1 June 2016 to 31 December 2017. The Investigational Review Board's (IRB) approval was obtained in February 2018.
Study Population
All of the adult patients who were 18 years old or older, had been diagnosed with a solid tumor, and received one of the novel targeted therapies that is available in our formulary (i.e., nivolumab, atezolizumab, everolimus, sorafenib, sunitinib, pazopanib, regorafenib, or abiraterone) were eligible for the study. The patients treated with novel targeted therapies for hematological malignancies were excluded from the study due to limited time allocated for the study.
Study Procedures
The patients were identified through the Information System Development (ISD) department, and the data were retrieved through the patients' electronic medical records. We reviewed the electronic medical records using hospital information systems to obtain the pertinent laboratory and medical data. For each patient, we recorded the following variables: demographic characteristics (age, gender, diagnosis, date of diagnosis and comorbidities) and the presence of previously reported adverse drug events (ADEs) that were secondary to novel targeted therapies treatment. Furthermore, the percentage of monitoring errors associated with the use of novel targeted therapies among the cancer patients occurred at the baseline and during the follow-up visits, and they were defined as a failure to use appropriate clinical or laboratory data (as fasting blood glucose, random blood glucose, HbA1c, T 3 , T 4 , Thyroid-stimulating hormone (TSH) and lipid profiles (LDL, triglyceride and total cholesterol) for the adequate assessment and appropriate monitoring of the patient's response to a prescribed therapy as per recommended. These monitoring errors were identified by using the Lexicomp program (online commercial drug information database) and the Food and Drug Administration (FDA) who disseminate information for each medication (refer to Appendix A). Moreover, we tracked the patients who received the novel targeted therapies at our institution and had any endocrinopathies based on the clinician assessment and the laboratory results. Then, we assessed the percentage of referrals to the endocrinologist for the indicated patients with endocrinopathies. The data were collected and analyzed over a period of 6 months.
Statistical Analysis
The Statistical Package for the Social Sciences (SPSS) version 24 was used for data analysis. The descriptive statistics are shown as the means (95% confidence intervals, CI), frequencies or percentages when it is appropriate. The data are organized and summarized in the tables.
Results
A total of 128 adult patients diagnosed with solid tumors were included in this study. The baseline characteristics of the participants are shown in Table 1. Among 128 oncology patients, 14 (11%) patients received immune checkpoint inhibitors (nivolumab, atezolizumab), while 114 (89%) patients were on TKIs, mTOR and abiraterone. We found the inadequate monitoring of thyroid function and blood glucose in thirteen (92.8%) and two (14.2%) patients on the immune checkpoint inhibitors, respectively, as they required thyroid function tests and blood glucose test at the baseline and periodically (Table 2). While 50 (92.6%) patients on TKIs and 52 (64.2%) patients on sunitinib, everolimus and abiraterone had inadequate monitoring of their thyroid function and blood glucose levels, respectively, as they required thyroid function tests and blood glucose tests at the baseline and periodically. We found that up to 31 (94%) patients had monitoring errors associated with the use of oral everolimus which required a lipid profile test at the baseline and periodically (Table 3). Furthermore, we found that novel targeted therapy had caused 63 incidents of endocrinopathies among adult solid tumor patients, including hyperglycemia in 32% (41) of the patients, hypothyroidism and/or hyperthyroidism in 15.6% (20) of the patients and dyslipidemia in 1.5% (2) of the patients. A total of 11 patients were lost in their follow-up due to their referral to other oncology center or disease progression. Moreover, we assessed the referral to the endocrinologist for the indicated patients, and found that nearly 27% (17 patients from a total of 63 patients) of those who had endocrinopathies were referred to an endocrine specialist (Table 4).
Discussion
Despite the efficacy of the new, targeted molecular therapies, several studies have highlighted their side effects such as endocrinopathies [3][4][5][6][7][8][9]. Endocrinopathies are among the most common side effects associated with the use of immune checkpoint inhibitors and targeted therapies, which have a major impact on the patients' quality of life in the long term [5][6][7][8][9]. Establishing clinical practice guidelines for the diagnosis and management of immunotherapy-related toxicities is essential to ensuring the patients' safety and possible efficacy [10][11][12]. Our study has shown a high incidence of endocrinopathies among the cancer patients who received a novel targeted therapy, which has been difficult to state precisely due to the variable methods of assessment, diagnosis and monitoring. The most common endocrinopathies are hyperglycemia, hypothyroidism and hyperthyroidism. These endocrinopathies, which are caused by novel targeted therapies, can be prevented through proper monitoring and early detection, as it is supported in many studies [19][20][21][22][23][24].
A delay in identifying the endocrinopathies associated with targeted molecular therapies can limit the opportunities to provide supportive care to minimize or adequately manage these side effects. Hence, any failure to review a prescribed regimen for appropriateness or to use appropriate clinical or laboratory data for an adequate assessment of the patient's response or safety could lead to a monitoring error [14]. In 2016, Marabelle and his colleagues came up with some collaborative institutional guidelines that emphasize the five pillars of immunotherapy toxicity management. These pillars include how to prevent, anticipate, detect, treat and monitor immunotherapy-related toxicity [25]. A systematic review of ten and thirty-eight randomized control trials published in 2014 and 2018, respectively, demonstrated that the use of immune checkpoint inhibitors and TKIs are associated with an increased risk of hypothyroidism, hyperthyroidism, adrenal insufficiency and hypophysitis [26,27].
Our study found the inadequate monitoring of the thyroid functions in patients who are on novel targeted medication (i.e., nivolumab, atezolizumab, sorafenib, sunitinib, pazopanib and regorafenib). We also found that 94% of the study patients had an inadequate monitoring of their lipid profiles whilst they were on everolimus, while 57% of the patients who were on novel targeted medications (i.e., nivolumab, atezolizumab, everolimus, sunitinib and abiraterone) had monitoring errors in their blood glucose levels.
Moreover, when we assessed the referral to the endocrinologist for the indicated patients, we found that only 27% of the affected patients were referred which is considered to be a medical malpractice and inconsistent with the current experts' recommendations [10][11][12][20][21][22]28]. Many studies recommend that oncologists should seek an endocrinologist specialist or internist support for two reasons: for oncologists to learn the proper management of specific endocrinopathies toxicities, and for the organ specialists to increase their knowledge about these new drug-related toxicities, thereby creating a virtuous circle for the patients' management [25,29,30].
To the best of our knowledge, this is the first published study of its kind that investigated the monitoring practice of endocrinopathies associated with the use of novel targeted therapies. Additionally, due to this established association among targeted molecular therapies and endocrinopathies, even with a small number of patients, it is now recommended that these toxicities should be identified as they may easily be overlooked among many patients.
Our study has several limitations such as we only observed a single institution and we included novel targeted therapies that were available during the study period so we were unable to assess if the patients had access to different healthcare systems in order to manage their chronic diseases which could contribute to the accuracy of the study results. Moreover, 11 patients were lost in their follow-up due to their referral to other oncology centers or disease progression, and as such, they could not be included to determine the true incidence of endocrinopathies, and this might have led to failure in monitoring. The sample size was not calculated as we include all of the cancer patients who received the selected medications for solid tumor indications at our center during the period from 1 June 2016 to 31 December 2017. Despite these limitations, the study demonstrated a high prevalence of monitoring errors associated with the use of targeted anticancer medications, and we identified the incidence of endocrinopathies that were secondary to novel targeted therapies among adult patients with solid tumors. The strengths of the study include the variety of sources that were used to identify the incidence of endocrinopathies. Rather than epidemiologically valid incidence estimates, these reports provide a descriptive account of a whole range of medication-related monitoring errors and facilitate the development of interventions to address the areas of risk and improve the patients' quality of life.
Conclusions
Our study demonstrated a high prevalence of monitoring errors associated with the use of targeted anticancer medications and identified the incidence of endocrinopathies that were secondary to novel targeted therapies among adult patients with solid tumors. The causes of these errors have been identified as being multifaceted, involving various members of a multidisciplinary team. Our findings emphasize the importance of establishing monitoring guidelines for novel targeted therapies and the need for collaborative efforts among health care providers to optimize the cancer treatment outcomes. To that end, several strategies should be considered and implemented to overcome monitoring errors, as novel chemotherapy now constitutes a new era in cancer treatment. Informed Consent Statement: Not applicable. Patient informed consent was not required, as no invasive procedure was applied to the patients.
Data Availability Statement: Not applicable.
Acknowledgments: The authors would like to recognize and thank all of the KAIMRC team members who participated and served in the proposal review and approval. We would also like to thank the reviewers, whose comments helped to improve this publication.
Conflicts of Interest:
The authors declare no conflict of interest.
|
v2
|
2022-11-23T16:10:23.424Z
|
2022-11-21T00:00:00.000Z
|
253791584
|
s2ag/train
|
Identifying Good Candidates for Active Surveillance of Ductal Carcinoma in Situ: Insights from a Large Neoadjuvant Endocrine Therapy Cohort
Ductal carcinoma in situ (DCIS) is a biologically heterogenous entity with uncertain risk for invasive ductal carcinoma (IDC) development. Standard treatment is surgical resection often followed by radiation. New approaches are needed to reduce overtreatment. This was an observational study that enrolled patients with DCIS who chose not to pursue surgical resection from 2002-2019 at a single academic medical center. All patients underwent breast magnetic resonance imaging (MRI) exams at three to six-month intervals. Patients with hormone-receptor positive disease received endocrine therapy. Surgical resection was strongly recommended if clinical or radiographic evidence of disease progression developed. A recursive partitioning algorithm incorporating breast MRI features and endocrine responsiveness was used retrospectively to stratify risk of IDC. 71 patients were enrolled, two with bilateral DCIS (73 lesions). 34 (46.6%) were premenopausal, 68 (93.2%) were hormone-receptor positive, and 60 (82.1%) were intermediate or high-grade lesions. Mean follow-up time was 8.5 years. Over half (52.1%) remained on active surveillance without evidence of IDC with mean duration of 7.4 years. Twenty patients developed IDC, of which six were HER2-positive. DCIS and subsequent IDC had highly concordant tumor biology. Risk of IDC was characterized by MRI features after six months of endocrine therapy exposure; low, intermediate, and high-risk groups were identified with respective IDC rates of 8.7%, 20.0%, and 68.2%. Thus, active surveillance consisting of neoadjuvant endocrine therapy and serial breast MRI may be an effective tool to risk-stratify patients with DCIS and optimally select medical or surgical management.
|
v2
|
2022-11-23T16:17:34.608Z
|
2022-11-21T00:00:00.000Z
|
253789908
|
s2ag/train
|
Association of Prostate Cancer and Lipid Profile: A Case-Control Study
There are inconsistent findings concerning the association between the serum concentrations of lipid parameters and prostate cancer (PCa), particularly in Caucasian men. There is limited data on men of African ancestry. The study examined the relationship between serum total cholesterol (TC) levels and its fractions and PCa in a hospital-based case-control study in Jamaica. The serum levels of TC, triglycerides (TG), HDL-cholesterol (HDL-C), and LDL-cholesterol (LDL-C) in 46 male patients (cases) who underwent prostate biopsy were measured over an eighteen month period. There were 32 patients without PCa who served as controls. The serum lipid concentrations between cases and controls were compared using an independent samples t-test. Multiple linear regression and binary logistic regression were used to assess the relationship between lipids and overall PCa, as well as disease severity. Based on the results, there were no significant differences between the concentrations of lipids for the cases and controls. The results of the regression analysis revealed that the serum lipid levels were not significant predicators of overall PCa. The outcomes of the binary regression analysis showed the same for PCa severity. The study concluded that there was no association between serum levels of lipids and overall PCa as well as disease severity at the time of diagnosis in the sample of Jamaican men.
|
v2
|
2022-11-26T14:56:26.822Z
|
2022-11-21T00:00:00.000Z
|
253985505
|
s2orc/train
|
Application of Indocyanine Green Fluorescence Imaging Combined with Laparoscopic Ultrasound in Laparoscopic Microwave Ablation of Liver Cancer
Background The aim of this study was to assess the effect of indocyanine green (ICG) fluorescence imaging combined with laparoscopic ultrasound in laparoscopic microwave ablation of liver cancer. Material/Methods This study retrospectively analyzed 61 patients who underwent laparoscopic microwave ablation of liver cancer, including laparoscopic microwave ablation with and without ICG fluoroscopy. Results The operative times, ablation times, postoperative hospital stay, postoperative complication rate, hospitalization cost, postoperative liver function changes, and postoperative overall survival were similar between the 2 groups, but there was a statistically significant difference in recurrence-free survival (P<0.05). A total of 5 lesions were found in the fluorescence laparoscopy group that were not found by preoperative imaging, while no new lesions were found in the ordinary laparoscopy group. Fluorescence laparoscopy has obvious advantages over ordinary laparoscopy in finding small lesions that were not found before surgery. In terms of complete ablation rate, 3 patients in the ordinary laparoscopy group and 1 patient in the fluorescence laparoscopy group were judged to be incompletely ablated and were ablated again at 1 month after the operation. Conclusions For small hepatocellular carcinoma with severe liver cirrhosis and located on the liver surface, fluorescence laparoscopy can better reveal the location and boundary of the tumor, and fluorescence laparoscopy can detect tiny lesions that cannot be detected by preoperative imaging. The combination of fluorescence laparoscopy and microwave ablation has a good effect on the treatment of small hepatocellular carcinoma located on the surface of the liver that is difficult to distinguish.
Background
Primary liver cancer is one of the most common cancers in China [1], ranking fourth in incidence and third in mortality rate [2]. Surgical resection remains the preferred treatment for liver cancer [3], but some patients are not candidates for hepatectomy due to cirrhosis, hepatic insufficiency, or poor general health. Local ablation, including microwave ablation, is a widely used minimally invasive treatment for small liver cancers, especially in patients with severe cirrhosis. Laparoscopic microwave ablation of liver cancers is a safer and more effective treatment for liver cancers in some anatomic sites [4], such as beneath the liver capsule, and lesions protruding beyond the liver capsule, in the diaphragmatic dome, or close to the gallbladder and gastrointestinal tract, and adjacent to great vessels and the hepatic portal [5,6]. Accurately locating tiny tumor tissue under laparoscopic direct vision has become a key factor to improve the success rate of laparoscopic microwave ablation. Indocyanine green (ICG) fluorescence imaging during laparoscopy has emerged as an auxiliary technique in recent years [7]. ICG injected into the human bloodstream binds to proteins and emits near-infrared light with a wavelength of 840 nm when excited by light at 750-810 nm. The emitted near-infrared light can penetrate connective tissues at a 5-10 mm thickness [8][9][10]. ICG can be used to indicate tumor position and delineate tumor boundaries based on this principle [11,12]. From 1 April 2019 to 31 December 2021, our department used the intraoperative navigation of ICG fluorescence imaging laparoscopy, combined with laparoscopic ultrasound, to perform fluorescence laparoscopic microwave ablation of liver cancer in 29 patients with small liver cancer. Microwave ablation of hepatocellular carcinoma was performed in 32 patients with small hepatocellular carcinoma. The data and results of the 2 groups of patients are reported as follows.
Clinical Data
We recruited 61 patients who underwent laparoscopic microwave ablation to treat liver cancers at the First Affiliated Hospital of the University of Science and Technology of China from 1 April 2019 to 31 December 2021 with a total of 67 tumors in 61 patients. The research was approved by the Ethics Committee of Anhui Provincial Hospital, He Fei, China (No. 2021-S(H)-017). All patients signed an informed consent form conforming to medical ethics requirements before surgery. Inclusion criteria were: (1) pathologically confirmed or clinically diagnosed with primary liver cancer using 2 or more imaging modalities (liver MRI, upper-abdomen contrast-enhanced CT, ultrasonography, and PET-CT) combined with tumor markers (AFP, DCP, and human prothrombin complex); (2) single or multiple liver cancers with a diameter £3 cm and totaling <3; (3) no evidence of vascular invasion, bile duct invasion, or extrahepatic metastases; and (4) assessed as class A or B according to the Child-Pugh classification system. The exclusion criteria were: (1) single or multiple liver cancers with a diameter >3 cm and totaling >3; (2) combined with extrahepatic metastases; (3) class C liver function according to the Child-Pugh classification system; and (4) severe underlying diseases, such as heart and renal failure. There were 35 males and 26 females with an average age of 58.91 years. Among the 61 patients, 54 and 7 had class A and B liver function, respectively, according to the Child-Pugh classification system. According to the 2020 China liver cancer (CNLC) staging system, 55 and 6 patients were classified as CNLC stage Ia and Ib, respectively. All patients were thoroughly evaluated before surgery to exclude surgical contraindications. All tumors located inside the liver parenchyma and positive iodine allergy test were in the general laparoscopy group.
Equipment and Instruments
We used a high-resolution fluorescence laparoscope manufactured by Guangdong Optomedic Technologies, Inc. (Guangdong, China). The microwave ablation system was the ECO-100C cold cycle microwave knife (Nanjing Yigao Microwave System Engineering Co., Ltd., Nanjing, China). We used an ECO-100AI8 disposable microwave ablation needle (2.0×150.0 mm).
Surgical Method
The patients were placed in supine position after endotracheal anesthesia was established. The right shoulder was slightly elevated in some patients who had cancers in the right posterior lobe of the liver. A pneumoperitoneum was established, with the pressure maintained at 10-14 mmHg and the head elevated. The laparoscope was inserted through a subumbilical incision to inspect the peritoneal cavity, the tumor position, number of tumors, and whether there was a need to dissociate the liver from the peri-adherent tissues. If necessary, 2 to 4 incisions were made in the following positions: below the xiphoid process, right midclavicular line, below the costal margin at the anterior axillary line, and below the costal margin at the left midclavicular line. The specific positions of the additional incisions were adjusted according to the patients' body size, tumor position, and intraoperative manipulations performed. The ablation fully covered and exceeded the liver tumor by 0.5-1 cm. For larger tumors, overlapping ablation was performed.
Timing and Dose of ICG Injection
All patients underwent iodine allergy testing before surgery. An ICG injection was contraindicated in patients with positive iodine allergy testing. Otherwise, ICG was injected into a peripheral vein 2-3 days before surgery at a dose of 0.3-0.5 mg/kg.
Postoperative Follow-Up
The patients were re-examined radiographically (by contrastenhanced MR, or contrast-enhanced CT, or CEUS of the liver) 1 month after surgery, and the ablation effect was assessed by tumor-related indicators (eg, AFP, DCP). Contrast-enhanced ultrasonography indicates that there is no contrast agent filling inside the tumor, and contrast-enhanced CT or MR indicates that the tumor is not enhanced, indicating that the ablation is complete. If there is still tumor residue, re-ablation can be performed until the tumor is completely inactivated. The patient was followed up every 2-3 months in the outpatient clinic thereafter. Color Doppler ultrasound or CT showed new lesions around the ablated tumor, in the liver, or in other parts, which was defined as recurrence.
Statistical Methods
Statistical analyses were performed using SPSS26.0 software. Quantitative data are expressed as c _ ±s, and the intergroup comparison was conducted using an independent-samples t test. Qualitative data are expressed as n (%), and the intergroup comparison was made using Pearson's c 2 test/adjusted c 2 test. P<0.05 was regarded as indicating a significant difference. The Kaplan-Meier method was used to compare the overall survival and recurrence-free survival between the 2 groups.
Results
The 2 groups of patients were comparable with respect to general preoperative information inpatient observation index, and postoperative complete ablation rate (P>0.05; Table 1). The surgery was successfully completed in all 61 patients. There was no postoperative bleeding or deaths. The 2 groups of patients did not differ significantly in length of hospital stay, operative time, ablation time, hospitalization cost, complications, and liver function ( Table 2). The patients began to consume a liquid diet on the first day after surgery. The average time to discharge was 4.27 days. All 61 patients completed follow-up evaluations as scheduled. By 18 March 2022, the patients had been followed for 3-41 months, with an average of 23.21±11.23 months. Metastasis or recurrence occurred in 7 patients (1 in the fluorescence laparoscopy group and 6 in the laparoscopy group), and 2 patients died (1 in the fluorescence laparoscopy group died 15 months after the operation, and 1 in the laparoscopy group died after 18 months). The overall survival time of the 2 groups of patients was not significantly different, but the recurrence-free survival time was significantly different (P<0.05) (Figures 1, 2).
Discussion
Local ablation and surgical resection had similar efficacy for a single lesion £5 cm or 2-3 lesions with a maximum diameter
e937832-4
£3 cm not combined with vascular, bile duct, and adjacent organ invasion or distant metastases, and class A and B liver function. These patients usually achieved a radical cure by local ablation [13]. Local ablation has the advantages of simplicity and effectiveness, less trauma, better safety and reliability, and wide indications, and has been widely used in the treatment of small hepatocellular carcinoma. Microwave ablation uses heat produced by high-speed rotational friction of polar molecules and collision movement of polar ions. This process usually creates considerable heating in the tissues. Protein denaturation can be induced in tumor cells when the tissue temperature is >60°C, resulting in irreversible necrosis. The tumor cells release heat shock proteins, which activate the immune response to inhibit tumor spread [14,15].
Microwave ablation can be delivered via 3 pathways: percutaneous, laparotomy, and laparoscopic. B-mode-guided, CT-guided, and guided microwave ablation of liver cancers have the benefits of convenience and minimal invasiveness [4]. This ablation procedure has been widely used clinically, but it requires highlevel physician skill in puncture techniques under the guidance of images such as B-ultrasound, CT, or MRI. In addition, intense pain may be induced because the liver capsule is irritated during the microwave ablation. Body position changes and respiratory movement can lead to inaccurate puncture or displacement of the ablation needle, thus resulting in incomplete ablation, as shown in Figure 3. Using B-mode-, CT-and MRI-guided percutaneous microwave ablation, the clinician is operating blindly. The ablation procedure is more likely to cause bleeding, tumor rupture, and incidental injury to the surrounding organs and tissues if the liver cancers are protruding outside the liver capsule, located in the diaphragmatic dome or near the gallbladder and the gastrointestinal tract, or near the great vessels and hepatic portal [16]. Lesions close to the portal vein are even more difficult to manage and are a potential risk factor for incomplete ablation [17]. Laparoscopic microwave ablation is a better treatment for liver cancers in these positions [18][19][20].
Liver cancer patients undergoing microwave ablation usually have severe cirrhosis and cirrhotic liver nodules of varying sizes on the liver surface [21,22]. Laparoscopic ultrasound cannot demonstrate the liver surface or accurately detect tumor position in these patients. The cancers are difficult to differentiate from cirrhotic liver nodules. Laparoscopic ultrasound with ICG fluorescence navigation can visualize the tumors, as shown in Figure 4. In patients undergoing conventional laparoscopy combined with laparoscopic ultrasonography, the tumor cannot be found and located at all, but the tumor can be clearly located under ICG fluorescence laparoscopy. The advantage of ICG fluorescence laparoscopy over conventional laparoscopy is the ability to detect micrometastases and satellite lesions during surgery [23,24]. Intraoperative detection and management of micrometastases other than primary lesions are very important for patient prognosis. Preoperative imaging examinations, including enhanced CT, MRI, and contrast-enhanced ultrasonography, cannot easily identify small liver cancer lesions, and small lesions are easily confused with cirrhotic nodules in imaging and cannot be located during surgery [25]. Intraoperative ultrasound has a low sensitivity for small hepatocellular carcinoma lesions, is difficult to distinguish from normal cirrhotic nodules, and has extremely high requirements for the surgeon [26]. ICG fluorescence imaging has high sensitivity for small liver cancer. It can detect and locate liver cancer lesions in real time during surgery, and can also detect small lesions not found by preoperative imaging examination, thereby reducing the residual risk of tumor and improving the prognosis of patients. The comparison of the 2 groups of patients in the present study shows that the ability of ICG fluorescence laparoscopy to detect small lesions that cannot be detected by preoperative imaging is significantly better than that of conventional laparoscopy, so the patients in the ICG fluorescence laparoscopy group had better recurrence-free survival. There was no significant difference in the overall survival of the 2 groups of patients, which may be related to the shorter follow-up time and the small number of cases. Multicenter studies with larger samples and longer follow-up are needed.
It is noteworthy that some cirrhotic liver nodules, hepatic cysts, and hyperplastic bile ducts are also capable of ICG uptake, thus leading to false-positive results [27]. Differentiation can be made by surgeons based on their experience and by combining laparoscopic ultrasound findings with preoperative imaging.
We identified 4 major benefits of laparoscopic microwave ablation of liver cancers with ICG fluorescence navigation through the comparative analysis of 29 cases of fluorescence laparoscopic microwave ablation of liver cancer and 32 cases of conventional laparoscopic microwave ablation of liver cancer undergoing this procedure in our department: (1) This procedure has high safety, and all manipulations are done under direct vision. In addition to dissociating the liver and the perihepatic adhesion, safe spaces can be created by artificial ascites or separation with wet gauze for cancers in special positions, such as lesions close to segments VII and VIII of the right liver and lesions close to the gallbladder and the gastrointestinal tract. In this way, undesired damage caused by needle puncture and burn injury caused by heat conduction can be avoided. (2) The procedure has high reliability and can be repeated under direct vision. The procedure also allows for puncture from different directions, thereby preventing incomplete ablation as much as possible. (3) The procedure achieves a more precise tumor localization by combining fluorescence navigation with laparoscopic ultrasound. The tumor position and the relationship between the tumor and the surrounding blood vessels and organs can be more accurately determined. A clearer delineation of tumor boundaries is conducive to manipulations during the ablation. Liver cancers can be difficult to differentiate from the cirrhotic liver nodules in some patients when combined with severe cirrhosis. This situation may still exist, even under laparoscopic ultrasound. ICG staining offers a good solution to this problem. (4) For detection of small lesions, ICG staining can reveal minimal lesions that are otherwise not visible using preoperative imaging modalities [28].
However, the application of fluorescence laparoscopy also has certain limitations. For tumors located inside the liver parenchyma, fluorescence cannot be localized and imaged. In addition, fluorescence cannot monitor the ablation effect in real time, and cannot judge whether the ablation is complete. In addition, in this study, tumors located deep in the liver parenchyma were more difficult to ablate than tumors located on the surface of the liver. The advantages of fluorescence laparoscopy in microwave ablation of liver cancer require further e937832-6 research with large samples, multicenter evaluation, and longer follow-up.
Conclusions
To conclude, ICG fluorescence navigation during laparoscopic microwave ablation of liver cancers achieved good efficacy. This procedure has advantages compared with conventional laparoscopic microwave ablation. The complete ablation rate was even higher for those patients with severe cirrhosis and small liver cancers on the liver surface. In addition, the new ablation procedure more accurately detected tumor position and differentiated between cancers and cirrhotic liver nodules, especially in patients with severe cirrhosis. More importantly, fluorescence laparoscopy can effectively detect small lesions that cannot be displayed by preoperative imaging. Therefore, laparoscopic microwave ablation with ICG fluorescence navigation has high clinical utility for liver cancers and is highly recommended.
Declaration of Figures' Authenticity
All figures submitted have been created by the authors, who confirm that the images are original with no duplication and have not been previously published in whole or in part.
|
v2
|
2022-12-06T06:17:16.955Z
|
2022-11-21T00:00:00.000Z
|
254245141
|
s2ag/train
|
[Bioinformatics analysis of core differentially expressed genes in hepatitis B virus-related hepatocellular carcinoma].
OBJECTIVE
To identify the core genes associated with the development and progression of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC), so as to provide insights into the elucidation of pathogenesis of HBV-related HCC.
METHODS
GSE55092 and GSE121248 datasets were downloaded from the Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) between HCC and peri-cancer tissues were screened using the R package, and the volcano map of DEGs were plotted. The DEGs were subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses, and a protein-protein interaction (PPI) network was created. The hub DEGs were screened using Molecular Complex Detection (MCODE) and cytoHubba plugins in the open-access platform Cytoscape 3.9.0. Then, the screened hub DEGs were validated for differential expression and survival analysis using clinical sample data captured from the UALCAN and Kaplan Meier-plotter databases.
RESULTS
A total of 1 148 and 686 DEGs were screened between HCC and peri-cancer tissues in GSE55092 and GSE121248 datasets, including 703 and 477 down-regulated genes and 445 and 209 up-regulated genes, respectively. A total of 557 common DEGs were screened between GSE55092 and GSE121248 datasets, including 384 down-regulated genes and 173 up-regulated genes. GO enrichment analysis showed that these DEGs were significantly enriched in biological processes of cell division, cell proliferation, redox process, immune response and proteolysis, cellular components of cell nucleus, cytoplasm, extracellular vesicle and endoplasmic reticulum membrane, and molecular functions of binding to calcium ion, protein kinase, DNA and heme. KEGG pathway analysis revealed that these DEGs were significantly enriched in pathways of cell cycle, oocyte meiosis, metabolic pathway, antibiotic biosynthesis and p53 signaling. PPI network analysis identified 10 DEGs, including CDK1, CCNB1, CCNA2, TOP2A, AURKA, CCNB2, KIF11, CDC20, KIF20A and BUB1B, and CDK1, KIF11 and KIF20A were found to be differentially expressed and correlate with poor prognosis among HBV-related HCC patients following clinical sample data validation.
CONCLUSIONS
CDK1, KIF11 and KIF20A may play a critical role in the development and progression of HBV-related HCC, which may be potential diagnostic biomarkers and therapeutic targets of HBV-related HCC.
|
v2
|
2022-11-21T16:08:45.482Z
|
2022-11-26T00:00:00.000Z
|
253728351
|
s2orc/train
|
Integrative analysis of platelet-related genes for the prognosis of esophageal cancer
BACKGROUND Every year, esophageal cancer is responsible for 509000 deaths and around 572000 new cases worldwide. Although esophageal cancer treatment options have advanced, patients still have a dismal 5-year survival rate. AIM To investigate the relationship between genes associated to platelets and the prognosis of esophageal cancer. METHODS We searched differentially expressed genes for changes between 151 tumor tissues and 653 normal, healthy tissues using the “limma” package. To develop a prediction model of platelet-related genes, a univariate Cox regression analysis and least absolute shrinkage and selection operator Cox regression analysis were carried out. Based on a median risk score, patients were divided into high-risk and low-risk categories. A nomogram was created to predict the 1-, 2-, and 3-year overall survival (OS) of esophageal cancer patients using four platelet-related gene signatures, TNM stages, and pathological type. Additionally, the concordance index, receiver operating characteristic curve, and calibration curve were used to validate the nomogram. RESULTS The prognosis of esophageal cancer was associated to APOOL, EP300, PLA2G6, and VAMP7 according to univariate Cox regression analysis and least absolute shrinkage and selection operator regression analysis. Patients with esophageal cancer at high risk had substantially shorter OS than those with cancer at low risk, according to a Kaplan-Meier analysis (P < 0.05). TNM stage (hazard ratio: 2.187, 95% confidence interval: 1.242-3.852, P = 0.007) in both univariate and multivariate Cox regression and risk score were independently correlated with OS (hazard ratio: 2.451, 95% confidence interval: 1.599-3.756, P < 0.001). CONCLUSION A survival risk score model and independent prognostic variables for esophageal cancer have been developed using APOOL, EP300, PLA2G6, and VAMP7. OS for esophageal cancer might be predicted using the nomogram based on TNM stage, pathological type, and risk score. The nomogram demonstrated strong predictive ability, as shown by the concordance index, receiver operating characteristic curve, and calibration curve.
INTRODUCTION
Esophageal cancer incidence is increasing globally, and western countries have a higher prevalence of adenocarcinomas. Over 95% of esophageal cancer patients in China are diagnosed with esophageal squamous cell carcinomas; males are more likely than females to have this kind of cancer [1]. The 5-year survival rate for esophageal cancer patients is a little less than 20% despite improvements in therapy [2]. Target therapy and immunotherapy have made significant advancements in the treatment of esophageal cancer, although only some patients may benefit from them. Therefore, biomarkers to guide therapy and predict prognosis are urgently needed for patients with esophageal cancer.
In recent years, several attempts have been undertaken to better forecast the biology of each individual in esophageal cancer and to find prognostic and predictive biomarkers. Despite this, individuals with esophageal cancer often have a dismal prognosis because there are currently no reliable biomarkers for prognosis prediction and early identification.
Platelets may help tumor cells escape the immune system if they abnormally concentrate in the peripheral circulation and survive for prolonged periods of time. Platelet-derived growth factors, such as platelet factor 4, thrombospondin, and vascular endothelial growth factor, aid in the adhesion, invasion, angiogenesis, and tumor development of hematogenous cancer [3]. These findings suggest that platelets may be crucial in the growth of tumors in a variety of cancer types. Platelet research is mostly focused on the relationship between platelet characteristics and the prognosis of malignant tumors [4,5]. It may be possible to find novel treatment targets and improve the prognosis of malignant tumors by better understanding platelet-related genes and the underlying processes.
The link between platelet-related genes and survival in esophageal cancer patients has never been addressed, and only a small number of studies have looked at this relationship. This study examined the expression patterns of platelet-related genes and their prognostic significance in esophageal cancer by bioinformatics analysis. Finally, using pathway enrichment analysis to uncover probable biological roles for the platelet-related gene signature, we further investigated its function. November 26, 2022 Volume 10 Issue 33
MATERIALS AND METHODS
Animals were not used in the study nor were there any human participants, data, or tissue. A public database served as the source for all the data.
Data collection
The Cancer Genome Atlas website at https://portal.gdc.cancer.gov/ allows users to obtain clinical and gene expression data of patients with esophageal cancer. Gene expression levels in normal tissue samples were extracted using Genotype-Tissue Expression databases (https://xenabrowser.net/). Using the "limma" package in R, raw data is normalized before further data processing. With the term "platelet" as the target keyword, 369 genes with a connection to platelets were found in AmiGO 2 ( http://amigo.geneontology.org/).
Identification of differentially expressed and prognostic genes
Using the "limma" package and the Wilcoxon signed-rank test in the R project, we found plateletrelated genes that were differently expressed between esophageal cancer and the healthy control group. By utilizing log fold change > 1 and false discovery rates < 0.05 as criteria, differentially expressed genes (DEGs) were chosen. The univariate Cox regression tests were used to evaluate the prognostic implications of overall survival (OS) for esophageal cancer. The intersection of DEGs and prognostic genes yielded the DEGs associated with the prognosis of esophageal cancer. Additionally, prospective risk factors based on genes connected to platelets were developed using least absolute shrinkage and selection operator Cox regression using the package "glmnet." The penalty regularization parameter lambda (λ) was established using ten-fold cross validation, and we selected the value where the partial likelihood deviance was the minimum in order to prevent overfitting effects in the model. We generated the risk score using the following formula: risk score[6] = Σgenes Cox coefficient × genes expression levels . Then, based on their median risk ratings, patients were divided into high-risk and low-risk groups.
Principal component analysis and t-distributed stochastic neighbor embedding
Using the R function "prcomp" from the "stats" package, principal component analysis was performed on each set of data. Additionally, the R package "Rtsne" was utilized to create the t-distributed stochastic neighbor embedding approach in order to visualize clustering. The optimal cutoffs for the survival analysis of each gene were determined using the "surv cutpoint" function of the "survminer" R package. Finally, a software package called "survival operating characteristic (ROC)" was adopted to analyze the time-dependent ROC curve in order to evaluate the gene signature's predictive ability for time-dependent cancer mortality.
Function enrichment analysis
We used the "ClusterProfiler" R package to distinguish between the functional pathways enriched by the DEGs in the high-risk and low-risk groups based on Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. Additionally, results from expression analysis and functional annotation enrichment analysis were displayed using the GO Circle and GO Chord plot methods in the "GO plot" R package.
Establishment of the nomogram
With the use of prognostic gene signatures and a combination of clinicopathological factors, we developed a nomogram to predict each person's probability of survival. Performance of the nomogram was evaluated using calibration plots and the concordance index. The dotted 45-degree line on the calibration graph represented the ideal forecast, while the X-and Y-axes on the graph represented the nomogram-predicted progression and the observed outcome, respectively. The model's performance was assessed using a bootstrapping approach, which could compensate for overly optimistic assumptions.
Statistical analysis
R (version 4.1.0, http://www.r-project.org/) was used to perform the statistical analysis. The gene expression in tumor tissues and healthy control tissues was compared using the Student's t test. The Fisher's exact test or the χ 2 test, as applicable, were used to compare proportions. The Kaplan-Meier method was also used to assess progression-free survival and OS using log-rank testing. Univariate and multivariate Cox proportional hazard regression analyses were used to identify OS independent factors. Less than 0.05 was considered statistically significant for two-tailed P values.
RESULTS
We recruited 653 healthy individuals as controls and 151 patients with esophageal cancer as the case group. The Cancer Genome Atlas database contained details on 151 esophageal cancer cases, whereas the Genotype-Tissue Expression database contained information on 653 healthy controls. All information came from freely accessible public databases. The clinical characteristics of 151 esophageal cancer patients are summarized in Table 1.
Identification of prognostic platelet-related DEGs
We discovered 144 DEGs when comparing the case group to the healthy control group, of which 77 genes were upregulated and 67 genes were downregulated ( Figure 1A). Using univariate Cox proportional regression analysis for 369 platelet-related genes, we identified 9 platelet-related prognostic genes, including AP3D1, APOOL, CYRIB, DDIT3, EP300, HPS4, PLA2G6, TMSB4X, and VAMP7, as being substantially correlated with OS ( Figure 1B, P < 0.05). We utilized a Venn diagram to discover the intersection between DEGs and platelet-related prognostic genes in order to further select plateletrelated DEGs. The findings demonstrate that four genes, including APOOL, EP300, PLA2G6 and VAMP7 , were connected to esophageal cancer prognosis ( Figure 1C).
Establishment of prognostic risk model
Additionally, we utilized the least absolute shrinkage and selection operator Cox analysis to choose the top four prognostic biomarkers, including APOOL, EP300, PLA2G6, and VAMP7, and we subsequently developed a classifier for predicting survival based on the four platelet-related prognostic genes ( Figure 2A and B, Table 2). The formula used was: risk score = 0.739 × expression level of APOOL -0.801 × expression level of EP300 -0.674 × expression level of PLA2G6 + 0.798 × expression level of VAMP7. Using the aforementioned formula, we calculated the risk scores for each sample. The median risk score was then used to classify the patients into low-risk and high-risk categories. The distributions of the high-and low-risk groups were marginally dispersed, according to the t-distributed stochastic neighbor embedding analysis ( Figure 2C). The expression of the four platelet-related genes was used to compare the differences between the low-and high-risk groupings using principal component analysis ( Figure 2D). In conclusion, high-risk and low-risk groups can be easily distinguished by the predictive characteristics of platelet-related genes.
Correlation between risk model and clinicopathologic characteristics
To ascertain if our risk score model correlated with clinicopathological parameters, squamous cell carcinoma and adenocarcinoma were examined as two clinicopathological findings. A similar outcome was shown in subgroups of adenocarcinoma and squamous cell carcinoma as a same consequence of the evaluation between subgroups and the entire population ( Figure 2E, F, and G). In contrast to PLA2G6 and EP300, which were strongly expressed in the low-risk group, APOOL and VAMP7 were substantially expressed in the high-risk group. Age, sex, grade, and stage were clinical factors that did not differ between the high-and low-risk groups in our research. However, there were still notable variances between the various pathological categories of the two groups. Patients with adenocarcinoma had a comparatively high-risk score (Table 3). Figure 3A and B show that patient survival times decreased as risk ratings rose, and largely all of the patients who passed away belonged to the high-risk category. The prognostic significance of DEGs in esophageal cancer was assessed using Kaplan-Meier survival analysis ( Figure 3C). Patient prognoses were considerably poorer in the high-risk group than in the low-risk group (P < 0.001). Additionally, the univariate Cox regression analysis was carried out to determine the prognostic significance of the features ( Figure 3D). The stage and risk score were statistically significant (P < 0.001), while the individuals' age, sex, and pathological type had no statistically significant impact on their survival. Furthermore, the multivariate analysis revealed that the stage and risk score were independent risk factors for OS ( Figure 3E). The area under the curve (AUC) for survival after 1, 2, and 3 years using ROC analysis was 0.620, 0.750, and 0.790, respectively ( Figure 3F). In other words, the platelet-related gene signature offered useful predictive value with clinical relevance for appropriately classifying OS patients.
Functional enrichment analysis of platelet-related genes
GO enrichment analysis revealed ten significantly enriched pathways, including those that control neuron death, cell growth, response to cytokine stimulus, positive regulation of phosphorylation, positive regulation of intracellular protein transport, positive regulation of cellular protein localization, positive regulation of intracellular transport, and regulation of cytokine-mediated signaling pathway ( Figure 4A). "Neuron death" is a biological process word with the greatest richness (GO: 0070997). Platinum drug resistance was the most enriched and important route, according to KEGG analysis. Additionally, apoptosis was significantly enriched ( Figure 4B). November 26, 2022 Volume 10 Issue 33
Nomogram construction based on 4-gene signature
As previously mentioned, statistical analysis showed that the two independent prognostic indicators linked to OS were TNM stage and risk score. In order to eliminate pathological type interference, two pathological types were added to the nomogram. Finally, to predict the prognosis of esophageal cancer more accurately, we created a novel prognostic nomogram based on TNM stage, pathological type, and risk score ( Figure 5A). The nomogram was used to evaluate each patient's probability of survival, and the ROC curve was used to assess how well the nomogram predicted outcomes. In 1-, 2-, and 3-year intervals, this nomogram's AUC was 0.722, 0.831, and 0.843, respectively ( Figure 5B). The nomogram's November 26, 2022 Volume 10 Issue 33 calibration curves for the probabilities of 1-, 2-, and 3-year survival demonstrated good agreement between prediction and observation ( Figure 5C). The nomogram's Harrell's concordance index value was 0.729. After additional verification, our nomogram performed quite well.
DISCUSSION
Esophageal cancer is mostly an aging illness, peaking in incidence in the eighth decade of life, and the global elderly population is expanding quickly [7]. Even after full resection of esophageal cancer, the prognosis remains dismal despite advancements in multimodal treatments that integrate surgery, chemotherapy, radiation, and chemoradiotherapy [8]. The main treatment for early esophageal cancer was surgical resection; however, patients who had recurrence or progression of the cancer had a difficult time getting treatment [9]. Although tumor grade offers useful prognostic information, other trustworthy factors are required for more accurate prognosis prediction. In terms of incidence, esophageal cancer is the eighth most common cancer worldwide, and it is the sixth most lethal [10]. The majority of those affected by this illness are elderly, and the average age of diagnosis is becoming older, peaking between 70-years-old and 75-years-old [11]. However, there were no age differences between the high-risk and low-risk esophageal cancer groups in our study. Numerous forms of solid tumors typically exhibit platelet-related characteristics, which are essential to the development and growth of tumors [12,13]. Platelets help tumor cells grow, survive, and migrate [14]. To shield tumor cells from immune responses, platelets may form complexes with tumor cells [15]. Conversely, tumor cells are known to both activate platelets and cause platelet aggregation [12]. These mechanisms of tumor cell-platelet interactions, however, are still poorly understood. In order to further investigate the relationship between platelets and esophageal cancer, 369 platelet-related genes were included. By using univariate Cox and multivariate Cox analysis, four differentially expressed prognostic genes were identified, and a prognostic risk model based on a four platelet-related gene signature was established.
Four gene signatures have been discovered in this study utilizing univariate Cox regression analysis and the least absolute shrinkage and selection operator Cox regression model to predict OS for esophageal cancer. Additionally, 151 patients with esophageal cancer were divided into high-and lowrisk groups based on the median risk model score using gene expression data. The prognosis between the high-risk group and the low-risk group differed significantly. As a result, the risk score method can accurately predict the results of esophageal cancer samples. GO and KEGG analyses were used to analyze biological processes and pathways in order to better investigate the mechanisms behind our risk model. Analyses of GO and KEGG data revealed that the majority of GO and KEGG enrichments were related to carcinogenesis and development. The risk model contained the four genes APOOL, EP300, PLA2G6 and VAMP7. While the expression levels of EP300 and PLA2G6 were positively connected with prognosis, the expression levels of APOOL and VAMP7 were adversely correlated with esophageal cancer.
According to Zhu et al[16], VAMP7 is implicated in the promotion of tumors since it is upregulated in high-risk individuals. Poor prognosis for esophageal squamous carcinoma was linked to EP300 epitopes as an oncogene [17]. The EP300 oncogene promotes tumor development, as was shown in earlier research in vitro using esophageal squamous cell carcinoma cell lines [17]. Studies on lung, colon, prostate, bladder, and breast cancer have also demonstrated this [18][19][20]. There are, however, few studies on APOOL and PLA2G6 in the growth and prognosis of tumors.
This study developed a nomogram for risk assessment based on multivariate Cox regression analysis to evaluate the OS of patients with esophageal cancer. The TNM stage and risk score are independent predictive factors for esophageal cancer, according to the univariate and multivariate Cox regression models. We further incorporated clinicopathological factors in addition to TNM stages and risk scores. The two histopathological types of primary esophageal cancer that are most frequently identified in clinics are squamous cell carcinoma and adenocarcinoma; they have the same potential risks, gene alterations, and treatments [21]. These two histopathological tumor types exhibit distinct behaviors in esophageal cancer, with squamous cell carcinoma exhibiting an earlier lymphatic migration and a poorer prognosis than adenocarcinoma [22]. Esophageal squamous cell carcinoma has a bad prognosis in November 26, 2022 Volume 10 Issue 33 terms of 5-year survival rates (15%-25%) since it is the most common subtype of esophageal cancer and has a higher occurrence in specific geographic areas [23]. The OS of esophageal cancer and the histopathological type were not significantly different in our study. But between the two pathological types, there was a statistically significant difference in the high-risk or low-risk groups. As a result, the nomogram also included pathological types. In order to verify the nomogram's effectiveness, its accuracy was assessed. The esophageal cancer cohort also demonstrated good calibration and discrimination; in particular, our high concordance index value (0.729) indicated the performance of the nomogram [24]. The nomogram had strong predictive power, as evidenced by the ROC curve, which had AUC values of 0.722 at 1 year, 0.831 at 2 years, and 0.843 at 3 years. The calibration curve's observation and prediction exhibited a fair amount of agreement, indicating that the nomogram may be utilized to predict the 1-, 2-, and 3-year survival rates for the cohort with esophageal cancer [25]. Overall, the findings of the calibration curve, ROC curve, and concordance index revealed that our developed prediction nomogram had excellent prediction performance.
This study had a number of advantages. First, our study was the first to describe platelet-related genes and the prognosis of esophageal cancer. Four prognostic differential genes combined with histopathological types and TNM stages were used to construct a prognosis nomogram. The individualized prediction of this nomogram may reflect the survival rate of esophageal cancer candidates more correctly, making up for the inadequacy of earlier TNM stages. Second, our analysis differed from earlier research of similar studies. We choose more control group samples than in earlier research in order to exclude the interference of a small sample of the control group in the differential analysis. Finally, we developed a prognostic model that predicted 1-, 2-, and 3-year survival with high AUC utilizing a 4-gene signature, histopathological types, and TNM stages.
This study had a number of limitations. First, there was not enough esophageal cancer in our study's sample to build prediction models based on adenocarcinoma and squamous cell carcinoma. Second, it was difficult to tell if the four DEGs differed between normal and malignant tissues due to a lack of additional histology specimens for external confirmation. Additionally, we utilized statistical analysis to find the related differential gene and GO and KEGG enrichment analysis to confirm the DEG's pathway and enrichment function. There was no mechanism research to confirm these findings. November 26, 2022 Volume 10 Issue 33
CONCLUSION
In conclusion, we constructed a novel esophageal cancer risk model based on four platelet-related genes. It was a great prognostic information tool that could be utilized to supplement the TNM staging system combined with histopathological types and risk scores.
Research background
The prognosis for esophageal cancer, one of the malignancies that responds least to cancer therapy, has not improved despite several breakthroughs in treatment. Improving patient outcomes depends on finding biomarkers and comprehending the molecular causes of esophageal cancer.
Research motivation
We wanted to create a risk score based on platelet-related gene signatures for prognosis prediction since the expression of platelet-related genes is strongly linked to patient prognosis.
Research objectives
To predict esophageal cancer prognosis, a risk model and nomogram constructed based on plateletrelated gene signatures and clinical factors associated with prognosis could be utilized.
|
v2
|
2022-11-21T16:17:30.814Z
|
2022-11-26T00:00:00.000Z
|
253725560
|
s2orc/train
|
Synchronous early gastric and intestinal mucosa-associated lymphoid tissue lymphoma in a Helicobacter pylori-negative patient: A case report
BACKGROUND Mucosa-associated lymphoid tissue (MALT) lymphoma occurs largely in the digestive tract, with the stomach being the most commonly affected organ, followed by the small intestine, large intestine, and esophagus. It is rarely found in both the stomach and colon. Helicobacter pylori (H. pylori) infection is strongly associated with gastric MALT lymphoma, although there is a small number of H. pylori-negative gastric MALT lymphomas. Diagnosis of MALT lymphoma is challenging because of nonspecific symptoms and diverse presentations of endoscopic findings. CASE SUMMARY We report a case of an asymptomatic patient who during screening endoscopy and was found to have stromal tumor-like submucosal uplift lesions in the stomach body and polypoid lesions in the rectum. After endoscopic resection, the patient was diagnosed with multiple early simultaneous gastrointestinal MALT lymphomas. CONCLUSION This study may help improve our understanding of MALT lymphomas and multifocal lesions treated using early endoscopy.
INTRODUCTION
Mucosa-associated lymphoid tissue (MALT) lymphoma is a subtype of non-Hodgkin's lymphoma classified by the World Health Organization as an extranodal marginal zone B-cell lymphoma, which accounts for approximately 5% of non-Hodgkin's lymphomas and has a good long-term prognosis with a 10-year survival rate > 90%. MALT lymphoma can occur at many sites, including the salivary glands, thyroid, orbits, lungs, breast, kidneys, skin, liver, and prostate, and most often involves the gastrointestinal tract. The lack of specificity in the endoscopic presentation of MALT lymphoma of the gastrointestinal tract, especially in its early stages, presents a significant risk of underdiagnosis and misdiagnosis, posing a clinical diagnostic challenge.
Helicobacter pylori (H. pylori) infection is the initial event in gastric MALT lymphoma. There are a variety of clinical approaches for diagnosing H. pylori infection, usually combining noninvasive and invasive methods as well as the need to exclude false-negative results caused by antacids [1]. Most gastric MALT lymphomas are H. pylori positive and sensitive to eradication therapy. However, recent studies have found that the pathogenesis of H. pylori-negative gastric MALT lymphoma is increasing annually and may be related to genetics, autoimmunity, or other microorganisms. The clinical features and endoscopic presentation lack specificity, and the occurrence of simultaneous MALT lymphoma in the stomach and intestine in an H. pylori-negative background has rarely been reported [2].
Here, we report a case of H. pylori-negative gastric MALT lymphoma mimicking a gastric stromal tumor, together with a rectal presentation of intestinal MALT with a polyp-like appearance, which were treated endoscopically with complete remission.
Chief complaints
A 46-year-old woman presented with a gastric submucosal uplift by screening endoscopy. She was admitted to our hospital for a further diagnose without any symptoms.
History of present illness
One week ago, the patient presented to the hospital for a screening endoscopy and gastroscopy revealed a submucosal bulge in the upper anterior wall of the gastric body. The possibility of a stromal tumor was considered, and rectal polyps were found by colonoscopy; therefore, the patient was admitted for further endoscopic treatment. The patient was lack of bowel habits change and other alarm symptoms.
History of past illness
The patient had no history of H. pylori infection or chronic infection.
Personal and family history
The patient denied any family history of malignant tumor. November 26, 2022 Volume 10 Issue 33
Physical examination
There were no obvious abnormalities during physical examination.
Laboratory examinations
Carbon-14 breath test results were negative, and antibodies against H. pylori types I and II were negative, indicating that the patient had no history of H. pylori infection. Routine blood examination showed normal white blood cells, lymphocytes, hemoglobin and platelets, normal liver and renal functions and electrolytes, and a negative fecal occult blood test.
Imaging examinations
Esophagogastroduodenoscopy revealed an isolated submucosal protrusion in the upper anterior wall of the gastric body, about 4 mm × 5 mm in size, and the surface was slightly faded. The blood vessels were slightly dilated, elongated, and thickened ( Figure 1A). Narrow-band imaging (NBI) revealed elongated and dilated marginal crypt epithelium and widened intervening space, similar to the pyloric gland structure, elongated and thickened blood vessels, and slightly thickened radial vessels around the area ( Figure 1B). The possibility of gastric fundus gland cancer could not be ruled out using endoscopy. To further characterize this suspicious lesion, endoscopic ultrasonography was performed, which showed thickening of the musculature of the mucosa and homogeneous hypoechoic changes in the lesion of the gastric body ( Figure 1C), and leiomyoma was suspected. At this point, the nature of the lesion examined by endoscopy and endoscopic ultrasonography remained controversial. Therefore, diagnostic endoscopic submucosal dissection was performed after communicating with the patient, but no intact tumor was found during the dissection. Therefore, the lesion was removed via endoscopic mucosal resection and sent for pathological examination. Results of hematoxylin and eosin staining showed massive lymphocytic infiltration ( Figure 2A). Immunohistochemistry was positive for CD20 ( Figure 2B) and MUM1. CD21 ( Figure 2C) showed expansion and destruction of follicular dendritic cells. Immunohistochemistry was negative for ( Figure 2D-F) CD3, CD10, Bcl-6, CK, cyclin-D and P53. The Ki-67 proliferation index was < 10%. Gene detection revealed clonal rearrangement of the IgH gene in B cells (Figure 3), and Giemsa staining confirmed the absence of H. pylori infection. Therefore, H. pylorinegative gastric MALT lymphoma was diagnosed. Rectal polypoid lesions were observed by colonoscopy ( Figure 4A), and electrosurgical treatment was performed. Lymphocyte sheet infiltration was observed on hematoxylin and eosin staining ( Figure 4B). Immunohistochemistry was positive for CD20 ( Figure 4C). CD21 ( Figure 4D) showed expansion and destruction of follicular dendritic cells. It was negative for Bcl-2 ( Figure 4E), Bcl-6 and MUM1. Therefore, rectal MALT lymphoma was considered. Systemic positron emission tomography/computed tomography showed no abnormal uptake in the stomach and other areas of the body.
FINAL DIAGNOSIS
The patient was diagnosed with synchronous gastrointestinal MALT lymphoma (stage I).
TREATMENT
The patient was referred to the Department of Hematology because of multiple simultaneous MALT lymphomas in the gastrointestinal tract. After a multidisciplinary discussion, the clinical manifestations of MALT lymphoma were considered to be indolent and H. pylori negative, and complete endoscopic resection was performed. Close follow-up monitoring was then performed.
OUTCOME AND FOLLOW-UP
Five months later, gastroenteroscopy showed no residual or recurrent MALT lymphoma. Currently, the patient is undergoing regular follow-up.
DISCUSSION
MALT lymphoma can occur anywhere in the gastrointestinal tract, but most cases occur in the stomach. Colorectal MALT lymphomas are rare, accounting for < 1% of malignant tumors of the large intestine.
The clinical presentation of gastrointestinal lymphoma varies and lacks specificity. gastrointestinal bleeding, but approximately one third of patients have no alarming symptoms; therefore, the diagnosis is often incidental, especially in the early stages. Histologically, the disease is characterized by a heterogeneous small B-cell infiltrate that usually shows lymphoepithelial lesions or follicular colonization and a typical immunophenotype of CD20(+), CD5(-), CD10(-) and cyclin D1(-), in marginal zone B cells. Restriction molecular techniques have revealed immunoglobulin light chain restriction or clonal IgH rearrangement. In this case, the patient was diagnosed with H. pylori-negative, early MALT lymphoma with gastrointestinal co-occurrence. We conducted a literature review based on the characteristics of this case. Recent studies have confirmed that the occurrence and development of most gastric MALT lymphomas are associated with H. pylori infection [3], and the main pathogenic mechanism may be that H. pylori leads to chronic inflammation and proliferation of T and B cells in the gastric mucosa. Long-term inflammation causes gastric mucosa without lymphoid tissue to produce MALT, which can lead to genetic abnormalities and malignant transformation, namely MALT lymphoma. However, recent studies have found that H. pylori-negative gastric MALT lymphoma is on the rise, and it is believed to be closely related to genes, autoimmunity, or other bacterial and viral infections. In a recent study of genetic alterations and somatic mutations in 57 patients with H. pylori-negative gastric MALT lymphoma, Kiesewetter Sjogren's syndrome, IgG4-related diseases, and obesity also increase the risk of primary MALT lymphoma [5]. Another possibility is infection with bacteria other than H. pylori, which could explain why the eradication of H. pylori can treat some H. pylori-negative MALT lymphomas [6]. Currently, there is no unified conclusion regarding the etiology of simultaneous gastrointestinal or multisite lymphomas. Clinical reports of simultaneous gastrointestinal MALT lymphoma are rare. We reviewed nine cases of simultaneous gastrointestinal MALT lymphoma reported in the literature (Table 1), and the analysis of the clinical characteristics of these cases showed that the incidence in males was higher than in females, which was consistent with the overall sex characteristics of MALT lymphoma. The median age of onset was 70 years (57-85 years), which is higher than that of single-site lymphoma (50-60 years) [15]. H. pylori infection was present in seven of the nine cases, but six failed to eradicate H. pylori infection, which was lower than the previously reported effective eradication rate of 70%-80% [16]. Most patients (5/9) presented with large tumor-like lesions associated with ulceration with lymphoma other than in the stomach and colon, and 3/9 patients had underlying diseases, including diabetes mellitus, celiac disease, and early gastric cancer. Analysis of the above clinical characteristics suggests that the therapeutic effect of H. pylori eradication in patients with homologous gastrointestinal lymphoma may be less than that in patients with a single site tumor, and most cases Endoscopic biopsy is the gold standard method for the diagnosis of lymphoma. The endoscopic manifestations of lymphoma are diverse, ranging from normal gastric mucosa to ulceration or masses. Studies have shown that superficial lesions are more common [15,17], similar to erosion, multifocal gastritis, and other malignant tumors, and are often indistinguishable from gastric cancer or gastritis. Nakamura et al [18] found that H. pylori-negative cases were more often located proximal to the stomach, invading the submucosa but rarely presenting with the common superficial type, and H. pylori-negative gastric MALT lymphomas were often clinically advanced. The endoscopic appearance of NBI magnifying glasses is characterized by a tree-like appearance [19]. This is helpful for the endoscopist's judgment in guiding the biopsy. In our case, the gastric lesion appeared in the upper third of the gastric body, was confined to the submucosa, and appeared as a small submucosal bulge. No typical dendritic vascular manifestations were observed under NBI, which was different from the endoscopic features of gastric MALT lymphoma reported in the past, and may provide a reference to endoscopists when similar cases are encountered in future. Colorectal MALT lymphoma is rare, and there is no consensus regarding colonic MALT lymphoma. A flat, raised, polypoid, or semi-pedicled appearance can be observed during endoscopy. In the rectum, polypoid lesions are more common, the tumors vary in size, with a median diameter of 20 mm [20]. In our case, since the diameter of the lesion was only 5 mm, it was difficult to recognize that it was lymphoma before pathology. This case was found incidentally in an H. pylori-negative setting, confined to the site that had been endoscopically resected, with no other organ metastases; therefore, we did not opt for further treatment, and no recurrence or progression of the lesion was detected during follow-up. There is no consensus or recommendation for H. pylori-negative gastric MALT lymphoma or intestinal MALT lymphoma. However, because of the low malignancy and slow progression of MALT lymphoma, endoscopic resection as a local treatment method has achieved ideal results.
CONCLUSION
In this case report, we have described the endoscopic presentation of early gastrointestinal MALT lymphoma in the asymptomatic stage, where endoscopic presentation is rare and easily misdiagnosed. The patient in this case was treated using endoscopic resection.
|
v2
|
2022-11-23T06:17:33.680Z
|
2022-11-22T00:00:00.000Z
|
253759701
|
s2ag/train
|
Undergraduate Medical Students' Perceptions of Nutrition Education at the Northern Ontario School of Medicine.
The objective of this evaluation was to determine Northern Ontario School of Medicine (NOSM) undergraduate medical education (UME) students' perceptions of the curriculum related to their nutrition knowledge, attitudes, and counseling self-efficacy/confidence. A 16-item electronic survey (Qualtrics©) was developed, and it included nutrition competency statements, adult and pediatric nutrition assessment and counseling confidence, and nutrition curriculum satisfaction. Students in Years 2, 3, and 4 (n = 192, 66%-73% female) were invited to answer the survey. Of the 61 respondents, 50.8% were Year 2, 34.9% Year 3, and 10.6% Year 4. Overall, 72.1% of the respondents were dissatisfied. Respondents perceived they were least competent in strategies to prevent and treat disease (72.1%), in applying basic dietary strategies (65.6%), and in ethics and nutrition management (62.3%), whereas 52.5% felt competent in the team approach to nutrition care. Respondents reported lowest confidence (less than 10%-15%) in specialized nutrition support, cancer care, renal nutrition, and mental health/eating disorders for both populations. NOSM medical learners reported curriculum dissatisfaction, nutrition incompetence, and poor levels of perceived confidence in nutrition management. Results will be used to inform nutrition curriculum enhancements and future outcome evaluations. Current and future physicians with enhanced nutrition knowledge, awareness of the Registered Dietitian (RD) roles, and an understanding of when to refer patients to an RD can provide better patient-centred care.
|
v2
|
2022-11-23T14:23:48.118Z
|
2022-11-22T00:00:00.000Z
|
253798292
|
s2orc/train
|
Evaluation of the efficacy of EU-TIRADS and ACR-TIRADS in risk stratification of pediatric patients with thyroid nodules
Background Pediatric thyroid nodules have a lower prevalence but a higher rate of malignancy (ROM) than those in adults. Ultrasound features suspected of malignancy lead to fine needle aspiration biopsy (FNAB) and subsequent cytological determination, upon which management is decided. Based on the characteristics of ultrasound, to standardize clinician decisions and avoid unnecessary FNAB, the European Thyroid Association and the American Radiology College have established guidelines for Thyroid Imaging, Reporting and Data System (EU-TIRADS and ACR-TIRADS) for ROM stratification of thyroid nodules. The aim of this study is to evaluate the diagnostic performance of ACR-TIRADS and EU-TIRADS in pediatric age. Materials and methods Subjects younger than 18 years of age with thyroid nodules greater than 0.5 cm observed in the 2000-2020 period were included. Results Data from 200 subjects were collected. The overall ROM was 13%, rising to 26% if nodules with a diameter >1 cm were considered. Patients with a malignant nodule were more likely to have a higher EU-TIRADS score (p=0.03). Missed cancer diagnoses were 26.9%. Using the EU-TIRADS system, 40% of FNABs could have been avoided, while this scoring system would have resulted in FNAB being performed in 12% of cases where the assessment of ultrasound features would not recommend FNAB. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 73.1%, 57.1%, 73.1%, and 50%, respectively. Even considering the ACR-TIRADS, a higher score correlated with a higher ROM (p<0.001). This system missed 6 diagnoses of cancer (23.1%). Using the ACR-TIRADS system, 45.3% of FNABs could have been avoided, while FNAB should have been performed in 12% of cases where it was not recommended by ultrasound characteristics. Sensitivity, specificity, PPV and NPV were 76.9%, 50%, 76.9%, and 42.9%, respectively. Conclusion The present study confirms the correspondence of the EU-TIRADS and ACR-TIRADS categories with respect to malignancy but indicates not entirely satisfactory performance compared to FNAB alone. However, the use of the two TIRADS systems should be encouraged in multicentre studies to increase their performance and establish paediatric-specific points in the scoring criteria.
Nodule size >1 cm, hypoechoic pattern, intranodal vascularization or microcalcifications, irregular edges and neck lymph nodes are the main ultrasound features that indicate malignancy (5,(22)(23)(24)(25)(26)(27)(28)(29). Once suspicious ultrasound features are present, fine needle aspiration biopsy (FNAB) is required to determine the cytological category, identified based on the most widely used cytological classifications, namely the Bethesda System for Reporting Thyroid Cytopathology (BSRTC), the British Thyroid Association (BTA) and, in Italy, the Guidance of the Italian Society of Anatomic Pathology and Cytology (SIAPEC) (30)(31)(32). The cytological category assignment leads to different clinical management that includes clinicalradiological follow-up or surgery, with some differences between these classifications. Despite the differences, all agree on a higher ROM in paediatric age for all categories, especially for indeterminate nodules (6,33,34).
Based on the ultrasound features, the European Thyroid Association and the American Radiology College have established guidelines for Thyroid Imaging, Reporting and Data System (EU-TIRADS and ACR-TIRADS respectively) for risk stratification of malignancy of thyroid nodules. Once the TIRADS category is assigned, both guidelines determine whether to perform FNAB or adopt an active surveillance strategy, depending primarily on the size of the nodule (35, 36). The main reasons that led to the definition of these guidelines were the need to standardize the ultrasound description of thyroid nodules as much as possible, provide selection criteria to perform FNAB, avoid unnecessary procedures, and provide clinicians with an additional tool for the management of thyroid nodules, especially in the category of the indeterminate cytology.
The purpose of this retrospective study is to evaluate the diagnostic performance of ACR-TIRADS and EU-TIRADS in risk stratification of paediatric thyroid nodules and determine whether extensive use of these tools can help the paediatric endocrinologist better manage thyroid nodules in pediatric age.
Materials and methods
The study included all subjects under the age of 18 with thyroid nodules greater than 0.5 cm followed at the Tertiary Center of Paediatric Endocrinology of the Regina Margherita Children's Hospital in Turin in the period 2000-2020. Patients with nodules less than 0.5 cm in diameter and with suspicious characteristics were also initially considered. However, none of these were then included in the study as no malignant features were found in any of these nodules. After approval by the Institute's Ethical Committee, clinical, laboratory and radiographic data were collected from electronic medical records. All patients underwent thyroid ultrasound evaluation, which assessed the diameter of the nodule and the ultrasound pattern; they were therefore classified as anechoic, hypoechoic, isoechoic, hyperechoic, or mixed nodules. All lymph node changes were then recorded, such as rounded swollen shape, irregular margins, increased size, absence of echogenic hilum, heterogeneous echo pattern, presence of calcifications or cystic areas, and irregular vascularization. Patients undergoing multiple ultrasound monitoring were considered as a single case. In patients with multiple nodules, the largest nodule was considered.
All ultrasound evaluations were performed in the same institution and the images were retrospectively evaluated by two independent radiologists blinded for the outcome. The TIRADS category was indicated according to both EU-TIRADS and ACR-TIRADS. Patients with inadequate ultrasound images to correctly assess the TIRADS category were excluded. In case of nodules >1 cm or suspicious features of malignancy on ultrasound evaluation, a cytological sample was obtained by fine needle aspiration biopsy (FNAB) within one month of the ultrasound finding. Histological specimens were also obtained from subjects undergoing lobectomy or total thyroidectomy. All specimens were evaluated by a single pathologist.
Statistical analysis and graphs construction were performed using Graphpad 7 (GraphPad Software, La Jolla, CA, USA). Sensitivity (number of true positives divided by the sum of true positives and false negatives), specificity (number of true negatives divided by the sum of true negatives and false positives), positive predictive value (number of true positive divided by the sum of true positive and false positive), negative predictive value (number of true negative divided by the sum of true negative and false negative) and diagnostic accuracy (sum of true positives and true negatives divided by the samples' number) were calculated based on the results of patients undergoing both FNAB and surgery. Differences between groups were established by t test to compare mean values of continuous variables. The calculations were considered statistically significant when the P-value was <0.05. Cohen's kappa coefficient was calculated to measure the inter-rater reliability among the radiologists assigning the TIRADS score.
Results
We collected clinical, laboratory and ultrasound retrospective data from 200 subjects (119 females and 81 males) aged less than 18 years with thyroid nodules ( Table 1). The observed overall rate of Regarding risk factors such as age, gender, family history of thyroid diseases, positive thyroid antibodies and radiation exposure for cancer previously treated with radiotherapy, no difference was observed between benign and malignant nodules.
All subjects had normal levels of fT4 and fT3, but the TSH level was significantly lower in subjects with a benign nodule than in subjects with a malignant nodule (p=0.01).
Bilateral and right lobe involvement was associated with a higher malignancy rate than left lobe localization (23.6% vs 18.5% vs 5.5% malignancy rate, respectively, p=0.01), as also observed for intranodal vascularization and calcification (p=0.003 and p=0.009, respectively), and lymph node involvement (p<0.0001). A larger nodule diameter was significantly more present in the malignant nodule than in the benign nodule group (mean diameter 24 mm vs 8 mm, respectively, p=<0.001). The echogenic pattern was not related to ROM.
Based on the EU-TIRADS score, missed cancer diagnoses would have occurred in 7 cases (26.9%), with 5 nodules classified in category 3 and 2 nodules in category 5 ( Table 5). All nodules in category 3 were < 20 mm and in category 5 < 10 mm. All missed diagnoses were assigned to the TIR5 cytological category. Using the EU-TIRADS system, 40% (30/75) of the FNABs performed could have been avoided, while this scoring system would have led to perform a FNAB in 12% (15/125) of the cases in which the assessment of the ultrasound features would not have recommended FNAB. Sensitivity, specificity, positive predictive value, and negative predictive value based on
Discussion
Thyroid nodules in paediatric age have a lower prevalence than in adulthood, but greater ROM (1-7). Considering only nodules >1 cm, the ROM rate of our cohort was 26%, in line with previous published studies. The overall ROM rate was 13%, probably underestimated as most patients did not have suspicious ultrasound features leading to FNAB.
The behaviour of pediatric thyroid cancer is different from that of adults, with higher rates of extrathyroid extension and disease recurrence, but much better prognosis and survival rates; to date their management therefore remains challenging. Giving the invasiveness of FNAB, to avoid unnecessary procedures and anxiety for children and their parents, the best follow-up strategy should include this procedure only when strictly necessary, in presence of certain clinical and ultrasound features. The most important, reported by the current guidelines for adults, is the size of the nodule greater than 1 cm. Other features include intranodal calcification or vascularization, lymph node involvement, marked hypoechoic pattern, bilateral or right lobe localization of the nodule, poorly defined nodule margins and some clinical risk factors, particularly radiation exposure for cancer treatment, increased TSH values, young age and male gender. In our cohort, TSH levels were correlated to malignancy, as previously reported (6,26). Considering the child's body size and the presence of microcarcinomas, in presence of multiple risk factors FNAB should be performed even if the nodule size is smaller than 1 cm (1-7).
For both the paediatric and adult populations, numerous efforts have been made to improve the selection criteria that lead clinicians to perform FNAB. To standardize the ultrasound description of thyroid nodules as much as possible and better select candidates for FNAB, the European Thyroid Association and the American Radiology College have established guidelines for Thyroid Imaging, Reports and Data System. Despite several limitations of both scoring system, previous studies in paediatric and adult cohorts have encouraged their use to increase the available data that can improve their performance. EU-TIRADS categories have been observed to be related to thyroid nodules malignancy, although the performance of such system should be improved and therefore a FNAB is currently recommended in all EU-TIRADS ≥4 nodules (50,(72)(73)(74)(75), as cancer underdiagnosis rate rises to 37.7% (55,(72)(73)(74)(75)(76)(77)(78)(79)(80). The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of EU-TIRADS in adulthood are between 70.6-83.5%, 51.2-94.1%, 11.8-76.1% and 85.4-94.9%, respectively (62, 73-75, 77, 78). The performance of EU-TIRADS in paediatric age has been evaluated in a few studies that showed lower efficacy than in adults, with sensitivity, specificity, PPV and NPV ranging between 41.7-100%, 25-75.9%, 41.7-44%, 75.9-100% respectively (63,78,79). The data from our study confirm the significant correlation of the EU-TIRADS category with malignancy. Sensitivity, specificity, PPV and NPV were 73.1%, 57.1%, 73.1% and 50 %, respectively, showing an underestimation of malignant lesions and a low ability to detect histologically determined benign nodules, which do not require FNAB. Lost cancer diagnoses in our cohort were 26.9%, while 40% of FNABs could have been avoided and 12% of patients who were not selected for needle-biopsy should have undergone FNAB.
Despite the limitations of TIRADS scores, we confirm that their use should be encouraged to improve their performance and have an additional tool in the management of paediatric thyroid nodules. The main limitation of TIRADS in children is the criterion of the size as a determinant for the execution of FNAB. We must be aware that the current guidelines have been established for adults and are not at all suitable for children, especially considering their body size. The EU-TIRADS score does not include lymph node involvement in the score but indicates the need for FNAB in case of suspicious ultrasound features; intranodal vascularization, described as a risk factor for malignancy, is also not included in the EU or in the ACR-TIRADS. Bilateral and right lobe localization should also be considered in the final score. To improve the diagnostic performance of ACR-TIRADS, some authors have indicated additional characteristics as risk profiles or PET activity (61, 62). The association of ultrasound data with clinical data could be an additional aid to performance improvement. The final score could also include an age <10 years, male gender, previous radiation exposure for cancer treatment, a higher TSH level, as well as a familial history or genetic predisposition to thyroid cancer (1)(2)(3)(4)(5).
The present study has several limitations. The retrospective nature of the study limits the statistical power of the data analysis. The number of histologically and cytologically determined malignant nodules is limited due to the low prevalence of pediatric thyroid nodules and restrictive criteria for FNAB, which can lead to underestimation, despite the case series being recruited in a tertiary centre of Paediatric Endocrinology over a 20-years period.
In conclusion, in the present study the correlation of the EU-TIRADS and ACR-TIRADS categories with malignancy was confirmed, even if their performance was not entirely satisfactory compared to FNAB alone. However, their use should be encouraged within multicentre studies, to increase the performance of both TIRADS systems and to allow for an update of the scoring criteria, including pediatric-specific points.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving human participants were reviewed and approved by Ethics Committee of the City of Health and Science University Hospital of Turin. Written informed consent to participate in this study was provided by the participants' legal guardian/next of kin.
Author contributions
GT and JM contributed to the study concept, the statistical analysis and to the first draft of manuscript. MS contributed to the data collection and literature research. FQ and LS contributed to the study concept and the revision con the final version of the manuscript. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-23T14:34:26.779Z
|
2022-11-22T00:00:00.000Z
|
253797590
|
s2orc/train
|
Improving the prediction for the response to radiotherapy of clinical tumor samples by using combinatorial model of MicroRNA expression
Purpose: Radiation therapy (RT) is one of the main treatments for cancer. The response to radiotherapy varies widely between individuals and some patients have poor response to RT treatment due to tumor radioresistance. Stratifying patients according to molecular signatures of individual tumor characteristics can improve clinical treatment. In here, we aimed to use clinical and genomic databases to develop miRNA signatures that can predict response to radiotherapy in various cancer types. Methods: We analyzed the miRNAs profiles using tumor samples treated with RT across eight types of human cancers from TCGA database. These samples were divided into response group (S, n = 224) and progressive disease group (R, n = 134) based on RT response of tumors. To enhance the discrimination for S and R samples, the predictive models based on binary logistic regression were developed to identify the best combinations of multiple miRNAs. Results: The miRNAs differentially expressed between the groups S and R in each caner type were identified. Total 47 miRNAs were identified in eight cancer types (p values <0.05, t-test), including several miRNAs previously reported to be associated with radiotherapy sensitivity. Functional enrichment analysis revealed that epithelial-to-mesenchymal transition (EMT), stem cell, NF-κB signal, immune response, cell death, cell cycle, and DNA damage response and DNA damage repair processes were significantly enriched. The cancer-type-specific miRNA signatures were identified, which consist of 2-13 of miRNAs in each caner type. Receiver operating characteristic (ROC) analyses showed that the most of individual miRNAs were effective in distinguishing responsive and non-responsive patients (the area under the curve (AUC) ranging from 0.606 to 0.889). The patient stratification was further improved by applying the combinatorial model of miRNA expression (AUC ranging from 0.711 to 0.992). Also, five miRNAs that were significantly associated with overall survival were identified as prognostic miRNAs. Conclusion: These mRNA signatures could be used as potential biomarkers selecting patients who will benefit from radiotherapy. Our study identified a series of miRNA that were differentially expressed between RT good responders and poor responders, providing useful clues for further functional assays to demonstrate a possible regulatory role in radioresistance.
Introduction
Radiotherapy (RT) plays a crucial role in cancer treatment and more than half of all cancer patients receive RT during the course of disease (Baskar et al., 2012). In last 20 years, the outcomes of RT have been improved dramatically due to the developments of highly conformal RT techniques, such as intensity-modulated RT (IMRT), intensity-modulated arc therapy (IMAT) and stereotactic RT (SRT) (Minniti et al., 2021). However, unfortunately the outcomes of therapy are not fully satisfactory. The response to radiotherapy varies widely between individuals and some patients are resistant to RT treatment. Radioresistance is considered as a main factor impeding efficacy of radiotherapy. Although radioresistance has been implicated to be associated with several biological alterations of the tumor cells, such as tumor metabolism (Tang et al., 2018), cell cycle arrest (Chen et al., 2017), oncogene and tumor suppressor alterations (Pitroda et al., 2009), microenvironment (TME) change (Suwa et al., 2021), autophagic regulation (Chang et al., 2014), cancer stem cells (CSCs) generation (Ning et al., 2013), and DNA damage response (DDR) and repair (Huang and Zhou, 2020;Sun et al., 2020), the mechanisms underlying resistance to radiation are still largely obscure. Therefore, uncovering the processes and charactering the molecules associated with regulation of radioresistance may lead to improved, efficient treatment for cancer patients.
Previous studies revealed that the difference in the status of mutation and expression profile of gene including microRNAs (miRNAs), is associated with radioresistance (Mercatelli et al., 2008;Huang et al., 2013;Mueller et al., 2013;Hatano et al., 2015;Mao et al., 2016;Xu et al., 2016;Tao et al., 2018;Xu et al., 2018). MicroRNAs are small non-coding single-stranded RNAs of 19-23 nucleotides in length, which play a critical role in posttranscriptional regulation by degrading or preventing the translation of their target messenger RNA (mRNA). They play important roles in tumor development and metastasis. Emerging evidences demonstrated that miRNAs may be involved in regulation processes associated with response to radiation. For examples, previous studies revealed that some microRNAs, such as miR-95 (Huang et al., 2013), miR-221, miR-222 (Mercatelli et al., 2008), and miR-106b (Hatano et al., 2015) enhanced radioresistance in cancer cells, while miR-30a (Xu et al., 2016), miR-16 (Tao et al., 2018), miR-449 (Mao et al., 2016), miR-17 , and miR-100 (Mueller et al., 2013) enhanced the radiosensitivity. Several miRNAs regulate DNA damage response (DDR) pathway. For example, miR-101 can regulate the expression of ataxia-telangiectasia mutated (ATM) gene and DNA-PK. ATM is a central regulator of DNA damage response (Maréchal and Zou, 2013). Mutation and inactivation of ATM can lead to increased instability of genome and impaired repair ability for DNA double-strand breaks. It has been experimentally demonstrated that up-regulation of miR-101 reduced the protein levels of DNA-PK and ATM, rendering tumor cells more sensitive to radiation (Yan et al., 2010). MiR-223 is a regulator of maturation and differentiation of hematopoietic stem cells. Up-regulation of miR-223 will reduce the expression of ATM and make U87 cells sensitive to radiation in vitro (Liang et al., 2014). MiR-375 is a negative regulator of p53 which is a key tumor suppressor gene that inhibits cell growth by activating cell cycle arrest or apoptosis (Vousden and Prives, 2009). Increase of miR-375 was detected in recurrent gastric cancer , and the increased miR-375 interacts with the 3′UTR of p53 gene that negatively regulates p53 and its downstream pathway genes, resulting in radioresistance of cells to radiation (Liu et al., 2013).
In the era of precision medicine, a biomarker of intrinsic radiosensitivity would be extremely valuable for selecting patients in whom will benefit from RT, adjusting individual dosing, and aiding decision making. Previous studies have demonstrated that several miRNAs were associated with radioresistance or radiosensitivity, implying they could serve as promising biomarkers for prediction of RT response. Besides, miRNAs serving as biomarkers have several advantages. Firstly, unlike mRNAs, miRNAs remain largely intact in routinely collected, formalin-fixed, paraffinembedded (FFPE) clinical tissues. Therefore, the detection for miRNA levels could be conveniently performed in clinical practice. Secondly, miRNAs have been called 'the master regulators' of gene expression since a single miRNA can regulate several hundreds of mRNA targets. Researches have indicated that many cell phenotype or subtype is likely governed by a miRNA regulatory network (Yang D. et al., 2013), suggesting some miRNAs may sever as the possible major determinants of cellular phenotype (including radioresistant phenotype) (McDermott et al., 2017). In additional, cell lines are usually used as models for radioresistance research in previous reports. However, tumors from patients are heterogeneous. Thus, the factors related to RT response are more complex than in cell lines. On the other hand, the results from cellular experiments also need to be confirmed with clinical samples. Therefore, we proposed that miRNA profile analysis based on clinical samples could provide a direct assessment for their performances as biomarkers for prediction of RT response. In this study, we performed a large-scale analysis of miRNA expression data collected from The Cancer Genome Atlas (TCGA) from those who treated with RT across eight cancer types. By analyzing the expressional difference between RT response and progressive disease samples, the miRNA signatures for predicting RT response were obtained and their performance for prediction of RT response were estimated.
Data collection
MiRNA expression data of cancer patients undergoing RT were downloaded from the website UCSC XENA -GDC TCGA (https:// xenabrowser.net/hub/). The expression profiles were presented as RPM (reads of exon model per million mapped reads). Clinical information including patient's response to RT and overall survival was downloaded from GDC TCGA website (https://portal.gdc. cancer.gov/). The tumor samples were categorized into complete response (S) and progressive disease (R) groups depending on their clinical response to RT treatment. Complete response and progressive disease was defined according to RECIST. Finally, we analyzed 358 clinical samples across eight different cancer types, including S group (n = 224) and R group (n = 134).
Differential expression analysis
The differential expression analysis of miRNAs was performed in each cancer type. The miRNA would not be analyzed further if the reads of these miRNAs were empty in more than 10% of the samples. Deseq2 package was used to normalize miRNA data and identify the difference of miRNA expression levels between S and R group (Love et al., 2014). Wald test and t-test was used to calculate the p-value. In this study, p < 0.05 and | logfc | > 1 were used as threshold criteria for screening DEMs between S and R group.
Identification of DEMs
The identification of differentially expressed miRNA (DEMs) was conducted using R language. The expression levels of DEMs were visualized using the ggpubr, ggplot2 and complexheatmap packages. The receiver operating characteristic (ROC) curve was drawn and the area under the curve (AUC) was calculated using the pROC package. Heatmap of miRNA expression levels for individual cancer types was drawn using pheatmap. Kaplan Meier curve of single gene was drawn by Survminer package. The samples were divided into high expression group and low expression group according to the median expression level of miRNA and p-value was calculated by log-rank test. Enrichment analysis of the 47 DEMs was performed by the online tool TAM 2.0 (http://www.lirmed.com/tam2/).
Combinatorial models of multiple miRNAs
A logistic regression model was developed by combining expressions of multiple miRNAs in each cancer type. In our study, the dependent variable of this model was the normalized expression level of miRNA; the independent variable was the RT response of samples (response or progressive disease). The method of combinatorial modelling was described in detail (Xu et al., 2022). The samples were randomly divided into a training set and a test set. The training set contained 80% of the total samples, and the test set contained 20% of the total samples. The K-fold cross-validation (K = 5) was used to fit the combinatorial model. We used different combinations of the data groups that were partitioned to train and test K different models, and then the performance was evaluated. In running the final model, we found that the independent variable were the most accurate predictors of the dependent variable. The formula of logistic regression is as follows: Y is the predictive index of RT response. The optimal threshold for dividing S and R groups in the training set was calculated, and then the test set was tested using the threshold. If the predictive index is more than the threshold, it indicates that the sample is more likely to be radiosensitive; otherwise, the sample is more likely to be radioresistant. Finally, the area under the receiver operating characteristic curve was calculated to evaluate the predictive ability.
Clinical information of cancer patients
To investigate the association between RT response and miRNA profiles, we analyzed miRNA expression data of patients undergoing RT from TCGA database. A total of 358 clinical samples across eight different types of human Frontiers in Genetics frontiersin.org cancers were analyzed in this study, including bladder urothelial carcinoma (BLCA, n = 23), esophageal carcinoma (ESCA, n = 47), lung adenocarcinoma (LUAD, n = 61), lung squamous cell carcinoma (LUSC, n = 38), pancreatic cancer (PAAD, n = 35), sarcoma (SARC, n = 56), skin cutaneous melanoma (SKCM, n = 34) and stomach adenocarcinoma (STAD, n = 64). The samples of cancer patients were categorized into complete response (S) group and progressive disease (R) group depending on their clinical response to RT treatment. Finally, 358 clinical samples were analyzed, including S group (n = 224) and R group (n = 134). The cancer types and groups of samples were shown in Figure 1A.
Identification of DEMs in group S and R
To identify potential miRNA biomarkers, it is essential to compare miRNAs differentially expressed from RT response samples (S) and disease progression samples (R). According to the screening criteria of p-value <0.05 and | logfc | > 1, the miRNAs differentially expressed (DEMs) between the groups S and R in each caner type were identified. Each cancer type consisted of 2-13 of these miRNAs, and total 47 miRNAs were identified in eight cancer types (Table 1). Hierarchical unsupervised clustering analysis were performed to visualize the expression patterns of these 47 miRNAs. The results showed that these DEMs were cancer type-specific. For example, 13 miRNAs were identified in BLCA, five of them tended to be highly expressed in R group, including miR-196a-1 (p = 0.0048), miR-196a-2 (p = 0.005), miR-130a (p = 0.0163), miR-376b (p = 0.0078) and miR-30d (p = 0.0016), and eight miRNAs including miR-10a (p = 0.021), miR-151a (p = 0.013), miR-101-1 (p = 0.0007), miR-101-2 ((p = 0.0007), miR-944 (p = 0.049), miR-378c (p = 0.0153), miR-378a (p = 0.0158), and miR-214 (p = 0.0162) tended to be highly expressed in S ( Figure 1B). Likewise, 4-miRNA signature was identified in LUSC. The high Mapping the differentially expressed miRNAs to signal transduction pathways is important toward understanding their significance in radioresistance. Therefore, we performed enrichment analysis to identify significantly enriched terms. As shown in Figure 1C, the identified pathways were primarily involved in regulation of epithelial-to-mesenchymal transition and stem cell, NF-κB pathway, immune response, cell death and cell cycle. EMT plays a critical role not only in tumor metastasis but also in tumor radioresistance. Epithelial-mesenchymal transition could increase radioresistance (Zhang et al., 2014). We found that EMT signalling pathway to be among the top enriched pathways, implying that the radioresistant phenotype of tumour could be largely explained by the enhancement of EMT pathway. Activation of DNA damage response and DNA damage repair pathway has been demonstrated to be involved in radioresistance (Vousden and Prives, 2009;Yan et al., 2010;Liu et al., 2013;Liang et al., 2014). Enrichment analysis revealed that several DEMs including miR-34a, miR-146a, miR-
FIGURE 2
The levels of miRNAs expression that were differentially expressed between the responding (complete response) versus non-responding (progressive disease) tumors. p values were calculated by t-test.
Frontiers in Genetics frontiersin.org 130a, miR-196a, miR-143, and miR-101 were involved in DNA damage response and DNA damage repair. We also found that several miRNAs (miR-1224, miR-33a, and miR-142) were involved in inflammatory process which was implicated to be a critical radioresponse in radioresistant lung cancer cells (Yang H. J. et al., 2013).
Distinguishing S and R group with a single miRNA
To evaluate the performance for distinguishing between S and R samples, we analyzed the specificity of each miRNA identified in DEMs list with the area under the curve (AUC). The expressed difference of individual miRNA for two groups was exhibited by boxplots. As shown in Figure 2, the expressed difference of individual miRNA was significant between RT responsive and non-responsive patients (p values <0.05, t-test). Particularly, miR-101-1 had the highest discrimination for the two groups in BLCA (p = 0.0007). Receiver operating characteristics (ROC) analyses were conducted to further assess the prediction performance of individual miRNA. As shown in Figure 3 several miRNAs had relatively high specificity for distinguish S and R samples, e.g., miR-101-1 (BLCA AUC = 0.889), miR-101-2 (BLCA, AUC = 0.889), miR-30d (BLCA, AUC = 0.889), miR-196a-1 (BLCA, AUC = 0.865), miR-196a-2 (BLCA, AUC = 0.865), and miR-592 (LUSC, AUC = 0.832). MiR-101 was known to be involved in the regulation of DNA damage repair and radiosensitivity. In our study, expression of
FIGURE 3
Receiver operating characteristics (ROC) analyses of the performance for the predictions for RT response using one single miRNA in different cancer types. The representative results were presented.
Frontiers in Genetics frontiersin.org miR-101 was up-regulated in RT responsive samples, hinting the high expression of miR-101 may be associated with radiosensitive phenotype. This is consistent with previous studies showing that up-regulation of miR-101 will enhance radiosensitivity (Yan et al., 2010). Our results implied that these miRNAs could be used as potential biomarkers for stratify patients treated with RT.
Using combinatorial model to predict RT response
Despite the most of differentially expressed miRNAs could be used to distinguish patients according to RT response (e.g., AUC ranging from 0.667 to 0.786 in ESCA), further improvement on discriminatory ability is still needed. Therefore, we developed the combinatorial models based on a binary linear regression model. The combinatorial models for each cancer type consisted of 2-13 miRNAs (Supplementary Table S1). By applying this model, the best combination of multiple miRNAs for RT response prediction in each cancer type was identified (Supplementary Table S1). As showed in Figure 4, the optimal cut-off value for distinguishing S and R samples was indicated by a dotted line, and the predictive index of each sample was calculated for classification. The results showed that the most of S samples were above the threshold line, while the R samples were below the threshold line, implying that the most samples could be correctly classified in both training set and test set. Receiver operating characteristic (ROC) analysis revealed that the predictive reliability was significantly increased in all caner types by using these combinatorial models ( Figure 5). For example, the AUC of single miRNA in LUSC was between 0.664 and 0.832. The AUC value was improved to 0.879 in of LUSC by applying the combinatorial models. Likewise, the combinatorial model improved the AUC to 0.900 in ESCA. These results indicated that the combinatorial model could significantly improv the AUC relative to the single miRNA.
Evaluation of survival rate with single miRNA signature
We also explored the potential link between these differentially expressed miRNAs and the overall survival of cancer patients. We found that the expression level of five miRNAs, including miR-378a Assessing the performance for distinguishing RT responsive samples in eight different cancer types using combinatorial models. Predictive indexes were calculated to classify samples. Patients are shown in columns, and predictive index was showed in rows. The optimal cut-off value for distinguishing S and R samples was indicated by a dotted line. Samples of training set were indicated by circles, and samples of test set were indicated by diamonds.
Frontiers in Genetics frontiersin.org (p = 0.017, in BLCA), miR-142 (p = 0.018, ESCA), miR-655 (p = 0.03, in STAD), miR-29a (p = 0.025, in SARC), and miR-150 (p = 0.019, SARC), were significantly associated with the overall survival time ( Figure 6). Among them, miR-378a, miR-29a, and miR-150 tended to highly expressed in S group. The patients with high expression of these miRNAs showed a trend of better survival. In contrast, miR-142 (p = 0.018, ESCA) and miR-655 (p = 0.03, STAD) tended to be highly expressed in R group, and the worse survival was observed in these high expression samples. The results showed that only five miRNAs of 47 miRNAs were predictive for patient survival. It is not surprised since the differentially expressed miRNAs were identified based on the response to RT treatment rather than the overall survival of cancer patients.
Discussion
Currently, only few biomarkers have been evaluated for their radiotherapy-specific predictive value. Several research groups have reported that miRNAs were involved in regulation of radioresistance, hinting they have the potential serving as biomarkers for prediction of RT response. However, most of them was identified based on cellular experiments in previous studies. The evaluation based on clinical samples should be performed. In this study, we evaluated the predictive value of miRNA signatures for predicting RT response by using data of clinical samples. Comparison of miRNA expression of radioresistant and radiosensitive tumors led to the identification of 47 miRNAs. Most of them showed to be predictive for RT response (AUC ranging from 0.606 to 0.889). Some of them showed the high specificity for the prediction. For instance, miR-101, miR-30d, and miR-196s has AUC of 0.889, 0.889, and 0.865, respectively. To further improve the prediction performance, we developed a combinatorial model in each cancer type. In these models, the best combinations of multiple miRNAs could be obtained, which leads to an improved discriminatory power. The results showed that the most of AUC in each cancer type was greater than 0.8 when the combinatorial models were applied. Particular, the AUC value reach to 0.992 by applying the combinatorial model in BLCA. The prediction models and miRNAs identified here have the potential clinical application. As these combinatorial models contained 2-13 miRNAs, it is convenient to develop digital PCR and real-time qPCR based clinical test in routine practice (Hindson et al., 2013).
Identifying genetic clues to the molecular basis of radioresistance is a major challenge. Our study identified a series of miRNA that were differentially expressed between RT good responders and poor responders, providing useful clues for further functional assays to demonstrate a possible regulatory role in radioresistance. Among DEMs, several miRNAs were known to be involved in the regulation of DNA damage repair or radiosensitivity. For example, our results showed that miR-34a was upregulated in S group and was downregulated in R group in SARC. MiR-34a is one of the
FIGURE 5
Receiver operating characteristics (ROC) analyses of the performance for the predictions for RT response using combinatorial models in eight different cancer types.
Frontiers in Genetics frontiersin.org family members of miR-34 (Misso et al., 2014), and is a key regulator of tumor suppression. The overexpression of miR-34a makes cells sensitive to radiation by inhibiting several targets of DNA damage repair pathway (Lacombe and Zenhausern, 2017), such as Bcl-2 (Liu et al., 2011), LyGDI (Duan et al., 2013), Notch-1 (Kang et al., 2013), Rad51 (Cortez et al., 2015). MiR-29a is a member of the miR-29 family (Patel and Noureddine, 2012), and miR-29 is a tumor suppressor that can promote cell senescence and differentiation (Martinez et al., 2011;Chuang et al., 2022). The studies found that overexpression of miR-29a enhanced radiosensitivity, and promoted apoptosis in radiation resistant CaSki and c33a cells (Martinez et al., 2011). Overexpression of miR-29a induces a significant decrease in cell migration speed through K-ras/c-raf/p38 signal pathway, and may reduce metastasis of lung cancer (Chuang et al., 2022). In this study, we found that miR-29a in SARC tended to be highly expressed in S group. MiR-214 involved in radioresistance have been demonstrated in multiple cancer types Hu et al., 2018;Li et al., 2019). In ovarian cancer, the expression level of miR-214 rises after ionizing radiation, which activates P13K/Akt pathway by targeting PTEN, resulting in the increased radioresistance of cell lines . In colorectal cancer, miR-214 is significantly down-regulated in cells after ionizing radiation, which resulting in the increased sensitivity of cells to radiation . MiR-214 is overexpressed in osteosarcoma tissues and is a negative regulator of phlda2, maintaining radioresistance of osteosarcoma cells to apoptosis (Li et al., 2019). We found that miR-214 tended to be highly expressed in the R group, implying it may play a role in radioresistance in BLCA. Besides, miR-101, miR-146a, miR-196a, miR-143, miR-222 (Shi et al., 2019), and mir-130a (Ha Thi et al., 2019), has been reported to be involved in the regulation of DNA damage repair or radiosensitivity in previous studies.
As several miRNAs previously reported to be associated with radiotherapy sensitivity (e.g., miR-34, miR-101, miR-29a, miR-214, miR-146a, miR-196a, miR-143, miR-222, and mir-130a), we hypothesized that the signatures identified in here would identify additional radiotherapy sensitivity-related miRNAs. Our gene function enrichment analysis showed the many miRNAs were involved in EMT pathway. EMT is associated with characteristics of cancer stem cells, including radioresistance and chemoresistance. Several miRNAs were reported to directly target multiple key components of EMT pathway. For Kaplan Meier overall survival curves for patients with one single miRNA stratified by high versus low miRNA expression. Results from five miRNAs with low p values (p values <0.05) were shown. The high and low expression was indicated with red and blue color, respectively. p values were derived from log-rank test.
Frontiers in Genetics frontiersin.org example, the miR-200 family inhibits EMT and tumor metastasis, inhibits self-renewal of cancer stem cells, and enhances radiosensitivity of several types of cancer (Baumann et al., 2008). Overexpression of miR-200c regulates oxidative stress response genes and increases radiosensitivity of lung cancer cells (Magenta et al., 2011). Zheng et al. found that the downregulation of miR-200c was related to radiation tolerance in esophageal squamous cell carcinoma (Zheng et al., 2017). can suppress EMT by targeting the EMTinducing transcription factor ZEB1 (Zhang et al., 2014). Function enrichment analysis indicated that 14 miRNAs, including miR-34a, miR-215, miR-221, miR-30d, miR-194-2, miR-let-7, miR-194-1, miR-129-1, miR-29b-2, miR-29b-1, miR-29a, miR-192, miR-150, and miR-143, were associated with EMT pathway in our study. This enrichment implied that these miRNAs might be involved in radioresistant phenotype through EMT pathway. Previous studies reported that miR192 regulated the EMT pathway (Khella et al., 2013). promoted EMT of gastric cancer, migration and invasion by targeting RB1 (Song et al., 2022). In additional, the previous study has demonstrated that miR-192 was significantly upregulated in cisplatin-resistant lung cancer cells, and miR-192 induced cisplatin resistance through activating the NF-κB pathway (Li et al., 2022). Moreover, miR-192 could influence 5fluorouracil resistance (Boni et al., 2010). Cisplatin and 5fluorouracil were genotoxic and their cytotoxic mode of action prominently involves the generation of DNA lesions followed by the activation of the DNA damage response and the induction of mitochondrial apoptosis. Zhai et al. found miR-143 suppressed epithelial-mesenchymal transition and inhibited tumor growth of breast cancer through down-regulation of ERK5 (Zhai et al., 2016). Up-regulating miR-143 enhances E-cadherin-mediated cell-cell adhesion ability, reduces mesenchymal markers, and decreases cell proliferation, migration, and invasion in vitro (Zhai et al., 2016). Yang et al. found that up-regulated miR-143 represses EMT in esophageal cancer cells (Yang et al., 2019). Our results showed that miR-143 was upregulated in S group and was downregulated in R group in esophageal cancer samples. We suggested that miR-143 might play a regulatory role in radiosensitivity through influencing EMT pathway in esophageal cancer. The further experiments will be required for function assay. We also noticed that miR-150 tended to be highly expressed in the RT response samples, implying it may play a role in radiosensitivity. Previous report showed that the expression of miR-150 decreased significantly in serum after irradiation in animal studies (Jia and Wang, 2022). Recently, miR-150-5p were confirmed to target ZEB1 and caused mRNA degradation, thus blocking EMT (Lu et al., 2017). Together, these multiple clues suggested that these miRNAs (miR-192, miR-143, and miR150) might serve as putative regulators of radiosensitivity through EMT pathway.
In conclusion, this work showed the miRNA signatures could serves as biomarkers to classify the RT response patients, and further investigations with larger numbers of patient samples are currently underway to validate the utility of using these biomarkers. Finally, additional work will be required to determine the role of these miRNAs in radioresistance of tumor.
Data availability statement
Publicly available datasets were analyzed in this study and are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.
Author contributions
All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.
Funding
This work was supported in part by projects (cstc2020jxjl130015, cstc2021jcyj-msxmX0767, cstc2019jcyj-msxmX0671, and cstc2020jcyj-msxmX1093), project (no. 82073347) from the National Natural Science Foundation of China, cstc2022ycjh-bgzxm0208, and the Integrated Innovation and Application of Key Technologies for Precise Prevention and Treatment of Primary Lung Cancer (no. 2019ZX002) from the Chongqing Municipal Health Committee.
|
v2
|
2022-11-23T14:34:59.466Z
|
2022-11-22T00:00:00.000Z
|
253797672
|
s2orc/train
|
Coexpression network analysis of human candida infection reveals key modules and hub genes responsible for host-pathogen interactions
Invasive fungal infections are a significant reason for morbidity and mortality among organ transplant recipients. Therefore, it is critical to investigate the host and candida niches to understand the epidemiology of fungal infections in transplantation. Candida albicans is an opportunistic fungal pathogen that causes fatal invasive mucosal infections, particularly in solid organ transplant patients. Therefore, identifying and characterizing these genes would play a vital role in understanding the complex regulation of host-pathogen interactions. Using 32 RNA-sequencing samples of human cells infected with C. albicans, we developed WGCNA coexpression networks and performed DESeq2 differential gene expression analysis to identify the genes that positively correlate with human candida infection. Using hierarchical clustering, we identified 5 distinct modules. We studied the inter- and intramodular gene network properties in the context of sample status traits and identified the highly enriched genes in the correlated modules. We identified 52 genes that were common in the most significant WGCNA turquoise module and differentially expressed genes in human endothelial cells (HUVEC) infection vs. control samples. As a validation step, we identified the differentially expressed genes from the independent Candida-infected human oral keratinocytes (OKF6) samples and validated 30 of the 52 common genes. We then performed the functional enrichment analysis using KEGG and GO. Finally, we performed protein-protein interaction (PPI) analysis using STRING and CytoHubba from 30 validated genes. We identified 8 hub genes (JUN, ATF3, VEGFA, SLC2A1, HK2, PTGS2, PFKFB3, and KLF6) that were enriched in response to hypoxia, angiogenesis, vasculogenesis, hypoxia-induced signaling, cancer, diabetes, and transplant-related disease pathways. The discovery of genes and functional pathways related to the immune system and gene coexpression and differential gene expression analyses may serve as novel diagnostic markers and potential therapeutic targets.
Introduction
Solid organ transplant (SOT) patients are exposed to various complications, e.g., invasive fungal infection and organ failure, which are the major challenge in SOT and affect the morbidity and mortality in transplant patients. The most prevalent invasive fungal infection in SOT is Candidiasis, which includes about 60% of infections, followed by aspergillosis accounts for up to 25% of fungal infections (Shoham and Marr, 2012).
An opportunistic fungal pathogen, Candida albicans, is part of healthy human gut microbiota. However, when immunity is compromised or suppressed, particularly in organ transplant individuals, AIDS patients, chemotherapy-treated patients, and neonates, the mucosal layer becomes more susceptible to fatal invasive C. albicans infections such as candidiasis (Sangeorzan et al., 1994;Rhodus et al., 1997;Revankar et al., 1998;Redding et al., 1999;Willis et al., 1999), (Sobel, 1985). C. albicans can switch from an avirulent commensal yeast form to a virulent invasive hyphal form in which hyphae invade through the mucosal layer and disseminate/propagate through the blood, infecting other organs as well as developing multidrug resistance (Klepser, 2006;Cowen et al., 2015;Arendrup and Patterson, 2017;Pendleton et al., 2017;Nishimoto et al., 2020). In the process of C. albicans infection, the first site of host-pathogen interactions is epithelial and endothelial cells (Barker et al., 2008;Liu et al., 2015a). The development of invasive fungal diseases relies on the synergy between the host immune response and fungal virulence. Comprehensive network analysis is vital to understanding the regulatory network and rewiring to respond to these infections.
In this work, we applied WGCNA to analyze 32 RNA-seq samples from in vitro infection of C. albicans on human endothelial and oral epithelial cells after 1.5, 5, and 8-h of infection and controls. We identified 5 modules in human endothelial cells (HUVEC) human cell lines in infection vs. control status and separately identified differentially expressed genes (DEG). We reported the common genes across the two methods (WGCNA and DEG). We then validated a subset of genes using differential gene expression analysis of candidainfected human cell lines OKF6. Finally, we performed protein-protein interaction network analysis and identified hub genes that could be novel targets to investigate C. albicans infection in humans. Through these central genes' biological and molecular functions, we gained insights into the signaling pathways previously not correlated with the fungal pathogen-host response and other diseases.
Data collection
All processed gene expression datasets were collected from publicly available NCBI Gene Expression Omnibus GSE56093 (Liu et al., 2015a). The raw sequence data was aligned to the human and candida reference genomes separately by Liu et al. (Liu et al., 2015a), and the resultant count matrices were utilized for the WGCNA and DEG analyses. This dataset was comprised of 88 samples from in vitro and in vivo experiments. Of those, we only utilized 32 in vitro samples of human cell lines (endothelial and epithelial) infected with C. albicans (SC5314 and WO1 strains) and their controls at three different time points. More information is given in Supplementary Table S1. The overall methodology steps are shown in Figure 1.
Data normalization and transformation
We performed normalization on the RPKM (reads per kilobase of transcript, per million mapped reads) values using the GCRMA limma package (Gautier et al., 2004) by first removing features with counts <10 in 90% of the samples, as these could be a potential cause of the noise. Then, we performed and compared three data transformation techniques, logarithmic, regularized logarithmic, and variance stabilizing transformation (Lin et al., 2008) , to stabilize the variance across sample mean values. We chose the regularized log transformation due to stability (Supplementary Figure S1).
Weighted gene coexpression network framework
We constructed the weighted gene coexpression network using the R WGCNA package (Langfelder and Horvath, 2008). The normalized data were used as input for network construction and gene module detection. It uses correlation to find functional modules of the highly correlated gene networks. First, we evaluated the soft threshold power (β) to convert coexpression into weight with a scale-free topology index of 0.9. We chose soft threshold powers of 8 to calculate the correlations between the adjacent genes (Supplementary Figure S2). Pearson correlations between each gene pair were calculated. We then converted this adjacency matrix into a topological overlap matrix (TOM) to define gene clusters that show the amount of overlap in shared neighbors of the gene network. The dissimilarity measure was determined for hierarchical clustering and module detection. Modules of clusters of genes with high topological overlap were selected using a dynamic tree-cut algorithm. Several modules were identified, and the modules with similar expression levels were merged by calculating their eigengenes corresponding to their correlations. We further determined the association of these modules with the external traits. We identified the genes with high gene significance (GS) and module membership (MM) in the turquoise and blue modules in HUVEC data. Last, intramodular connectivity was analyzed in human modules using MTR>0.35 and p-value < 0.05. All the categorical variables were binarized for the analyses.
Identification of differentially expressed genes
Differentially expressed genes were identified using DESeq2 R Bioconductor package (Love et al., 2014). We used raw counts that were fed to the DESeq2 since it corrects for library size. The variance stabilizing transformations (VST) function estimated the sample differences (Lin et al., 2008). The statistical significance for the differentially expressed genes was set to q-value < 0.05 and log2 fold change (log2FC) > 1.
FIGURE 1
The schematic representation of the overall methodology: The discovery dataset was analyzed using two independent methods (WGCNA and DESeq2). Their intersecting genes were overlapped with the DEG list from the independent validation set to build the PPI network and identify the hub genes.
Frontiers in Genetics frontiersin.org
Functional enrichment analysis of genes
We performed Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses to study the role of the genes and identify their biological functions and pathways. Gene Ontology analysis was performed to determine the biological process. We considered an adjusted p-value threshold of ≤ 0.05 and a minimum gene count of 3 for the KEGG pathways and GO functional terms. As the contribution of all the genes is not the same, we identified hub genes and further investigated their function. (D) Venn diagram representing the overlapping genes from the turquoise module genes and differentially expressed genes. (E) Venn diagram representing the common genes between the discovery set genes (overlapping genes from WGCNA turquoise module and DEG genes) and differentially expressed genes from the validation set.
Statistical analysis and data visualization
The R programming language (Horgan, 2012) was used to normalize the RNA-seq data. We conducted Fisher's exact tests to identify the statistically significant Gene Ontology terms and functional classes. Enrichment analysis based on a hypergeometric test was implemented, and Benjamini Hochberg multiple testing was used to correct the p-value. Data visualization to show differentially expressed genes between infected and uninfected groups for top selected genes was plotted using the complex Heatmap function in R. The data visualization was performed using the cluster profiler package in R (Yu et al., 2012).
Protein-protein interaction network analysis
The validated genes are uploaded into the STRING database, and high confidence interaction score ≥ 0.7 was used to reduce false-positive interactions (Bozhilova et al., 2019). The resultant network output was loaded into Cytoscape. CytoHubba (Chin et al., 2014) was used with the Maximal Clique Centrality (MCC) algorithm to discover the hub genes in the PPI network (Li and Xu, 2019).
Network construction and module identification
Weighted Gene Correlation Network Analysis was conducted on HUVEC data. We performed hierarchical clustering of genes using a topological overlap matrix and merged modules with similar expression profiles ( Figure 2A). Each leaf corresponds to a gene, and branches correspond to the cluster of highly coexpressed genes. After cutting tree branches, we identified five different modules, turquoise, yellow, black, blue, and green, with 1,365, 459, 261, 1829, and 755 genes in HUVEC (Supplementary Table S2). A total of 4,669 genes were identified from the HUVEC data set, and in each module, the number of genes ranged between 261 and 1829.
Module association with external traits
We further analyzed the module trait relationship (MTR) between the module eigengene and clinical traits, where each cell represents the correlation strength (red is positively correlated, and green is negatively correlated) with their corresponding p-value ( Figure 2B). We demonstrate that some module eigengenes are highly correlated with infection (status traits). We observed that the turquoise (r = 0.55, p = 0.03) and blue (r = −0.52, p = 0.04) modules were highly correlated with the infection status in HUVEC cells. Since the turquoise module, with 1,365 genes, is the most significantly correlated with the clinical trait, we focused on this module for further analysis.
Intramodular connectivity using gene significance and module membership
We quantified genes with high significance for the trait status of HUVEC and high module membership by comparing their similarities in every module. There was a highly significant correlation between gene significance and module membership in the turquoise module. Figure 2C represents the correlation between turquoise module membership and gene significance (r = 0.45, p = 5.1e-69).
Differentially expressed genes and intersection with WGCNA
We used DESeq2 as a second independent method on the entire HUVEC dataset to identify 54 genes that were differentially expressed in the HUVEC (infection vs. control) samples (q-value < 0.05 and log2FC > 1). From the WGCNA analysis, we identified 1,365 genes in the most significantly correlated turquoise module ( Figure 2B). When we further investigated the intersection of WGCNA and DEGs, 52 genes were common between the turquoise module and the DEG list ( Figure 2D). The list of turquoise module genes, DEGs, and intersecting genes is given in Supplementary Table S4.
Integrating network analysis with functional enrichment analyses
To understand the biological roles of these 30 validated genes, we performed GO and KEGG pathway analyses to identify the biological pathways that were significantly enriched (FDR ≤ 0.05) in these modules. KEGG analyses revealed that the genes were highly enriched in the HIF-1 signaling pathway, microRNAs in cancer, renal cell carcinoma, and AGE-RAGE signaling pathway in diabetes complications, as shown in Figure 3A (Detailed information is provided in Supplementary Table S3). Gene Ontology analyses elucidated that these genes were enriched in response to hypoxia, monosaccharide metabolic process, angiogenesis regulation, vasculature development, reproductive process, and epidermis development, as shown in Figure 3B. Additional details are given in Supplementary Table S3.
Protein-protein interaction network analysis
From the 30 validated genes, we first performed the proteinprotein interaction (PPI) analysis using the STRING database ( Figure 4A). The resultant data is then imported to the Cytoscape plugin CytoHubba, and the top 8 genes with the highest Maximal Clique Centrality (MCC) score were considered hub genes: JUN, ATF3, VEGFA, SLC2A1, HK2, PTGS2, PFKFB3, and KLF6 ( Figure 4B and Supplementary Table S3).
Discussion
The interaction between host cells and Candida is central to the immunopathology of candidiasis in transplant patients; a comprehensive understanding of this synergy will identify new treatment strategies. Here, we investigate how human epithelial and endothelial cells communicate with different Candida species during infection. In this study, we constructed a weighted gene correlation network and performed differential gene expression analysis to identify genes that are important in host-candida interactions.
Comparative network analysis could rank genes for further investigation of their connectivity (Schadt et al., 2005). A distinct advantage of WGCNA is that it considers modules or gene clusters for constructing interactions, and the genes in a module are likely to be connected by the same regulatory pathways. Therefore, in this study, we aim to discover novel genes and molecular pathways in human-candida infection and to understand the regulation due to cell dynamics using the WGCNA and DESeq2 algorithms. Network depictions provided immediate insight into the relationships between the correlated modules. The construction of a gene coexpression network and differential gene expression analysis of the discovery and validation data set facilitated the identification of genes with similar biological functions by GO and KEGG analyses. According to the results of functional enrichment analysis, the top 3 GO terms and topmost KEGG pathway were a response to hypoxia, response to decreased oxygen, response to oxygen levels ( Figure 3B), and hypoxia-inducible factor 1 (HIF-1) signaling pathway ( Figure 3A). HIF-1 is a transcription factor that functions as a master regulator of oxygen homeostasis. It has been shown that suppressing HIF-1 helps treat cancer and ischemia (Ziello et al., 2007). All organs during the process of transplantation undergo hypoxic and ischemic injury. Low oxygen levels trigger the colonization of candida infection in the human host, resulting in complications like allograft rejection in SOT patients (Akhtar et al., 2014). We identified eight hub genes using PPI network analysis. Four hub genes (HK2, PFKFB3, SLC2A1, and VEGFA) are involved in the HIF-1 signaling pathway. The hexokinase isoenzyme (HK2) elevates innate immunity in hepatocellular carcinoma (Perrin-Cocon et al., 2021). HK2 and PFKFB3 are involved in glycolysis which affects the immune response against fungal infection (Perrin-Cocon et al., 2021); specifically, after transplantation, the PFKFB3 gene increase the risk of invasive pulmonary aspergillosis (Gonçalves et al., 2021). Huang et al. showed in their omics analysis that SLC2A1 is involved in ischemic reperfusion injury in liver transplant patients and forms the core gene network (Huang et al., 2019a). Vascular Endothelial Growth Factor A (VEGFA) is associated with an increased risk of chronic kidney disease (Anderson et al., 2018) but induces vasculogenesis. Kidney vasculature comprises vascular smooth muscle and endothelial cells (Udan et al., 2012). One of the most challenging components to handle during a kidney transplant is through vasculogenesis and angiogenesis processes (Munro and Davies, 2018;Lebedenko and Banerjee, 2021). HIF-1 stimulates the VEGFA to maintain oxygen delivery and protect the kidney (Hunga et al., 2013).
The other top enriched KEGG pathways in our analysis were microRNAs in cancer (hsa05211) and renal cell carcinoma (hsa05206). MicroRNAs play a diverse role in cancer and infections (Yong and Dutta, 2009). Recent advances in microRNA therapeutics have shown the extensive use of microRNAs for cancer and infections (Rupaimoole and Slack, 2017). There has been increased support for microRNA therapeutics in solid organ transplantation, including kidney (Wilflingseder et al., 2015;Jin et al., 2017;Ledeganck et al., 2019), lung (Benazzo et al., 2022), and heart transplantation (Hamdorf et al., 2017). Candida albicans have been linked to cancerous processes by taking advantage of the compromised immune system (Ramirez-Garcia et al., 2016;Chung et al., 2017;Sultan et al., 2022). Our PPI network analysis identified four hub genes (SLC2A1, VEGFA, JUN, and PTGS2) enriched in the cancer-related pathways. SLC2A1 belongs to a glucose transporter family and has been reported to be associated with Frontiers in Genetics frontiersin.org HCC (Kim et al., 2017). SLC2A1 is also essential to IRI during liver transplantation (Huang et al., 2019b) and a diagnostic biomarker for colorectal cancer (CRC) (Liu et al., 2022). In CRC, the SLC2A1 gene infiltrates the CD4 + T cell, neutrophil, dendritic cells, and B cells (Liu et al., 2022). Candidiasis is one of the risk factors for Oral squamous cell carcinoma (OSCC). The transcriptomics data analysis revealed that VEGFA and JUN are highly regulated in OSCC invasion and metastasis (Vadovics et al., 2022). JUN is a member of the activator protein-1 family of oncogenic transcription factors, which is involved in various cancer-related and cell signaling pathways such as tumorigenesis, cell differentiation, and angiogenesis (Brennan et al., 2020). Post renal transplantation, the activation of c-JUN affects acute humoral rejection and acute T-cell-mediated rejection (Kobayashi et al., 2010). c-JUN is also associated with reduced graft function and plays an important role in renal pathophysiological events (Kobayashi et al., 2010). Prostaglandin E2 (PGE2) is an inflammatory mediator produced by the Prostaglandin-endoperoxide synthase (PTGS2) enzyme, and PGE2 promotes candida morphogenesis. In response to candida infection, PTGS2 activation promotes NF-kB and MAPK signaling pathways (Deva et al., 2003). In OSCC, PTGS2 involves an inflammatory response to infection by promoting tumorigenesis (Cacina et al., 2018) and activating transcription factor 3 (ATF3), one of the 8 hub genes that regulate the PTGS2 during acute inflammation (Hellmann et al., 2015) and helps in the homeostasis of the metabolism and immune system (Sha et al., 2017). Zhu et al. also showed that ATF3 is one of the top hub genes in samples infected with 4 different candida species (Zhu et al., 2022). Using bioinformatics omics analysis, ATF3 and Kruppel-like factor 6 (KLF6, hub gene) are shown to be the central players in ischemic reperfusion injury in liver transplant patients (Huang et al., 2019b). KLF6 promotes inflammation and oxidative stress by regulating HIF-1 expression in macrophages (Kim et al., 2020).
Another enriched KEGG pathway was the AGE-RAGE signaling in diabetes complications (hsa04933). Endoplasmic reticulum stress due to AGE-RAGE plays an essential role in renal inflammation, diabetic nephropathy (Pathomthongtaweechai and Chutipongtanate, 2020) and early-stage renal disease (Meerwaldt et al., 2009). Advanced glycation end products (AGEs) may also play a role in the hardening of arteries after renal transplantation (Liu et al., 2015b). Our two hub genes, JUN and VEGFA, showed enrichment in the AGE-RAGE signaling pathway in diabetes complications. Poorly controlled diabetes increases the risk of fungal infections (Rodrigues et al., 2019). Some diabetes-related complications include cardiovascular disease, kidney disease, neuropathy, hearing loss, vision loss, Alzheimer's, liver disease, etc. (Deshpande et al., 2008;Prasad et al., 2016). VEGFA and JUN were identified as the central players in diabetic nephropathy (Oltean et al., 2015;Wang et al., 2021) and Alzheimer's disease (Zu et al., 2021) whereas, VEGFA was associated with diabetic retinopathy (Bucolo et al., 2021), cardiac autonomic neuropathy (Ravichandran et al., 2019), and non-alcoholic fatty liver disease-hepatocellular carcinoma (Shen et al., 2022). Each hub gene plays a vital and diverse role in the pathways and biological processes. Therefore, more research is warranted on the divergent roles of these genes' signaling and regulatory mechanisms during infection, cancer, and transplantation.
Limitations
WGCNA lacks resolution as it decomposes a group of genes into a single eigenvalue that may not correctly represent a single gene's expression profile or pathway changes. Further analysis may be needed to detect changes in the expression of individual processes. Another limitation of the study is the small sample size; therefore, we present this study as a proof of concept to be validated in a larger cohort. The current study used cell lines from epithelial and endothelial cells; thus, the identified gene markers should be validated from the peripheral blood transcriptome of candidiasis patients for non-invasive clinical relevance.
Data availability statement
Publicly available datasets were analyzed in this study. This data can be found here: https://www.ncbi.nlm.nih.gov/geo/ query/acc.cgi?acc=GSE56093.
Author contributions
AM conceived the study design. SN developed the methodology and performed data analysis, and wrote the manuscript. SN and AM revised and edited the manuscript.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's note All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
|
v2
|
2022-11-23T14:39:36.203Z
|
2022-11-22T00:00:00.000Z
|
253797547
|
s2orc/train
|
CAR-T cell therapy in triple-negative breast cancer: Hunting the invisible devil
Triple-negative breast cancer (TNBC) is known as the most intricate and hard-to-treat subtype of breast cancer. TNBC cells do not express the well-known estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 (HER2) expressed by other breast cancer subtypes. This phenomenon leaves no room for novel treatment approaches including endocrine and HER2-specific antibody therapies. To date, surgery, radiotherapy, and systemic chemotherapy remain the principal therapy options for TNBC treatment. However, in numerous cases, these approaches either result in minimal clinical benefit or are nonfunctional, resulting in disease recurrence and poor prognosis. Nowadays, chimeric antigen receptor T cell (CAR-T) therapy is becoming more established as an option for the treatment of various types of hematologic malignancies. CAR-Ts are genetically engineered T lymphocytes that employ the body’s immune system mechanisms to selectively recognize cancer cells expressing tumor-associated antigens (TAAs) of interest and efficiently eliminate them. However, despite the clinical triumph of CAR-T therapy in hematologic neoplasms, CAR-T therapy of solid tumors, including TNBC, has been much more challenging. In this review, we will discuss the success of CAR-T therapy in hematological neoplasms and its caveats in solid tumors, and then we summarize the potential CAR-T targetable TAAs in TNBC studied in different investigational stages.
Introduction
Breast cancer is the malignancy arising from the breast tissues. It is the most common malignancy worldwide and is estimated to be detected in more than 280,000 women in the United States in 2022 (1). According to the estimations, more than 43,000 women will die because of breast cancer only in 2022 (1). The molecular process that leads to the emergence of breast cancer entails the active proliferation of epithelial cells of the breast tissue resulting in the formation of malignant cells in the ductal or lobular compartment of the breast (2,3). There are various types of classifications for categorizing breast cancer. Histologic grade, disease stage, and expression of classic hormone and growth factor receptors are three well-known criteria used for breast cancer classification and determining the aggressive capacity of a breast tumor (2)(3)(4)(5)(6)(7). The hormone and growth factor receptors employed for breast cancer classification include estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2) (2)(3)(4)(5)(6)(7). Based on the expression or lack of expression of these receptors and the expression rate of Ki-67, there are four subtypes of breast cancer including Luminal A, Luminal B, HER2 amplified, and triple-negative breast cancer (TNBC) (2)(3)(4)(5)(6)(7). The first three subtypes are easier to manage since the patients with these subtypes are eligible for antibody-based therapies such as anti-HER2 monoclonal antibodies (mAbs) or antibody-drug conjugates (ADCs), tyrosine kinase inhibitors, and endocrine therapies (including ER degraders) (8)(9)(10). Dissimilar from these subtypes, TNBC is a heterogeneous type of basal-like tumor that does not express the abovementioned receptors and it represents about 15 to 20% of all breast cancer cases (11).
Due to the unresponsiveness of TNBC patients to anti-HER2 mAb-based and endocrine therapies, traditional anticancer treatments including surgery, radiotherapy, and chemotherapy continue to be the commonly available therapies for these patients (12,13). Of note, these traditional cancer treatment modalities manage to eradicate malignant cells in a nonspecific manner resulting in mild to severe treatment-related side effects (13). Additionally, poly (ADP-ribose) polymerase (PARP) inhibitors are a class of pharmacological inhibitors that have been approved by the US Food and Drug Administration (FDA) for the treatment of TNBCs with BRCA mutations (14). Recently, many bodies of research have indicated that the immune system pathways are remarkably involved in the emergence and progression of TNBC (15,16). In this case, researchers have utilized immune checkpoint blockade therapies such as programmed death-1 (PD-1)-or programmed deathligand 1 (PD-L1)-specific antibodies for suppressing the outgrowth of TNBC (16,17). Moreover, T-cell-redirecting bispecific antibodies (TRBAs) have also been under investigation as candidates for the treatment of TNBC (16,17). Recently, the US FDA approved the PD-L1-specific mAb atezolizumab in combination with the chemotherapeutic agent nab-paclitaxel for the treatment of patients with PD-L1proficient unresectable TNBC (18). Such progress in the field of TNBC immunotherapy has encouraged researchers to employ other types of cancer immunotherapy such as cancer vaccines and adoptive cell therapy using genetically engineered immune cells for a more selective and successful fight against TNBC (16,19,20).
The importance of cancer-specific immune cells in TNBC was elucidated when researchers discovered that the presence of tumor-infiltrating lymphocytes (TILs) in residual disease following neoadjuvant chemotherapy is correlated with superior prognosis in TNBC patients (21). In a population of TILs, there are tumor-reactive T cells that recognize cancerspecific antigens and are responsible for the antitumor activity of a TIL population (22,23). However, there are several limitations in regards to the application of TILs for adoptive cell therapy of various types of tumors. These caveats include the exhausted phenotype of TILs resulting from multiple target antigen encountering, and difficulties in regards to the isolation and ex vivo expansion of TILs (24,25). Genetically engineered T cells expressing chimeric antigen receptors (CAR-Ts) have recently changed the face of immune cell-based cancer immunotherapy (26,27). CAR-Ts possess antibody-derived targeting domains linked to T-cell activating domains which grant them the ability to recognize cancer cells via target cell membrane-expressed antigens independent of major histocompatibility complexes (MHCs) (26,27). Moreover, CAR-Ts can be easily expanded to clinically relevant scales; therefore, they are considered an ideal alternative to TILs (26,27). The major purpose of this review is to discuss CAR-T therapy hurdles in solid tumors with a specific focus on TNBC, shine a light on the crucial stratagems to tackle these caveats, and discuss the potential and promising target antigens under investigation in various stages, from early developmental stages to clinical settings, for CAR-T therapy of TNBC ( Figure 1).
CAR-T therapy fundamentals
About twenty years ago, when a person was diagnosed with cancer, surgery, chemotherapy, and radiotherapy were the only available treatment options. However, these treatment modalities could not mediate complete disease remission in most cases (28-30). Additionally, disease relapse, shortly after the completion of the treatment, was always a challenge (28-30). Alongside the mentioned limitations, the severity of the treatment-related side effects was another factor affecting the patients' general well-being (28-30). Years later, with the emergence of immunotherapy and the development of novel treatment modalities, cancer treatment became more selective and efficient resulting in prolonged survival and reduced adverse events in patients.
CAR-Ts are the ultimate result of years of research and experience in various aspects of immunotherapy and cell therapy. The process of CAR-T generation starts with obtaining peripheral blood mononuclear cells (PBMCs) from patients (in the case of autologous CAR-T therapy) or thirdparty healthy donors (in the case of allogeneic CAR-T therapy) (31, 32). From the population of the isolated PBMCs, T lymphocytes are isolated. These T cells undergo activation and genetic manipulation steps, usually using retroviruses, to express a synthetic chimeric antigen receptor (CAR) on their surface (32). After expansion to a number of cells required for clinical applications, the engineered T cells are infused into recipients. It is worth mentioning that before CAR-T administration, patients usually undergo lymphodepletion chemotherapy (32).
A CAR construct is made of three important sections. An extracellular domain, a transmembrane domain, and an endodomain (32). The ectodomain of CARs is derived from a tumor antigen-specific fragment of a mAb such as a single-chain fragment variable (scFv) transmembrane domain exhibited prolonged in vitro expansion (up to 3 months) after a single TCR stimulation and without any IL-2 supplementation (37). Other researchers have added that CAR-Ts whose hinge and transmembrane domain are both based on CD28 are capable of producing target antigen-dependent inflammatory cytokines more than CAR-Ts whose hinge and transmembrane domain are both based on CD8 (38). Moreover, it has been demonstrated that these CAR-Ts require a lower antigen density for activation (studied only in CD19-redirected CAR-Ts with 4-1BB as the costimulatory domain) (39). Conclusively, broader investigation into the impact of CAR hinge and transmembrane domains on the persistence, tumoricidal efficacy, and phenotypic characteristics of CAR-Ts can further help the development of CAR-Ts with greater therapeutic benefits.
The endodomain of CARs is responsible for activating the effector cell upon target antigen encountering. This part of CARs has a CD3z signaling domain derived from the CD3 complex of the T-cell receptor (TCR) fused to intracellular costimulatory domains such as CD27, CD28, 4-1BB, OX40, and/or ICOS (32). Of note, dissimilar from first-generation CAR-Ts which did not possess any intracellular costimulatory domains, second-and third-generation CAR-Ts have one and two intracellular costimulatory domains, respectively (40-43). According to scientific evidence, the addition of intracellular costimulatory domains to the construct of CARs resulted in superior postinfusion expansion and persistence of second-and thirdgeneration CAR-Ts in comparison with their first-generation counterparts (40-43). Recently, the construct of CARs has been further modified to achieve certain aims during CAR-T therapy. For instance, fourth-generation CAR-Ts possess an intracellular expression inducer of a cytokine of interest resulting in CAR-Tmediated tumor site delivery of a cytokine of interest leading to more enhanced and safer antitumor responses, especially in solid tumor CAR-T therapy (44-46). In detail, one of the well-known characteristics of solid tumors is the heterogeneity of tumor cells. Various populations of tumor cells might express different antigens in which case a single type of CAR-Ts redirected against a particular type of target antigen may not be clinically beneficial. However, expanding the antitumor responses by triggering the immune responses by other types of endogenous immune cells can help mediate more efficient tumoricidal reactions. In this regard, CAR-Ts have been engineered to specifically deliver a transgenic product, which can be chemokines or cytokines, to the targeted tumor sites (44, 46, 47). In detail, Chmielewski et al. generated CAR-Ts harboring an engineered intracellular expression inducer of IL-12 which can recruit macrophages (47). These CAR-Ts, which are also regarded as "T-cell redirected for universal cytokine-mediated killing (TRUCKs)" or "armored CAR-Ts", are engineered to produce and release inducible IL-12 upon encountering CAR target antigen (47). This mechanism results in simultaneous antitumor attacks against both tumor cells expressing the CARredirected target antigen and tumor cells deficient in its expression (47). This strategy demonstrated that local delivery of IL-12 by CAR-Ts in the solid tumor tissues can be applied for targeting both target antigen-negative and hard-to-reach tumor cells through recruiting and activating innate immune components without the risk of systemic cytokine administration-related toxicities (47). Such CAR-Ts can be clinically valuable for targeting different target antigens of TNBC.
Moreover, fifth-generation CAR-Ts have also been developed which harbor an intracellular domain of a cytokine receptor, for example, interleukin (IL) 2 receptor subunit beta (IL-2Rb) (44-46). In detail, proper activation and expansion of T cells are highly dependent on various signals including T-cell receptor (TCR), co-stimulatory, and cytokine-mediated signals (48). In the case of CAR-Ts, T cells can only benefit from TCR signaling which is provided through their CD3z domain and costimulatory domains (48). Therefore, researchers have generated CAR-Ts with novel constructs which are capable of inducing cytokine signaling once the effector cells encounter the target antigen (48). Fifth-generation CAR-Ts harbor truncated cytoplasmic domain of the IL-2Rb and a STAT3-binding tyrosine-X-X-glutamine (YXXQ) motif alongside a primary activation domain and a co-stimulatory domain (48). Kagoya et al. reported that these CAR-Ts demonstrated target antigenspecific activation of the JAK kinase and the STAT3 and STAT5 transcription factor signaling pathways (48). These signaling pathways prevented terminal phenotypic differentiation of the effector cells and mediated their efficient expansion in vitro (48). Moreover, in vivo assessments proved that these CAR-Ts function efficiently in terms of persistence and tumoricidal reactions in comparison with their conventional counterparts, both in hematologic and solid tumors (48). Such novel CAR constructs can be evaluated in the case of TNBC CAR-T therapy to help achieve more efficient and less toxic antitumor responses. The constructs of different generations of CARs have been illustrated in Figure 2.
In regards to the application of scFvs as the targeting domain of CARs, it has been discovered that target antigen-independent CAR scFv aggregation occurs which is a result of CAR CD3z phosphorylation and/or the intrinsic instability of the V H and V L chains of the scFv (49). This phenomenon, known as "tonic signaling", can mediate premature CAR-T activation (before encountering the intended target antigen) and render these effector cells exhausted leading to their impaired antitumor activity (49). Long et al. have demonstrated that the 4-1BB costimulatory domain in CAR-Ts better manages to minimize the phenotypic characteristics of CAR-T exhaustion mediated by the aggregation of CAR scFvs in comparison with CAR-Ts harboring CD28 as their co-stimulatory domains (49). These findings are consistent with the reports by Frigault and colleagues who reported that antigen-independent signaling cascades in CAR-T with CD28 as their co-stimulatory domains culminate in their exhaustion (based on animal studies) (37). Such findings might explain why CAR-Ts with the 4-1BB co-stimulatory domain demonstrate superior postinfusion persistence over their counterparts with the CD28 costimulatory domain (49).
Alongside the mentioned findings, other researchers have proposed the replacement of particular framework residues of the CAR scFv as an attempt to minimize the aggregation tendency of these scFvs and circumvent spontaneous tonic signaling (50). In detail, Landoni et al. employed in silico techniques to find the framework residues of a CSPG4-specific scFv (known as 763.74 scFv) that contributed to its stability and whose substations would result in decreasing the tendency of scFv aggregation, and consequently tonic signaling and CAR-T exhaustion (50). Moreover, the importance of scFv is also accentuated in regard to the on-target off-tumor effects of CAR-Ts. In the context of solid tumors, target antigens are usually expressed by healthy tissues as well (however, at physiological levels), which results in the CAR-T-mediated cytolytic reactions against healthy cells (51). Since the discovery of tumor-specific antigens is not always a feasible task, researchers have suggested targeting the tumor-specific glycoforms (called T, Tn, or sialyl Tn glycoforms) of commonly known TAAs (such as MUC1) (51). mAbs against such glycoforms are called cancer-specific mAbs (CasMabs) and scFvs derived from these CasMabs could be applied as the targeting domain of CAR-Ts for the development of cancerspecific CAR-Ts (Cas-CAR-Ts) (51). Such CAR-Ts have been developed and assessed in different investigational stages, and The structure of a CAR and its five generations. CARs are the result of meticulous protein engineering. The targeting domain of CARs is usually derived from the single-chain fragment variable (scFv) of a monoclonal antibody. scFvs are made from the variable light chain (V L ) and variable heavy chain (V H ) of a monoclonal antibody fused together through a synthetic linker peptide. A spacer called hinge connects the targeting domain of CARs to their transmembrane domain, which connects the ectodomain to the endodomain. Currently, the endodomain of CARs consists of one or two costimulatory domains and an activation domain. An interleukin expression inducer domain and an interleukin intracellular receptor are also located on the endodomain of the fourth-and fifth-generation CARs, respectively. Of note, the first-generation CARs lacked a costimulatory domain which led to their inadequate in vivo persistence and weak antitumor responses. CAR, chimeric antigen receptor; Co-S, costimulatory domain; IL, interleukin; ITAM, immunoreceptor tyrosine-based activation motif; mAb, monoclonal antibody; scFv, single-chain fragment variable; TCR, T-cell receptor; TM, transmembrane domain.
have proven to be efficient and safe (51). As another scFv-related strategy to minimize off-tumor toxicities, researchers have demonstrated that CAR-Ts equipped with targeting domains whose affinity towards their target antigen is moderate (micromolar affinity range), rather than high (nanomolar affinity range), manage to efficiently target tumor cells with high target antigen density while sparing normal cells which express the target antigen at a lower level (52). All of the mentioned strategies accentuate the importance of CAR targeting domains, particularly scFvs, and how their engineering can culminate in the development of safer and more efficient CAR-Ts (52).
CAR-T therapy challenges in solid tumors
The tremendous success of CAR-T therapy in relapsed/ refractory (R/R) B-cell hematologic malignancies has rendered this type of cancer treatment approach a considerable treatment option (53-55). Based on such favorable clinical outcomes especially in patients non-responsive to the other types of treatments, the US FDA has approved six CAR-T products for the treatment of patients with failed previous lines of treatments. The list of the FDA-approved CAR-T products has been summarized in Table 1. However, patients with several hematologic malignancies, including T-cell neoplasms, as well as solid tumor patients have not yet benefited from the anticancer capability of this unique type of therapy (68)(69)(70). There are various limitations attributed to the inability of CAR-Ts to mediate tumor rejection and disease remission in solid tumors. In detail, CAR-Ts are redirected towards TAAs or TSAs (51, 71). Therefore, ideal CAR-T targets are antigens that are absent or have a low-level expression on the healthy cells of normal tissues but are overexpressed by malignant cells (51, 71). Targeting such target antigens can result in minimal unwanted off-tumor toxicities toward healthy tissues. However, finding target antigens with such characteristics is hardly feasible in solid tumors (51, 71). Additionally, another limitation of solid tumor CAR-T therapy is intratumor antigen heterogeneity described as tumor cells of a particular type of tumor expressing different levels of a CAR-redirected target antigen on their surface or not expressing the target antigen at all (72)(73)(74). This phenomenon results in impaired detection of malignant cells by CAR-Ts further leading to disease escape and relapse (72)(73)(74).
Another limitation of CAR-T therapy in solid tumors is associated with the immunosuppressive tumor microenvironment (TME). The TME consists of various types of immune suppressor cells such as regulatory T cells (Tregs), myeloid-derived suppressor cells (MDSCs), cancer-associated fibroblasts (CAFs), and tumorassociated macrophages (TAMs) (75-78). There are also various suppressive factors such as chemokines, cytokines, and extracellular matrix (ECM) present in the TME. The immune suppressor cells residing in the TME are beneficial for tumors in terms of supporting their progression, angiogenesis, and metastasis by providing them with various types of growth factors, chemokines, and cytokines such as ILs, transforming growth factor beta (TGF-b), indoleamine 2,3-dioxygenase (IDO), and vascular endothelial growth factor (VEGF) (75-78). Moreover, the expression of immune checkpoint molecules (such as CTLA-4 and PD-1) on tumorresiding T lymphocytes is another factor that hampers antitumor reactions (79-82). Overall, the mentioned immunosuppressive factors in the TME remarkably suppress CAR-T-mediated antitumor reactions in solid tumor CAR-T therapy.
In addition to all of the abovementioned hurdles, the inadequate trafficking and infiltration of CAR-Ts into the tumor tissues is known as another major limitation impairing Idecabtagene vicleucel the functionality and tumoricidal activity of CAR-Ts (83)(84)(85)(86). Dissimilar to CAR-T therapy in hematologic malignancies where CAR-Ts encounter target cells in the bloodstream and the lymph nodes, in the case of solid tumors, CAR-Ts must cross the vascular endothelium and penetrate the tumor tissues (86).
There are various tumor-associated action mechanisms that remarkably limit the accessibility of CAR-Ts to tumor cells (83)(84)(85)(86). In brief, CAR-T tumor site penetration and trafficking are substantially correlated with the presence of various types of chemokines (87). Tumor tissues downregulate the expression of such chemokines leading to restricted tumor tissue CAR-T infiltration (87). Additionally, researchers demonstrated that the dense nature of the ECM of the tumor tissues is an important obstacle to CAR-T penetration and infiltration into the TME (88). Tumor ECM also protects the TME from the antitumor activity of CAR-Ts as it has been demonstrated that its degradation can result in superior CAR-Tmediated tumor cell eradication (89). In a nutshell, profound knowledge of solid tumor mechanisms can better help us develop counterstrategies to overcome the mentioned hurdles for a more efficient and safer CAR-T therapy in patients with various types of solid tumors. A simplified view of the principal limitations of CAR-T therapy in solid tumors and examples of counterstrategies for tackling these hurdles have been brought together in Figure 3 and Table 2, respectively.
Targets antigens investigated for the CAR-T therapy of TNBC
Chondroitin sulfate proteoglycan 4 (CSPG4) CSPG4 is a cell membrane-spanning protein with a high level of glycosylation (137). Various roles have been proposed for CCSPG4 which include involvement in the regulation of neuronal networks, replacement of epithelial keratinocytes, and homeostasis of epidermal stem cells (137). CSPG4 has been studied as a target antigen in cancer immunotherapy based on its limited expression level in normal tissues and overexpression and supporting roles in cancer progression and invasion in various types of neoplasms, including TNBC (137). Researchers demonstrated that antibodymediated targeting of CSPG4 using an scFv fused to Tau, which is a negative regulator of protein translation, results in efficient cytotoxicity against CSPG4-proficient TNBC-derived cell lines including MDA-MB-231 and Hs 578T (138).
In 2014, Geldres et al. generated second-generation CSPG4redirected CAR-Ts, and reported that these cells significantly suppressed the growth of various CSPG4-expressing cell lines (including SENMA, CLB, UACC-812, MILL, MDA-MB-231, PHI, and PCI-30) (139). Moreover, these researchers reported that these CAR-Ts suppressed tumor cell growth in human melanoma, head and neck squamous cell carcinoma, and breast carcinoma preclinical mouse models established using SENMA, PCI-30, and UACC-812 cell lines, respectively (139). In the same year, Beard et al. generated second-generation CSPG4-redirected CAR-Ts using murine-based scFvs and reported target antigendependent cytotoxicity and cytokine secretion of these CAR-Ts against glioblastoma, breast cancer, mesothelioma, osteosarcoma, and melanoma CSPG4-expressing cell lines (A1207, MDAMB231, Mill, MgG-63, and mel938, respectively) and glioblastoma stem cells (140). Such data may support the applicability of CSPG4-redirected CAR-Ts in TNBC and various other neoplasms. However, more preclinical and clinical data are critically required for more reliable conclusions in this regard.
Intracellular adhesion molecule-1 (ICAM-1)
ICAM-1 is a transmembrane glycoprotein that mediates the transmigration of leukocytes through the endothelium of various cell types (141). In 2014, Guo et al. reported that ICAM-1 is upregulated and overexpressed in TNBC cell lines and tissues and it might act as a possible biomarker and target antigen for the diagnosis and treatment of TNBC (142). These researchers demonstrated that antibody-assisted targeting of ICAM-1 is feasible and efficient even after systemic administration into xenograft TNBC tumor models (142). In 2017, Park et al. generated ICAM-1-redirected CAR-Ts with micromolar affinity against ICAM-1 (instead of nanomolar affinity) to prevent CAR-T-mediated cytotoxicity in non-malignant cells with normal ICAM-1 expression levels (52). These researchers demonstrated that these affinity-tuned CAR-Ts demonstrated superior tumoricidal efficacy and safety index compared to their higher affinity counterparts (52). Moreover, in 2019, using preclinical models, it was demonstrated that micromolar affinity-tuned ICAM-1-redirected CAR-Ts can target tumors with a high level of ICAM-1 expression while sparing normal tissues with lower and basal ICAM-1 expression levels with significant efficacy (143). The researchers of this report also suggested the investigation of micromolar affinity-tuned ICAM-1-redirected CAR-Ts in a Phase I clinical trial for assessing the safety and possible efficacy of these cells against R/R thyroid cancers (143). Moreover, Yang et al. generated ICAM-1redirected CAR-Ts and investigated the efficacy of these cells in vitro and in vivo (144). These researchers demonstrated that in vitro CAR-T-mediated cytotoxicity against HeLa and MDA-MB-231 cell lines was specific and ICAM-1 expressiondependent (144). However, the ongoing preclinical assessments of this study, from which no results have yet been reported, will further elucidate the applicability and safety index of targeting ICAM-1 using CAT-Ts for the treatment of TNBC (144).
Natural killer group 2, member D ligand (NKG2DL)
Malignant cells, such as TNBC cells, exhibit upregulated levels of stress-induced ligands, which are naturally recognized by molecules such as natural killer group 2, member D (NKG2D) (145). NKG2DL has been considered an immunotherapy target antigen in various studies (146). DAP10 costimulation (146). These researchers demonstrated that T cells expressing these CARs reacted to NKG2DLexpressing tumor cells by secreting cytokines and chemokines and exhibiting cytotoxicity (146). Moreover, in vivo results also supported the tumor suppression ability of these CAR-Ts (146). Other researchers have also reported that NKG2DL-redirected CAR-Ts mediated disease remission in a patient with acute myeloid leukemia (AML) (NCT02203825) (147,148).
In 2018, Han et al. evaluated the antitumor activity of NKG2DL-redirected CAR-Ts against TNBC cell lines and cell line-established preclinical TNBC mouse models (149). These researchers generated different NKG2DL-redirected CAR-Ts by fusing the extracellular domain of human NKG2D to the TCR CD3z alone or with CD27 or 4-1BB co-stimulatory domains (149). They reported that the in vitro expansion of CAR-Ts without any co-stimulatory domain was dependent on the high CD25 expression and the presence of IL-2 (149). Moreover, it was reported that NKG2DL-redirected CAR-Ts efficiently recognized TNBC NKG2DL-expressing MDA-MB-231 and MDA-MB-468 cell lines and eliminated them (149). In vivo experiments demonstrated tumor growth suppression ability of NKG2DL-redirected CAR-Ts in MDA-MB-231-derived TNBC preclinical mouse models paving the way for more preclinical and early phase clinical evaluations of this potential TNBC treatment approach (149). It is worth mentioning that a Phase I clinical trial (NCT04107142) has investigated the safety and tolerability of NKG2DL-redirected CAR-Ts in patients with R/R solid tumors including TNBC; however, the results are yet to be reported.
Receptor tyrosine kinase AXL
AXL is a well-known member of the TAM family of receptor tyrosine kinases. Growth arrest-specific protein 6 (GAS6) is known as a high-affinity ligand for AXL (150, 151). AXL signaling operates as a critical pathway mediating tumor cell survival, expansion, migration, and invasion suggesting the potential of AXL as a suitable cancer treatment target antigen (150, 151). Normally, AXL has a low expression level during adulthood; however, its abnormal expression has been observed in various types of neoplasms including breast cancer (152).
Ye et al. developed an anti-AXL mAb with the ability to target both human and murine AXL (153). These researchers demonstrated that this mAb suppressed tumor outgrowth, reduced distant organ tumor cell metastasis, and had enhancing effects on anti-VEGF treatment in MDA-MB-231established breast cancer xenograft models (153). In 2018, Wei et al. reported the overexpression of AXL in multiple tumor cell lines (including MDA-MB-231, but not MCF-7) and patientderived samples (154). These researchers also generated AXLredirected CAR-Ts using a novel AXL-specific scFv, and demonstrated that these CAR-Ts exhibit target antigendependent cytotoxicity and cytokine secretion against the AXL-expressing TNBC cell line MDA-MB-231 (154). In vivo evaluations of these CAR-Ts in MDA-MB-231-established xenograft models also indicated significant tumoricidal reactions and in vivo persistence (154). Moreover, Zhao et al. designed AXL-redirected CAR-Ts co-expressing a constitutively activated IL-7 receptor (C7R), and reported that these CAR-Ts
Limitations
Examples of strategies (71) Immunosuppressive TME * Tackling the hypoxic TME nature (101-104) * Metabolic reprogramming of CAR-Ts (105) In vivo CAR-T exhaustion * Immune checkpoint blockade therapy (106) * CAR cells with PD-1 disruption (107 Tumor endothelial marker 8 (TEM8) TEM8, alternatively known as anthrax toxin receptor 1 (ANTXR1), is an integrin-like transmembrane protein with roles in the migration and invasion of endothelial cells (156). TEM8 is overexpressed in invasive and TNBC breast cancer tissue correlating with elevated possibility and potential of disease relapse in basal breast cancer (157)(158)(159). Researchers have demonstrated that overexpressing TEM8 in preclinical breast cancer models resulted in the amplified capability of the tumor for expansion and metastasis whereas blocking or knocking out TEM8 expression hindered tumor progression in various preclinical models (159-162). Byrd et al. generated TEM8-redirected CAR-Ts and assessed the cytotoxicity of these cells in vitro and in vivo (162). They demonstrated that TEM8-redirected CAR-Ts managed to efficiently inhibit the growth of various TNBC lines (Hs 578T, MDA-MB-231, MDA-MB-436, and MDA-MB-468), a human breast tumorassociated endothelial cell line (HC 6020), and murine tumorassociated endothelial cell lines (2H11 and bEND.3) (162). It is worth mentioning that all of these tumor-associated endothelial cell lines tested positive for variable levels of TEM8 expression (162). Moreover, according to in vivo experiments, these researchers indicated that their TEM8-redirected CAR-Ts mediated tumor growth suppression and prolonged survival in localized patient-derived xenograft (PDX) and lung metastatic TNBC cell line LMD231-established xenograft preclinical models and via eliminating TEM8-expressing TNBC tumor cells and targeting the tumor endothelium to prevent further tumor neovascularization (162). Moreover, Petrovic et al. generated different types of TEM8-redirected CAR-Ts including one type generated using the same antibody utilized in the discussed study by Byrd et al. (163). The in vivo results reported by Petrovic and co-workers indicated that these CAR-Ts rapidly and selectively disappeared from the circulation and caused rapid toxicity after administration into healthy C57BL6 and NSG mice (163). Using TEM8-knockout mouse models, Petrovic et al. demonstrated that the selective clearance of these CAR-Ts from the circulation was because of their cytotoxicity toward normal tissue-expressed TEM8 (163). Such data may speculate critical safety concerns regarding the potential ontarget off-tumor toxicity in CAR-T-mediated TEM8 targeting in TNBC (163). Therefore, more comprehensive investigations are required before moving on to clinical trials.
Integrin alpha V beta 3 (a v b 3 )
Integrins are adhesion receptors with critical roles in intercellular signal transduction; however, they are also involved in tumor cell migration, tissue invasion, and survival (164). a v b 3 is a well-known integrin in neoplasm-related studies. The expression of this integrin is normally observed in newly forming endothelial cells; however, accumulating evidence confirms its expression in various types of malignancies (165-168). Targeting a v b 3 using antagonists has been evaluated in clinical settings but has not been found effective (169). Moreover, antibody-assisted targeting of this integrin also did not result in beneficial therapeutic effects (170)(171)(172). In this regard, researchers have recently contemplated targeting a v b 3 using CAR-Ts.
In 2018, Wallstabe et al. investigated the applicability of targeting a v b 3 using CAR-Ts (173). These researchers used two codon-optimized scFvs as the targeting domain of two different CARs. Of note, these scFvs had been previously humanized (174). The a v b 3 -redirected CAR-Ts developed by Wallstabe et al. were also equipped with a truncated epidermal growth factor receptor (EGFRt) enabling the eradication of the infused CAR-Ts by administering the anti-EGFR mAb cetuximab (173). In terms of detecting the expression of a v b 3 on human cell lines, these researchers confirmed the high-level expression of this target antigen on the human TNBC cell line MDA-MB-231 as well as on various melanoma cell lines (173). Their data further confirmed the suitability of integrin a v b 3 as a target antigen by being only overexpressed on malignant cells in hematologic and non-hematologic cancers (173). In vitro functionality assessments demonstrated that a v b 3 -redirected CAR-Ts exhibited potent and exclusive tumoricidal activity by producing IFN-g and IL-2 upon encountering a v b 3 -expressing tumor cells (173). It is worth mentioning that the in vivo preclinical assessments of this study did not involve mouse models of TNBC, and the preclinical models in which a v b 3redirected CAR-Ts demonstrated promising tumoricidal activity were A-375 cell line-established melanoma models (173). Overall, these data only demonstrate that a v b 3 might be a target antigen for CAR-T therapy of TNBC and they cannot guarantee that targeting a v b 3 is safe and effective, and does not result in unwanted toxicities towards normal tissues (173). It is worth noting that CAR-Ts redirected against a v b 3 have also been evaluated for targeting other malignancies such as glioma and it has been indicated that this target antigen holds promising immunotherapeutic value with a low risk of on-target off-tumor toxicity due to its restricted expression on normal tissues (175).
Receptor tyrosine kinase-like orphan receptor 1 (ROR1) ROR1 is a type 1 membrane-spanning tyrosine kinase receptor with significant roles in embryonic and fetal development (176,177). Normally, ROR1 expression is observed during embryogenesis but not in normal adult tissues except for adipocytes and a subset of immature B-cell precursors (176,177). Malignancy-associated expression of ROR1 has been observed in B-cell chronic lymphocytic leukemia (B-CLL), mantle cell lymphoma (MCL), breast cancer, and ovarian cancer (178).
Various researchers have generated ROR-1-redirected CAR-Ts, and have investigated their various aspects in solid tumors and hematologic malignancies (179)(180)(181)(182). In 2019, Wallstabe et al. developed microphysiologic three-dimensional (3D) MDA-MB-231-established breast cancer and A549-established lung cancer models, and reported that ROR1-redirected CAR-Ts efficiently infiltrated into the tumor tissues and mediated tumoricidal reactions against multiple layers of malignant cells (183). These researchers suggested that such 3D models may act as reliable platforms for evaluating the safety and efficacy of CAR-Ts before preclinical and clinical assessments (183). (182). In this case, these researchers developed ROR1-redirected CAR-Ts harboring synthetic Notch (synNotch) receptors specific for EpCAM or B7-H3 (expressed by ROR1-expressing tumor cells but not ROR1-expressing stromal cells) and reported that these CAR-Ts safely mediated efficient tumoricidal activity without toxicity (182).
Moreover, a Phase I clinical trial (NCT02706392) has investigated the safety of second generation ROR1-redirected CAR-Ts in patients with various types of neoplasms including metastatic TNBC (184). So far, the reported results of 4 enrolled TNBC patients indicated that dose-limiting toxicities, severe neurotoxicity, or severe cytokine release syndrome (CRS) were not detected at dose levels 1 and 2 (184). It is worth mentioning that half of the patients experienced grade 1 CRS (184). Also, post-CAR-T administration tumor tissue biopsy demonstrated infiltration of CD3-positive T cells and macrophages proposing efficient tumor site CAR-T trafficking (184). In regards to clinical outcomes, 2 out of 4 patients achieved stable disease (one at 15 weeks and the other at 19 weeks post infusion) (184). One patient achieved stable disease after the first CAR-T administration and established partial response following the second CAR-T administration which has prolonged for 14 weeks as of the time of the report (184). Of note, the results of this trial are expected to be updated (184).
Receptor tyrosine kinase c-Met c-Met, alternatively known as MET or hepatocyte growth factor receptor (HGFR), is a membrane-bound tyrosine kinase known to have critical roles in organogenesis and cancer development (185). c-Met is involved in the development of TNBC (186). c-Met inhibitors have been studied in various solid tumors and these inhibitors have resulted in encouraging results in lung and ovarian cancer (187,188). c-Met overexpression has been observed in more than 50% of TNBC patients correlating with unfavorable overall survival (OS) of these patients (185,186). In this regard, Kim et al. reported that c-Met expression is high in TNBC cell lines, and siRNA-mediated silencing of c-Met decreases the proliferation and migration capacity of certain TNBC cell lines (186). Such data can support the critical role and applicability of c-Met as a target for cancer immunotherapy.
Tchou et al. developed mRNA electroporation-generated second generation c-Met-redirected CAR-Ts and demonstrated the antitumor activity of these cells in in vitro killing assays against BT20 (a TNBC-derived breast cancer cell line) and TB129, both with similar c-Met expression levels (189). These CAR-Ts also managed to suppress tumor growth in xenograft preclinical mouse models established using the ovarian cancer cell line SK-OV-3 (189). These researchers also conducted a Phase I clinical trial (NCT01837602) to study the safety and feasibility of intratumoral delivery of these CAR-Ts for treating metastatic breast cancer (189). The results of this trial demonstrated that CAR-T administration was well-tolerated as no CAR-T-related adverse events (>grade 1) were observed (189). Moreover, immunohistochemical analysis of tumors treated with intratumoral CAR-T delivery demonstrated significant administration site tumor necrosis, cellular debris, and the presence of macrophages around the necrotic areas (189). It is worth mentioning that Tchou et al. used mRNAbased CAR-Ts based on the concerns regarding the tumoricidal effects of c-Met-redirected CAR-Ts for targeting non-malignant c-Met-expressing cells (189). Moreover, Shah et al. reported the results from a Phase I clinical trial (NCT03060356) evaluating the safety and feasibility of intravenously delivered mRNA electroporation-generated second generation c-Met-redirected CAR-Ts in patients with metastatic or unresectable melanoma or TNBC with more than 30% c-Met expression (190). According to this report, 5 patients experienced grade 1 or 2 CAR-T infusion-related toxicity (no grade 3 or CRS was observed) but one patient terminated the therapy course due to these toxicities (190). Among the patients with TNBC who received the c-Met-redirected CAR-Ts (4 patients), 2 (50%) achieved stable disease and 2 (50%) experienced partial disease (190). Ultimately, these researchers suggested using lentiviralgenerated CAR-Ts alongside lymphodepleting chemotherapy in future clinical studies (190).
Folate receptor alpha (FRa)
FRa, alternatively termed FOLR1 or folate binding protein (FBP), is a glycosylphosphatidylinositol (GPI)-anchored transmembrane protein with a high affinity for binding the active folate form and managing its transportation (191,192). The overexpression of FRa has been detected in various solid tumors such as ovarian and breast cancers remarkably correlating with disease grade and stage (193,194). TNBC is also among solid malignancies in which FRa upregulation has been observed (195). It has been demonstrated that overexpression of FRa provides malignant cells with the privilege of proliferation (195). Song et al. generated FRaredirected CAR-Ts, and reported target antigen engagementdependent proinflammatory cytokine secretion by these cells after co-culturing with FRa-expressing human ovarian cell lines in culture (SKOV3, A1847, and OVCAR3) (196). Moreover, these researchers reported that these CAR-Ts induced tumor growth suppression in cell line-established FRa-positive human ovarian cancer preclinical models (196). Years later, Song et al. evaluated FRa-redirected CAR-Ts in breast cancer cell lines and preclinical mouse models (197). These researchers reported that their FRa-redirected CAR-Ts secreted significant levels of IFN-g upon co-cultivation with TNBC cell lines expressing FRa (197). These researchers also added that their engineered effector cells also mediated remarkable tumor outgrowth suppression in cell line-established preclinical xenograft mouse models of TNBC (197). However, Song et al. concluded that the tumoricidal activity of their FRa-redirected CAR-Ts was not as robust as it was in the case of ovarian cancer xenograft models, where the expression levels of the target antigen were higher (197). Moreover, they demonstrated the same CAR-Ts induced improved tumor suppression in preclinical models established using MDA-MB-231 cells engineered to overexpress FRa (197).
Lanitis et al. used a strategy for generating FRa-redirected CAR-Ts with a reduced level of capability for mediating ontarget off-tumor toxicity (198). In detail, they generated a transsignaling CAR in which the activation domain is physically separated from the co-stimulatory domain in two separate CARs, one set of which targets mesothelin and the other set targets FRa (198). These CAR-Ts exhibited ineffective cytokine secretion upon encountering target cells expressing only one of the target antigens but exhibited significantly improved cytokine secretion against tumor cells expressing both target antigens in vitro (198). Furthermore, these CAR-Ts demonstrated effective tumoricidal activity and persistence in vivo (198). Conclusively, these researchers suggested this method as a potent strategy for reducing the possibility of on-target off-tumor activity of CAR-Ts towards non-malignant tissues (198). In addition to this strategy, other researchers have also demonstrated that folate-FITC bispecific molecules can manage to mediate the redirection of FITC-redirected CAR-Ts activity against folate receptor (FR)positive malignant cells (199).
Epidermal growth factor receptor (EGFR)
EGFR is a membrane-spanning glycoprotein that is a member of the ERBB receptor tyrosine kinase family (200). It has been known to be involved in malignant cell proliferation and metastasis (200). TNBC is among solid tumors in which EGFR overexpression has been detected (201). EGFR targeting using CAR-Ts has been investigated in TNBC as well as other solid tumors (202)(203)(204). In detail, Li et al. generated EGFRredirected CAR-Ts using the non-viral piggyBac transposon system, and reported that these cells demonstrated tumoricidal activity against EGFR-positive cells in vitro and suppressed tumor growth in human lung cancer xenografts (203). Moreover, a Phase I clinical trial (NCT03182816) has investigated PiggyBac transposon-generated EGFR-redirected CAR-Ts in patients with advanced R/R non-small cell lung cancer (202).
In the case of TNBC, Liu et al. investigated EGFR-redirected CAR-Ts in vitro and in vivo (205). These researchers reported the overexpression of EGFR in TNBC Hs 578T, MDA-MB-468, and MDA-MB-231 cell lines in comparison with the non-TNBC cell line MCF-7 (205). In vitro co-cultivation assay of EGFRredirected CAR-Ts with the mentioned TNBC cell lines demonstrated the target antigen-dependent cytokine secretion and antitumor activity of these CAR-Ts (205). Moreover, these researchers reported tumor growth suppression by EGFRredirected CAR-Ts in cell line-established and PDX preclinical mouse models (205). Recently, Xia et al. reported the overexpression of EGFR in TNBC cell lines MDA-MB-231, MDA-MB-468, Hs 578T, and HCC1860 (201). These researchers generated third-generation EGFR-redirected CAR-Ts and reported that these effector cells demonstrated specific cytokine secretion, antitumor activity, and upregulation of T-cell activation markers (including CD69 and CD25) upon cocultivation with EGFR-positive TNBC cells lines (201). In vivo assessments using severe combined immunodeficient (SCID) mice subcutaneously implanted with the MDA-MB-231 cell line demonstrated the capability of these CAR-Ts to inhibit TNBC tumorigenesis in preclinical mouse models with "minimal" levels of off-tumor cytotoxicity (201). Overall, both of the discussed studies proposed that EGFR may be a potent CAR-T target antigen for TNBC treatment; however, more preclinical and clinical investigations are critically required (201).
Tumor-restricted variants of EGFR such as EGFR variant III (EGFRvIII) can be utilized in case of safety concerns regarding the expression of EGFR on normal tissues (206). However, so far, such variants have only been investigated in a limited number of solid tumors such as glioma (206). Moreover, modulating the immunosuppressive nature of the TME using immune checkpoint blockade approaches has been shown to improve CAR-T functioning in solid tumors (207). Several clinical trials (NCT03182816, NCT02873390, NCT02862028, and NCT03170141), some of which are completed and some are still ongoing, have aimed to investigate the effects of EGFRredirected CAR-Ts with the ability to secrete anti-CTLA-4, anti-PD-1, or anti-PD-L1 antibodies in EGFR-positive advanced solid tumors (207). However, no reports regarding the results of such trials in patients with TNBC have yet been published (207).
Mesothelin
Mesothelin is a tumor differentiation glycoprotein involved in cell adhesion (208, 209). It has a normally restricted expression on the mesothelial surfaces of the body, but it is remarkably overexpressed in a wide range of solid cancers including TNBC (208, 209). There is scientific evidence that mesothelin is engaged in oncogenesis through various cellular signaling pathways including NF-kB, PI3K, and MAPK (208-210). TNBC is one of the solid tumors in which mesothelin overexpression has been detected. In detail, one study has demonstrated that the majority of TNBC cases (67%) exhibit mesothelin overexpression as confirmed using immunohistochemical analysis (211). The limited normal expression of mesothelin and its high-level expression in a large proportion of TNBC cases have rendered it an appealing target antigen for various types of cancer immunotherapy. In 2019, Del Bano et al. investigated bispecific antibodies with mesothelin targeting and CD16 engagement domains, and demonstrated that this construct mediated the recruitment and penetration of natural killer (NK) cells into tumor spheroids and provoked robust dose-dependent cell-mediated cytotoxicity of mesothelin-expressing TNBC cell lines (212). CAR-T-mediated mesothelin targeting has also been studied in the context of TNBC. In detail, Hu et al. assessed the expression of mesothelin on three TNBC cell lines including MDA-MB-231, BT-549, and Hs 578T (107). They reported that only BT-549 cells (but not MDA-MB-231 and Hs 578T cells) expressed mesothelin as verified by both Western blot and flow cytometry (107). These researchers generated second-generation mesothelin-redirected CAR-Ts and evaluated their performance in vitro and in vivo (107). Of note, Hu et al. disrupted the PD-1 gene locus in T cells before CAR transgene introduction. In detail, these CAR-Ts demonstrated remarkably increased cytokine production and antitumor activity against PD-L1-expressing cancer cells in culture (107). Accumulating evidence suggests that PD-L1 is remarkably overexpressed in TNBC cells (213). Therefore, the strategy proposed by Hu et al. might overcome the suppressive effects of PD-1-PD-L1 interaction on CAR-Ts (107). Moreover, Hu et al. added that PD-1-deficient mesothelin-redirected CAR-Ts exhibited improved tumor outgrowth suppression and disease recurrence prevention in BT-549-established preclinical TNBC mouse models in comparison with CAR-Ts with or without PD-1-specific antibody blockade (107). Overall, even though more preclinical and clinical data are required for safe conclusions on the applicability of using mesothelin as a CAR-T therapy target antigen for TNBC, the study by Hu et al. highlights the potential of checkpoint blockade alongside mesothelin-redirected CAR-T therapy for targeting TNBC. It is worth mentioning that a Phase I clinical trial (NCT02792114) is currently investigating the safety and tolerability of mesothelin-redirected CAR-Ts in patients with metastatic/advanced mesothelin-proficient breast cancer including TNBC. Moreover, another Phase I/II clinical trial (NCT02414269) is also investigating second generation mesothelin-redirected CAR-Ts in patients with lung cancer or breast cancer. Additionally, two other clinical trials (NCT01355965 and NCT02580747) have been completed but no official reports in regards to the results of these trials in TNBC patients have been published yet.
Disialoganglioside GD2
GD2 is a surface antigen with normal expression limited to peripheral pain fibers, neurons, and melanocytes (214). GD2 expression has been documented in neuroectoderm-originated neoplasms such as neuroblastoma and melanoma (214). The prevalent tumor cell-restricted expression of GD2 has made it a suitable target antigen for various types of cancer immunotherapy. GD2 is mostly known as a target antigen for the treatment of neuroblastoma. In 2011, Louis et al. conducted a Phase I clinical trial (NCT00085930) by generating GD2redirected CAR-Ts using Epstein-Barr virus (EBV)-specific cytotoxic T lymphocytes or blood T cells for investigating the effectiveness and prolonged persistence of these cells (215). Back in 2008, these researchers also reported the results of a Phase I clinical trial in 11 pediatric patients with neuroblastoma (216). In 2015, the US FDA approved the anti-GD2 mAb dinutuximab (also known as ch14.18) for the treatment of pediatric patients with high-risk neuroblastoma (217). This mAb has recently been investigated in TNBC (218, 219). Ly et al. investigated dinutuximab application for targeting GD2-expressing breast cancer stem cells (BCSCs) and suppression of cancer progression (219). BCSCs are a type of cells in the early tumor that are chemotherapy resistant, capable of metastases, and remarkably tumorigenic (218). Therefore, BCSC targeting is known as a critical approach for precluding cancer metastases and rendering tumors susceptible to chemotherapy (219, 220). Ly et al. first demonstrated the upregulation of GD2 in TNBC cell lines, PDX preclinical models, and primary TNBC samples (219). In detail, Hs 578T and HCC1395 were two TNBC cell lines in which more than 90% of cells were GD2-proficient (219). Moreover, GD2 expression was also documented in about 60% of primary TNBC tumors (even though with fluctuating levels) correlating with poorer OS (219). Dinutuximab treatment meaningfully reduced adhesion and migration of MDA-MB-231 and SUM159 TNBC cells and suppressed GD2-upregulated mTOR signaling in BCSCs (219). This mAb also exhibited tumor outgrowth suppression capability in MDA-MB-231-established TNBC xenograft models suggesting the applicability of dinutuximab for TNBC control in the preclinical stage (219).
There are not many studies investigating CAR-Ts targeting GD2 in TNBC. In 2020, Seitz et al. used the scFv derived from ch14.18 to generate GD2-redirected CAR-Ts (221). In detail, TNBC cell lines including MDA-MB-468, MDA-MB-231, Hs 578T, and BT-549 were screened for GD2 expression and it was demonstrated that MDA-MB-231 expressed GD2, even though at very low levels, and Hs 578T and BT-549 exhibited uniform GD2 expression (221). In vitro cytotoxicity assay demonstrated these CAR-Ts did not mediate any specific tumoricidal activity towards MDA-MB-231 (221). However, these CAR-Ts induced specific cytotoxicity and cytokine secretion upon co-cultivation with the Hs 578T and BT-549 cell lines (221). Moreover, GD2redirected CAR-Ts mediated tumor growth suppression in the preclinical orthotopic model of TNBC (established using MDA-MB-231) and inhibited lung metastasis (221). Conclusively, these researchers suggested CAR-T-mediated GD2 targeting as an approach to eradicate disseminated malignant cells and inhibit metastasis (221). However, more preclinical and clinical findings may be required.
Mucin 1 (MUC1)
MUC1 is a heavily glycosylated membrane-spanning mucin protein expressed on glandular epithelial cells (222)(223)(224). The extracellular domain of MUC1 has a variable number tandem repeats (VNTR) region which is rich in serine and threonine residues acting as a platform for the attachment of O-glycans (222)(223)(224). O-glycosylation of MUC1 leads to the generation of a tumor-associated aberrantly glycosylated form of MUC1 known as tMUC1 (51, [222][223][224]. tMUC1 is overexpressed in all subtypes of breast cancer, including in 95% of TNBCs (225). Moreover, this antigen has no detectable expression on the cells of normal breast tissues (225). Such characteristics render tMUC1 a great CAR-T therapy target antigen in various types of malignancies including TNBC. To this day, various studies have demonstrated that CAR-T-mediated targeting of the aberrantly glycosylated tumor forms of surface antigens is a safe and feasible approach (51). In this regard, Zhou et al. generated second-generation CAR-Ts using a mAb named TAB004, which is capable of tMUC1 recognition (225). These tMUC1-redirected CAR-Ts demonstrated target antigendependent cytotoxicity against nine different TNBC cell lines and released cytokines, chemokines, and granzyme B (225). Moreover, these CAR-Ts inhibited HCC70 tumor outgrowth and maintained their prolonged antitumor activity for HCC70 tumor burden reduction in vivo (225). In a nutshell, this study indicated that tMUC1-redirected CAR-Ts harbor significant therapeutic value against tMUC1-positive TNBC with minimal on-target off-tumor toxicity toward normal breast epithelial cells (225).
In addition to t-MUC1, TnMUC1 is another aberrant glycoform of MUC1 that is also highly expressed by TNBC cells (226). Posey et al. generated TnMUC1-redirected CAR-Ts using a Tn-MUC1-specific antibody named 5E5, and reported target antigen-specific antitumor activity of these CAR-Ts against a panel of human primary cells and several cell lines (226). Moreover, these researchers reported that these CAR-Ts managed to suppress tumor growth in T-cell leukemia and pancreatic cancer xenograft in vivo models (226). Recently, Zhai et al. reported that their TnMUC1-redirected CAR-Ts generated using Vg9Vd2 T cells exhibited "similar or stronger" antitumor effects, than CAR-expressing ab T cells, against breast cancer cell lines with various levels of Tn-MUC1 expression (including T47D, MDA-MB-231, and MDA-MB-468) (227). Similar results were observed upon the co-cultivation of these CAR-Ts with gastric cancer cells (227). However, these researchers indicated that Vg9Vd2 TnMUC1-redirected CAR-Ts demonstrated persistence deficiencies for addressing which IL-2 is critical (227). Moreover, Zhai also reported that Vg9Vd2 TnMUC1-redirected CAR-Ts had significant tumoricidal activity in gastric cancer preclinical mouse models (227).
It is worth mentioning that clinical trials (NCT04020575 and NCT04025216) are currently evaluating second generation CAR-Ts redirected against cancer glycoforms of MUC1. In an early report from one of these trials (NCT04025216), the researchers indicated that since this study is still in the doseescalation step and it has completed only 2 of 6 planned dose levels, safety issues or on-target off-tumor toxicity of this platform are not yet up for debate (228). Moreover, another clinical trial (NCT02587689) investigating MUC1-redirected CAR-Ts in patients with advanced refractory solid tumors, including TNBC, has been finished. However, no official report of the results of this trial has been published yet.
Trophoblast cell-surface antigen 2 (TROP2)
TROP2 is a transmembrane protein expressed on the surface of human trophoblast cells (229). The expression of this antigen has been often detected in various epithelium tissue malignancies mediating their tumor-associated behavior and correlating with poor prognosis (229, 230). Such characteristics have rendered TROP2 an attractive target antigen for various types of cancer immunotherapy, especially in TNBC (229). In this regard, a recent study by Liu et al. has reported that TRBAs targeting TROP2 and CD3 suppress tumor growth in both TNBC cell lines and primary tumor cells (231). In detail, the TROP2-CD3 TRBAs developed by these researchers were capable of recruiting T lymphocytes to TROP2-positive tumor cells in vitro and into tumor tissues in xenograft TNBC preclinical models (231). The data presented by this study supports the potential of TROP2 for the immunotherapy of TNBC in patients with advanced/metastatic neoplasms (231). Moreover, in 2021, the US FDA approved sacituzumab govitecan, a TROP2-redirected mAb conjugated to a topoisomerase I inhibitor drug, for the treatment of patients with R/R unresectable locally advanced or metastatic TNBC who have received two or more prior systemic therapies (232). FDA granted medical use permit to this ADC based on a clinical trial report (NCT01631552) by Bardia et al. which demonstrated that sacituzumab govitecan mediated a response rate of 33% (3 complete and 33 partial responses) and a median duration of response of 7.7 months in patients with TNBC who had received a range of 2 to 10 previous lines of therapies (232). In the case of CAR-T-mediated TROP2 targeting, Zhao et al. developed bispecific TROP2-and PD-L1-redirected CAR-Ts and evaluated their antitumor activity in vitro (using TROP2-positive and PD-L1-positive gastric cancer cell lines) and in vivo (233). According to the in vitro results obtained by these researchers, the tumoricidal activity of bi-specific CAR-Ts was higher than that of mono-specific CAR-Ts (233). These bi-specific CAR-Ts also managed to secrete pro-inflammatory cytokines upon encountering TROP2-positive and PD-L1-positive gastric cancer cells (233). Additionally, in vivo results demonstrated that these CAR-Ts remarkably suppressed tumor growth via intratumoral injection in preclinical mouse models established using human gastric tumors (233). However, these results only support the suitability of TROP2 as a CAR-T target but not specifically in TNBC CAR-T therapy. In the case of TROP2redirected CAR-Ts for targeting TNBC, Bedoya et al. generated TROP2-redirected CAR-Ts and reported that these cells demonstrated antitumor activity against TROP2-negative cells in the presence of TROP2-positive cells (234). As claimed by these researchers, this phenomenon has also been observed in the case of CD19-redirected CAR-Ts but it was resolved by the blockade of death receptor ligands; however, Bedoya et al. indicated that this approach only resulted in the partial blockade of TROP2-redirected CAR-T cytotoxicity in this case (234). Moreover, these researchers transferred TROP2-positive tumor exosomes to TROP2-negative tumor cells to elevate the percentage of tumor cells targetable by TROP2-redirected CAR-Ts and to tackle the limitation of antigen heterogeneity in solid tumors (234). In a nutshell, even though CAR-T-mediated TROP2 targeting in TNBC has not been comprehensively investigated, the data discussed here might highlight the potential applicability of this target antigen for the treatment of TNBC.
The potential toxicities of CAR-T therapy in TNBC Identifying suitable target antigens is one of the most important steps of CAR-T cell therapy in solid tumors. "Ontarget off-tumor" effect is a CAR-T cell-mediated adverse event that takes place when a CAR-T-redirected target antigen is present (even at very low levels) on normal tissues, especially vital organs. The best strategy to prevent this adverse event is to select target antigens that are completely absent from normal tissues or have undetectable expression levels. However, this strategy is quite unpractical since every target antigen, which is overexpressed on malignant tissue cells, can also be detected in normal tissues. Therefore, researchers have focused on strategies that can be beneficial for preventing or reducing the occurrence of such events. For instance, "safety switches" which include suicide genes have been utilized in the construct of CAR-Ts for targeting TNBC (173). In detail, Wallstabe et al. have studied the applicability of a v b 3 -redirected CAR-Ts equipped with EGFRt which allows the eradication of the infused CAR-Ts using the anti-EGFR mAb cetuximab (173). Furthermore, other novel CAR designs such as dual CARs and synNotch CARs have also been investigated in solid tumor CAR-T therapy for preventing off-tumor toxicities (182). Researchers have claimed that ROR1-expressing stromal cells are simultaneously targeted by ROR1-redirected CAR-Ts in the context of targeting ROR1expressing malignant cells (182). This "on-target off-tumor" toxicity can lead to lethal bone marrow failure; therefore, an efficient strategy is required for preventing this adverse event. Srivastava et al. generated ROR1-redirected CAR-Ts with synNotch receptors specific for EpCAM or B7-H3 (182). Conclusively, these researchers added that such CAR-Ts mediated efficient antitumor responses without causing toxicity (182). In addition to these strategies, trans-signaling CARs are also considered innovative in terms of preventing ontarget off-tumor toxicities (198). In the context of TNBC, Lanitis et al. generated trans-signaling FRa-redirected CAR-Ts in which the CAR activation domain was physically detached from the CAR co-stimulatory domain and they were incorporated into two separate CARs one targeting mesothelin and the other targeting FRa (198). These CAR-Ts were evaluated in preclinical assessments, and it was proposed that this strategy can decrease the risk of on-target off-tumor toxicity against normal tissues (198).
In the context of CAR-T therapy of hematologic malignancies, such as those targeting CD19, due to the expression of CD19 by both normal cells and malignant blasts, CAR-T therapy results in the elimination of both of these cells leading to a phenomenon known as "B-cell aplasia" (235). This occurrence is a marker of an efficient CAR-T therapy of hematologic malignancies, even though it makes the relative patients susceptible to opportunistic infections to tackle which the patients should undergo immunoglobulin replacement (235).
As mentioned earlier, the application and efficiency of CAR-Ts whose scFv targeting domains have been subjected to affinitytuning have also been studied (52). Such CAR-Ts can be favorable in distinguishing between tumor cells overexpressing the target antigen and normal cells expressing the antigen at physiologic levels (52). In this regard, Park et al. demonstrated that affinity-tuned ICAM-1-redirected CAR-Ts are capable of targeting malignant cells with ICAM-1 overexpression while sparing normal cells expressing ICAM-1 at basal physiologic levels (52). Moreover, mRNA-based CAR-Ts have also been proposed as a platform for preventing on-target off-tumor toxicities. In this regard, Tchou et al. developed mRNA-based c-Met-redirected CAR-Ts to avoid CAR-T-mediated targeting of c-Met-expressing normal cells (189). They reported that these cells demonstrated efficient antitumor activity against TNBC both in vitro and in vivo (189). These researchers also evaluated the safety and efficacy of these CAR-Ts in a Phase I clinical trial (NCT01837602) (189).
CRS is well-known CAR-T cell therapy adverse event which is more likely to occur in CAR-T therapy of hematologic malignancies (235, 236). CRS results from the rapid activation of various immunological pathways. Its critical damages can include cardiac-related toxicities and hyponatremia. CRS has also been detected in patients with solid tumors undergoing CAR-T therapy (236). As discussed throughout the article, there are not many clinical trials investigating CAR-T therapy for the treatment of patients with TNBC. Some of these trials have been completed and some are still ongoing. Among the trials that have been completed, only a few of them have reported their findings. In detail, in a Phase I clinical trial (NCT02706392) studying ROR1-redirected CAR-Ts in patients with various solid tumors including metastatic TNBC, it was reported that out of 4 TNBC patients, half of them experienced grade 1 CRS (184). According to a report by Tchou et al. from a Phase I clinical trial (NCT01837602) assessing the safety and feasibility of intratumoral delivery of c-Met-redirected CAR-Ts in patients with metastatic breast cancer, it was reported that the CAR-Ts were well-tolerated and no CAR-T-associated toxicities (> grade 1) were documented (189). Moreover, in another a Phase I clinical trial (NCT03060356) investigating the same target antigen, it was reported that 5 patients experienced grade 1 or 2 CAR-T delivery-associated adverse events (no grade 3 or CRS were reported) (190).
CRS is a medical condition that demands rapid clinical intervention to prevent the condition from worsening.
Generally, antihistamines or corticosteroids are recommended for low-grade CRS management (236). However, in the case of CAR-T therapy-mediated CRS, more efficient strategies are required, especially in the case of hematologic malignancies such as B-cell acute lymphoblastic leukemia (B-ALL) (235). IL-1 and IL-6 blockade, GM-CSF blockade, antibody-based immunotherapy pretreatment, therapeutic plasma exchange, hemofiltration, and fractionated CAR-T infusion are among strategies proven to be efficient in the management of severe lethal CRS after CAR-T therapy (235, [237][238][239][240][241][242][243][244][245]. Such strategies can also be applied in the case of solid tumor CAR-T therapy including TNBC CAR-T therapy.
Conclusion
CAR-T products have been available on the market as a treatment option for R/R hematologic malignancies in the past recent years. However, CAR-T-mediated targeting of solid tumors or some hematologic malignancies, such as T-cell neoplasms, face additional unexpected challenges that remarkably restrain the antitumor potential of these therapeutics. TNBC is known as a heterogeneous breast cancer subtype that is mainly resistant to conventional therapies. The immunogenic nature of this neoplasm has proven that immunotherapy-based therapeutics can lead to favorable clinical benefits. For instance, the promising clinical outcomes of the checkpoint inhibitor atezolizumab in combination with nab-paclitaxel led to its approval by the US FDA for the treatment of locally or metastatic advanced unresectable TNBC (18). However, CAR-T therapy of TNBC is in an emerging field as it is focused on discovering the suitable and targetable TAAs, mostly in preclinical and early clinical stages. Of note, to overcome the hurdles of CAR-T therapy in TNBC, various challenges of solid tumor CAR-T therapy should be tackled beforehand. Various critical stratagems should be employed to ensure the safety and effectiveness of CAR-T therapy in solid tumors. Combinatorial CAR-T therapy with other types of therapies can result in improved clinical outcomes of CAR-T therapy in solid tumors (70). For instance, ECM-or CAF-targeting agents can be applied for maximizing CAR-T antitumor effects (70). Also, macrophage-or monocyteeliminating agents are beneficial for amplifying the tumoricidal impact of CAR-Ts, or anti-angiogenic agents can be leveraged for enhancing the tumor site trafficking of CAR-Ts (70). Moreover, other types of effector cells have also been investigated for the expression of CARs. In this regard, gd-CAR-Ts and CAR-expressing NK cells (CAR-NKs) have been investigated for the treatment of both hematologic and solid malignancies, including TNBC (68,(246)(247)(248). Such alternative CAR-expressing effector cells might be beneficial for tackling various types of CAR-T therapy challenges (68,(246)(247)(248).
Throughout this article, we reviewed various novel CAR-T target antigens for the selective targeting of TNBC. A number of these target antigens have only been evaluated in in vitro and in vivo studies while some of them have found their way to inhuman clinical investigations, as summarized in Table 3. Moreover, in addition to the discussed target antigens, new targets can also be discovered. For instance, stage-specific embryonic antigen-4 (SSEA-4) is among more novel antigens detected in a proportion of TNBCs, and CAR-Ts against this target antigen have been investigated in vitro and in vivo (249). However, such antigens are not comprehensively studied as TNBC CAR-T-targetable antigens and require more indepth assessments.
In a nutshell, it is safe to mention that considerable efforts have been made in the field of TNBC CAR-T therapy as demonstrated by several ongoing preclinical and clinical studies. However, there is still a long way to go for some of the proposed strategies (for enhancing the efficacy of CAR-T therapy in solid tumors) and several TNBC target antigens discussed in this article. Successful CAR-T therapy in TNBC is substantially reliant on improving the specificity, safety, and efficacy of CAR-T therapy in solid tumors by both choosing the most appropriate target antigens and addressing the unmet restricting challenges. Until then, fingers are crossed for CAR-T therapy to be able to act as an efficient treatment approach alongside the conventional TNBC treatment modalities.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
|
v2
|
2022-11-23T15:00:17.909Z
|
2022-11-22T00:00:00.000Z
|
253764386
|
s2orc/train
|
Survival outcomes after breast-conserving surgery plus radiotherapy compared with mastectomy in breast ductal carcinoma in situ with microinvasion
Ductal carcinoma in situ with microinvasion (DCIS-MI) is a subtype of breast cancer with a good prognosis, for which both breast conserving surgery plus radiotherapy (BCS + RT) and mastectomy are feasible surgical methods, but no clear conclusion has been made on the choice of these treatments. We used the Surveillance, Epidemiology and End Results database to extract 5432 DCIS-MI patients. Participants were divided into the BCS + RT group and the mastectomy group. We compared the overall survival (OS) and breast cancer-specific survival (BCSS) of the two groups using the Kaplan–Meier method and Cox regressions before and after propensity score matching (PSM). Before PSM, both univariate and multivariate analyses showed that BCS + RT group had significantly higher OS and BCSS compared with patients in the mastectomy group (P < 0.001). After PSM, the multivariate analysis showed that compared with mastectomy, the BCS + RT showed significantly higher OS and BCSS (HR = 0.676, 95% CI = 0.540–0.847, P < 0.001; HR = 0.565,95% CI = 0.354–0.903, P = 0.017). In addition, the subgroup analysis showed that BCS + RT is at least equivalent to mastectomy with respect to OS and BCSS in any subgroup. For patients with DCIS-MI, the prognosis of BCS + RT was superior to mastectomy.
Statistical analysis. Propensity score matching (PSM) was applied to create a matched pair between the two groups to eliminate the selection bias of this study population 19 . We performed PSM for all the variables included in the study. Landmark analysis was used to eliminate a lead time bias among the propensity-matched cohort 20 . With the landmark, analysis was restricted to patients who survived to 6 months without death. X 2 test was used to compare the distribution of the clinical and pathological features between the two groups before and after PSM. The OS and BCSS survival curves were plotted through the Kaplan-Meier method and compared by the log-rank test. The cox regression model was used for the univariate and multivariate analyses of the BCSS and OS. All P values were two-sided, and P < 0.05 was considered significant. The SPSS 20.0 (IBM SPSS Statistics, Chicago, IL, US) was used for these analyses.
Ethics approval and consent to participate. All patients were collected from the SEER database, and all of them have given prior informed consent to being registered in it. The study was approved by the Ethics Committee of The First Affiliated Hospital of Chengdu Medical College and was complied with the Declaration of Helsinki.
Results
Baseline characteristics. In total, 5432 patients with DCIS-MI from 2000 to 2014 were included in the study through the SEER database. We divided the patients into two groups: BCS + RT group (2834,52.17%) and mastectomy group (2598,47.83%). Table 1 summarizes the patient clinical characteristics of the two groups. Compared with mastectomy group, the patients in the BCS group were older (78.9% vs. 64.2%; P < 0.001) and had a lower histological grade (grade I + II, 65.6% vs. 56.4%; P < 0.001), less lymph node metastasis (N0, 96.7% vs. 87.5%; P < 0.001). Further, the BCS group had a higher ER (76.7% vs. 67.4%; P < 0.001) and PR (62.9% vs. 55.4%; P < 0.001) positive rates and were less likely to receive chemotherapy (6.2% vs. 14.8%; P < 0.001). After PSM, the two groups consisted of 1902 pairs. There was no significant difference in clinicopathological characteristics between the two groups.
Prognostic factors associated with OS and BCSS. Before PSM, the median follow-up time for these patients was 101 months. The 5-year and 10-year OS for patients in BCS + RT and mastectomy groups were 97.3% vs. 95.4% and 91.2% vs. 88.5% respectively (log-rank P = 0.001, Fig. 1A). The 5-year and 10-year BCSS for patients in BCS + RT and mastectomy groups were 99.1% vs. 97.8% and 98.0% vs. 95.9% (log-rank P < 0.001, Fig. 1B). After adjusting for the prognostic variables in the univariate analysis (Supplementary Table 1), the multivariate analysis indicated that black race and patients with more lymph node metastases are associated with poor OS and BCSS (all P < 0.05). Besides, patients at a younger age and not married had better OS relatively while patients without chemotherapy had lower BCSS (all p < 0.05). The BCS + RT group showed significantly higher OS and BCSS compared with patients in the mastectomy group (HR = 0.686, 95% CI = 0.571-0.825, P < 0.001; HR = 0.596, 95% CI = 0.411-0.865, P = 0.007) ( Table 2).
Subgroup analysis of OS and BCSS.
To further explore possible factors affecting the overall survival time for patients who had undergone two types of surgery, we performed a subgroup analysis of all patients after PSM. BCS + RT group showed significantly higher OS than the mastectomy group for patients aged between 50-79 years, patients married or unmarried, the white race group, patients with grade III + IV, patients with lymph nodal negative, patients with ER positive, patients with PR-positive or negative and those who did not receive chemotherapy (all P < 0.05). There was no difference significantly observed in OS in other subgroups (Fig. 3). The BCS + RT group also showed BCSS benefits in patients who were not married, patients with lymph nodal negative, patients with ER-negative, and those who did not receive chemotherapy (Fig. 4). Further, the OS and BCSS outcomes of mastectomy were not better than BCS + RT in any subgroup.
Discussion
DCIS-MI is a special type of breast cancer, and there is little evidence on the prognosis of patients with DCIS-MI undergoing BCS + RT and mastectomy. We found that the prognosis of patients with DCIS-MI after mastectomy is not better than those of BCS + RT in any subgroup by using the SEER database.
In the NCCN guidelines, DCIS-MI is classified as early-stage invasive breast cancer. All surgical options for early-stage invasive breast cancer are unified. There is no special explanation for the surgical options for DCIS-MI. In our study, 71.9% of patients with DCIS-MI were older than 50 years. Besides, patients with DCIS-MI had few lymph node metastases (7.7%), low histological grade (61.3% in GI + II), and high positive rates of ER and PR (72.3% and 59.3%), which was consistent with other studies 21,22 . These results indicate that most of the DCIS-MI have a good prognosis. www.nature.com/scientificreports/ www.nature.com/scientificreports/ Although the clinicopathological features of DCIS-MI indicate a good prognosis, at present, there are still a large number of DCIS-MI patients undergoing a mastectomy. In our study, 47.83% of the patients received mastectomy. In Eastern countries, the proportion of patients with DCIS-MI undergoing mastectomy is higher, even as high as 80% 23,24 . At present, there are few studies on the prognosis of DCIS-MI after surgery. Mamtani et al. investigated the prognosis of DCIS ± MI after mastectomy. It proved that distant disease-free survival after mastectomy for DCIS ± microinvasion is excellent among all age groups, and overall rates of locoregional recurrence after mastectomy for DCIS with or without microinvasion are low. Even in the age group with the highest recurrence rate, 10-year locoregional recurrence remains low at 4.2% 25 . Park et al. conducted a study on 3648 patients with DCIS younger than 40 years old, and the results showed that mastectomy does not offer survival benefits over BCS + RT 26 . Mamtani et al. also confirmed this 15 . The Yale School of Medicine retrospective clinical study included 72 patients with DCIS-MI and 321 patients with DCIS, all of whom received BCS + RT. There was no difference in regional recurrence rates after 10 years between the DCIS-MI group and the DCIS group (8.3% vs. 6.8%) 18 . DCIS-MI often has multiple minimally invasive foci, associated with a higher risk of ipsilateral recurrence 27,28 . The study by Si et al. showed that 35.1% of DCIS-MI Patients have multiple foci, which had a worse disease-free survival rate compared with one-focus patients (98.29 vs. 93.01%, P = 0.032) 24 . The safety of BCS for DCIS-MI with multiple minimally invasive foci is worth exploring. Rakovitch compared the local recurrence rate after BCS in DCIS-MI patients with one-focus and multiple foci 17 . The results showed that multiple foci of MI are associated with an increased risk of invasive local recurrence in women with DCIS treated with BCS, but treatment with the whole breast and boost RT can mitigate this risk. At present, there is no study comparing the prognosis of BCS + RT and mastectomy in DCIS-MI patients with multiple foci.
There are few studies comparing the prognosis of patients with DCIS-MI after BCS + RT and mastectomy. Bartova et al. compare the prognostic difference between BCS and mastectomy in DCIS-MI. They followed up on 41 patients with DCIS or DCIS-MI after BCS and mastectomy, and finally, only 27 patients completed the followup. There is no local recurrence occurred 16 . However, the sample size of this study was small, and no survival www.nature.com/scientificreports/ www.nature.com/scientificreports/ rate was reported. In our study, we observed that 95.5% of patients received mastectomy without RT. Thus we think that BCS + RT showed a better prognosis than mastectomy may due to RT. Studies have confirmed that RT can reduce the local recurrence of breast cancer after BCS. Fisher et al. 12 showed that adjuvant radiotherapy after BCS could reduce the risk of recurrence by approximately 50%. The EBCTCG study also demonstrated this 29 . Rakovitch et al. proved that postoperative radiotherapy could reduce the local recurrence rate in patients with DCIS-MI 17 . Li et al. compared the difference in survival between DCIS-MI patients treated with BCS + RT (n = 74) and mastectomy without RT (n = 221). No survival difference was observed between the two groups 30 .
In their study, none of the patients in the mastectomy group received radiotherapy, and the sample size of this study was small. We believe that further studies are needed to investigate the prognosis of DCIS-MI after different surgical methods. Similar to the result of another study, chemotherapy cannot improve the survival of DCIS-MI in our study. Pu et al. proved that postoperative chemotherapy did not improve DFS in patients with DCIS-MI after mastectomy (HR = 1.50, 95% CI 0.29-7.87, P = 0.63) 31 . Chen et al. analyzed 3198 DCIS-MI patients and concluded that chemotherapy was an independent factor for worse BCSS (P = 0.008), and there was no statistical significance for OS (P = 0.248) in patients with DCIS-MI 32 . However, further studies are needed to verify whether chemotherapy is beneficial to patients with DCIS-MI.
Our study had several limitations. Firstly, the SEER database did not provide detailed information on breast multiple lesions and lacks data on the size of the DCIS in DCIS-MI and postoperative local recurrence. Secondly, there is no information on endocrine therapy and targeted therapy in the SEER database. Despite these limitations, the sample size of our study was large and the follow-up time was long. In the research method, we also www.nature.com/scientificreports/ used the propensity-matched landmark analysis to minimize the confounding factors. All these guarantee the reliability of our research results. We not only analyzed the OS and BCSS of the two groups but also performed subgroup analysis. We found that in any subgroup, the OS and BCSS results of mastectomy were not better than BCS + RT. There is little evidence on the prognosis of patients with DCIS-MI undergoing BCS + RT and mastectomy at present. The sample sizes of the studies were all small, and one of the studies did not report the survival rate. Therefore, our research is still very valuable and can provide a theoretical basis for the selection of surgical methods for DCIS-MI.
Conclusion
This population-based study revealed that the prognosis of patients who were diagnosed with DCIS-MI receiving mastectomy was not better than those receiving BCS + RT. We think that BCS + RT should be considered preferentially in DCIS-MI. However. BCS + RT is appropriate in patients with a limited extent of disease. The surgical method should be selected carefully when the tumor has multiple foci or with a large mass.
|
v2
|
2022-11-23T15:00:33.123Z
|
2022-11-22T00:00:00.000Z
|
253764130
|
s2orc/train
|
Long noncoding RNA DIO3OS induces glycolytic-dominant metabolic reprogramming to promote aromatase inhibitor resistance in breast cancer
Aromatase inhibition is an efficient endocrine therapy to block ectopic estrogen production for postmenopausal estrogen receptor (ER)-positive breast cancer patients, but many develop resistance. Here, we show that aromatase inhibitor (AI)-resistant breast tumors display features of enhanced aerobic glycolysis with upregulation of long noncoding RNA (lncRNA) DIO3OS, which correlates with poor prognosis of breast cancer patients on AI therapies. Long-term estrogen deprivation induces DIO3OS expression in ER-positive breast tumor cells, which further enhances aerobic glycolysis and promotes estrogen-independent cell proliferation in vitro and in vivo. Mechanistically, DIO3OS interacts with polypyrimidine tract binding protein 1 (PTBP1) and stabilizes the mRNA of lactate dehydrogenase A (LDHA) by protecting the integrity of its 3’UTR, and subsequently upregulates LDHA expression and activates glycolytic metabolism in AI-resistant breast cancer cells. Our findings highlight the role of lncRNA in regulating the key enzyme of glycolytic metabolism in response to endocrine therapies and the potential of targeting DIO3OS to reverse AI resistance in ER-positive breast cancer.
therapy to inhibit estrogen signaling-dependent tumor growth 10,11 . Aromatase inhibitors (AIs) can inactivate the aromatase, a rate-limiting enzyme that converts androgen to estrogen, therefore reducing the estrogen level and suppressing ER-positive breast cancer cell proliferation. In postmenopausal patients, the third-generation AI drugs (letrozole, anastrozole, and exemestane) have become first-line adjuvant therapy for ER-positive breast tumors with therapeutic benefits superior to tamoxifen, once a classical endocrine drug for breast cancer [12][13][14] . However, acquired resistance to aromatase inhibition accounts for the majority of failure in adjuvant endocrine therapy, making it vital and urgent to decipher the molecular mechanism of AI resistance in ER-positive breast cancer. In current studies, most of the findings focus on the dysregulation of ER and other growth factor receptor signaling, while metabolic reprogramming, one of the key hallmarks of malignant tumors, is rarely discussed in the acquisition of AI resistance [15][16][17] .
Long noncoding RNAs (LncRNAs) are transcripts of over 200 nucleotides in length, lacking protein-coding capacity. To date, numerous studies have unraveled lncRNAs as pivotal molecules regulating gene expression at the transcriptional, post-transcriptional and epigenetic levels, thereby affecting multiple steps of tumor development [18][19][20] . Specifically, lncRNA lincRNA-p21 21 , LINK-A 22 , and GLCC1 23 have been reported to modulate aerobic glycolysis in cancer cells. Previously, we have shown that an exosomal lncRNA HISLA can enhance aerobic glycolysis and induce chemoresistance in breast cancer cells 6 . Despite that, the role of lncRNAs in glycolytic alteration of AI-resistant ER-positive breast cancer and the underlying mechanisms need further elucidation.
In this work, we compare lncRNA expression profiles in the clinical samples of AI-treated ER-positive breast cancer patients and identify DIO3OS as an important RNA regulator in reprograming glucose metabolism in AI-resistant breast tumors.
AI-resistant breast cancer exhibits enhanced aerobic glycolysis
To explore the molecular features of AI-resistant tumors, we performed high-throughput RNA sequencing to examine the transcription profiles of ER-positive breast cancer tissues from patients with adjuvant AI treatment (Fig. 1a). Gene set enrichment analysis (GSEA) showed that the glycolytic pathway was positively enriched in AIresistant tissues, compared with the AI-sensitive group ( Fig. 1b and Supplementary Fig. 1a, b), suggesting the glucose metabolism reprogramming after AI treatment.
Long-term estrogen deprivation (LTED) has been widely used to mimic clinical AI resistance in ER-positive breast cancer cell 24 . To investigate whether metabolic changes occur in AI-resistant ER-positive breast cancer cells, we examined the uptake of glucose and production of extracellular lactate, a crucial metabolite of glycolysis, in two cell models MCF-7 LTED and T47D LTED, which were insensitive to either estrogen deprivation or AI drugs such as letrozole and anastrozole ( Supplementary Fig. 1c, d). We found that both glucose uptake and lactate production were markedly upregulated in LTED cells, which were cultured in estrogen-deprived media for long period (at least 10 months), comparing to parental control cells, which were cultured in estrogen-deprived media for short term (7 days) (Fig. 1c, d). Additionally, glycolytic flux measured by extracellular acidification rates (ECARs) demonstrated that the overall glycolytic flux and glycolytic capacity were enhanced in LTED cells in comparison with parental MCF-7 and T47D cells (Fig. 1e, f).
To further identify the differential metabolites between parental and LTED MCF-7 cells, we screened glycolytic metabolites using liquid chromatography-mass spectrometry (LC-MS). Phosphoenolpyruvate and lactate were significantly upregulated in LTED cells, as compared to parental MCF-7 cells (Fig. 1g). To investigate other potential metabolic alterations in AI-resistance cells, we examined the situations of oxidative phosphorylation and tricarboxylic acid (TCA) cycle in MCF-7 and T47D cells upon estrogen deprivation. However, neither the concentrations of acetyl-CoA nor the oxygen consumption rate (OCR) showed significant difference between the parental and LTED MCF-7/ T47D cells ( Supplementary Fig. 1e-g). Hence, LTED induces enhanced aerobic glycolysis, rather than oxidative phosphorylation or TCA cycle, in ER-positive breast cancer cells.
Considering the crucial roles of ER signaling in endocrine resistance, we interrogated whether these glycolytic changes in LTED cells were linked to ER expression and activation. Strikingly, parental MCF-7 cells expressed much more ER messenger RNA (mRNA) and protein than the LTED cells did, as revealed by real-time quantitative PCR (qRT-PCR) and immunofluorescence (IF), respectively ( Supplementary Fig. 1h, i). Western blot assay further confirmed that both total ER and phosphorylated ER were consistently downregulated in MCF-7 LTED cells in comparison with their parental counterparts ( Supplementary Fig. 1j). Meanwhile, a markedly lower basal transcriptional level of ER in MCF-7 LTED cells was determined using the luciferase reporter assay ( Supplementary Fig. 1k). To further determine whether ER loss may increase cellular glycolysis following LTED, specific siRNAs and fulvestrant, a well-known ER antagonist, were employed to suppress ER expression in MCF-7 LTED cells, and ER expression was detected by qRT-PCR ( Supplementary Fig. 1l) and Western blot ( Supplementary Fig. 1m). Subsequent functional assays demonstrated that neither siRNA-nor fulvestrant-mediated ER downregulation in these cells affected the lactate production ( Supplementary Fig. 1n) or glycolytic capacity measured by ECARs ( Supplementary Fig. 1o). Together, our data suggested that LTED in ER-positive breast cancer cells induces an ER-independent increase in aerobic glycolysis.
LncRNA DIO3OS is upregulated in AI-resistant breast cancer and clinically associated with AI resistance
To identify the potential regulator driving the glycolytic changes induced by AI resistance, we initially analyzed the mRNA expression profiles and found that the differentially expressed coding genes in the glycolytic pathway were not top listed. Therefore, we focused on the differentially expressed lncRNAs in the top listed panel (Fig. 1h), most of which had unknown biological functions. A total of 60 differentially expressed lncRNA genes were identified between AI-sensitive or -resistant tumor tissues, among which 33 lncRNAs were significantly upregulated, whereas 27 lncRNAs were downregulated in the AIresistant group (Supplementary Table 1). To narrow down these potential regulators, we performed qRT-PCR to examine the expression level of the top ten upregulated lncRNAs in MCF-7 LTED and T47D LTED cells. DIO3OS was the only concordantly upregulated lncRNA in both LTED cells, with an approximately three-fold increase in comparison to their parental cells (Fig. 1i). Moreover, among the four differentially expressed lncRNAs with statistical significance in cell models, only DIO3OS was markedly increased in the AI-resistant breast cancer tissues compared with those without clinical relapse ( Fig. 1j and Supplementary Fig. 2a). Besides, Northern blot confirmed the higher expression of DIO3OS in LTED cells, comparing to their parental counterparts ( Supplementary Fig. 2b).
LncRNA DIO3OS is located at human chromosome 14 and transcribed in opposite directions from the DIO3 gene locus. Using 5'-and 3'-rapid amplification of cDNA ends (RACE) and Sanger sequencing, we identified DIO3OS as a 3128 nt intronless transcript (Ensembl/GEN-CODE transcript ID: ENST00000554735) ( Supplementary Fig. 2c, d). To elucidate whether this DIO3OS transcript possesses protein-coding potential, we performed epitope tagging of green fluorescent protein (GFP)-fused putative protein assay and confirmed that DIO3OS-GFP fusion did not encode any protein product ( Supplementary Fig. 2e). To detect the subcellular localization of DIO3OS, we extracted the nuclear and cytoplasmic RNA fractions in parental and LTED MCF-7 cells. Subsequent qRT-PCR demonstrated that most DIO3OS was located in the nucleus, while a small portion of DIO3OS was distributed in the cytoplasm ( Supplementary Fig. 2f). This finding was also validated by fluorescence in situ hybridization (FISH) of DIO3OS ( Supplementary Fig. 2g), implying that DIO3OS might exert its biological function in the nucleus.
To further evaluate the clinical and pathological significance of DIO3OS, we collected 257 paraffin-embedded tumor tissue samples from ER-positive breast cancer patients who received postoperative AI treatment. In situ hybridization (ISH) assay showed that DIO3OS expression was significantly higher in tumor tissues from patients who Article https://doi.org/10.1038/s41467-022-34702-x developed AI resistance than in those without clinical relapse (P < 0.0001, Fig. 1k, l). More importantly, high DIO3OS level was associated with poor relapse-free survival (P < 0.0001) and overall survival (P = 0.0008) of these patients (Fig. 1m). Moreover, DIO3OS expression positively correlated with both tumor size (P = 0.0142) and Ki67 index (P = 0.0010), but not with lymph node or distant metastasis ( Fig. 1n and Supplementary Table 2). Collectively, these data indicated that DIO3OS is highly expressed in AI-resistant tumors and clinically relevant to breast cancer resistance to AI treatment.
DIO3OS regulates the proliferation and aerobic glycolysis of LTED cells
To investigate the functional role of DIO3OS in AI resistance, we performed loss-/gain-of-function assays. Due to the nuclear localization of DIO3OS, we employed locked nucleic acids (LNAs), instead of siRNAs, to silenced its expression 25 .With the effectiveness confirmed by qRT-PCR ( Supplementary Fig. 3a) and Northern blot (Fig. 2a), two LNAs were applied for further knockdown experiments. As revealed by cell counting and MTT assays, silencing DIO3OS significantly impaired the growth of MCF-7 LTED and T47D LTED cells ( Fig. 2b and Supplementary Fig. 3b). More specifically, DIO3OS knockdown in LTED cells reduced the proportion of proliferating cells, as evidenced by EdU incorporation assay ( Fig. 2c and Supplementary Fig. 3c) as well as Ki67 staining ( Supplementary Fig. 3d). In parallel, the clonogenicity of LTED cells were decreased upon DIO3OS silencing in the colony formation assay (Fig. 2d). It was worth noting that LNA-mediated DIO3OS knockdown ( Supplementary Fig. 3e) hardly affected the growth and proliferation of parental MCF-7 cells, as detected by cell counting, MTT, EdU incorporation and clone formation assays (Supplementary Fig. 3f-i). As DIO3OS expression was extremely low in parental T47D cells, we speculated that downregulating DIO3OS would not affect the biological behavior of T47D cells.
Given that aerobic glycolysis was enhanced in AI-resistant breast cancer, we next evaluated the effect of silencing DIO3OS on the glycolytic metabolism in these cells. Indeed, knocking down DIO3OS dramatically reduced the ECAR level as well as glycolytic capacity in MCF-7 LTED and T47D LTED cells, as compared to their control cells (Fig. 2e, f). In addition, glucose consumption assay and lactate production assay accordantly validated that DIO3OS silencing in LTED cells significantly impaired their glucose uptake capacity (Supplementary Fig. 3j) and lactate levels (Fig. 2g).
Moreover, we determined the contribution of DIO3OS in tamoxifen-resistant model as well as HER2-positive and triple-negative breast cancer (TNBC). Unlike AI-resistant models, DIO3OS expression decreased in tamoxifen-resistant MCF-7 cells compared to their parental line ( Supplementary Fig. 3k). Among the TNBC lines, only BT549 cells exhibited moderate elevation of DIO3OS expression, by twofold higher than MCF-7 cells, but the lncRNA was almost not detected in MDA-MB-231, MDA-MB-468, Hs578T, and MDA-MB-436 cells ( Supplementary Fig. 3l). As for HER2-positive breast cancer cells, except that SKBR3 revealed similar DIO3OS expression as MCF-7, both BT474 and MDA-MB-361 cells expressed lower level of DIO3OS (Supplementary Fig. 3l). Since BT549 cells and SKBR3 cells exhibited moderate DIO3OS expression, we then explored the potential function of DIO3OS in these two cell lines. Both cell counting and MTT assays revealed that with DIO3OS knockdown by LNAs ( Supplementary Fig. 3m), the growth of BT549 cells was slightly impaired, while SKBR3 cells remained largely unaffected ( Supplementary Fig. 3n, o), indicating a less important role of DIO3OS in these ER-negative or HER2positive breast cancer.
DIO3OS interacts with PTBP1 in the nucleus
To dissect the molecular mechanism of DIO3OS in cancer metabolism, we applied RNA pulldown assay with an in vitro transcribed biotinlabeled DIO3OS RNA to screen DIO3OS-interacting proteins. A specific band between 55-70 kDa pulled down by DIO3OS, but not the antisense control RNA, was sent to mass spectrometric analysis (Fig. 3a). Among all the RNA pulldown proteins, polypyrimidine tract binding protein 1 (PTBP1), a nuclear alternative splicing regulator, was the most enriched one with a highest protein sequence coverage (Fig. 3b, Supplementary Tables 3-4, Supplementary Data 1 and Supplementary Note 1). Western blot further confirmed that PTBP1 bound specifically to DIO3OS (Fig. 3c). Also, RNA immunoprecipitation (RNA-IP) assay using an anti-PTBP1 antibody was performed, which showed that DIO3OS was retrieved with~15-fold enrichment in the anti-PTBP1 immunoprecipitates, compared with control IP reactions using IgG in LTED cells (Fig. 3d). To exclude indirect interactions of DIO3OS and PTBP1, we further performed crosslinking immunoprecipitation (CLIP) assay in LTED cells by using formaldehyde to stabilize the physiological interactions of RNA-protein complexes. DIO3OS was enriched more than 20-fold in the anti-PTBP1 immunoprecipitates comparing to the Fig. 1 | AI-resistant ER-positive breast cancer exhibits enhanced aerobic glycolysis with lncRNA DIO3OS upregulation. a Heatmap of expression profile sequencing for breast cancer tissues obtained from patients that were either sensitive or resistant to AI treatment. n = 7 biological replicates. b GSEA results of the differential gene expression profiles between AI-resistant and AI-sensitive breast cancer. c, d Glucose uptake (c) and lactate production (d) of MCF-7/T47D LTED cells in comparison with their parental counterparts. e, f ECAR values and calculated glycolytic capacity of MCF-7 LTED (e) and T47D LTED (f) cells compared with their parental counterparts. g LC-MS analysis of MCF-7 LTED cells in comparison with parental MCF-7 cells. h Volcano plot of lncRNA sequencing data for breast cancer tissues from patients that were either sensitive or resistant to AI treatment. i qRT-PCR of the top ten lncRNA candidates in two AI-resistant cell models. j qRT-PCR of lncRNA DIO3OS in breast tumor tissues from patients that were sensitive or resistant to AI treatment. Means ± s.d. of n = 13 biological replicates are shown. k Representative ISH images of DIO3OS in paraffin-embedded breast cancer tissues and the correlation analysis (mean ± s.e.m.) between DIO3OS level and AI treatment response. Tumor tissues were obtained from ER-positive breast cancer patients that were either sensitive (n = 133) or resistant (n = 124) to AI treatment. Scale bars, 50 μm. l Statistical analysis of AI-resistant and AI-sensitive breast tumor tissues under different DIO3OS expression status in cohort. m Kaplan-Meier survival curve for relapse-free survival and overall survival in AI-treated ER-positive breast cancer patients with high (n = 112) or low (n = 145) DIO3OS expression in tumor tissues. HR hazard ratio. n Heatmap of the association of the different clinical characters (tumor size, Ki67 index, and metastatic status) with DIO3OS high-expressing and DIO3OS low-expressing AI-treated breast tumors. Means ± s.d. of n = 3 (c, d, g, i) or n = 4 (e, f) independent experiments yielding similar results are presented, and P-values were determined by two-tailed Student's t-test (c-g, i, j), Mann-Whitney test (k), χ2 test (l, n), or log-rank test (m). Source data are provided as a Source Data file.
IgG control ( Supplementary Fig. 4a). Thus, an enhanced retrieval of DIO3OS was obtained in CLIP assay, relative to RNA-IP assay. In accordance with the nuclear localization of DIO3OS, confocal microscopy for FISH and immunofluorescence staining demonstrated the colocalization of DIO3OS and PTBP1 in the nucleus of MCF-7 LTED and T47D LTED cells (Fig. 3e). Given that PTBP1 is a hnRNP protein (hnRNP I) containing four RNA recognition motifs (RRMs) connected by three linkers, its specific binding site with DIO3OS was further identified by using full-length and truncated PTBP1 constructs tagged with FLAG ( Fig. 3f, upper). RNA pulldown assays demonstrated that the RRM3 of PTBP1 protein, rather than other RRM-type RNA-binding motifs, specifically interacted with the in vitro transcribed DIO3OS RNA (Fig. 3f, bottom). In consistence with previous reports [26][27][28] , RRM3 is a recognized binding domain that contributes to RNA-binding specificity of PTBP1. Correspondingly, the PTBP1-targeting pyrimidine-rich motifs (e.g., UCUU, UCUUC, UCUCU, UUCUCU, CUCUCU) were found widespread in the DIO3OS RNA. By comparing the secondary structures of different DIO3OS variants predicted by RNAfold 29 (Supplementary Fig. 4b-g), we found that 60% PTBP1-binding motifs clustered in the region of nt 850-1930 of ENST00000554735, which only accounted for a third of its length and formed multiple stem-loop structures ( Supplementary Fig. 4h). These data suggested that DIO3OS specifically binds to PTBP1 in the nucleus of LTED breast cancer cells.
PTBP1 modulates the alternative splicing of many genes under certain contexts, resulting in the production of miscellaneous RNA transcripts with distinct functions 28,30 . To assay the potential regulation of PTBP1 on DIO3OS, we knocked down PTBP1 by two siRNAs ( Supplementary Fig. 4i, j) and found that there were no significant changes in DIO3OS expression, as determined by qRT-PCR (Supplementary Fig. 4k) and Northern blot ( Supplementary Fig. 4l). Since PTBP1 has been reported to participate in various processes of cancer progression, involving cell proliferation and metabolic reprogramming 31 , we then explored the functional role of PTBP1 in LTED cells. As revealed by cell counting and MTT assays, LTED cell growth was significantly suppressed by siRNA-mediated PTBP1 knockdown (Fig. 3g, h). EdU incorporation assay further substantiated that PTBP1 knockdown inhibited LTED cell proliferation (Supplementary Fig. 4m). We next tested whether silencing PTBP1 influenced the aerobic glycolysis in LTED cells. Consistent with the function of DIO3OS in LTED cells, knocking down PTBP1 decreased the ECAR levels as well as glycolytic capacity in MCF-7 LTED and T47D LTED cells (Fig. 3i, j). Both glucose consumption and lactate production were also impaired in PTBP1-knockdown LETD cells (Fig. 3k, l). Therefore, DIO3OS specifically interacts with nuclear PTBP1, a splicing regulator that supports the growth and aerobic glycolysis of LTED breast cancer cells.
DIO3OS regulates LTED cell proliferation and glycolysis through PTBP1
To validate that DIO3OS exerts its function via direct interaction with PTBP1, we conducted a series of rescue experiments. First, we transfected MCF-7/T47D LTED cells with control LNA or DIO3OS LNA, along with control vector, full-length PTBP1 construct or PTBP1 truncation variant deleting the RRM3, and then measured their lactate production as well as ECAR levels. Compared with control LNA/vector-transfected MCF-7/T47D LETD cells, co-transfection with DIO3OS LNA and fulllength PTBP1 construct rescued the lactate production that was impaired in cells with DIO3OS downregulation and without PTBP1 overexpression (Fig. 4a), while co-transfection of DIO3OS LNA and PTBP1 RRM3-deleted construct resembled the result of DIO3OS knockdown ( Supplementary Fig. 5a). Similar results were obtained from the glycolytic seahorse assay (Fig. 4b, c and Supplementary Fig. 5b). On the other hand, enhanced lactate production and ECAR levels in DIO3OS-overexpressing parental MCF-7 cells were inhibited by PTBP1 siRNAs, comparing to control siRNAs (Figs. 4d, left, and 4e).
These results were also readily detected in T47D cells (Figs. 4d, right, and 4f).
In the cell growth experiments, comparing to control sets, DIO3OS LNA transfection led to decreased cell number in both MCF-7 LTED and T47D LTED cells, whereas PTBP1 overexpression rescued cell growth almost back to the control levels ( Fig. 4g). Specifically, MTT assay showed that contrary to the full-length PTBP1, PTBP1 RRM3deleted construct failed to rescue the cell viability that were dampened in cells co-transfected with DIO3OS LNA ( Supplementary Fig. 5c). In parallel, PTBP1 siRNAs abolished the stable DIO3OS overexpressionincreased cell growth in parental MCF-7 and T47D cells (Fig. 4h, i). Similar result of cell proliferation was obtained from the EdU incorporation assay ( Fig. 4j and Supplementary Fig. 5d-f). Moreover, PTBP1 upregulation rescued the colony-forming capacity of LTED cells with DIO3OS knockdown ( Fig. 4k and Supplementary Fig. 5g), and vice versa ( Supplementary Fig. 5h). Together, DIO3OS regulates the aerobic glycolysis as well as proliferation in LTED breast cancer cells via PTBP1.
DIO3OS interacts with PTBP1 to upregulate LDHA mRNA stability
To explore how DIO3OS functions through interacting with PTBP1, we employed two sets of high-throughput sequencing to compare the transcription profiles of MCF-7 LTED cells with and without LNAmediated DIO3OS knockdown, as well as siRNA-mediated PTBP1 knockdown, respectively (Fig. 5a). Given that PTBP1 was a suppressor of alternative splicing (AS) 32 and DIO3OS potentially exerted an inhibitory effect on gene splicing by binding with PTBP1, we focused on the splicing events concordantly increased in DIO3OS and PTBP1knockdown cells. Twelve basic types of AS patterns in each sample were analyzed with ASprofile program and the differential AS events were identified in four sets of samples (Knockdown vs Control) ( Fig. 5a and Supplementary Data 2). By overlapping all four sets of the differential AS events, we identified 1457 events in 659 genes from concurrent DIO3OS and PTBP1-knockdown cells relative to control ( Fig. 5b and Supplementary Data 2). Among these events, AS occurring at transcription start site (TSS) was the predominant type (27%), followed by skipped exon (SKIP) (22%) and transcription terminal site (TTS) (17%) (Fig. 5c). The remaining types of AS events, including multiple skipped exon (MSKIP), alternative exon ends (AE), intron retention (IR), approximate intron retention (XIR), approximate skipped exon (XSKIP), approximate multiple skipped exon (XMSKIP), multiple intron retention (MIR), approximate alternative exon ends (XAE), and approximate multiple intron retention (XMIR), occurred at lower frequencies (Fig. 5c). With the DAVID Functional Annotation Tool 33 , KEGG pathway analysis revealed that many of differentially spliced genes were clustered in metabolic pathways, especially the glycolysis/gluconeogenesis pathway (Fig. 5d). These genes included LDHA, ALDH3A1, ALDOC, PFKP, ENO3, ALDH3B2, and PCK2. As such, our data suggested that DIO3OS and PTBP1 might regulate the alternative splicing of certain key glycolysis-related genes. a Northern blot of DIO3OS-knockdown efficiency in MCF-7/T47D LTED cells by the indicated LNAs. ACTB serves as loading control RNA. nt, nucleotide. b Growth curves of MCF-7/T47D LTED cells transfected with control or DIO3OS-targeting LNAs. c Representative immunofluorescence images and quantification of EdUincorporated MCF-7 LTED cells transfected with control or DIO3OS-targeting LNAs. Scale bars, 50 μm. d Representative images and quantification of plate clone formation of MCF-7/T47D LTED cells transfected with control or DIO3OS-targeting LNAs. e, f ECAR values and calculated glycolytic capacity of MCF-7 LTED (e) and T47D LTED (f) transfected with control or DIO3OS-targeting LNAs. g Lactate production of MCF-7/T47D LTED cells transfected with control or DIO3OS-targeting LNAs. h Northern blot of DIO3OS overexpression efficiency in parental MCF-7/ T47D cells. ACTB serves as loading control RNA. nt, nucleotide. i Growth curves of parental MCF-7/T47D cells transfected with control or DIO3OS overexpression plasmids. j Representative immunofluorescence images and quantification of EdUincorporated parental MCF-7 cells transfected with control or DIO3OS overexpression plasmids. Scale bars, 50 μm. k Representative images and quantification of plate clone formation of parental MCF-7/T47D cells transfected with control or DIO3OS overexpression plasmids. l Lactate production of parental MCF-7/T47D cells transfected with control or DIO3OS overexpression plasmids. m, n ECAR values and calculated glycolytic capacity of parental MCF-7 (m) and T47D (n) cells transfected with control or DIO3OS overexpression plasmids. Means ± s.d. of experimental triplicates (b-g, i-n), one representative experiment out of three that were similar (a, h) are shown, and P-values were assessed with two-sided one-way ANOVA with Dunnett's multiple-comparisons test. Source data are provided as a Source Data file.
To narrow down the candidate genes, LC-MS was performed to screen the glycolytic metabolites affected by DIO3OS and PTBP1. We found both D-glucose 6-phosphate (G6P) and lactate were significantly decreased upon DIO3OS and PTBP1 knockdown ( Fig. 5e and Supplementary Fig. 6a). Hexokinase-2 (HK2) and lactate dehydrogenase A (LDHA) are the two crucial metabolic enzymes that catalyze the production of G6P and lactate, respectively. Backtracking the alternatively spliced genes in DIO3OS and PTBP1-knockdown cells, besides LDHA, ALDH3A1, ALDOC, PFKP, and ENO3 also affect lactate production. Therefore, we used RT-PCR to examine the splicing events of these genes upon DIO3OS and PTBP1 knockdown, and found that AS events occurred Article https://doi.org/10.1038/s41467-022-34702-x only in the mRNA of LDHA (Fig. 5f), but not other glycolytic enzymes.
LDHA is the key enzyme which can catalyze the conversion of pyruvate into lactate, a crucial process of glycolysis. According to the ISOexpresso database, the most common LDHA transcript in breast cancer cells is LDHA-203, which harbors a 357-nt 5' untranslated region (UTR) as well as a 567-nt 3'UTR. The LDHA isoform encoded by this transcript has also been chosen as the canonical sequence, as shown in UniProtKB database. The AS event in LDHA mRNA from DIO3OS and PTBP1-knockdown cells was identified as TSS by the ASprofile software, which occurred at chr11:18,396,554-18,396,968, pointing to another transcript variant LDHA-220. As revealed by the UCSC Genome Browser, LDHA-220 not only harbors a 5'UTR differing from that of LDHA-203, but also features a complete deletion of 3'UTR (Supplementary Fig. 6b). To confirm these alterations of UTR sequences in LDHA mRNA, specific primers targeting the 3'UTR of LDHA-203, 5'UTRs of LDHA-203 and LDHA-220 were designed and subjected to RT-PCR assay ( Supplementary Fig. 6b). The products were further visualized in agarose gels, which showed that both 3'UTR and 5'UTR of LDHA-203 were strikingly reduced while the 5'UTR of LDHA-220 was upregulated in LTED cells with DIO3OS or PTBP1 knockdown ( Fig. 5f and Supplementary Fig. 6c), indicating an increase in LDHA-220 transcript featuring 3'UTR deficiency after DIO3OS or PTBP1 downregulation. To further demonstrate the impacts of UTR sequence changes on mRNA and protein expression, qRT-PCR was performed and revealed that only LDHA, but not PFKP, ALDOC, ENO3, or HK2, exhibited decreased mRNA level in DIO3OS and PTBP1-knockdown LTED cells ( Fig. 5g and Supplementary Fig. 6d). And the protein level of LDHA upon DIO3OS or PTBP1 knockdown was consistently reduced, as determined by western blot (Fig. 5h). Thus, we focused on LDHA for further investigation.
To elucidate the potential regulation of LDHA by DIO3OS in BT549 and SKBR3 cells, we also detected the LDHA expression in these DIO3OS-knockdown cells. In consistence with the phenotypic assays, silencing DIO3OS in BT549 and SKBR3 cells did not affect the LDHA mRNA level (Supplementary Fig. 6e). Also, qRT-PCR revealed that both LDHA-203 and LDHA-220 expression remained unchanged in DIO3OS-knockdown BT549 cells and SKBR3 cells ( Supplementary Fig. 6f), indicating that DIO3OS might not affect LDHA splicing in these cells.
5'UTR and 3'UTR harbor important regulatory elements for mRNA processing. Particularly, 5'UTR has been reported to regulate the translation initiation 34 while 3'UTR is more likely to be involved in the mRNA transport and stabilization 35 . To test whether DIO3OS and PTBP1 stabilize LDHA mRNA, we used actinomycin D (ActD) to inhibit mRNA transcription with the proto-oncogene c-Myc as an unstable control, whose mRNA level was strikingly decreased 2 h following ActD treatment ( Supplementary Fig. 6g). As revealed in Fig. 5i, a significant reduction of the LDHA mRNA stability was detected in MCF-7 LTED cells with either DIO3OS or PTBP1 knockdown. To further dissect the interaction between PTBP1 and LDHA mRNA, as well as the role of DIO3OS therein, we examined whether DIO3OS directly regulated the PTBP1 expression. However, PTBP1 protein level was not changed after DIO3OS knockdown (Supplementary Fig. 6h). Then we performed the RNA-IP assay to detect whether PTBP1 interacted with LDHA mRNA, and whether DIO3OS affected the interaction. In parental MCF-7 cells transfected with control vector, LDHA mRNA harboring 3'UTR was enriched~5 folds in the anti-PTBP1 immunoprecipitates, compared to the IgG immunoprecipitates, whereas in DIO3OS-overexpressing MCF-7 cells, the retrieval of LDHA mRNA with 3'UTR by anti-PTBP1 antibody was significantly increased to~15 folds (Fig. 5j, left). These RNA-protein (RNP) complexes were also readily detected in MCF-7 LTED cells, where the anti-PTBP1 immunoprecipitates retrieved much more LDHA mRNA with 3'UTR in sh-Ctrl cells than in DIO3OSknockdown cells (Fig. 5j, right). These data suggested that DIO3OS promotes the interaction between PTBP1 and LDHA mRNA that harbors 3'UTR.
To further validate that DIO3OS modulates LDHA expression by virtue of PTBP1 binding to LDHA 3'UTR, we transfected reporter plasmids containing luciferase gene along with LDHA-220 5'UTR, LDHA-203 5'UTR, or LDHA-203 3'UTR ( Fig. 5k) into MCF-7 LTED cells with or without DIO3OS knockdown or PTBP1 overexpression. Luciferase reporter assay showed that partial silencing of DIO3OS by LNA only caused reduction in the luciferase activity of cells with luc-3'UTR (LDHA-203) expressed, while in the presence of PTBP1-overexpressing plasmid, the luciferase activity in luc-3'UTR (LDHA-203) construct was Fig. 3 | DIO3OS interacts with PTBP1 in the nucleus. a Biotin-RNA pulldown assay using full-length DIO3OS transcript (sense) and the antisense RNA, followed by silver staining. Red box indicates the differential band. MW molecular weight. b Mass spectrometry profile of DIO3OS-binding protein PTBP1 with high score. c Biotin-RNA pulldown assay followed by western blot demonstrating the interaction of DIO3OS and PTBP1. MW, molecular weight. d RNA immunoprecipitation assay with the anti-PTBP1 antibody demonstrating the interaction of DIO3OS and PTBP1 in MCF-7/T47D LTED cells. MW, molecular weight. e Fluorescence assessment of the subcellular colocalization of DIO3OS and PTBP1 in MCF-7/T47D LTED cells. DIO3OS and PTBP1 were detected respectively by RNA FISH (green) and immunofluorescence (red) staining. Scale bars, 10 μm. f Biotin-RNA pulldown assay followed by western blot demonstrating the interaction of DIO3OS and the RNA recognition motifs (RRMs) of PTBP1. Schematic diagram for PTBP1 truncation variants deleting the RRM1, RRM2, RRM3, or RRM4 (upper). Representative western blot for in vitro binding of FLAG-tagged PTBP1 truncation variants with DIO3OS (bottom). MW, molecular weight. g, h Cell number counting (g) and MTT assay (h) of MCF-7 LTED cells transiently transfected with control or PTBP1 siRNAs. i, j ECAR values and calculated glycolytic capacity of MCF-7 LTED cells (i) and T47D LTED cells (j) transiently transfected with control or PTBP1 siRNAs. k Glucose uptake of MCF-7/T47D LTED cells transiently transfected with control or PTBP1 siRNAs. l Lactate production of MCF-7/T47D LTED cells transiently transfected with control or PTBP1 siRNAs. Means ± s.d. of experimental triplicates (d, g-l), one representative experiment out of three that were similar (a-c, e, f) are shown, and P-values were determined using two-tailed Student's t-test (d), two-sided one-way ANOVA with Dunnett's multiple-comparisons test (g-l). Source data are provided as a Source Data file.
dramatically restored (Fig. 5n). However, similar phenomena didn't occur in contexts with constructs harboring 5'UTR (Fig. 5n). Collectively, our findings indicated that DIO3OS promotes the binding of PTBP1 to LDHA 3'UTR, therefore stabilizing LDHA mRNA and resulting in more LDHA expression and lactate production. LDHA-203 5'UTR and LDHA-220 5'UTR showed that LDHA-220, harboring the peculiar 5'UTR and lacking 3'UTR, mainly existed in parental MCF-7 and T47D cells, whereas LDHA-203, harboring longer 5'UTR and 3'UTR, significantly increased in LTED cells (Fig. 6a), suggesting that LDHA-220 in parental cells might be replaced by LDHA-203 following LTED treatment. To figure out whether such sequence variations in UTRs affect the stability of LDHA mRNA in parental and LTED cells, we inhibited mRNA transcription through ActD treatment, which was proven to be effective using c-Myc mRNA as a positive control (Supplementary Fig. 6g). Subsequent qRT-PCR revealed an over 10-h halflife of LDHA mRNA in LTED cells, whereas the stability of LDHA mRNA in parental MCF-7 or T47D cells was markedly shortened (Fig. 6b). Meanwhile, LTED cells exhibited a higher LDHA protein level in contrast to their parental cells (Fig. 6c).
Furthermore, we detected the sequence alterations of LDHA mRNA in breast tumor tissues obtained from AI-sensitive and AIresistant patients using qRT-PCR. In consistence with the result of cell lines, AI-resistant breast cancer samples featured a generally higher expression of LDHA-203 (with 3'UTR) than AI-sensitive groups, while the latter ones exhibited more upregulated expression of LDHA-220 (with unique 5'UTR) (Fig. 6d, e).
To further demonstrate the functionalities of these two LDHA transcripts in ER-positive breast cancer cells, we overexpressed them in parental MCF-7 and T47D cells. Although qRT-PCR with specific primers targeting either 5'UTR of LDHA-203 and LDHA-220 confirmed that both transcript variants were successfully overexpressed 24 h post transfection ( Supplementary Fig. 6i), western blot revealed different LDHA protein levels 48 h after ectopic expression of two transcripts (Supplementary Fig. 6j), indicating that LDHA-220 featuring 3'UTR absence might be prone to decay, resulted in lowered protein level than LDHA-203 did. In the phenotypic assays, when compared to the control context, overexpression of LDHA-203, but not LDHA-220, markedly promoted the growth of parental MCF-7 and T47D cells (Fig. 6f). LDHA-203expressing cells showed enhanced lactate production (Fig. 6g) as well as higher ECAR levels (Fig. 6h, i). Instead, LDHA-220 transfection didn't enhance the glycolysis of parental MCF-7 and T47D cells, as determined by lactate production and ECAR measurement (Fig. 6g-i). Therefore, LDHA-203, upregulated in AI-resistant breast cancer cells, could increase the lactate production during the glycolytic processes.
Collectively, lncRNA DIO3OS directly interacts with PTBP1 protein in the nucleus and stabilizes the mRNA of LDHA by protecting the integrity of its 3'UTR, which consequently upregulates LDHA expression and activates glycolytic metabolism in AI-resistant breast cancer cells (Fig. 6j).
DIO3OS promotes ER-positive breast tumor progression in vivo
To further explore whether DIO3OS regulates ER-positive breast cancer development in vivo, we orthotopically injected parental MCF-7 cells with or without DIO3OS overexpression in ovariectomized nude mice. These mice were previously divided into three groups by treating with different doses of estrogen tablets. In the low-dose (0.18 mg) estrogentreated group, the tumor growth of DIO3OS-overexpressing MCF-7 xenografts was significantly promoted, compared with controls ( Fig. 7a-c). Similarly, enforced DIO3OS expression accelerated the growth of tumor when conditioned with high-dose (0.72 mg) estrogen (Fig. 7a-c). Furthermore, under the condition of complete estrogen deprivation by ovariectomy and without estrogen tablet treatment, DIO3OS overexpression strikingly boosted the tumorigenicity in nude mice (6/10), in comparison to the control group (2/10) (Fig. 7e, f). Moreover, enhanced 18 F-fluorodeoxyglucose ( 18 FDG) accumulation was observed in tumor xenografts from mice with enforced expression of DIO3OS, regardless of estrogen administration, as assessed by positron emission tomography and computed tomography (PET-CT) scanning (Fig. 7d, g). In addition, xenografts with DIO3OS upregulation exhibited more intensive Ki67 staining than those with control DIO3OS expression, supporting the pro-tumor role of DIO3OS in vivo (Fig. 7h, i). Meanwhile, higher LDHA expression was observed in xenografts with DIO3OS overexpression, as detected by immunohistochemistry (Fig. 7h, i), which was consistent with the enhanced glycolytic level upon DIO3OS overexpression. These results showed that DIO3OS might promote the estrogen-independent growth of ER-positive breast tumors in vivo.
Conversely, DIO3OS knockdown suppressed the growth of MCF-7 LTED xenografts (Fig. 7j-l), reduced 18 FDG accumulation (Fig. 7m) and decreased the expression of LDHA along with Ki67 in xenografts of ovariectomized nude mice with low-dose estrogen treatment (Fig. 7n, o). Together, DIO3OS positively regulates the estrogenindependent breast cancer cell proliferation and glucose metabolism in vivo.
Discussion
Aerobic glycolysis is one of the hallmarks of metabolic reprogramming in human cancers. Here, we demonstrate that lncRNA DIO3OS directly interacts with PTBP1 protein in the nucleus and further stabilizes the mRNA of LDHA by protecting the integrity of its 3'UTR, ultimately activates glycolytic metabolism in AI-resistant breast cancer cells. LDHA is a pivotal glycolytic enzyme that converts pyruvate to lactate. As a crucial metabolite in glycolysis pathway, lactate not only fuels the energy supply for cancer growth, but also serves as a signaling molecule in promoting cancer angiogenesis, invasion and migration, as well as immune escape [3][4][5][6] . It has been documented that LDHA is upregulated in various human cancers, which promotes glycolytic metabolism and cancer development. Inhibition of LDHA activity can resensitize breast tumor cells to anti-HER2 therapy (trastuzumab) or chemotherapy (taxol) 36,37 .
Recently, lncRNAs have been reported to participate in LDHA regulation. HULC and LINC00973 can directly bind to LDHA protein and enhance its enzymatic activity in cancers 38,39 . LncRNA GLCC1 stabilizes c-Myc through direct interaction with HSP90 chaperon and transactivates target genes such as LDHA, thereby reprograming glycolytic metabolism for cancer cell proliferation 40 . Different from these studies, we discover that DIO3OS interacts with PTBP1 to modulate the stability of LDHA mRNA. Specifically, DIO3OS and PTBP1 can bind to and presumably protect the 3'UTR of LDHA mRNA from splicing-induced deletion, and thus maintain its integrity and stability. Most eukaryotic mRNA has a polyadenylic acid (poly A) tail downstream of 3'UTR, which is involved in the regulation of mRNA stability and transport 41,42 . Selective usage of alternative poly(A) sites residing in 3'UTR has been reported to mediate 3'UTR shortening, without altering the coding capacity of the mRNA 43 . As the well documented types l Biotin-RNA pulldown followed by western blot demonstrating the interaction of PTBP1 and different LDHA variants with or without intact 3'UTR. MW molecular weight. m RNA immunoprecipitation assay with the anti-PTBP1 antibody revealing the interaction of PTBP1 and LDHA 5'/3'UTR in MCF-7 LTED cells transfected with different LDHA-UTR-fused luciferase reporter plasmids. n Luciferase reporter assay in MCF-7 LTED with indicated treatment. Means ± s.d. of n = 3 (e, g, j, m, n) or n = 4 (i) independent experiments yielding similar results, one representative experiment out of three that were similar (f, h, l) are shown, and P-values were analyzed using two-tailed Student's t-test (e, j, m), two-sided one-way ANOVA with Dunnett's multiple-comparisons test (g, i, n). Source data are provided as a Source Data file.
of alternative poly(A) signals are located within the 3'UTR of LDHA mRNA, alternative polyadenylation (APA) seems not adequate to account for the complete absence of 3'UTR in LDHA-220. In fact, AS within UTRs has been found in the transcript variants of at least 13% of genes 44,45 . Alternative 3'UTR formation can also be mediated by a dual mechanism, which involves both APA and AS 44,46 . Reciprocally, lacking 3'UTR may hinder the formation of poly A tail and cause aberrant mRNA processing 45,47 . In our study, knockdown of DIO3OS or PTBP1 contributes to a splicing switch from canonical LDHA transcript to an LDHA variant lacking 3'UTR, which is proven to be unstable. Overexpression of canonical transcript LDHA-203, but not the unstable transcript LDHA-220, markedly enhances lactate production and promotes breast cancer cell growth. Thus, our study uncovers a mechanism of regulating LDHA at the mRNA level which depends on splicing switch induced by lncRNA DIO3OS expression.
The third-generation AI therapy has become the standard of care for ER-positive postmenopausal breast cancer patients, yet resistance to AI treatment remains an important clinical challenge [12][13][14] . Previous studies have revealed mechanisms of intrinsic and acquired resistance to AI therapy, most of which involves the deregulation of the ER pathway. It has been reported that MCF-7 LTED cells display higher aerobic glycolytic activity than their parental counterparts, which is mechanistically regulated by ER-dependent miR-155 expression 48 . Although we also detected enhanced aerobic glycolysis in breast cancer cells that have undergone LTED, the mechanism was not linked to ER. Instead, we have characterized a lncRNA DIO3OS that remodels glycolytic metabolism in AI-resistant cancer cells. As we and others have found that AI-resistant breast cancer may rely more on aerobic glycolysis for tumor growth, DIO3OS confers growth advantages to AIresistant cells by regulating splicing switch to enhance aerobic glycolysis. These results highlight a critical role of DIO3OS in inducing AI resistance by activating an alternative proliferating pathway independent of ER.
Although estrogen deprivation can mimic the hormone withdrawal that occurs during AI treatment in clinic, this in vitro model cannot simulate the adaptive changes within cancer cells chronically treated with an AI, nor can it be employed to detect the regulatory effects of AI drugs on aromatase expression and activity. This is because there is low or no endogenous aromatase expressed in breast cancer cell lines. To solve this problem, other research groups have used breast cancer cells stably overexpressing the human aromatase gene to study AI response. However, this model cannot account for the fact that it is the stromal cells of breast cancers that predominantly express aromatase to convert androgens to estrogens 49,50 . As such, we have detected the expression level of DIO3OS in tumor samples from ER-positive breast cancer patients receiving AI therapy, and have demonstrated its prognostic value in these patients, further confirming the clinical relevance of above-mentioned findings.
Nuclear lncRNAs often exert their biological functions by regulating gene expression in cis or in trans. We demonstrate that lncRNA DIO3OS functions through the interaction with PTBP1 protein in the nucleus. PTBP1 is a splicing suppressor, which can compete with spliceosome for RNA binding and thereby inhibits alternative splicing 28,30,32 . As lncRNAs can act as molecular scaffolds to bring together larger complexes and localize splicing regulators, the DIO3OS-PTBP1 interaction may be important for the nuclear localization and function of PTBP1. Various studies have reported that PTBP1 regulates oncogenic splicing switch in pyruvate kinase from splice variant PKM1 towards PKM2, thereby contributing to energy metabolism remodeling from oxidative phosphorylation to aerobic glycolysis, as well as resistance to cancer therapy 51,52 . A neural-specific lncRNA named Pnky can also interact with PTBP1 to regulate the alternative splicing of a core set of transcripts which are involved in the neuronal differentiation and neurogenesis 31 . To explore the alteration of splicing events induced by DIO3OS and PTBP1 in LTED cells, we identified the concordantly occurred AS events in DIO3OS-knockdown and PTBP1knockdown cells, which occurred in genes pointed to glycolysis pathway. Screen of the glycolytic metabolites affected by DIO3OS and PTBP1 and further verification of the splicing events in glycolytic enzymes identified LDHA as a key downstream target of DIO3OS and PTBP1 in AI-resistant breast cancer cells.
Recently, genetic polymorphisms (rs10420407) in the PTBP1 gene were found to affect the patient response to androgen-deprivation therapy in prostate cancer 53 . So far, no breast cancer-associated single nucleotide polymorphisms (SNPs) of PTBP1 have been reported, however, this finding may have implications for the role of PTBP1 SNP in endocrine therapies in cancers.
Targeting cancer metabolism emerges as an important strategy to improve anti-cancer therapy. As the expression of DIO3OS increases in breast cancer cells undergoing AI resistance, the aberrant activation of DIO3OS/PTBP1/LDHA glycolysis cascade may endow tumor cells with metabolic adaptation that allows them to survive from AI-caused estrogen deficiency. Future translational studies will focus on exploring effective approach for specific DIO3OS knockdown to re-sensitize breast cancer cells to AI treatment.
Study approval
All tumor samples from breast cancer patient used in this study were obtained under informed written consent and approval of the Internal Review and Ethics Board of Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University, according to ethical regulations. In vivo experiments in murine model were reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) guidelines at Sun Yat-Sen University (SYSU-IACUC-2020-000062).
Cell culture and treatment BT549, MDA-MB-231, MDA-MB-468, Hs578T, MDA-MB-436, SKBR3, BT474, MDA-MB-361, MCF-7, and T47D human breast cancer cell lines were obtained from the American Type Culture Collection (ATCC) and cultured according to standard protocols. All the cell lines were authenticated by short tandem repeat profiling prior to use and tested negative for mycoplasma contamination. For the establishment of the AI-resistant cell lines, MCF-7 and T47D cells were grown in steroiddeprived phenol red-free RPMI 1640 medium (Gibco™) with 10% dextran charcoal-stripped (DCC) fetal bovine serum (Hyclone) for at least 10 months to gain AI resistance, termed as LTED cells. For functional experiments, cells were cultured for 3-7 days in DCC medium. For the assessment of AI drug sensitivity, LTED cells were changed with fresh medium supplemented with 10 −8 -10 −5 mol/L letrozole (Sigma) and anastrozole (Sigma) every 2 days and counted at 8 th day using an automated cell counter (Countstar IC1000). Fulvestrant (Selleck) at a final concentration of 1 μM in the ordinary culture medium was used to induce ER downregulation in LTED cells.
Seahorse assay
In vitro cellular metabolic alterations were accessed by Seahorse XF24 extracellular flux analyzer (Seahorse Bioscience), following the replicates. f, g Cell number counting (f) and lactate production (g) of parental MCF-7/T47D cells transiently transfected with different LDHA variant-expressing plasmids or control vector. h, i ECAR values and calculated glycolytic capacity of parental MCF-7/T47D cells with indicated treatments. j Graphical illustration of the DIO3OS/PTBP1/LDHA glycolysis cascade acting as a switch regulating the glucose metabolism in breast cancer adaptation to LTED by stabilizing LDHA mRNA. Means ± s.d. of experimental triplicates (c, f-i), one representative experiment out of three that were similar (a, e) are shown, and P-values were analyzed by two-tailed Student's t-test (b-d), two-sided one-way ANOVA with Dunnett's multiplecomparisons test (f-i). Source data are provided as a Source Data file.
manufacturer's instructions. Briefly, MCF-7/T47D cells with indicated treatments were seeded in XF24-well culture plates at a density of 3000 (MCF-7) or 5000 (T47D) cells per well overnight, followed by a 24 h serum starvation. After 1 h incubation with basal medium containing 2 mM L-glutamine at 37°C, cells were subjected to assessment of the extracellular acidification rates (ECAR) or oxygen consumption rate (OCR) at 8 min intervals. For the measurement of ECAR, glucose, oligomycin, and 2-deoxyglucose (2DG) was sequentially injected to the wells at every interval time of 3 measurement, with a final concentration of 10 mM, 1 μM, and 50 mM, respectively. In terms of the OCR detection, oligomycin, carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone (FCCP), and rotenone/antimycin A were in turn added to the wells after every 3 measurements at intervals of 8 min, with a final concentration of 1 μM, 0.5 μM, or 0.5 μM, respectively. The reading difference between the ECAR following oligomycin addition and the basal ECAR was represented as the glycolytic capacity.
Glucose uptake, lactate production, and acetyl-CoA level measurement Cells with indicated treatments in 96-well plates were grown to about 40% confluence and replenished with fresh medium. Twenty-four hours later, Glucose Uptake Assay Kit (#ab136955, Abcam), Lactate Colorimetric Assay Kit II (#K627-100, BioVision) or PicoProbe™ Acetyl CoA Assay Kit (#K317-100, BioVision) was employed to measure the glucose uptake, lactate production or acetyl-CoA level following the manufacturer's instructions, respectively.
Liquid chromatography-mass spectrometry (LC-MS) analysis
For metabolites analysis by LC-MS, cells with indicated treatments were seeded in six-well plates at~80% confluence. With the medium aspirated and the plates rinsed thrice with ice-cold saline, the plates were placed on liquid nitrogen to quench cell metabolism, followed by addition of 1 mL of pre-cooled (−80°C) 80% (vol/vol) methanol containing internal standards and incubation at −80°C for 20 min. The cell lysate/methanol mixture was scraped and collected into a 1.5 mL conical tube, followed by addition of 500 μL of pre-cooled (−80°C) acetonitrile/water (80:20, vol/vol) on the culture plates, which was transferred to the same conical tube. After a 10-min centrifugation (18,000 × g at 4°C), the supernatant was transferred to another 1.5 mL conical tube and dried with SpeedVac. The analysis of all and samples were performed on ultra-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry platform (UPLC-ESI-MS/MS; Agilent 1290 Infinity II with Agilent G6495A). Mobile phase A and mobile phase B were LC/MS-grade water and 85% acetonitrile respectively that both contained 20 mM ammonium acetate and 0.1% ammonium hydroxide (35% (wt/vol). Dried samples were resuspended in mobile phase (A/B, 4/6) and 3 μL was injected and chromatographically separated on an amide XBridge HPLC column (4.6 mm inner diameter × 100 mm length; 3.5 μm; Waters) maintained at 40°C, the target metabolites were eluted with a 92-50% gradient of mobile phase B at flow rate of 0.6 mL/min. Mass spectral data were acquired in a positive/negative ion switching mode with multiple reaction monitoring of precursor and characteristic product ions specific to each metabolite. These metabolites were identified according to standard compounds and integrated using the Agilent MassHunter workstation QQQ quantitative analysis software (Version B.07.00). Results of each sample were normalized based on total protein concentration.
Patient tissue specimens
Tumor samples (n = 257) were obtained from female patients diagnosed with ER-positive breast cancer, receiving AI treatment post surgical resection at Sun Yat-Sen memorial Hospital, Sun Yat-Sen University ranging from 2011 to 2018 with informed written consents. All study-related procedures were conducted under the approval and guidance from the Internal Review and Ethics Board of Sun Yat-Sen Memorial Hospital. Acquired endocrine resistance of breast cancer refers to those recurrence and metastasis after receiving more than two years of postoperative endocrine therapy, or recurrence and metastasis within one year after finishing adjuvant endocrine therapy, or disease progression of metastatic breast cancer after receiving the first-line endocrine therapy more than six months.
RNA extraction and quantitative real-time PCR (qRT-PCR)
RNA was extracted by TRIzol agent (Invitrogen), homogenized in chloroform, and then purified by isopropanol and ethanol. Frozen tissues in TRIzol were subjected to mechanical homogenization by a TissueLyserII (Qiagen), with a 5-mm stainless steel bead per sample. For qPCR, RNA concentration and quality were assessed by the NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific). At most 1 μg of total RNA per sample was reverse transcribed with a Pri-meScript™ RT Reagent Kit (#RR014A, Takara), subsequent real-time qPCR amplification was performed on a LightCycler 480 instrument (Roche) with SYBR Premix Ex Taq TM (#RR420L, TAKARA). Gene expression level was normalized by GAPDH or ACTB. All primer sequences in this study are included in Supplementary Table 5.
High-throughput RNA sequencing
Total RNA was extracted using TRIzol (Invitrogen) according to the manufacturer's instructions. Both integrity and concentration of RNA was assessed with the RNA Nano 6000 Assay Kit of the Bioanalyzer 2100 system (Agilent Technologies, USA). A total of 2 μg RNA per sample was used and the sequencing libraries were generated with NEBNext® Ultra™ RNA Library Prep Kit for Illumina® (NEB, USA). Sequencing was performed on an Illumina platform (Annoroad Gene Technology Co., Ltd., China) and 150 bp paired-end reads were generated, mapped to the human reference genome (hg38 sourced from UCSC Genome Browser) with HISAT2 v2.1.0. Reads counts for genes in each sample were assessed by HTSeq v0.6.0, and fragments per kilobase per million mapped reads (FPKM) was calculated to estimate gene expression in each sample. Genes with P-value ≤ 0.05 and |log2_-ratio| ≥ 1 are identified as differentially expressed genes (DEGs). Heatmap and volcano plot were generated by using R package. The sequencing data have been deposited in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) public database under accession code GSE206199 and GSE206142. For Gene Set Enrichment Analysis (GSEA), software of GSEA v.4.2.3 was accessible from the GSEA website (http://www.broadinstitute.org/ gsea). For KEGG analysis, the DAVID online bioinformatics system (https://david.ncifcrf.gov) was utilized. a-c Growth curves of tumor xenografts (a), tumor size (b), and tumor weight (c) in nude mice bearing parental MCF-7 cells stably transfected with control or DIO3OSoverexpressing plasmid. The mice were pretreated with low or high dose of estrogen tablets. d Representative PET-CT scan images of the glucose accumulation in tumor xenografts arisen from parental MCF-7 cells stably transfected with control or DIO3OS-overexpressing plasmid under low-or high-dose estrogen treatment. e, f Tumor size (e) and tumor volume (f) in mice orthotopically injected with parental MCF-7 cells harboring control or DIO3OS-overexpressing plasmid without estrogen tablet. g Representative PET-CT scan images of the glucose accumulation in tumor xenografts arisen from parental MCF-7 cells stably transfected with control or DIO3OS-overexpressing plasmid without estrogen treatment. h, i Representative ISH staining of DIO3OS and IHC staining for LDHA/Ki67 expression in indicated tumor xenografts arisen from parental MCF-7 cells (h) and the quantitative analyses (i). Scale bars, 50 μm. j-l Growth curves of tumor xenografts (j), tumor size (k), and tumor weight (l) in mice that were orthotopically injected with MCF-7 LTED cells transducing lenti-shDIO3OS or lenti-control. The mice were pretreated with low dose of estrogen tablets. m Representative PET-CT scan images of the glucose accumulation in tumor xenografts arisen from MCF-7 LTED cells transducing lenti-shDIO3OS or lenti-control. n, o Representative ISH staining of DIO3OS and IHC staining for LDHA/Ki67 expression in indicated tumor xenografts arisen from MCF-7 LTED cells (n) and the quantitative analyses (o). Scale bars, 50 μm. For a-c, h-o, n = 6 mice per group. For e, f, n = 10 mice per group. Each dot represents one mouse (c, f, i, l, o). Means ± s.e.m. (a, c, f, j, l) are presented, and P-values were analyzed using two-sided one-way ANOVA with Sidak's multiplecomparisons test (a, c), two-tailed Student's t-test (f, j, l). For i, o, means ± s.d. are shown, and P-values were determined using two-tailed Student's t-test (for DIO3OS expression and Ki67 index), Mann-Whitney test (for IHC score of LDHA). For d, g, m, colored circles denoted the xenografts with indicated treatments. E2 estrogen. Source data are provided as a Source Data file.
Rapid amplification of cDNA ends (RACE)
The 5'/3' RACE of DIO3OS RNA was determined using the SMARTer RACE 5'/3' Kit (#634860, Clontech) following the manufacturer's guidance. The RNA template for DIO3OS was extracted from MCF-7 LTED cells and all gene-specific primers used in RACE are listed in Supplementary Table 5.
RNA decay assay
Cells with indicated treatments were plated into six-well plates. The mRNA decay rates were estimated by inhibiting the mRNA transcription with 10 µg/mL actinomycin D (APExBio) in culture medium. The equal volume of DMSO was used for negative control. RNA was extracted at indicated time points for further qRT-PCR analysis, as the relative amount of specific mRNA remaining in each sample was linked to mRNA degradation. Results were normalized to the expression level of reference gene GAPDH in each sample.
In vitro translation assay
For the generation of putative DIO3OS-GFP fusion protein, the fulllength sequence of DIO3OS was amplified by reverse transcription PCR and then cloned into the mammalian expression vector pcDNA3.1 with a GFP fused to the carboxyl-terminus. After validation by sanger sequencing, the fusion plasmids were used for cell transfection.
Transfection and transduction of tumor cells
The LNAs antisense oligonucleotides (Exiqon) and siRNA/shRNA (Genepharma) sequences are listed in Supplementary Table 6. DIO3OS and GFP-DIO3OS fusion proteins, FLAG-tagged full-length and truncated PTBP1 constructs, LDHA variants, luciferase-LDHA 5'/3'UTR chimeric PCDNA3.1 plasmid as well as pGL3 luciferase reporter plasmid used for ectopic expression in cell lines were constructed by Generay Biotech. Cells were transfected with indicated LNAs, siRNAs, or plasmids using Lipofectamine 3000 (#L3000008, Invitrogen) following the manufacturer's recommendations. For the screening of stable DIO3OS-overexpressed cells, cells were changed fresh medium supplemented with 1 mg/ml G418 every 2 days for at least two weeks. For transduction of tumor cells, cells were seeded into six-well plates tõ 40% confluence and transduced with lentiviral particles at an estimated multiplicity of infection of 5 with 8 µg/mL polybrene (Sigma). To select stable DIO3OS-knockdown cells, 5 µg/mL puromycin (Gibco) was used 72 h after transfection.
EdU incorporation assay
The EdU incorporation assay was conducted using a Cell-Light EdU Apollo®567 In Vitro Imaging Kit (#C10310-1; Guangzhou RiboBio Co., Ltd.). Briefly, cells with indicated treatment were incubated with diluted EdU reagent (1:5000 in complete culture medium) at 37°C overnight. After fixation with 4% paraformaldehyde (PFA), cells were subjected to 30 min incubation with Apollo Staining reaction liquid at room temperature, protected from light. Hoechst 33342 reagent was applied to counterstain the nuclei for another 30 min and the images were obtained from an automatic inverted fluorescence microscope (Axio Observer Z1, Zeiss).
RNA pulldown assay
Biotin-16-UTP labeled DIO3OS RNA was in vitro transcribed with a MEGAscript™ T7 Transcription Kit (#AM1333, Thermo Fisher Scientific) following the manufacturer's guidance. Briefly, 3 ug of folded biotin-RNA was incubated with 1 mg of total protein extract in IP lysis buffer (#87788, Thermo Fisher Scientific) under 1 h gentle rotation. Afterward, the mixture was incubated with 50 µL of prewashed Dyna-beads® M-280 Streptavidin (#11205D, Thermo Fisher Scientific) at room temperature for another hour. After extensive washing with IP lysis buffer using a magnetic stand, the retrieved protein was detected by either polyacrylamide gel-electrophoresis (PAGE) or immunoblot. The specific silver-stained protein bands cut out of PAGE gel were subjected to further identification by liquid chromatographic-tandem mass spectrometric.
RNA immunoprecipitation and crosslinking immunoprecipitation (CLIP)
RNA immunoprecipitation assay was conducted with Magna RIP™ RNA-Binding Protein Immunoprecipitation Kit (#17-700, Millipore), following the manufacture's protocol. In brief, cells were lysed by RIP lysis buffer supplemented with RNase inhibitor (#N2111S, Promega) and protease inhibitor (#8444, Thermo Fisher Scientific), and incubated at 4°C with antibody-conjugated Protein A/G Dynabeads overnight. Specific anti-PTBP1 antibody (#57246, CST, dilution 1:100) and negative control rabbit IgG (#17-700, Millipore, 5 μg per reaction) were used for the retrieval of RNA. After rigorous washes with RIP lysis buffer, co-precipitated RNA was detected by qRT-PCR analysis. For CLIP assay, before cell lysis, 0.3% formaldehyde (in PBS) was added to the cell culture plates for 10 min, shaking at room temperature, followed by incubation with 0.125 M glycine for 5 min to stop crosslinking.
For RNA ISH, sections of the paraffin-embedded tissue samples were dewaxed and rehydrated following standard protocols. After 30 min digestion with 5% trypsin, the tumor sections were subjected to fixation in 4% PFA at room temperature and 2 h prehybridization with hybridization solution at 52°C. Afterward, a 25 nM DIG-labeled probe targeting DIO3OS was used for hybridization at 52°C overnight in a humidified chamber. With the nuclei counterstained with Nuclear Fast Red solution, ISH signals were detected as blue-purple-colored staining under an Olympus BX51 microscope (Olympus, Tokyo, Japan). The staining index (SI) of DIO3OS was calculated as the multiplication of the proportion and intensity scores of positively stained cells by counting at least 10 random fields (objective ×20) 54 . In detail, the proportion of DIO3OS positively stained tumor cells was graded into five levels (0, no positive cells; (1) <25% DIO3OS-positive cells; (2) 25-50% DIO3OS-positive cells; (3) 50-75% DIO3OS-positive cells; (4) >75% DIO3OS-positive cells). The staining intensity was scored on a four-point scale (0, no DIO3OS staining; 1, weak DIO3OS staining; 2, moderate DIO3OS staining; 3, strong DIO3OS staining). The expression of DIO3OS was recorded from 0 to 12 according to the SI, with an optimal cutoff value of <3 (low) versus ≥ 3 (high).
Immunohistochemistry staining
Sections of the paraffin-embedded tumor samples were deparaffinized, rehydrated, followed by antigen retrieval. Endogenous peroxidase was eliminated by 3% hydrogen peroxide. The tissue sections were blocked with goat serum for 30 min at room temperature and then incubated with primary antibody against LDHA (#3582, CST, dilution 1:200) or anti-Ki67 (#ZM-0166, ZSGB-BIO, ready to use) reagent overnight at 4°C. The immunodetection was performed with DAB (GK500710, Gene Tech) following the manufacturer's instructions. The immunostaining was observed under an Olympus BX51 microscope (Olympus, Tokyo, Japan). The quantification of Ki67positive cells was assessed using ImageJ software (v.1.43) with at least 5 ×200 magnification images per section 55 . The staining scores of LDHApositive cells were calculated according to their proportion and intensity in 10 random fields (objective ×20) 6 . The evaluation of automated measurements was reviewed independently by two pathologists.
Animal experiments
Four-week-old female athymic BALB/c nude mice purchased from the Vital River Laboratories (Beijing) were raised under standard conditions (20-26°C temperature, 40-60% humidity) with a 12 h light/12 h dark cycle at the specific-pathogen-free (SPF) animal facility in the Laboratory Animal Resource Center of Sun Yat-Sen University until six weeks old. Mice were initially ovariectomized and allowed to acclimatize for a period of seven days before subcutaneous estrogen tablet implantation. Six to ten mice were randomly assigned to each group for different treatments. Another week later, a total of 0.1 mL sterile cell suspension containing 1 ×10 7 cells with or without altered DIO3OS expression in Matrigel (BD Biosciences) were injected orthotopically into the fourth pair mammary fat pads of nude mice. The monitoring of tumor volumes (mm 3 ) was performed every week according to the formula width 2 × length / 2. Maximum tumor size (1500 mm 3 ) permitted by ethics committee was not exceeded. Humane endpoints in this study were also referred to cachexia, body-weight reduction of 10% or infection, etc. At the experimental endpoint, mice were euthanized by cervical dislocation following overdose anesthesia.
Positron emission tomography/computed tomography imaging
Glucose accumulation of nude mice bearing breast tumor xenografts at 8 th week was detected using the Inveon micro-PET/CT Scanner (Siemens). Prior to micro-PET scanning, all mice were fasted for at least 8 h, followed by anaesthetization, and then received 5 µCig −1 [ 18 F] fluorodeoxyglucose ( 18 F-FDG) in saline (100 μL) intravenously via tail vein injection. Forty minutes later, a static 15 min PET scan was performed for the static acquisition. Collected images were further corrected for attenuation, scatter, normalization and camera dead time, followed by co-registration with micro-CT images using the Inveon Research Workplace software under the manufacturer's instructions. Tumor uptake of 18 F-FDG measured in the three-dimensional regions of interest (ROIs) was determined according to the standardized uptake value (SUV). Statistical differences between two groups were analyzed with twotailed Student's t-tests or Mann-Whitney test (for the quantification of ISH and IHC staining), while comparisons among more than two groups were performed using two-sided one-way ANOVA with Dunnett's or Sidak's multiple-comparisons test. All clinical and pathologic data were retrospectively collected from medical records and analyzed. The ISH/IHC staining of tumor tissue samples was quantified by research staff blinded to the clinical data and experimental design. Cutoff value for clinical sample grouping based on the DIO3OS staining scores in AI-treated breast cancer tissues was determined by X-tile software (Version 3.6.1, Yale University). Survival data were plotted as Kaplan-Meier curves and analyzed with log-rank test. The chi-square (χ2) test was used for comparing difference of frequencies among groups. P-value <0.05 was considered statistically significant.
Data availability
The high-throughput sequencing data generated in this study have been deposited in the NCBI-GEO public database under accession code GSE206199 and GSE206142. The isoform-level expression analysis of LDHA, and related sequences were obtained from the ISOexpresso database (http://wiki.tgilab.org/ISOexpresso/) and the UniProtKB database (https://www.uniprot.org/uniprotkb/P00338/entry), respectively. A reporting summary for this article is available as a Supplementary Information file. The remaining data are available within the Article, Supplementary Information or Source Data file. Source data are provided with this paper.
|
v2
|
2022-11-23T15:00:40.171Z
|
2022-11-22T00:00:00.000Z
|
253764344
|
s2orc/train
|
Structural informatic study of determined and AlphaFold2 predicted molecular structures of 13 human solute carrier transporters and their water-soluble QTY variants
Solute carrier transporters are integral membrane proteins, and are important for diverse cellular nutrient transports, metabolism, energy demand, and other vital biological activities. They have recently been implicated in pancreatic cancer and other cancer metastasis, angiogenesis, programmed cell death and proliferation, cell metabolism and chemo-sensitivity. Here we report the study of 13 human solute carrier membrane transporters using the highly accurate AlphaFold2 predictions of 3D protein structures. In the native structures, there are hydrophobic amino acids leucine (L), isoleucine (I), valine (V) and phenylalanine (F) in the transmembrane alpha-helices. These hydrophobic amino acids L, I, V, F are systematically replaced by hydrophilic amino acids glutamine (Q), threonine (T) and tyrosine (Y), thus the QTY code. Therefore, these QTY variant transporters become water-soluble without requiring detergents. We present the superposed structures of these native solute carrier transporters and their water-soluble QTY variants. The superposed structures show remarkable similarity with RMSD ~ 1 Å–< 3 Å despite > 46% protein sequence substitutions in transmembrane alpha-helices. We also show the differences of surface hydrophobicity between the native solute carrier transporters and their QTY variants. Our study may further stimulate designs of water-soluble transmembrane proteins and other aggregated proteins for drug discovery and biotechnological applications.
research, but should also be the targets for new therapeutic strategies to treat pancreatic cancer and perhaps other cancer metastasis 1-4 . There are 46 distinct gene families of SLC transporters comprising 384 genes in the human genome 2,5 . Some members of these SLC transporters have recently been found to be involved in several key aspects of cancer including fatal pancreatic cancer ( Table 1). One of the key characteristics of cancer is deregulation of cellular energetics, namely the insatiable energy demand including sugars and nutrients through upregulating the transporters 1,3,6 . By effectively blocking the key logistics of transport systems to the cancer cells, we may be able to add another tool to more effectively treat caners.
SLC transporters have multiple transmembrane (TM) helices. Most have 10TM-12TM helices depending on their transporting substrates with the longest having 13TM (SLC5A8 for short chain fatty acids). Several of these SLC transporters have large N-termini and extracellular loops (Figs. 1, 2) of ~ 30-60 amino acids. Those with large extracellular loops are more likely to be better targets for generating therapeutic monoclonal antibodies since these extracellular loops may stimulate robust immune responses during immunizations.
On 15 July 2021, Google DeepMind announced the AlphaFold2, and at the same time David Baker's lab introduced RoseTTAFold as machine learning revolutionary tool for the very accurate prediction of protein structures [28][29][30] . AlphaFold2 and, to a lesser extent, RoseTTAFold, have already made an enormous impact on our understanding of 350,000 protein structures. On July 28, 2022, DeepMind released 214 million protein structures, nearly all known protein structures. AlphaFold2 predicts the structures with very high accuracy for 35% of all protein structures (~ 75 million), and with high confidence for another 45% (~ 96.3 million). AlphaFold2 has truly started a new era of digital biology. Nevertheless, academic investigators, the pharmaceutical and biotech companies must still ultimately study the physical structures of proteins, including SLC membrane transporters since the structures are vital to understanding how substrates including sugars, amino acids, essential ions, organic molecules, drugs, or other essential nutrients are transported across highly regulated and controlled cell membranes.
We previously applied the QTY (Glutamine, Threonine, Tyrosine) code to design several detergent-free transmembrane (TM) protein chemokine receptors and cytokine receptors for various uses using conventional computing programs to simulate several G protein-coupled receptors. Each took ~ 5 weeks to complete the simulation [33][34][35] . The expressed and purified water-soluble proteins exhibited predicted characteristics and retained ligand-binding activity [31][32][33][34][35] . In July 2021, we prepared QTY variant protein structure predictions using Alpha-Fold2, achieving better results in 1-2 hours 36,37 , rather than ~ 5 weeks for each molecular simulation using GOMoDo, AMBER and YASARA programs [31][32][33] . We also produced a program and website for generating the membrane protein water-soluble QTY variants 38 .
Here, we report using AlphaFold2 to design water-soluble QTY variants of the 13 solute carrier transporters, and to directly compare with their counterpart native structures. In addition to targeting the key nutrients and ion uptake activity of cancer cells, these QTY variant water-soluble SLC transporters can prospectively find many additional applications. Working with water-soluble QTY variants may substantially accelerate the discovery and development of therapeutic and diagnostic biologicals. Table 2). This is because Q, T, Y amino acids do not introduce any charges, they only introduce water-soluble side chains. Q (glutamine) side chains form 4 water hydrogen bonds, 2 donors through -NH2, and 2 acceptors through oxygen on -C=O; the sidechains -OH of T (threonine) and Y (tyrosine) form 3 water hydrogen bonds, 1 donor from H (hydrogen) and 2 acceptors from O (oxygen). www.nature.com/scientificreports/ Since the electron density maps share remarkable structure similarities between leucine (L) versus glutamine (Q); isoleucine (I), valine (V) versus threonine (T); and phenylalanine (F) versus tyrosine (Y), the QTY code selects 3 neutrally polar amino acids: glutamine, threonine and tyrosine to replace 4 hydrophobic amino acids leucine, isoleucine, valine and phenylalanine. After applying the QTY code, the hydrophobic amino acids in the transmembrane segments are replaced by Q, T, and Y, therefore the transmembrane segments have significantly reduced hydrophobicity. For example, SLC39A3 and SLC29A1 differ > 54 and > 45% from their water-soluble QTY variants, respectively, in their transmembrane alpha-helical segments. (Figs. 1, 2, Table 2).
Other characteristics are also notable. The pIs (isoelectric-focusing points) vary, some in the acidic and some in the basic range. For example, native SLC7A11 has a basic pI of 9.29. On the other hand, SLC1A5 has an acidic pI of 5.34. Others including SLC9A1 have a near neutral pI of 6.74 (Table 2). It is noted that the pIs are identical for the native and QTY variants for SLC4A4 (pI 6.35), SLC39A3 (pI 6.39), SLC4A7 (pI 6.26), and SLC41A1 (pI 5.11) despite the large number of QTY substitutions. The reason is that Glutamine, Threonine, Tyrosine (Q, T, Y) have neither positive nor negative charges at neutral pH. Therefore, substitutions of Q, T, Y do not change the pIs. This is significant because altered pIs could cause non-specific interactions.
Moreover, while there are between > 45-> 54% QTY substitutions in the transmembrane helices, the molecular weights of the native and QTY variants differ by only a few hundreds of Daltons. This is due to a) the substitutions of CH 3 -on Lue and Val, by -OH groups to Gln (Q) and Thr (T), and b) the addition of OH-on Tyr (Y). These increase the protein molecular weights (Figs. 1, 2, Table 2).
The experimentally-determined native structures and their in-silico-determined water-soluble QTY variants superposed within a few Å. Their RMSDs are as follows: SLC2A1 versus SLC2A1 QTY (2.281 Å); for SLC7A11 (Fig. 3, Table 2). It can be seen from Fig. 3, these molecular structures, experimentally-determined and AlphaFold2-predicted visibly superpose very well. These results show that despite > 45% QTY substitutions in the transmembrane alpha-helices in the water-soluble QTY variants, their structures share rather similar 3-dimentional folds. These closely superposed structures perhaps confirm that the AlphaFold2's predictions are highly accurate, since the predicted native Table 3. RMSD between native solute carrier transporters, their water-soluble QTY variants, and crystal structures. Residue mean-square distance (RMSD) in Å, -= not applicable. All RMSD values are below 3 Å and show good superposition between structures. All models and scripts for structure preparation and superposition in PyMOL are presented in the GitHub repository: https:// github. com/ eva-smoro dina/ slc. Because the X-ray crystal or CryoEM structures of the other 7 native SLCxxxx are not yet available, including SLC39A4, SLC39A6, SLC39A3, SLC9A1, SLC6A14, SLC4A7, SLC5A8 and SLC41A1, AlphaFold2 tool is used for the structural predictions. The RMSD in Å (residue mean-square distances) for these superposed structures are displayed (Fig. 4, Table 2 Table 2). The AlphaFold2-predicted structures of both natural SLCs and their water-soluble SLC variants superpose very well implying that they share comparable structures despite the significant substitutions in the transmembrane alpha-helices (45-54%).
Analysis of the hydrophobic surface of native transporters and the water-soluble QTY variants.
It is known that the natural SLCxxxx have high hydrophobicity content, especially in the transmembrane alpha-helical segments. They are inherently insoluble in water and require surfactants to solubilize and stabilize them. These natural transporters quickly self-associate to form unstructured aggregation, precipitation, and no longer biologically functional without the highly selected surfactants.
In the natural SLCxxxx, the 6TM-12TM alpha-helices are directly embedded in the hydrophobic lipid bilayer. The hydrophobic side chains of phenylalanine, isoleucine, leucine, and valine interact with the hydrophobic lipid bilayers. Thus, the 6TM-12TM alpha-helices exhibit water-repelling hydrophobic surfaces (Figs. 5, 6).
After QTY substitutions of hydrophobic amino acids L, I, V, F, with hydrophilic amino acids Q, T and Y, the hydrophobic surfaces are decreased (Figs. 5, 6). The QTY changes hydrophobic 6TM-12TM into hydrophilic 6TM-12TM, without, however, significantly changing the alpha-helical molecular structures, as shown in Fig. 3. Analogous reproducible experimental results reported for chemokine and cytokine receptors in our previous publications [31][32][33][34][35] . However, our experimental results showed that the structure integrity, stability, and ligandbinding activities have been retained from the water-soluble QTY-variant chemokine receptors and cytokine receptors 31-35 . There are three chemically distinct alpha-helix types. Type I: the water-soluble hydrophilic alpha-helix, commonly water-soluble enzymes in the cellular cytosols and extracellular circulating proteins including antibodies, protein and peptide hormones and more; Type II: the water-insoluble hydrophobic alpha-helix commonly in www.nature.com/scientificreports/ transmembrane proteins including hormone receptors, transporters, ion channels, G protein-coupled receptors, photosynthesis systems; and Type III: the amphiphilic alpha-helix, like a Janus that have a hydrophobic face on one side and a hydrophilic face on the other side. These three chemically distinct alpha-helical types have very similar structures, regardless their hydrophobicity and hydrophilicity [45][46][47][48][49] . This is the molecular foundation of the QTY code.
AlphaFold2 predictions.
For over 65 years, scientists have made great efforts to predict protein folding.
It was the dream of structural biologists and protein scientists to predict protein folding rapidly and accurately.
With the advent of AlphaFold2 through machine learning, the tool is now available to predict protein structure by all scientists, almost free of charge. We can now study previously unattainable protein structures, particularly membrane-embedded transmembrane proteins. Systematic bioinformatic studies revealed that in most organisms, ~ 20-30% genes code for membrane proteins 50 . It is known that the human genome codes for ~ 24% membrane proteins 51 . But structural determination of a single transmembrane protein is an extremely difficult process, traditionally requiring even decades of endeavor. There are many barriers, from gene expression, protein production, detergent selection, purification, detergent exchange, to maintaining their long-term stability and integrity as well as maintaining functionality to avoid irreversible aggregation. The numbers of integral transmembrane protein structures lag far behind water-soluble proteins. Recently, several groups have systematically analyzed the known structures and at least 16 different folds of ~ 400 members, 65 families of the solute carrier transporter 52,53 . These studies further provide insights into molecular structures and functions of these transporters.
Applying AlphaFold2 accurate protein structure predictions, we can directly compare the native structure with an AlphaFold2 predicted water-soluble QTY variant. With the QTY variant, expressed in various cells, it becomes possible to overcome the high barriers of studying membrane embedded transmembrane proteins.
We previously reported 36,37 use of the AlphaFold2 tool to predict structures of the water-soluble variants of G protein coupled receptor chemokine receptors and glucose transporters, and compared them to the known experimentally-determined crystal or CryoEM structures.
One of the questions concerns what would be the utility of these water-soluble transporters that can no longer transport molecules cross the membrane since the water-soluble QTY variants can no longer insert themselves into the lipid bilayer membrane. It is plausible that after these transporters are rendered water-soluble, they can be used: (1) as water-soluble antigens to generate useful monoclonal antibodies in animals since they have many extracellular loops including few large loops, and (2) these anti-SLCxxx antibodies could be useful as research reagents for an assay system to study the transporters in tissue cell cultures and in vivo. These specific monoclonal www.nature.com/scientificreports/ antibodies can perhaps also be useful (a) as therapeutics to treat diseases including pancreatic cancer, (b) as diagnostic reagent to monitor cancer treatments and perhaps (c) for early pancreatic cancer detection. Our current study using the AlphaFold 2 demonstrates that the water-soluble QTY-variant structures of SLC transporters are substantially similar to the native structures. AlphaFold 2 is a very useful approach to predict other membrane embedded transmembrane proteins. QTY is a useful approach for working with difficult-tostudy hydrophobic proteins. The SLC transporter water-soluble QTY variants not only could be used for designs of molecular machines, but also as water-soluble antigens for generating therapeutic monoclonal antibodies and for accelerating drug discovery.
Methods
Protein sequence alignments and other characteristics. The native protein sequences for SLC transporters and their QTY-variant sequences are aligned using the same methods previously described 33,34 . The website Expasy (https:// web. expasy. org/ compu te_ pi/) was used to calculate the molecular weights (MW) and pI values of the proteins.
Ethical approval. (1) All methods were carried out in accordance with relevant guidelines and regulations.
(2) All experimental protocols were approved by a named institutional and licensing committee. (3) Neither human biological samples, nor human subjects were used in the study. This is a completely digital structural biology study using the publicly available AlphaFold2 machine learning program.
|
v2
|
2022-11-23T15:25:28.542Z
|
2022-11-22T00:00:00.000Z
|
253798162
|
s2orc/train
|
MIR4435-2HG in exosomes promotes gastric carcinogenesis by inducing M2 polarization in macrophages
Gastric cancer (GC) is a cancer with a high mortality rate. lncRNAs play a role in regulating GC tumorigenesis. In this paper, we analyzed differentially expressed lncRNAs between GC and adjacent normal tissues using multiple bioinformatics tools to identify new potential targets in GC. Cell viability and migration ability were detected using the Cell Counting Kit-8 (CCK-8) and transwell assays, MIR4435-2HG was negatively correlated with the survival rate of GC patients, and by inhibiting the activity of MIR4435-2HG, the viability and migration ability of GC cells could be reduced. In addition, RT- qPCR and western blot to detect gene and protein level expression, transmission electron microscopy and nanoparticle tracking analysis (NTA) to study the efficiency of exosome isolation, and flow cytometry to observe cell differentiation were employed, delivery of MIR4435-2HG shRNA via MKN45 cell-derived exosomes significantly reversed the MKN45 exosome-induced M2 polarization in macrophages. Furthermore, the low expression of MIR4435-2HG in MKN45 cell-derived exosomes inhibited the Jagged1/Notch and JAK1/STAT3 pathways in macrophages; MIR4435-2HG downregulated exosomes were found to significantly inhibit GC tumor growth in vivo by establishing a mouse model. In short, MKN45 cell-derived exosomes deliver lncRNA MIR4435-2HG, which promotes gastric carcinogenesis by inducing macrophage M2 polarization.
Introduction
Gastric cancer is one of the most common high-mortality cancers. Accounting for 8% of all cancer types in terms of both cancer incidence and mortality (1,2). The main cause of its poor prognosis is distant metastasis, which accounts for approximately 90% of cancer-related deaths (3). Surgical resection with or without chemotherapy is usually used clinically for GC, but treatment outcomes have not met expectations so far (4). Therefore, it is an important and urgent matter to find the pathogenesis of GC-related molecules and promote targeted therapy.
Long-stranded noncoding RNAs are noncoding RNAs greater than 200 bp in length with limited protein-coding potential (5), and several existing findings have shown that lncRNAs play a role in epigenetic, transcriptional, and posttranscriptional levels, as well as in cancer development (6,7), and lncRNAs play a key role in gastric carcinogenesis (8,9).
Exosomes are a type of vesicle secreted from the mammalian intracellular to the extracellular, consisting of membranes of multivesicular bodies, which can participate in cellular communication by transferring proteins and nucleic acids (10- 13) and are an efficient drug carrier in clinical applications (14). Exosomes can play a role in tumorigenesis through communication with tumor cells and macrophages and exhibit immunosuppressive effects to some extent (15,16). In contrast, macrophages (especially M2 macrophages) can suppress the immune response in tumor development and thus can promote drug resistance in cancer therapy (17,18), which is widely used in clinical practice. Small interfering RNA (siRNA) is a therapeutic agent that can treat a variety of diseases, but safe, efficient, and targeted delivery of siRNA is still one of the main challenges (19). This paper focuses on exosome-targeted delivery of siRNA.
In this study, we investigated the differentially expressed lncRNAs between GC and adjacent normal tissues to find GC-associated lncRNAs and investigated the exosomal transport of siRNA, which could represent a breakthrough for GC clinical treatment and research. We found that MIR4435-2HG in exosomes is a key factor in regulating the promotion of macrophage M2 polarization and mapped the macrophage M2 polarization pathway based on this theory (Figure 1).
Materials and methods
GC lncRNA differential expression analysis R software limma package. The results are presented in volcano plots with a threshold of |log2FC| > 1, P < 0.05. The most significantly differentially expressed gene was selected as the key lncRNA. The R software ggplot2 package was employed to analyze the differential expression of this lncRNA in tumor nd normal samples. The results are illustrated by box plots. P < 0.05 was statistically significant.
Cell transfection
MIR4435-2HG shRNA1 and shRNA2, nontargeting sequences from Hanbio Biotechnology Co., Ltd, and OE from Shanghai GenePharma Co., Ltd were packaged into lentiviral vectors. The lentiviral vector DNA was transfected into 293T cells and incubated at 37°C after transfection. The supernatant was filtered into pellets, and the lentiviral pellets were infected with GC cells. qPCR was used to verify the transfection efficiency.
Extraction and identification of exosomes
Cells were first cultured in a complementary medium, and when the cells reach 80% fusion with the culture medium, the medium was replaced with an FBS-free medium for 2 days, the supernatant was collected by centrifugation and filtration (Millipore, USA), and the supernatant was collected by ultracentrifugation (Beckman Coulter) to obtain exosomes. The subgroups were named KN45-Exo-NC, MKN45-Exo-MIR4435-2HG OE or MKN45-Exo-MIR4435-2HGshRNA2, respectively. Exosomes were identified using transmission electron microscopy and immunoblotting.
Reverse transcription-quantitative PCR (RT-qPCR)
The total RNA of samples was extracted from cells using TRIzol reagents and reverse transcribed into cDNA using the PrimeScript RT kit, and qPCR was performed using the SYBR premix Ex Taq II kit to detect gene expression. The expression of MIR4435-2HG, arginase-1, and iNOS genes was measured using b-actin as an internal reference ( Table 1). The relative expression of the genes was calculated using the 2-DDCT method.
Immunofluorescence
Cells were inoculated into 24-well plates overnight, prefixed with 4% paraformaldehyde for 10 min, and then fixed in methanol for 10 min. Cells were stained with Phalloidin-iFluor 488 reagent (ab176753) and the PKH26 Red Cell Membrane Staining Kit (D0030, Solarbio), respectively. Phalloidin-iFluor 488 stains the cytoskeleton and PKH26 stains the cell membrane, enabling localization of the exosome.
Cell migration assay
GC cells were inoculated in the upper chamber of the medium containing 1% FBS and the density was adjusted to approximately 1.0×10 6 cells per chamber. RPMI 1640 medium containing 10% FBS was added to the lower chamber. After incubation at 37°C for 24 h, the transfer chamber was rinsed twice (5 min each time) with PBS. The cells were fixed with 5% glutaraldehyde at 4°C, stained with 0.1% crystal violet for 30 min, and the transwell chambers were washed twice with PBS, followed by observation under a microscope. The number of migrating cells was considered to be a reflection of migratory capacity.
Statistical analysis
Three independent replicate experiments were performed for each group of samples, and the results are presented as mean ± standard deviation (SD). Statistical methods included the t-test (with 2 groups) and one-way ANOVA (with 3 or more groups), and Tukey's test was used. P < 0.05 was statistically significant.
GC lncRNA differential expression results
Through differential analysis of GC and normal samples next to cancer, 847 genes were obtained, including 636 upregulated genes and 211 downregulated genes (Figure 2A). The expression results of MIR4435-2HG in tumors compared with normal samples showed that MIR4435-2HG was highly expressed in GC tissues ( Figure 2B).
Overexpression of MIR4435-2HG significantly promoted the proliferation of gastric cancer cells
The expression of MIR4435-2HG in various cells was detected using the RT-qPCR method. The results showed that MIR4435-2HG expression was significantly higher in MKN-45 and AGS cells than in GES-1 cells ( Figure 3A). In addition, MIR4435-2HG OE significantly upregulated MIR4435-2HG expression in MKN-45 and AGS cells ( Figure 3B) but was suppressed in the presence of MIR4435-2HG shRNA ( Figure 3C), and GC cells were more sensitive to MIR4435-2HG shRNA2. MIR4435-2HG shRNA2 was chosen as the subject for the follow-up study. High MIR4435-2HG expression could promote the proliferation of MKN-45 and AGS cells (Figures 3D, E). In summary, overexpression of MIR4435-2HG significantly promoted the proliferation of GC cells.
Overexpression of MIR4435-2HG increases the migratory ability of GC cells
Cell migration was examined using the transwell assay, and the results showed that overexpression of MIR4435-2HG significantly increased the migration of GC cells, whereas downregulation of MIR4435-2HG inhibited the migration of GC cells (Figures 4A, B). In addition, the expressions of N-cadherin and vimentin in GC cells were significantly upregulated by MIR4435-2HG overexpression but downregulated in the presence of MIR4435-2HG shRNA. In contrast, silencing of MIR4435-2HG significantly upregulated the E-cadherin protein level, whereas upregulation of MIR4435-2HG had an inhibitory effect on E-cadherin ( Figure 4C). In conclusion, overexpression of MIR4435-2HG increased the migration of GC cells by promoting the EMT process.
Results of exosome extraction and identification
Round particles with diameters of 30 to 150 nm were observed by TEM, and nanoparticle tracking analysis (NTA) yielded a size distribution close to that of TEM ( Figure 5A). The expression of exosomal proteins (CD63 and TSG101) was significantly higher in GC cell exosomes compared with GC cells ( Figure 5B). The expression of MIR4435-2HG in MKN45-Exo was significantly higher than that of GES-1-Exo ( Figure 5C), whereas MIR4435-2HG overexpression or knockdown in MKN45 cell exosomes had a limited effect on exosomal protein levels ( Figure 5D). MIR4435-2HG OE significantly upregulated MIR4435-2HG expression in MKN45 cells and MKN45 cell exosomes, and MIR4435-2HG shRNA2 significantly suppressed its expression ( Figure 5E). Tumor-derived exosomes labeled with fluorescent PKH26 were internalized by unstained macrophages when co-cultured with macrophages ( Figure 5F). Exosomes from MKN45 cells delivered MIR4435-2HG to PMA-treated THP-1 cells ( Figure 5G). Exosomes were successfully isolated from GC cells.
Overexpression of exosomal MIR4435-2HG promotes M2 polarization in macrophages
To find the immunomodulatory effect of MKN45 cellderived exosomes on macrophages, THP-1 cells were treated with 50 mg/ml of exosomes excreted from MKN45 cells. The distribution of CD86 (M1 phenotype marker) in THP-1 cells was significantly reduce, and THP-1 cells and macrophages exhibited the CD86low/CD206 high phenotype when cells were incubated with exosomes containing high levels of MIR4435-2HG. Additionally, exosomes from THP-1 cells significantly upregulated the distribution rate of CD206 (M2 marker) in THP-1 cells, which was partially reversed by exosomes downregulated with MIR4435-2HG ( Figures 6A, B). Exosomes of MKN45 origin significantly upregulated arginase-1 levels, whereas exosomes downregulated with MIR4435-2HG reversed the effect of exosomes from MKN45 cells, and the exosomes had a limited effect on iNOS expression ( Figure 6C). The expression of M2 markers (IL-10 and TGF-b) was significantly increased in macrophages incubated with MKN45 cell-derived exosomes compared to the PBS group, which was further exacerbated by MIR4435-2HG upregulated exosomes, and conversely, MIR4435-2HG downregulated exosomes partially reversed the effect of MKN45 cell-derived exosomes ( Figure 6D). In summary, overexpression of exosomal MIR4435-2HG promotes M2 polarization in macrophages. FIGURE 4 Overexpression of MIR4435-2HG increased the migration of gastric cancer cells. (A, B) The migration of gastric cancer cells was measured by the transwell assay. (C) The protein levels of E-cadherin, N-cadherin, and vimentin in GC cells were detected by western blot. **P < 0.01 compared with the control group.
Exosome-induced M2 macrophages promote the migration of GC cells by facilitating the EMT process
Tumor cell-derived exosome-induced M2 macrophages significantly increased the migration of GC cells. Meanwhile, MIR4435-2HG upregulated exosomes aggravated this phenomenon, whereas MIR4435-2HG downregulated exosomes reversed it ( Figure 7A). MKN45 cells exhibited a spindle-shaped morphology when co-cultured with M2 macrophages, which was reversed by MIR4435-2HG downregulated exosomes ( Figure 7B). N-cadherin and vimentin expression was significantly increased in MKN45 cells after co-culturing with exosome-induced M2 macrophages, whereas MIR4435-2HG downregulated exosomes caused significant downregulation of N-cadherin and vimentin. The opposite data were obtained on E-cadherin expression ( Figure 7C). In conclusion, exosome-induced M2 macrophages promote the migration of GC cells by increasing the EMT process.
Upregulation of MIR4435-2HG in exosomes significantly promotes tumor growth in GC
Results from xenograft mouse models showed that exosomes from MKN45 cells significantly increased tumor size in nude mice, which was partially reversed by exosomes downregulated by MIR4435-2HG (Figures 9A, B). Exosomes from MKN45 cells greatly increased tumor weight in nude mice, which was partially reversed by exosomes downregulated by the MIR4435-2HG phenomenon ( Figure 9C). MKN45 cell-derived exosomes significantly upregulated N-cadherin and vimentin expression in mouse tissues, whereas MIR4435-2HG downregulated exosomes reversed this phenomenon. Additionally, MKN45 cell-derived exosomes inhibited E-cadherin expression in nude mice, whereas MIR4435-2HG OE upregulated exosomes further enhanced the role of exosomes ( Figure 9D). Taken together, MIR4435-2HG upregulation in exosomes induced M2 polarization in macrophages, which significantly promoted the growth of GC tumors.
Discussion
RNA interference is a phenomenon that can reverse the silencing of any gene, and clinical delivery materials are generally selected to transport siRNA to sites of action in target tissue cells (20). Exosomes, also known as signalosomes, can be expressed on exosomal membranes or packaged through ligands and adhesion molecules, thereby performing signaling (21). In this study, we focused on the role and potential value of MIR4435-2HG in GC, starting from the exosomes of GC cells.
The results showed that MIR4435-2HG upregulation could promote tumorigenesis in GC, and a previous study found that B A FIGURE 8 Exosomes from MKN45 cells promote M2 polarization in macrophages by regulating the Jagged1/Notch and JAK1/STAT3 axes. (A) The Jagged1, Notch1, Notch2, Hes1, and Hes5 protein levels in macrophages were detected by WB. (B) WB detection of p-JAK1, JAK1, p-STAT3, and STAT3 protein levels in macrophages.
FIGURE 7
Exosome-induced M2 macrophages promote the migration of gastric cancer cells by inhibiting the EMT process. (A) MKN45 cells were cocultured with macrophages, ExoNC-treated macrophages, Exo-MIR4435-2HGOE-treated macrophages, and Exo-MIR4435-2HGshRNA2treated macrophages for 24 h, and the migration of MKN45 cells was detected by the transwell method. (B) The morphology of MKN45 cells was observed under a microscope. (C) The E-cadherin, N-cadherin, and vimentin protein levels in macrophages were detected by western blot. **P < 0.01. (22), which was consistent with our findings. In addition, we found that exosomal delivery of MIR4435-2HG from GC cells promotes the polarization of M2 macrophages, which in turn promotes GC development, and we speculate that exosomal MIR4435-2HG might act as an oncogene in GC and can be considered a marker of cancer tumorigenesis. Some studies have found that MIR4435-2HG is associated with various cancers, such as cervical cancer, GC, and renal clear cell carcinoma (23), which is consistent with our speculation. We also found that exosomes might increase migration of GC cells in vitro and in vivo when they induce selective activation of macrophages toward the M2 phenotype, and we hypothesized that macrophage M2 polarization might promote gastric carcinogenesis. One study found that macrophage M2 polarization can induce colorectal carcinogenesis by secreting CXCL13 (24). In addition, exosomes can be considered important mediators of the tumor microenvironment, and they can act as important messengers to regulate the cross-talk between different cells. Jagged1 (JAG1) is an important Notch ligand, and JAG1/Notch signaling controls oncogenic processes in multiple cell types and is related to poor clinical prognosis (25). We found that exosomes from MKN45 cells can promote M2 polarization in macrophages by regulating the Jagged1/Notch or JAK1/STAT3 axis.
MIR4435-2HG could promote tumorigenesis in GC cells
The JAK1/STAT3 axis is associated with phosphate metabolism, phosphorylation, and nuclear accumulation of STAT3 when JAK1 expression is upregulated (26), and the JAK1/STAT3 signaling pathway has been found to promote M2 polarization in macrophages (27), which is similar to our findings. The JAK1/STAT3 signaling pathway might be an M2 macrophage polarization-mediating factor.
In conclusion, we found that MKN45 cell-derived exosomes could induce M2 macrophage polarization and promote GC tumorigenesis; therefore, MIR4435-2HG could be used as a biomarker for GC and a new target for GC therapy.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.
|
v2
|
2022-11-23T15:32:42.051Z
|
2022-11-22T00:00:00.000Z
|
253798257
|
s2orc/train
|
Macular hole and vitreous hemorrhage subsequent to stereotactic hypofractionated radiotherapy for choroidal melanoma: A case report and review of the literature
Choroidal melanoma is the leading primary intraocular tumor with potentially fatal outcomes in adults. The coexistence of choroidal melanoma and a macular hole is extremely rare, and treatment strategies and information on the prognosis of associated complications are currently lacking. We report the first case of choroidal melanoma complicated with a macular hole and vitreous hemorrhage after stereotactic hypofractionated radiotherapy in Japan, and review the relevant literature in relation to the possible mechanisms, treatment strategies, and outcomes. An 83-year-old male with choroidal melanoma was treated with stereotactic hypofractionated radiotherapy in January 2021. Five months later, a full-thickness macular hole developed, followed by an acute massive vitreous hemorrhage about 2 weeks later. Following confirmation of tumor regression, the patient underwent a pars plana vitrectomy and internal limiting membrane peeling. The macular hole was closed postoperatively and the patient’s best-corrected visual acuity improved to 20/125. There was no evidence of intraocular tumor dissemination or distant metastases during follow-up. A systematic literature search only identified 10 previous cases of choroidal melanoma with a macular hole in eight reports worldwide, mainly in females. Macular edema may be the primary cause of macular hole formation in these cases. Most patients who underwent vitrectomy for complications after tumor regression achieved a good prognosis. The development of a macular hole is a rare complication associated with choroidal melanoma. Anterior-posterior traction of posterior vitreous detachment and secondary macular edema may have contributed to the formation of the macular hole in the current case.
Introduction
Choroidal melanoma is the leading primary intraocular malignancy among adults (1), with a low incidence of 0.6 cases per million per year in Japan (2). However, considering the high mortality rate of malignant metastases, this lifethreatening disease should be diagnosed and treated promptly. Radiation therapy, including plaque brachytherapy, proton beam radiotherapy, and stereotactic radiotherapy, is an alternative to enucleation and has become the first-line treatment for choroidal melanoma (3)(4)(5). However, the tumor may be accompanied by complications, such as vitreous hemorrhage, rhegmatogenous retinal detachment, and macular hole (MH). Care is therefore needed to prevent intraocular or extraocular tumor dissemination during therapy for these complications (6).
Choroidal melanoma coexisting with a MH is extremely rare. To the best of our knowledge, only 10 previous cases have been reported worldwide (6-13), none of which occurred after stereotactic hypofractionated radiotherapy, and with limited information on the treatment of associated complications. Herein, we report on a patient who was diagnosed with asymptomatic choroidal melanoma with atypical presentation, and who developed a full-thickness MH and vitreous hemorrhage during follow-up, which was eventually repaired by pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peeling. We also reviewed the relevant literature regarding the possible mechanisms of MH formation in patients with choroidal melanoma, and the corresponding treatment management and outcomes.
Case presentation
An 83-year-old man was referred to our hospital with suspected serous retinal detachment in his left eye. The patient's clinical course is presented in Figure 1. The best-corrected visual acuity (BCVA) was 20/20 in his right eye and 20/17 in his affected left eye. The intraocular pressure was normal (14 mmHg in the right eye and 13 mmHg in the left eye), and there were no appreciable findings in the anterior segments. Ultra-wide-field fundus photography (Figure 2A) of the left eye revealed an elevated choroidal mass with a central dark brown speckle in the nasal quadrant, about 4 disc diameters from the optic disc, along with concomitant posterior vitreous detachment (PVD). Fluorescein angiography showed that the choroidal mass had early diffuse hyperfluorescence with a central area of hypofluorescence ( Figure 2B). Indocyanine green angiography showed blocked fluorescence due to the choroidal mass and a small hyperfluorescent area at the margin in the late stage ( Figure 2C). Magnetic resonance imaging demonstrated a tumor measuring 5.9×5.7 mm in basal dimensions and 4.1 mm thick, with a hyperintense signal toward the vitreous cavity on axial T1 imaging-fast spin-echo ( Figure 2D). Iodine-123 isopropyl iodoamphetamine brain single-photon emission computed tomography revealed high focal uptake in his left eye, corresponding to the choroidal tumor ( Figure 2E). The patient underwent integrated positron emission tomography/ computed tomography, and a transaxial section across the left eye revealed no fluorodeoxyglucose activity and no evidence of distant metastases. There were no abnormalities in the fellow eye.
The patient received a course of stereotactic hypofractionated radiotherapy (60 Gy in 5 fractions) for 5 consecutive days after the clinical diagnosis of choroidal melanoma. Five months later, the patient complained of visual deterioration with a BCVA of 20/50 and distortion in his left eye. Fundus examination and optical coherence tomography showed a full-thickness MH (stage 4) with cystic cavities ( Figure 3A). Approximately 2 weeks later, his BCVA had decreased to 20/2000, attributed to an acute massive vitreous hemorrhage ( Figure 3B). Repeat positron emission tomography/computed tomography examination showed no significant abnormalities or metastases. We therefore performed a vitrectomy and inverted ILM peeling. During surgery, we found Clinical course of the patient with choroidal melanoma. UWF, ultra-wide-field fundus photography; OCT, optical coherence tomography; FA, fluorescein angiography; ICGA, indocyanine green angiography; MRI, magnetic resonance imaging; PET-CT, positron emission tomographycomputed tomography; SPECT, single-photon emission computed tomography; CFP, color fundus photography; BCVA, best-corrected visual acuity; LP, light perception; PPV, pars plana vitrectomy; ILM, internal limiting membrane. a massive subretinal hemorrhage, abundant fibrin, and retinal fragility but no obvious tears in his left eye. Gas-fluid exchange was completed at the end of surgery using 20% sulfur hexafluoride. After the first vitrectomy, the MH was closed on optical coherence tomography examination. However, the vitreous hemorrhage reappeared 2 weeks later and we performed a second vitrectomy after 4 months of observation. The MH remained closed after the two procedures ( Figure 3C), the retina remained attached, the tumor displayed marked regression, and the BCVA had improved to 20/125.
Review of the literature
We conducted a literature review by searching the PubMed, Cochrane Library, and Web of Science databases using the keywords ("choroidal melanoma" OR "uveal melanoma") AND ("macular hole" OR "retinal tear"), ("uveal neoplasms" OR " choroidal neoplasms") AND ("macular hole" OR "retinal tear"), for articles published from December 1951 to March 2022. The search was limited to publications in English. We reviewed the abstracts and full texts of the identified articles and the related references. Simultaneous occurrence of choroidal melanoma and MH was reported in 10 patients in eight publications (6-13), after excluding one case with MH in which it was difficult to determine the intraocular tumor type (14) and one case in which MH was considered a postvitrectomy complication (15). The findings of the literature review and the current case are presented in Table 1.
The total of 11 patients included four males and seven females, with an average age of 68.1 ± 13.2 years (range, 45-84 years). Five cases were reported in America, three in Europe (England, Italy, Switzerland), and two in West Asia (India, Israel). The current case was first documented in Japan (East Asia). The mean tumor-base diameter was 10.2×9.6 mm and the mean thickness was 4.9 mm. Most cases had a melanotic appearance, but the present case appeared amelanotic. Regarding the treatment of the tumor, one case underwent enucleation due to the large size of the tumor (height 13 mm) and total retinal detachment (10), one case underwent transscleral local resection (13), and the remaining cases were t r e a t e d wi t h r a d i o t h e r a p y , i nc l u d i n g st e r e o t a c t i c hypofractionated radiotherapy (the only treatment in the current case). MH developed in five cases after radiotherapy and in one case after resection, and was observed in four cases at melanoma diagnosis. The most common concomitant manifestations during the follow-up period were retinal detachment and macular edema; the present case was the only one in which PVD and massive vitreous hemorrhage were observed. Five cases with MH were repaired with PPV and ILM peeling and achieved good results, except for the absence of detailed information on the macular prognosis in one case (6) and one case that was unsuccessfully repaired by PPV and eventually underwent enucleation (13). Two cases that underwent observation (7) and subtenon triamcinolone acetonide injection (11), respectively, ended in failure of macular closure. One case reported by Gold et al. died of suspected metastatic disease (7), but no recurrence or metastases were reported in the remaining cases.
Discussion
Uveal melanoma is a severe intraocular malignancy with an elevated gray or gray-brown appearance, which predominantly occurs in Caucasians and is generally complicated with exudative retinal detachment and occasional vitreous hemorrhage (1,16). Rarely, MH can develop, and appropriate procedures must be followed to avoid metastasis (7). We presented a case with an atypical manifestation of choroidal melanoma with MH and subsequent vitreous hemorrhage in a patient who achieved a good prognosis after treatment.
Full-thickness MH represents an anatomical defect in the fovea involving interruption of all the retinal layers from the ILM to the retinal pigment epithelium (17). Notably, MH is more prevalent in women over the age of 60 years, due to hormonal influences (18), which may explain the high proportion of women in the above case series. There are several hypotheses regarding the co-occurrence of melanoma and MH. Typically, the development of MH in the retina has been attributed to anterior-posterior traction, commonly induced by PVD (19). In our case, PVD was detected by fundus observation at the first diagnosis, with no obvious retinal break. Although the specific etiology was unclear, we presumed that the PVD might have been caused by chronic tumor-related vitreous inflammation or floating blood cells, which could induce vitreous condensation, liquefaction, and final separation from the retina. In elderly patients, PVD may be caused by an inevitable, complex series of events such as synchysis and syneresis in the vitreous (17). It is therefore also possible that normal age-related PVD with secondary MH may have occurred coincidentally with melanoma in our case. In addition, tangential traction possibly resulting from an epiretinal membrane or increasing tumor height/thickness causing a lateral shift of vitreomacular traction may also play an important role in the pathogenesis of MH (20), as in a prior case report (7).
However, degenerative etiologies such as cystoid macular edema (CME) and secondary rupture of cysts may be responsible for the formation of MH, as seen in four previous FIGURE 3 Fundus appearance in the patient with choroidal melanoma before and after vitrectomy. (A) Five months after radiotherapy, a full-thickness macular hole with cystic cavities was shown on optical coherence tomography. (B) A further 2 weeks later, a massive vitreous hemorrhage appeared on ultra-wide-field fundus photography. (C) The macular hole was successfully closed after pars plana vitrectomy with internal limiting membrane peeling. reports (7,8,10,11). Retinal degeneration overlying the tumor, intraretinal edema secondary to chronic exudative retinal detachment, or an inflammatory cellular reaction to the necrotic tumor in the vitreous may lead to the development of CME in patients with melanoma (21). In one case in which peripheral melanoma was associated with CME, trypsin digest preparation of the intraretinal space demonstrated abnormal capillary architecture, which was thought to account for vascular leakage (22). CME may have been caused by an increase in capillary permeability during inflammation, resulting in MH (8).
Similarly, we detected cystic cavities in the current case, suggesting that tractional forces followed by retinal tissue degeneration at the macula may have facilitated the formation of the MH. Stereotactic hypofractionated radiotherapy, iodine-125 plaque brachytherapy, and transpupillary thermotherapy (TTT) are possible options for globe-salvaging treatment in patients with peripherally located tumors (3, 23-25). However, ocular complications such as cataract, glaucoma, maculopathy, and optic neuropathy require prompt attention (26). Mashayekhi et al. reported that most cases of atrophic retinal holes in the TTT-treated area occurred within 6 months after treatment, while retinal atrophy was much less prominent in patients treated with plaque radiotherapy or stereotactic hypofractionated radiotherapy (27,28). Heat-induced vitreous changes in TTT may lead to vitreoretinal traction or retinal atrophy, which may in turn explain the formation of a retinal hole (27). Balestrazzi et al. described a case of MH that occurred 3 months after TTT in a patient with melanoma. However, given the distance between the tumor and the macula, TTT was considered unlikely to have caused the MH in this case (9). In another case report, Beykin et al. observed an atrophic MH in close proximity to a melanoma after plaque radiotherapy (12). More than 50% of patients in one study cohort suffered lateonset radiation retinopathy 5 years after stereotactic hypofractionated radiotherapy (28). Another study found that the distance from the fovea to the tumor was the primary determinant of maculopathy in patients undergoing radiotherapy (26). In the current case, MH was observed 5 months after stereotactic hypofractionated radiotherapy, and the melanoma site was distant from the macula, thus ruling out the possibility of radiation-induced MH. However, long-term complications still require cautious evaluation in this case.
The incidences of vitreous hemorrhage in patients with uveal melanoma treated with plaque radiotherapy were 15.1% at 5 years and 18.6% at 10 years (29). Radiation can lead to fibrosis and necrosis of tumor tissue, as well as thinning and fragility of the retina, thus increasing the risk of bleeding. Radiationinduced tumor necrosis is most commonly linked to vitreous hemorrhage in melanoma-affected eyes after radiotherapy, followed by proliferative radiation retinopathy and PVD (29). Combined with the surgical finding of bleeding from the tumor surface, we considered that the hemorrhage in the present case was a consequence of acute ischemic shrinkage of the tumor after stereotactic hypofractionated radiotherapy and vascular rupture within the tumor. However, it is important to note that the occurrence of vitreous hemorrhage before melanoma treatment should raise concerns about possible tumor invasion through Bruch's membrane and diffuse intraocular tumor dissemination (6).
Limited information is available on the treatment of MH and vitreous hemorrhage in eyes with choroidal melanoma. Hypofractionated stereotactic radiotherapy with 50-70 Gy in five fractions or plaque brachytherapy has recently proven sufficient to preserve the eyeball and achieve excellent local tumor control in patients with choroidal melanoma (28,30). Among these 11 cases we reviewed, two cases did not explicitly state whether a metastasis occurred and one case was unclear whether the macular hole developed before or after radiation treatment, but was lost to follow-up and metastatic disease was presumed. In the remaining cases, no metastases were found during follow-up. Besides, almost all eyes with treated melanoma had a good prognosis in terms of the MH after PPV and ILM peeling treatment. Beykin et al. retrospectively evaluated six patients with radiation-treated choroidal melanoma who developed retinal detachment and one who developed MH, all of whom underwent PPV and ultimately had attached retinas (12). During a 5-year follow-up period, Bianciotto et al. revealed that the resolution rate of vitreous hemorrhage in regressed melanoma eyes was as high as 72% after vitrectomy, and PPV did not increase the risk of tumor recurrence or distant metastasis, with low rates of 3% and 5%, respectively (29). Exceptionally, Foster et al. reported one patient with vitreous hemorrhage before tumor treatment who unfortunately developed intraocular tumor spread after PPV, while the remaining eight patients with tumor regression developed complications including vitreous hemorrhage, MH, or retinal detachment, but showed no tumor spread following PPV (6). Therefore, we consider the complication that happened before tumor treatment will increase the risk of metastasis. In contrast, metastasis is comparatively low if a complication occurs following tumor remission. The timing of vitrectomy in our case was 6 months after tumor radiotherapy and the shortest interval was 3 months in a previous case; with no evidence of tumor dissemination in either case during follow-up (9). The conservative interval for vitrectomy after tumor treatment is unclear, but definite tumor regression should be confirmed before carrying out vitrectomy or other intraocular surgery. Furthermore, direct contact with the tumor or direct instrument interaction should be minimized and all steps should be carried out carefully during surgery.
The expectation of visual improvement also needs to be considered, especially in patients with chronic MH. Two cases of choroidal melanoma still had poor vision after MH repair surgery (7,12). Conversely, another case who developed MH 3 months after TTT had improved visual acuity from hand motion to 20/80 after timely vitrectomy (9), similar to the current case. PPV after confirmed tumor regression thus seems to be a feasible and effective treatment in terms of anatomical and visual success, and may be more beneficial in cases with newly developed MH. A recent case report notably demonstrated extraocular extension of a brachytherapy-treated choroidal melanoma following PPV and scleral buckle for rhegmatogenous retinal detachment (31). Combined with previous reports, PPV, especially with ILM peeling, might increase the risk of tumor recurrence or migration of tumor cells, but the risks of these surgical complications in patients with regressed tumors are probably low (6). However, minimizing visual loss and preventing metastasis of malignant tumors still require careful assessment to balance the risks and benefits.
In conclusion, there have been few reports of choroidal melanoma complicated with MH and vitreous hemorrhage in the literature. Vitrectomy seems to be feasible for repairing MH in patients with regressed tumors. However, the occurrence of complications after intraocular tumor treatment and the safety of vitrectomy for these complications require longer follow-up and cautious management.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
|
v2
|
2022-11-23T15:33:34.385Z
|
2022-11-22T00:00:00.000Z
|
253797901
|
s2orc/train
|
Using inflammatory indexes and clinical parameters to predict radiation esophagitis in patients with small-cell lung cancer undergoing chemoradiotherapy
Objective Radiation esophagitis (RE) is a common adverse effect in small cell lung cancer (SCLC) patients undergoing thoracic radiotherapy. We aim to develop a novel nomogram to predict the acute severe RE (grade≥2) receiving chemoradiation in SCLC patients. Materials and methods the risk factors were analyzed by logistic regression, and a nomogram was constructed based on multivariate analysis results. The clinical value of the model was evaluated using the area under the receiver operating curve (ROC) curve (AUC), calibration curves, and decision curve analysis (DCA). The correlations of inflammation indexes were assessed using Spearman correlation analysis. Results Eighty-four of 187 patients (44.9%) developed grade ≥2 RE. Univariate analysis indicated that concurrent chemoradiotherapy (CCRT, p < 0.001), chemotherapy cycle (p = 0.097), system inflammation response index (SIRI, p = 0.048), prognostic-nutrition index (PNI, p = 0.073), platelets-lymphocyte radio (PLR, p = 0.026), platelets-albumin ratio (PAR, p = 0.029) were potential predictors of RE. In multivariate analysis, CCRT [p < 0.001; OR, 3.380; 95% CI, 1.767-6.465], SIRI (p = 0.047; OR, 0.436; 95% CI, 0.192-0.989), and PAR (p = 0.036; OR, 2.907; 95% CI, 1.071-7.891) were independent predictors of grade ≥2 RE. The AUC of nomogram was 0.702 (95% CI, 0.626-0.778), which was greater than each independent predictor (CCRT: 0.645; SIRI: 0.558; PAR: 0.559). Calibration curves showed high coherence between the predicted and actual observation RE, and DCA displayed satisfactory clinical utility. Conclusion In this study, CCRT, SIRI, and PAR were independent predictors for RE (grade ≥2) in patients with SCLC receiving chemoradiotherapy. We developed and validated a predictive model through these factors. The developed nomogram with superior prediction ability can be used as a quantitative model to predict RE.
Introduction
Lung cancer is one of the most common malignant tumors and the leading cause of cancer death globally (1). Small cell lung cancer (SCLC) accounts for 15% of all lung cancers characterized by rapid doubling time, early metastasis, and poor prognosis (2). According to the Veterans Administration Lung Cancer Study Group Classification, SCLC can be divided into limited-stage small cell lung cancer (LS-SCLC) and extensive-stage small cell lung cancer (ES-SCLC). Because small cell lung cancer is difficult to diagnose early, only 2-5% of patients can be treated by surgery. And SCLC is sensitive to chemotherapy and radiotherapy, so the primary treatment for most patients is chemoradiotherapy (3). Nevertheless, radiation esophagitis (RE) is a common adverse reaction in SCLC patients who receive radiation therapy. In radiotherapy for lung cancer, it is impossible to avoid esophageal irradiation completely because of several factors: large, irregular shape and central location of lung cancer, often involving mediastinal lymph nodes (LN), and the central location and length of the esophagus (4,5). Despite advances in radiotherapy technology, RE is still a common side effect in patients receiving chemoradiotherapy.
RE usually occurs within two months from the beginning of radiotherapy to the end of radiotherapy. The typical symptoms of RE are dysphagia, retrosternal pain, burning, and other symptoms. Some patients may also have esophageal perforation or esophageal tracheal fistula and other serious complications. The development of RE will affect the quality of life of patients and the efficacy of local tumor control and treatment. Therefore, it may lead to hospitalization or interruption, or early termination of treatment, and the inability to eat requires a parenteral feeding tube in severe cases (6). In a randomized trial, 21% of patients receiving concurrent chemoradiotherapy (CCRT) stopped treatment because of severe RE (7). Predicting RE allows clinicians to take appropriate preventive measures in advance, such as medication, dietary guidance, rehydration, or tube feeding. Identifying low-risk patients with RE provides an opportunity to increase the dose of radiotherapy to improve tumor control. Hence, identifying predictive factors before the onset of RE may contribute to early intervention to decrease or avert the occurrence of severe RE (8).
Although some predictors of RE have been identified, including dosimetric factors and patient characteristics (9)(10)(11), it remains unclear how these factors can be used in routine clinical practice. Moreover, predictive models should at least perform better than doctors themselves to aid treatment decisions. However, this has been poorly studied in small cell lung cancer to the best of our knowledge. Consequently, we decided to build a nomogram model using clinical, dosimetric factors, and inflammation index to predict RE in SCLC patients receiving chemoradiotherapy. Patients had previously received chest radiotherapy for some reason; (B) Underwent pneumonectomy; (C) Failed to participate in the whole course of radiotherapy. (D) Hyperfractionated radiotherapy was administered. Finally, a total of 187 patients were enrolled. The Ethics Committee approved this study at Fujian Provincial Cancer Hospital (K2021-115-01).
Treatment schedules
All patients were scheduled to receive intensity-modulated radiation therapy (IMRT) or 3-dimensional conformal radiation therapy (3D-CRT). The prescribed RT dose was 46 -70 Gy, 23 -35 fractions (once-daily), 5 days per week. Patients were positioned by computed tomography (CT) simulation with a postural fixation device. Contrast-enhanced CT scans cover the entire chest from the cricoid cartilage to the costal diaphragmatic angle (the range can be increased according to the tumor) with a thickness of ≤ 5 mm. Radiotherapy (RT) was performed using a 6MV medical linear accelerator. According to the ESTRO ACROP guidelines, the gross tumor volume (GTV) includes lung primary tumors and involved or elective lymph nodes. The clinical tumor volume (CTV) is obtained by expanding 5 mm in all directions based on GTV. The planned tumor volume (PTV) is obtained by expanding 5-8 mm in all directions based on CTV (12). Dose and volume limits for organs at risk (OARs) were based on Radiotherapy and Oncology Group (RTOG) guidelines.
The 187 eligible patients received individualized concurrent or sequential chemoradiotherapy. The chemotherapy regimens included etoposide, irinotecan, or paclitaxel with cisplatin, carboplatin, lobaplatin, or nedaplatin. Most patients had received etoposide 100 or 120 mg/m2 day1-3 with cisplatin 75 or 60 mg/m2 day1 or carboplatin 80 or 60 mg/m2 day1 chemotherapy regimens. Every three weeks was a cycle. The chemotherapy regimens followed the National Comprehensive Cancer Network (NCCN).
Blood and biochemical parameters
Inflammation and nutritional index were calculated based on blood and biochemical parameters. For example, systemic immune-inflammation index (SII): absolute neutrophil count times absolute platelet count divided by absolute lymphocyte count. system inflammation response index (SIRI): absolute neutrophil count times absolute monocyte count divided by absolute lymphocyte count. neutrophil-lymphocyte ratio (NLR): absolute neutrophil count divided by absolute lymphocyte count. platelets-lymphocyte radio (PLR): absolute platelet count divided by absolute lymphocyte count. prognosticnutrition index (PNI): serum albumin level plus five times absolute lymphocyte count. platelets-albumin ratio (PAR): absolute platelet count divided by serum albumin level. The blood biochemical data were collected five days before therapy.
Evaluation of radiation esophagitis
Toxic side effects were assessed weekly for each patient during RT and monthly follow-up for three months after completion of radiotherapy. Radiation esophagitis (RE) was diagnosed by radiation oncologists based on clinical symptoms and imaging evidence. RE was assessed and graded according to the Radiation Therapy Oncology Group scale (RTOG)/ European Organization for Research and Treatment of Cancer (EORTC). The highest grade of esophagitis was recorded during treatment and follow-up. In this study, only grade ≥ 2 radiation esophagitis was considered an endpoint event.
Statistical analysis
In this study, all continuous variables are converted into classification variables according to the optimal cut-off value of the Receiver Operating Curve (ROC). The risk factors for grade 2 or higher radiation esophagitis (RE) were identified by univariate logic regression analysis. Risk factors with p < 0.10 in univariate analysis were incorporated into the multivariate logistic regression analysis to determine independent predictors of the occurrence of RE. The area under the ROC curve (AUC), calibration curve (with 1000 bootstrap resamples), and decision curve analysis (DCA) were used to assess the clinical utility of the nomogram. The AUC value of the nomogram was greater than that of each independent predictor, indicating a preferable discrimination ability. The calibration curve was used to assess the agreement between the actual and predicted probability of occurrence of RE. DCA was used to evaluate the clinical benefit of the nomogram by quantifying net benefits at different threshold probabilities. The correlations of inflammation indexes were assessed using the Spearman correlation analysis. All statistical analyses were performed using SPSS software (version 25.0) and R software (version 4.0.2). All p values were double-tailed, and p < 0.05 was considered statistically significant.
Patient characteristics and incidence of RE
Baseline characteristics of 187 eligible patients are presented in Table 1. One hundred and seventy-four patients (93.0%) were male, and thirteen patients (7.0%) were female. The median age of patients was 60 years. A total of 94 patients were smokers, and 93 patients were non-smokers. Of all patients receiving chemoradiotherapy, seventy-seven patients (41.2%) received concurrent chemoradiotherapy. The median chemotherapy cycle of a platinum-based chemotherapy regimen was four cycles. Most of the patients (68.4%) were limited-stage small
Development and validation of a nomogram
Based on multivariate analysis results, concurrent chemoradiotherapy, SIRI, and PAR were included in the (Figure 2A)]. It is proved that the model had preferable discrimination ability. The calibration curve displayed good consistency between the actual and predicted probability of occurrence of RE ( Figure 2B). Finally, the DCA demonstrated favorable positive net benefits of the nomogram in the threshold probabilities, indicating a satisfactory clinical benefit of the model ( Figure 2C).
Discussion
Chemoradiotherapy is the primary treatment regimen in patients with small cell lung cancer (SCLC), especially limitedstage small cell lung cancer (LS-SCLC) (13). In extensive-stage small cell lung cancer (ES-SCLC), if patients had a complete or partial response to the initial systemic treatment, thoracic radiotherapy can be performed (14). Surgery is available in only 2-5% of patients, and a study has shown that radiotherapy has better survival outcomes than patients who underwent surgery (15). Therefore, chemoradiotherapy has become the treatment of choice for most patients. Although SCLC is sensitive to radiotherapy and chemotherapy, local recurrence rates and survival prognosis are still unsatisfactory (16). Several clinical studies suggested that shortening overall treatment duration or increasing radiation dose may lead to better survival outcomes (17)(18)(19). However, this improved survival rate comes at a cost, namely increased toxicity, especially radiation esophagitis (RE) (17,20). Consequently, it is essential to recognize some of the factors associated with severe RE.
In this study, we established a nomogram model for grade ≥2 RE in SCLC patients who underwent chemoradiotherapy. We investigated 14 factors associated with the risk of grade ≥2 RE, including gender, age, smokers, CCRT, chemotherapy cycle, stage, duration of radiotherapy, RT dose, SII, SIRI, NLR, PNI, PLR, and PAR. Our data showed that CCRT, SIRI, and PAR were significantly correlated with grade ≥2 RE. As far as we know, the combination of CCRT, SIRI, and PAR for chemoradiotherapy in SCLC patients is the first nomogram model reported to assess the occurrence of RE. The model demonstrated remarkable good consistency and discriminative ability between the predicted risks and observed results. Bootstrap validation proved the stability of the model for similar populations in the future. Decision curve analysis (DCA) also showed potential clinical utility for future clinical practice.
The relationship between concurrent chemoradiotherapy (CCRT) and RE has been documented (11,(21)(22)(23). Compared with patients who received sequential chemoradiotherapy or radiotherapy alone, patients who received CCRT had an approximately five-fold increased risk of developing acute RE (11). Similarly, CCRT was significantly associated with grade ≥2 RE in our multivariate analysis. Due to the rapid doubling time and high aggressiveness of SCLC, CCRT had a favorable survival outcome in patients receiving chemoradiotherapy. However, for malignant tumors involving the esophagus in the thoracic radiation field, CCRT increased the incidence and degree of acute esophageal injury (24). One possible reason is that the daily administration of platinum keeps the concentration of platinum in the tissue above the threshold level for radiation enhancement, which leads to a relatively high frequency of RE. And we must keep in mind that RE may damage the patients' condition later. Esophageal stricture has been reported after a long incubation period (25,26). It is still an unsolved problem to achieve satisfactory therapeutic effects while reducing toxic side effects. One article reported that conserving the esophageal technique can limit the occurrence of RE in patients with non-small cell lung cancer receiving CCRT without compromising local control (27). Therefore, we hope to develop a more comprehensive individualized treatment plan for chemoradiotherapy patients to reduce or avoid treatmentrelated toxicity.
A large amount of radiation toxicities has been assumed to be caused by radiation-induced inflammatory responses (28,29). And previous evidence has shown that some serum inflammatory markers, including interleukin (ILs) and transforming growth factor (TGF-b) were correlated with RE (30)(31)(32). These studies emphasize the role of inflammation in the toxicity of RT. Unfortunately, because these indicators were not routinely monitored in the clinic, they were not widely utilized in clinical practice. Whereas, the inflammatory indicators including SII, SIRI, NLR, PLR, and PAR were simply measured by neutrophils, platelets, monocytes, lymphocytes, and albumins, which could be easily and conventionally measured during treatment. At present, the role of inflammatory index and incidence of RE in SCLC undergoing chemoradiotherapy has not been reported. To the best of my knowledge, this is the first study to assess the association between RE and inflammatory index in patients with SCLC. Our results demonstrated that SIRI and PAR were independent predictive factors of the grade ≥2 RE. SIRI < 0.3 and PAR > 8.0 were significantly associated with the occurrence of grade ≥2 RE. Interestingly, the result was consistent with other studies of lung cancer patients treated with chemoradiotherapy. A study has indicated that concurrent chemoradiotherapy and neutropenia were significantly associated with grade ≥2 RE (33). One study suggested that high platelet counts and low hemoglobin levels before radiotherapy were closely related to the occurrence of RE (34). The decrease of neutrophils and platelets could reliably reflect the systemic inflammation of cancer patients (35). Albumin synthesis can be inhibited by reduced protein intake or acute phase reaction, and inflammation was an important factor leading to the decrease of albumin synthesis (36). In this study, we preliminarily demonstrated a close relationship between inflammation index and RE.
It should not be ignored that radiation esophagitis (RE) is a common adverse effect of thoracic radiotherapy, affecting the patient's therapeutic effect and quality of life. Therefore, it is necessary for SCLC patients receiving radiotherapy to identify this toxicity as early as possible. Although there are a few predictive models based on clinical and dosimetric factors with good discriminative ability, the addition of new biomarkers can improve the predictive power of RE. More importantly, find biomarkers that are readily available and clinically useful. The accurate prediction of RE is critical for facilitating individualized radiation doses and maximizing therapeutic benefits.
A few shortcomings should be pointed out here. First of all, there might be a selection bias in our research due to the retrospective research. Additionally, many patient factors were not included in the study in this heterogeneous population. Some of these features may be associated with the occurrence of RE in individuals. Third, toxicity analysis of different grades of RE was not performed in our study. Finally, the sample size of this study was small, so a large number of queues were required to further construct and verify the nomogram to predict RE.
Conclusion
To sum up, our research showed that CCRT, SIRI, and PAR were independent predictive factors for grade ≥2 RE in SCLC patients who underwent chemoradiotherapy. We developed and validated a predictive model using these factors. The developed nomogram with superior prediction ability can be used as a quantitative model to predict RE. Any newly developed predictive model will need further validation before it can be advanced to clinical use.
Data availability statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Author contributions
JL and JQ designed this study. JQ and DK contributed to the data collection. HCL, YY, and HL analyzed the data. JL supervised the study. JQ, YY, QZ, LL, HZ, and HL wrote the manuscript. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-24T06:17:17.915Z
|
2022-11-22T00:00:00.000Z
|
253800163
|
s2ag/train
|
Co-assembled Nanocarriers of De Novo Thiol-Activated Hydrogen Sulfide Donors with an RGDFF Pentapeptide for Targeted Therapy of Non-Small-Cell Lung Cancer.
Hydrogen sulfide releasing agents (or H2S donors) have been recognized gasotransmitters with potent cytoprotective and anticancer properties. However, the clinical application of H2S donors has been hampered by their fast H2S-release, instability, and lack of tumor targeting, despite the unclear molecular mechanism of H2S action. Here we rationally designed an amphiphilic pentapeptide (RGDFF) to coassemble with the de novo designed thiol-activated H2S donors (CL2/3) into nanocarriers for targeted therapy of non-small-cell lung cancer, which has been proved as a one-stone-three-birds strategy. The coassembly approach simply solved the solubility issue of CL2/3 by the introduction of electron-donating groups (phenyl rings) to slow down the H2S release while dramatically improving their biocompatible interface, circulation time, slow release of H2S, and tumor targeting. Experimental results confirmed that as-prepared coassembled nanocarriers can significantly induce the intrinsic apoptotic, effectively arrest cell cycle at the G2/M phase, inhibit H2S-producing enzymes, and lead to mitochondrial dysfunction by increasing intracellular ROS production in H1299 cells. The mouse tumorigenesis experiments further confirmed the in vivo anticancer effects of the coassembled nanocarriers, and such treatment made tumors more sensitive to radiotherapy then improved the prognosis of tumor-bearing mice, which holds great promise for developing a new combined approach for NSCLC.
|
v2
|
2022-11-24T06:17:18.257Z
|
2022-11-22T00:00:00.000Z
|
253800881
|
s2ag/train
|
Identification of a Novel Potent CYP4Z1 Inhibitor Attenuating the Stemness of Breast Cancer Cells through Lead Optimization.
Pharmacological targeting cancer stem cells are emerging as a novel therapeutic modality for cancer treatment and prevention. Human cytochrome P450 enzyme CYP4Z1 represents a promising target for its potential role in attenuating the stemness of breast cancer cells. In order to develop potent and selective CYP4Z1 inhibitors, a series of novel N-hydroxyphenylformamidines were rationally designed and synthesized from a pan-CYP inhibitor HET0016. CYP4Z1 inhibitory activities of the newly synthesized derivatives were evaluated, and the structure-activity relationships (SARs) were summarized. Among them, compound 7c exhibited the best inhibitory activity with an IC50 value of 41.8 nM. Furthermore, it was found that 7c decreased the expression of stemness markers, spheroid formation, and metastatic ability as well as tumor-initiation capability in a concentration-dependent manner in vitro and in vivo. Altogether, compound 7c might be a potential lead compound to develop CYP4Z1 inhibitor with more favorable druggability for clinical application to treat breast cancer.
|
v2
|
2022-11-24T06:17:18.274Z
|
2022-11-22T00:00:00.000Z
|
253800206
|
s2ag/train
|
Marker-free characterization of full-length transcriptomes of single live circulating tumor cells.
The identification and characterization of circulating tumor cells (CTCs) are important for gaining insights into the biology of metastatic cancers, monitoring disease progression, and medical management of the disease. The limiting factor in the enrichment of purified CTC populations is their sparse availability, heterogeneity, and altered phenotypes relative to the primary tumor. Intensive research both at the technical and molecular fronts led to the development of assays that ease CTC detection and identification from peripheral blood. Most CTC detection methods based on single-cell RNA sequencing (scRNA-seq) use a mix of size selection, marker-based white blood cells (WBC) depletion, and antibodies targeting tumor-associated antigens. However, the majority of these methods either miss atypical CTCs or suffer from WBC contamination. We present unCTC, an R package for unbiased identification and characterization of CTCs from single-cell transcriptomic data. unCTC features many standard and novel computational and statistical modules for various analyses. These include a novel method of scRNA-seq clustering, named Deep Dictionary Learning using k-means clustering cost (DDLK), expression-based copy number variation (CNV) inference, and combinatorial, marker-based verification of the malignant phenotypes. DDLK enables robust segregation of CTCs and WBCs in the pathway space, as opposed to the gene expression space. We validated the utility of unCTC on scRNA-seq profiles of breast CTCs from six patients, captured and profiled using an integrated ClearCell FX and Polaris workflow that works by the principles of size-based separation of CTCs and marker-based WBC depletion.
|
v2
|
2022-11-24T14:10:45.846Z
|
2022-11-22T00:00:00.000Z
|
253803059
|
s2ag/train
|
Comprehensive glycoprofiling of oral tumours associates N-glycosylation with lymph node metastasis and patient survival
While altered protein glycosylation is regarded a trait of oral squamous cell carcinoma (OSCC), its heterogeneous glycoproteome and dynamics with disease progression remain unmapped. To this end, we here employ an integrated multi-omics approach comprising unbiased and quantitative glycomics and glycoproteomics applied to a valuable cohort of resected tumour tissues from OSCC patients with (n = 19) and without (n = 12) lymph node metastasis. While all tumour tissues displayed uniform N-glycome profiles suggesting relatively stable global N-glycosylation during lymph node metastasis, glycoproteomics and advanced correlation analysis notably uncovered altered site-specific N-glycosylation and previously unknown associations with several key clinicopathological features. Importantly, focused analyses of the multi-omics data unveiled two N-glycans and three N-glycopeptides that were closely associated with patient survival. This study provides novel insight into the complex OSCC tissue N-glycoproteome forming an important resource to further explore the underpinning disease mechanisms and uncover new prognostic glyco-markers for OSCC. Teaser Deep survey of the dynamic landscape of complex sugars in oral tumours paves a way for new prognostic disease markers.
|
v2
|
2022-11-24T16:04:28.596Z
|
2022-11-22T00:00:00.000Z
|
253827268
|
s2ag/train
|
Incorporating cross-voxel exchange for the analysis of dynamic contrast-enhanced imaging data: pre-clinical results
Tumours exhibit abnormal interstitial structures and vasculature function often leading to impaired and heterogeneous drug delivery. The disproportionate spatial accumulation of a drug in the interstitium is determined by several microenvironmental properties (blood vessel distribution and permeability, gradients in the interstitial fluid pressure). Predictions of tumour perfusion are key determinants of drug delivery and responsiveness to therapy. Pharmacokinetic models allow for the quantification of tracer perfusion based on contrast enhancement measured with non-invasive imaging techniques. An advanced cross-voxel exchange model (CVXM) was recently developed to provide a comprehensive description of tracer extravasation as well as advection and diffusion based on cross-voxel tracer kinetics (Sinno et al 2021). Transport parameters were derived from DCE-MRI of twenty TS-415 human cervical carcinoma xenografts by using CVXM. Tracer velocity flows were measured at the tumour periphery (mean 1.78–5.82 μm.s−1) pushing the contrast outward towards normal tissue. These elevated velocity measures and extravasation rates explain the heterogeneous distribution of tracer across the tumour and its accumulation at the periphery. Significant values for diffusivity were deduced across the tumours (mean 152–499 μm2.s−1). CVXM resulted in generally smaller values for the extravasation parameter Kext (mean 0.01–0.04 min−1) and extravascular extracellular volume fraction ve (mean 0.05–0.17) compared to the standard Tofts parameters, suggesting that Toft model underestimates the effects of inter-voxel exchange. The ratio of Tofts’ extravasation parameters over CVXM’s was significantly positively correlated to the cross-voxel diffusivity (P < 0.0001) and velocity (P = 0.0005). Tofts’ increased ve measurements were explained using Sinno et al (2021)’s theoretical work. Finally, a scan time of 15 min renders informative estimations of the transport parameters. However, a duration as low as 7.5 min is acceptable to recognize the spatial variation of transport parameters. The results demonstrate the potential of utilizing CVXM for determining metrics characterizing the exchange of tracer between the vasculature and the tumour tissue. Like for many earlier models, additional work is strongly recommended, in terms of validation, to develop more confidence in the results, motivating future laboratory work in this regard.
|
v2
|
2022-11-24T16:25:42.515Z
|
2022-11-22T00:00:00.000Z
|
253821170
|
s2orc/train
|
Adjuvant chemoradiotherapy versus chemotherapy or radiotherapy in advanced endometrial cancer: a systematic review and meta-analysis
Background Endometrial cancer is one of the most common gynecological cancer in the world. However, the available adjuvant therapies, chemotherapy (CT) and radiotherapy (RT), demonstrated several limitations when used alone. Therefore, we conducted a meta-analysis to investigate the clinical effectiveness of chemoradiotherapy (CRT) based on overall survival (OS) and disease-free survival (DFS). Methods A literature search was performed on five databases and one clinical trial registry to obtain all relevant articles. Search for studies was completed on September 9, 2021. A meta-analysis was conducted to determine the overall hazard ratio with the 95% Confidence Interval. Results A total of 17 articles with 23,975 patients in the CRT vs RT group and 50,502 patients in the CRT vs CT group were included. The OS Hazard Ratios (HR) of CRT compared to RT was 0.66 (95% CI [0.59–0.75]; P < 0.00001). Compared to CT, the OS HR was 0.70 (95% CI [0.64–0.78]; P < 0.00001). CRT also significantly improved the DFS compared to CT only (HR 0.79, 95% CI [0.64–0.97]; P = 0.02) However, CRT did not improve the DFS compared to RT only, with HR of 0.71 (95% CI [0.46–1.09]; P = 0.12). Conclusion Adjuvant CRT can significantly improve OS compared to CT or RT alone and improve the DFS compared to CT alone in patients with advanced endometrial cancer. Further research is needed to identify the optimal CRT regimen, and to whom CRT will be most beneficial.
INTRODUCTION
Accounts for 417,367 new cases and 97,370 deaths just in 2020, endometrial cancer is one of the most diagnosed gynecological cancer in the world (World Health Organization, 2020). Obesity, older age (≥55 years), exposure to estrogen, early menarche, and late menopause are among the risk factors (World Health Organization, 2020). While many other cancer incidences have decreased, endometrial cancer incidence has been rising by 21% from the past decade (Sorosky, 2012).
Since 1988, the International Federation of Gynecology and Obstetrics (FIGO) has decided to change the classification of endometrial cancer from clinical staging to surgical staging (Creasman, 2009). Most patients were initially presented to healthcare facilities with postmenopausal bleeding as primary complaint (Stubert & Gerber, 2016). Although it accounts for a smaller proportion, a higher FIGO stage is associated with poorer prognosis. The 5-year survival rate was around 30% to 60% in patients with advanced stages, in contrast to 80% to 97% in patients with early stages (Kosary, 2007).
The current treatment for advanced-stage endometrial cancer is surgery followed by adjuvant therapy. Through surgery, prognosis stratification and identification of appropriate adjuvant therapy was done. Adjuvant therapy is important to reduce the likelihood of cancer recurrence and increase overall survival (OS) (Concin et al., 2021). Radiotherapy (RT) has been used as adjuvant therapy for its ability to control local recurrence. Because RT demonstrated limited impact on distal recurrence, physicians have considered adjuvant chemotherapy (CT) (DeLeon, Ammakkanavar & Matei, 2014). However, CT alone has been shown to have a limited impact on preventing pelvic recurrence, which accounts for 18% as the first relapse site in the advanced stage (Randall et al., 2006). Currently, the clinical effectiveness of chemoradiotherapy (CRT) in advanced-stage endometrial cancer has not yet been established. Recently, several studies showed that CRT has a benefit on OS in advanced-stage endometrial cancer (Lee & Viswanathan, 2012;Lester-Coll et al., 2016;Secord et al., 2013). A meta-analysis on the effect of CRT on endometrial cancer has been previously done by Park et al. (2013). However, the study focused more on high-risk endometrial cancer and only compared the effect of CRT to RT. The number of studies used in that meta-analysis for advanced stage endometrial cancer was relatively small. Since there have been several studies published after 2013, including two large clinical trials, an update on the effect of CRT to RT or CT in advanced stage is needed to confirm the benefit.
We conducted a systematic review and meta-analysis to determine CRT's clinical effectiveness compared to RT or CT only as adjuvant treatment for women with advanced-stage endometrial cancer. We evaluate the clinical effectiveness according to the OS and disease-free survival (DFS) of each treatment modality.
Protocol and registration
A protocol was established before writing and registered in the International prospective register of systematic reviews (PROSPERO) on June 1, 2021 (CRD42021252529). We report this review according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist (Moher et al., 2009) (Table S1).
Eligibility criteria
Studies were included if they fulfilled the following criteria: (a) the subject was advanced stage endometrial cancer patients who underwent surgery, (b) RCT or observational studies, (c) the study compared the result between CRT and CT or RT alone, and (d) the OS or DFS hazard ratio (HR) with 95% confidence interval (CI) and p-value were stated or can be calculated from the Kaplan-Meier Curve. The overall survival was defined as the time from the starting time point to the date of death and disease-free survival was defined as the time from the starting time point until the disease reoccurred. The starting time point for RCTs was randomization. For observational studies, the starting point was the date of diagnosis. However, due to limited number of studies, we also included studies that did not specify the starting time point. The HR can be calculated if the following data can be estimated from the curve: (1) event-free at the beginning, (2) censored and at risk throughout the interval, and (3) the number of events from every interval (Tierney et al., 2007). If the last criteria were not fulfilled, we inquire about the data through e-mail. The studies were included if the author had provided the required information. Studies were excluded if the full text is not available, or they were not written in English.
Search strategy
We started the search for all published studies from electronic databases on April 10, 2021. We conducted preliminary search before PROSPERO registration to determine whether the volume of relevant studies is sufficient to conduct a systematic review and metaanalysis. Study search was completed on September 9, 2021. We searched for those studies from several databases, including SCOPUS, Medline (PubMed), EBSCO, Embase, Cochrane Central Register of Controlled Trials (CENTRAL) on the Cochrane Library, Web of Science, and ClinicalTrials.gov. We used key terms including "advanced stage endometrial cancer", "chemoradiotherapy", "chemotherapy", "radiotherapy", "outcome" and "survival" (Table S2). We also searched for additional studies through manual hand-searching and tracing of citations from related studies. Author NAAI, FDSA, and YM conducted the search for eligible studies.
Study selection
All studies were exported to Mendeley software and duplicates were removed. Four reviewers (NAAI, FDSA, YM and HW) screened the titles and abstracts independently. The remaining articles will be assessed from their full text for their eligibility by the same three reviewers independently. Any disagreements will be resolved through discussion with a fifth reviewer (EDS).
Quality assessment
The included studies were critically appraised by four independent reviewers (NAAI, FDSA, YM and HW) using the Cochrane Collaboration's tool for assessing the risk of bias in randomized trials for RCTs (Higgins et al., 2011) and the Newcastle Ottawa Scale (NOS) for retrospective studies (Wells et al., 2000). For RCTs, the risk of bias were assessed through selection, performance, detection, attribution, reporting, and other domains.
We then categorized the risk of bias as 'low risk,' 'high risk,' or 'unclear risk' of bias for each domain. For retrospective studies, total NOS score of 0 to 4 was categorized as low quality (high risk of bias), 5 to 7 as moderate quality (moderate risk of bias), and 8 to 9 as high quality (low risk of bias). Any discrepancy was resolved through discussion with a fifth reviewer (EDS) to reach an agreement.
Data extraction and synthesis
We developed a data extraction form for this review. The data extracted included: first author, year of publication, study location and period, study design, the total number of enrolled subjects, baseline population characteristics (age and performance score), details of diagnosis (FIGO stage, tumor grade, types, and extension), the total number of intervention groups and details of intervention (modalities, dose, cycle length), risk of bias, follow up duration, and outcomes (OS and DFS). We presented the outcomes in the HR with the 95% CI. If the articles did not report the HR, we estimated the HR from the Kaplan-Meier curves using the method purposed by Tierney et al. (2007). We contacted the author if there was missing or incomplete data.
Statistical analysis
The quantitative data were exported to Review Manager 5.4 and pooled in a meta-analysis only if appropriate. We used the inverse variance method to obtain the pooled HR. Studies were considered to have moderate heterogeneity if I2 >30%, substantial heterogeneity if I2 >50%, and considerable heterogeneity if I2 >75% (Higgins et al., 2021). Sources of heterogeneity were assessed for any studies with substantial heterogeneity or more. To detect any risk of publication bias, we constructed funnel plots in Review Manager 5.4 and performed Egger's Test in Stata 17. Symmetrical funnel plots indicated low risk of publication bias. The studies were considered to have potential risk of publication bias if p-value on Egger's Test is less than 0.05.
Sensitivity analysis
We conducted a sensitivity analysis to identify the possible contribution of specific clinical or methodological differences between the included studies. Sensitivity analysis was performed by omitting studies with a high risk of bias leaving only studies with low risk of bias. Possible study characteristics that contribute to high risk of bias are inadequate follow-up period, different study type, small sample size, etc. Sensitivity analysis on specific treatment regimens was not conducted due to the lack of data.
We also extracted DFS data from six studies with a total of 1,283 patients. The pooled DFS HR was 0.71 (95% CI [0.46-1.09]; 1,283 patients; Fig. 2B). Substantial heterogeneity was found with I2 of 55% (P = 0.05). Due to small number of included studies (less than 10 studies), we did not create a funnel plot for the DFS data. However, the Egger's test showed potential risks of publication bias (P value = 0.0370).
CRT vs CT alone in advanced endometrial cancer
A total of 13 studies were included in this category with 50,502 patients (Table 2). However, the one RCT included (GOG-258) only reported DFS and not OS. The OS was improved significantly in patients with advanced stage endometrial cancer receiving adjuvant CRT compared to adjuvant CT only. The OS HR was 0.70 (95% CI [0.64-0.78]; 49,933 patients; P < 0.00001; Fig. 3A) with substantial heterogeneity (I2 = 71%; P < 0.0001). The funnel plot (Fig. S4) and the Egger's test result (P value = 0.4686) indicated low risk of publication bias. Five studies enrolled less than 100 subjects for each treatment groups (Secord et al., 2013;Pichatechaiyoot et al., 2014;Secord et al., 2007;Tai et al., 2019;Nakayama et al., 2010). Because of the relatively small numbers of subjects enrolled, the result of those studies might not be representative. Therefore, both studies were excluded in the sensitivity analysis. After the exclusion, the adjusted OS HR was still similar with an increase in heterogeneity (0.71; 95% CI [0.64-0.79]; I2 = 81%; P < 0.0001; Fig. S5). Furthermore, three studies (Secord et al., 2013;Pichatechaiyoot et al., 2014;Secord et al., 2007) that reported 3-years OS were also excluded. After the exclusion of the three studies, the OS HR became 0.71 (95% CI [0.65-0.79]; Fig. S6) and the heterogeneity remained high (I2 = 76%; P < 0.0001).
The DFS data were obtained from four studies with a total of 1,561 patients. The pooled DFS HR was 0.79 (95% CI [0.64-0.97]; 1,561 patients; Fig. 3B) with no heterogeneity found (I2 = 0%; P = 0.51). The funnel plot for DFS data was also not created due to small number studies. The Egger's test resulted in low risk of publication bias (P value = 0.7670).
Summary of main results
The included studies showed different significant confounding factors, however multivariate analysis of the HR was provided in all studies except for Pichatechaiyoot et al. (2014) and Nakayama et al. (2010). Multivariate analysis from the other studies allowed correction for other covariates, therefore provide more accurate correlation between the adjuvant therapy and the outcome.
The result (Figs. 2A and 3A) implicates that a combination of adjuvant CT and RT improves the OS of women with advanced-stage endometrial carcinoma. However, only DFS of CRT vs CT showed favor to CRT (Fig. 3B), whereas DFS of CRT vs RT showed no statistical significance, even though the pooled HR point toward CRT (Fig. 2B).
It is important to note that there was a discordant result on DFS of CRT vs CT from GOG-258 and included retrospective studies (Fig. 3B). Although, ultimately pooled DFS favors CRT. Contributing factors to this difference might be firstly in GOG-258, both stage III and IV were included in the study. Meanwhile, only one retrospective study (Pichatechaiyoot et al., 2014) that included stage IV patients. In which, higher disease stage is known to be associated with worse disease prognosis. Second, GOG-258 carboplatin and paclitaxel combination was use as CT regimen in monotherapy, and an addition of cisplatin in CRT. Meanwhile, in the included retrospective studies other combination of CT regimens was used, although was not specified.
Agreements with other studies Hogberg et al. (2010) conducted two RCTs that included MaNGO ILIADE-III. MaNGO ILIADE-III aimed to compare the outcome of women who received RT and CRT in patients with stage IIB to stage III with a total of 156 subjects (Lee & Viswanathan, 2012). In accordance with our study, the result also showed that adjuvant CRT had better OS than women who received RT only. However, this outcome was not statistically significant (HR 0.64,; P = 0.07). One of the possible explanations was the small number of subjects in the trial. Another meta-analysis, Park et al. (2013), also included a combination of observational studies and RCTs. Compared to RT only, the study concluded that CRT had a significant effect on the OS of advanced-stage endometrial cancer (OS HR 0.53, 95% CI [0.36-0.80]) (Lester-Coll et al., 2016). This conclusion supports the result of our study and provides more reliable evidence to implement the use of combination adjuvant therapy after surgery in women with advanced-stage endometrial cancer. Therefore, the conclusions from the studies above further support the result of our review.
Overall completeness and quality of evidence
The sensitivity analysis result was consistent and did not differ from overall CRT vs RT and CRT vs CT group analyses. In which all result favors combination therapy rather than monotherapy. The overall risk of bias of the included studies was low to moderate. The result of our analysis was consistent with other studies, resulting in no to moderate heterogeneity.
We considered this result applicable to most patients with advanced-stage endometrial cancer. The survival benefit of CRT may vary depending on additional risk factors, such as age, tumor grade, and histological types. However, we were not able to perform subgroup analysis according to those risk factors due to insufficient data. Most studies did not include the performance status of their patients before receiving treatments on their data. The RCT study (de Boer et al., 2019;Matei et al., 2019) only included patients with a performance score of 0-2. It is important to note because not all patients with advanced disease have a high-performance score and patients with low-performance status are more susceptible to CT toxicity (Azam et al., 2019;Sargent et al., 2009). Further research is necessary to determine to whom CRT will be most beneficial and well-tolerated.
Biological plausibility
In advanced stage endometrial cancer, where the cancer cells have spread outside the uterus, some physicians have considered combining CT and RT to improve OS. The hypothesis was that radiation therapy damages DNA and creates genomic instability leading to cellular death (Galluzzi et al., 2007). However, this cytotoxic effect exerts only in the area treated with radiotherapy without influencing disease progression outside the treated area. Adding CT reduces the risk of relapse on pelvic and other occult metastatic as it exerts cytoreductive and cytostatic effects systemically (Schwandt et al., 2011).
Applicability of evidence and implication for practice
As reported in the forest plot, OS favors combination therapy. However, it is not the only important parameter regarding cancer patient's treatment. Hence, we also calculated the DFS, which more reflects the patient's quality of life compared to just OS.
We can conclude from the result before that the DFS of CRT vs RT yielded in no statistical difference, unlike CRT vs CT. This result showed that the role of RT as an adjuvant treatment for advanced stage endometrial cancer is valuable. Previous studies have shown that radiotherapy lowers both the rate of local and regional recurrence (Mundt et al., 2001;Klopp et al., 2009). However, the availability of RT centers in some South-East Asian countries are much lower compared to the United States and European countries. For instance, in Indonesia, there are only 33 RT centers. Which means that on average, there is one RT center per 7.4 million of the Indonesian population. The low number of RT specialists has also been an issue for years, as cancer incidence continues to increase (Hiswara, 2017). The scarcity of RT treatment in many countries should be addressed, as we found in this study that RT alone is comparable to combination therapy in terms of quality of life.
Limitation and potential bias in the review process
The result of OS in CRT vs CT yielded moderate heterogeneity (I2 = 71%). In addition, we combined the results of the RCT study with observational studies into one forest plot and only included observational studies for DFS result of CRT vs RT. Therefore, the result has lower level of evidence credibility compared to if we included RCTs. Conducting future trials regarding this matter are essential to improve our current knowledge.
In this review, funnel plot was not created for DFS result due to the inadequate number of studies (<10 studies) as suggested by Ioannidis & Trikalinos (2007). However, the results of Egger-test indicate low impact of small-study effect except for DFS of CRT vs RT. Contributing factors might be due to the small number of included studies (six studies) and the type of study which only comprised of observational studies. Retrospective observational studies in general have lower level of evidence compared to RCT. Observational study is not randomized and not blinded. Also, in retrospective study, researcher formulated study hypothesis before data collection. Hence, these factors contribute to the higher level of bias and lower level of evidence compared to RCT. We suggest that to obtain higher certainty of evidence, more RCT should be conducted comparing monotherapy and CRT. In addition, subgroup analysis on specific treatment regimens might also be beneficial.
CONCLUSIONS
Given the limitations of monotherapy, adjuvant CRT becomes a reasonable treatment option in advanced-stage endometrial cancer. There is a moderate quality of evidence with low risk of bias that adjuvant CRT can significantly improve OS compared to CT or RT alone in patients with advanced endometrial cancer. Further research is needed to identify the optimal CRT regimen and whom CRT will most benefit regarding toxicity and quality of life.
|
v2
|
2022-11-25T16:30:48.646Z
|
2022-11-22T00:00:00.000Z
|
253863098
|
s2orc/train
|
Dihydromyricetin Attenuates Depressive-like Behaviors in Mice by Inhibiting the AGE-RAGE Signaling Pathway
Depression is a complex mental disorder, affecting approximately 280 million individuals globally. The pathobiology of depression is not fully understood, and the development of new treatments is urgently needed. Dihydromyricetin (DHM) is a natural flavanone, mainly distributed in Ampelopsis grossedentata. DHM has demonstrated a protective role against cardiovascular disease, diabetes, liver disease, cancer, kidney injury and neurodegenerative disorders. In the present study, we examined the protective effect of DHM against depression in a chronic depression mouse model induced by corticosterone (CORT). Animals exposed to CORT displayed depressive-like behaviors; DHM treatment reversed these behaviors. Network pharmacology analyses showed that DHM’s function against depression involved a wide range of targets and signaling pathways, among which the inflammation-linked targets and signaling pathways were critical. Western blotting showed that CORT-treated animals had significantly increased levels of the advanced glycation end product (AGE) and receptor of AGE (RAGE) in the hippocampus, implicating activation of the AGE-RAGE signaling pathway. Furthermore, enzyme-linked immunosorbent assay (ELISA) detected a marked increase in the production of proinflammatory cytokines, interleukin-1 beta (IL-1β), IL-6 and tumor necrosis factor-alpha (TNFα) in the hippocampus of CORT-treated mice. DHM administration significantly counteracted these CORT-induced changes. These findings suggest that protection against depression by DHM is mediated by suppression of neuroinflammation, predominantly via the AGE-RAGE signaling pathway.
Introduction
Depression is a common and severe psychiatric disease with a prevalence of 5% in adults, affecting more than 350 million individuals across the world [1]. It is a complex mental disorder, associated with psychological, environmental and genetic factors [2]. Although there have been a great many studies, the underlying molecular mechanisms of depression formation and progression are poorly understood. Currently, a variety of antidepressants are available, e.g., tricyclics and monoamines, which have slow onset, low efficacy and undesirable side effects. Furthermore, less than two-thirds of patients respond to pharmacotherapy, while 70% of treated patients did not achieve full remission [3]. Therefore, there is an urgent need to develop more natural antidepressants with reliable efficacy and fewer side effects.
Animal Treatment
The timeline of drug treatment is shown in Figure 1. Animals were randomly divided into three groups: control group (Control, n = 10), corticosterone group (CORT, n = 10) and corticosterone + dihydromyricetin group (CORT+DHM, n = 10). Animals in the control group received an injection (s.c.) of 0.9% saline containing 0.1% DMSO and 1% Tween 80 for 23 consecutive days (day 1-day 23) and an additional injection (i.p.) of 0.9% saline containing 0.2% DMSO for 7 consecutive days (day 18-day 24); The animals in the CORT group received 23 consecutive s.c. injections of CORT (20 mg/kg body weight, day1-day 23); animals in the CORT + DHM group received 23 consecutive s.c. injections of CORT (day 1-day 23) and 7 consecutive i.p. injections of DHM (20 mg/kg body weight, day 18-day 24). For behavior experiments, an open field test (OFT) and forced swimming test (FST) were conducted from 10:00 to 14:00 on day 22, while a sucrose preference test (SPT) was carried out at all times on days 23 and 24. 23); animals in the CORT + DHM group received 23 consecutive s.c. injections of CORT (day 1-day 23) and 7 consecutive i.p. injections of DHM (20 mg/kg body weight, day 18day 24). For behavior experiments, an open field test (OFT) and forced swimming test (FST) were conducted from 10:00 to 14:00 on day 22, while a sucrose preference test (SPT) was carried out at all times on days 23 and 24.
Open Field Test (OFT)
For the OFT, mice were randomly placed into open field apparatus (50 × 50 × 70 cm) for 5 min. The total distance traveled was measured using behavioral analysis software (Anymaze 6.16).
Forced Swimming Test (FST)
For the FST, mice were individually placed into a plexiglass cylinder (25 cm high, 15 cm in diameter) containing 17 cm deep water (27 ± 1 °C) for 6 min. The behavior of these animals was video-recorded, and the immobility time was analyzed.
Sucrose Preference Test (SPT)
The SPT was performed according to a previous description [9]. Briefly, the SPT was divided into two phases (habituation and test phases). During the habituation phase, the mice were exposed to two identical bottles containing 1% sucrose solution for 24 h. The next 24 h was the test phase, during which mice were presented with two identical bottles, one containing 1% sucrose solution and the other containing only water. Sucrose preference was designated as the volume of sucrose intake/(sucrose intake + water in take) × 100% during the test phase.
Western Blotting
To avoid the possible influence of the behavioral tests on expression of target proteins, another cohort of animals was set up and received same treatment as described in Section 2.3 without the behavior task. These animals were killed 1 h after the final DHM treatment on day 24. The entire hippocampus tissues were carefully dissected and lysed in the RIPA buffer with protease inhibitors and phosphatase inhibitors. The concentration of lysed samples was measured using a commercial kit following the manufacturer's protocol. An amount of 50 µ g protein of individual samples was separated using sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto PVDF (polyvinylidene difluoride) membranes. After blocking with 3% BSA at room temperature for 1 h, the membranes were exposed to anti-AGE, anti-RAGE (Abcam, Cambridge, UK, 1:1000) or anti-β-actin (Proteintech, Rosemont, IL, USA, 1:5000) antibodies at For the OFT, mice were randomly placed into open field apparatus (50 × 50 × 70 cm) for 5 min. The total distance traveled was measured using behavioral analysis software (Anymaze 6.16).
Forced Swimming Test (FST)
For the FST, mice were individually placed into a plexiglass cylinder (25 cm high, 15 cm in diameter) containing 17 cm deep water (27 ± 1 • C) for 6 min. The behavior of these animals was video-recorded, and the immobility time was analyzed.
Sucrose Preference Test (SPT)
The SPT was performed according to a previous description [9]. Briefly, the SPT was divided into two phases (habituation and test phases). During the habituation phase, the mice were exposed to two identical bottles containing 1% sucrose solution for 24 h. The next 24 h was the test phase, during which mice were presented with two identical bottles, one containing 1% sucrose solution and the other containing only water. Sucrose preference was designated as the volume of sucrose intake/(sucrose intake + water intake) × 100% during the test phase.
Western Blotting
To avoid the possible influence of the behavioral tests on expression of target proteins, another cohort of animals was set up and received same treatment as described in Section 2.3 without the behavior task. These animals were killed 1 h after the final DHM treatment on day 24. The entire hippocampus tissues were carefully dissected and lysed in the RIPA buffer with protease inhibitors and phosphatase inhibitors. The concentration of lysed samples was measured using a commercial kit following the manufacturer's protocol. An amount of 50 µg protein of individual samples was separated using sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto PVDF (polyvinylidene difluoride) membranes. After blocking with 3% BSA at room temperature for 1 h, the membranes were exposed to anti-AGE, anti-RAGE (Abcam, Cambridge, UK, 1:1000) or anti-β-actin (Proteintech, Rosemont, IL, USA, 1:5000) antibodies at room temperature for 2 h, followed by exposure to secondary antibodies for 2 h. The signals were detected using the ChemiDoc XRS imaging system (Bio-Rad, Hercules, CA, USA) and the intensity of targeted bands was quantified and normalized to β-actin.
Enzyme-Linked Immunosorbent Assay (ELISA)
IL-1β (Cat. KE10003), IL-6 (Cat. KE10007) and TNFα (Cat. KE10002) ELISA kits were purchased from Proteintech (https://www.ptglab.com/, accessed on 5 July 2022). The procedure was performed according to the manufacturer's guidance. Briefly, 100mg mouse hippocampus tissue from individual animals was homogenized in 1 mL extraction reagent containing 1 mM phenylmethylsulfonyl fluoride and centrifuged at 10,000× g for 10 min at 4 • C. The supernatants were collected, and the concentration was determined using a commercial kit. An amount of 100 µL of standards or samples was added to the wells of capture antibody-coated microplate and incubated for 1 h at 37 • C. The liquid was removed, and the plate was washed four times with 1× washing buffer. An amount of 100 µL 1× detection antibody was added to each well and incubated for 1hr at 37 • C. The excess antibodies were washed away, and the plate was washed four times with 1× washing buffer. An amount of 100 µL of 1× horseradish peroxidase (HRP)-conjugated antibody was added to each well and was incubated for 40 min at 37 • C. Excess antibody was washed away with 1× washing buffer. An amount of 100 µL of 3,3',5,5'-Tetramethylbenzidine (substrate for HRP) was added to each well and incubated in darkness for 15 min at 37 • C. Additionally, 100 µL of stop solution was added to each well and the signal was determined using a microplate reader.
Construction of Venn Diagram and DHM Targets Depression Network
The common targets between DHM and depression were determined and visualized using the online tool Venny 2.1 (https://bioinfogp.cnb.csic.es/tools/venny/, accessed on 18 August 2022). Subsequently, the common targets were imported into Cytoscape software (version 3.7.2, https://cytoscape.org/, accessed on 18 August 2022) to construct a DHM targets depression network.
Construction of Protein-Protein Interaction (PPI) Network
The 75 intersection targets between DHM and depression were input to the STRING database (https://string-db.org/ver.11.0, accessed on 18 August 2022) to identify the PPI relationships. In the current study, the data analysis mode was set as "multiple proteins", the species were selected as "Homo Sapiens", the combined scores ≥ 0.4 and the disconnected nodes were hidden. The network graphics were downloaded, and the analyzed results were saved as TSV format files and imported into Cytoscape 3.7.2 software for the network construction.
Gene Ontology (GO) Function and Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway Enrichment
The intersecting targets were also imported into the R Studio software (https:// www.rstudio.com/, accessed on 22 August 2022) and the "clusterProfiler" R package (https://bioconductor.org/packages/release/bioc/html/clusterProfiler.html, accessed on 22 August 2022) to annotate the gene ontology (GO) enrichment and KEGG pathways. The related biological processes and signaling pathways were identified via GO enrichment and KEGG pathway enrichment analyses. The p-value < 0.05 was set as the screening threshold of enrichment molecular analysis. The top ten GO terms in biological process, function, and cellular component and top twenty KEGG pathways were imported into the R Studio software (https://www.rstudio.com/, accessed on 22 August 2022) and visualized.
Statistical Analysis
All the data in the present experiment are expressed as mean ±SEM and analyzed by Sigma Plot 12.5 software using a one-way ANOVA (analysis of variance) test, followed by a post hoc Fisher LSD test. Data were regarded as significant if the p-value < 0.05.
Dihydromyricetin Alleviated CORT-Induced Depressive-like Behaviors in Mice
We first investigated the effect of DHM on depressive-like behavior in animals receiving chronic injections. As shown in Figure 2, animals in the control group, the CORT group, or the CORT+DHM group had no significant differences in distance traveled in the open field test (OFT) (p > 0.05, Figure 2A). In the FST, the CORT group showed significantly increased immobility time compared with the control group, while DHM-treated animals had significantly less immobility time (p < 0.001, Figure 2B) compared to the CORT group; in the SPT, the animals in the CORT group had a significant decrease in sucrose preference, compared to the control group (p < 0.001, Figure 2C). Animals in the CORT + DHM group showed significantly increased sucrose preference (p < 0.001, Figure 2C) compared to the CORT group.
The related biological processes and signaling pathways were identified via GO enrichment and KEGG pathway enrichment analyses. The p-value < 0.05 was set as the screening threshold of enrichment molecular analysis. The top ten GO terms in biological process, function, and cellular component and top twenty KEGG pathways were imported into the R Studio software (https://www.rstudio.com/, accessed on 22 August 2022) and visualized.
Statistical Analysis
All the data in the present experiment are expressed as mean ±SEM and analyzed by Sigma Plot 12.5 software using a one-way ANOVA (analysis of variance) test, followed by a post hoc Fisher LSD test. Data were regarded as significant if the p-value < 0.05.
Dihydromyricetin Alleviated CORT-Induced Depressive-like Behaviors in Mice
We first investigated the effect of DHM on depressive-like behavior in animals receiving chronic injections. As shown in Figure 2, animals in the control group, the CORT group, or the CORT+DHM group had no significant differences in distance traveled in the open field test (OFT) (p > 0.05, Figure 2A). In the FST, the CORT group showed significantly increased immobility time compared with the control group, while DHM-treated animals had significantly less immobility time (p < 0.001, Figure 2B) compared to the CORT group; in the SPT, the animals in the CORT group had a significant decrease in sucrose preference, compared to the control group (p < 0.001, Figure 2C). Animals in the CORT + DHM group showed significantly increased sucrose preference (p < 0.001, Figure 2C) compared to the CORT group.
Target Prediction of DHM in Depression
To elucidate the underlying mechanisms of DHM against depression, a network pharmacology approach was applied. A total of 72 genes from the Swiss Target Prediction
Target Prediction of DHM in Depression
To elucidate the underlying mechanisms of DHM against depression, a network pharmacology approach was applied. A total of 72 genes from the Swiss Target Prediction platform and 114 genes from the TargetNet database were identified as DHM targets. After removing duplicate genes, the final number of targeted genes was 148 (Table S1). A total of 3231 genes from the GenecCards database, 560 genes from the OMIM database and 53 genes from the TTD database were identified. All collected genes were combined and deduplicated, and 3659 target genes were predicted (Table S2). Subsequently, 75 potential DHM targets for depression were predicted in the Venny database (http://bioinfogp.cnb. csis.es/tools/venny/index.html, accessed on 18 August 2022) (Figure 3, Table S3).
ter removing duplicate genes, the final number of targeted genes was 148 (Table S1). A total of 3231 genes from the GenecCards database, 560 genes from the OMIM database and 53 genes from the TTD database were identified. All collected genes were combined and deduplicated, and 3659 target genes were predicted (Table S2). Subsequently, 75 potential DHM targets for depression were predicted in the Venny database (http://bioinfogp.cnb.csis.es/tools/venny/index.html, accessed on 18 August 2022) (Figure 3, Table S3).
PPI Network of DHM Targets in Depression
We constructed a PPI network by importing 75 potential targets into the STRING database; the network was visualized using Cytoscape software Version 3.9.1. Figure 4 shows the 73 nodes and their interactions. Two targets, CMP-N-acetylneuraminate-beta-1,4-galactoside alpha-2,3-sialyltransferase (ST3GAL3) and galanin receptor type 3 (GALR3) are not presented in Figure 4, due to the fact that they were not connected to this network. In the network, the node represents the relevant gene, with the size and color of the node representing the value of the free degree. Proteins with a "degree" value higher than 31 were selected as the significant targets, which included proto-oncogene tyrosineprotein kinase (SRC), heat shock protein 90 alpha family class A member 1 (HSP90AA1), estrogen receptor 1 (ESR1), hypoxia inducible factor 1 subunit alpha (HIF1A), vascular endothelial growth factor A (VEGFA), silent information regulator 1 (Sirt1) and prostaglandin endoperoxide synthase 2 (PTGS2), all of which may play an important role in DHM's action against depression.
PPI Network of DHM Targets in Depression
We constructed a PPI network by importing 75 potential targets into the STRING database; the network was visualized using Cytoscape software Version 3.9.1. Figure 4 shows the 73 nodes and their interactions. Two targets, CMP-N-acetylneuraminate-beta-1,4-galactoside alpha-2,3-sialyltransferase (ST3GAL3) and galanin receptor type 3 (GALR3) are not presented in Figure 4, due to the fact that they were not connected to this network. In the network, the node represents the relevant gene, with the size and color of the node representing the value of the free degree. Proteins with a "degree" value higher than 31 were selected as the significant targets, which included proto-oncogene tyrosine-protein kinase (SRC), heat shock protein 90 alpha family class A member 1 (HSP90AA1), estrogen receptor 1 (ESR1), hypoxia inducible factor 1 subunit alpha (HIF1A), vascular endothelial growth factor A (VEGFA), silent information regulator 1 (Sirt1) and prostaglandin endoperoxide synthase 2 (PTGS2), all of which may play an important role in DHM's action against depression.
GO and KEGG Pathway Enrichment Analyses for DHM-Targeted Gene Function and Signaling Pathways
Related biological processes and signaling pathways of DHM targets were identified using GO enrichment and KEGG pathway enrichment analyses. A total of 1503 GO terms, including 1316 of biological process (BP), 65 of cellular component (CC), and 122 of molecular function (MF), were acquired. The top 10 terms of BP, CC, and MF are shown in Figure 5. The data demonstrate that these potential targets are predominantly associated with the activities of neurotransmitter receptors, transmitter-gated ion channel and nuclear (steroid) receptors, which mediate depression [9][10][11][12][13][14]; these target genes are also possibly involved in the binding of transcription factors and transcription coactivators, which may mediate the development of depression [15]. These potential targets function mainly in the (post)synaptic membrane, GABA-ergic membrane and (organelle) outer membrane, where receptor-mediated signal transductions may regulate the development of depression [16,17]. The targets possibly play a critical role in response to steroid hormone levels, regulation of inflammation and response to oxygen levels, which are known to be involved in the development of depression [18][19][20]. Additionally, the KEGG pathway enrichment analysis was performed to uncover DHM-mediated functional mechanisms. Among the mapped top 20 signaling pathways (Figure 6), the retrograde endocannabinoid signaling, Rap1 signaling, AGE-RAGE signaling, Estrogen signaling, oxygen reactive species signaling and VEGF signaling pathways are particularly associated with the development and progression of depression [21][22][23][24][25][26].
GO and KEGG Pathway Enrichment Analyses for DHM-Targeted Gene Function and Signaling Pathways
Related biological processes and signaling pathways of DHM targets were identified using GO enrichment and KEGG pathway enrichment analyses. A total of 1503 GO terms, including 1316 of biological process (BP), 65 of cellular component (CC), and 122 of molecular function (MF), were acquired. The top 10 terms of BP, CC, and MF are shown in Figure 5. The data demonstrate that these potential targets are predominantly associated with the activities of neurotransmitter receptors, transmitter-gated ion channel and nuclear (steroid) receptors, which mediate depression [9][10][11][12][13][14]; these target genes are also possibly involved in the binding of transcription factors and transcription coactivators, which volved in the development of depression [18][19][20]. Additionally, the KEGG pathway enrichment analysis was performed to uncover DHM-mediated functional mechanisms. Among the mapped top 20 signaling pathways (Figure 6), the retrograde endocannabinoid signaling, Rap1 signaling, AGE-RAGE signaling, Estrogen signaling, oxygen reactive species signaling and VEGF signaling pathways are particularly associated with the development and progression of depression [21][22][23][24][25][26].
Inhibition of AGE-RAGE Signaling Mediated Inflammation in CORT-Exposed Mouse Hippocampus
We next examined one of the mapped top signaling pathways, AGE-RAGE, in the hippocampus of mice chronically treated with CORT. We found that the levels of AGE and RAGE proteins were significantly increased in the CORT-treated animals, compared to the control animals; DHM-treated animals had significantly lower levels of AGE and RAGE proteins compared to animals treated with CORT alone (Figures 7A,B and S1). Ac-
Inhibition of AGE-RAGE Signaling Mediated Inflammation in CORT-Exposed Mouse Hippocampus
We next examined one of the mapped top signaling pathways, AGE-RAGE, in the hippocampus of mice chronically treated with CORT. We found that the levels of AGE and RAGE proteins were significantly increased in the CORT-treated animals, compared to the control animals; DHM-treated animals had significantly lower levels of AGE and RAGE proteins compared to animals treated with CORT alone (Figure 7A,B and Figure S1). Activation of the AGE-RAGE signaling pathway can lead to upregulation of proinflammatory cytokines [27]. Consequently, we examined the production of IL-1β, IL-6 and TNFα in the hippocampus of these animals by ELISA. The levels of these proinflammatory cytokines were markedly increased in CORT-treated mice compared to the control animals; mice co-treated with DHM had a marked decrease in the expression of these cytokines ( Figure 7C) compared to those treated with CORT alone.
Discussion
In the current study, we demonstrated that DHM relieved chronic CORT-induced depression-like behaviors in mice, supporting the notion that DHM has therapeutic potential for patients with depressive disorders. Additionally, network pharmacology analyses predicted that DHM's efficacy against depression is possibly mediated by different targets and signaling pathways, of which the AGE-RAGE was further examined. We confirmed that the AGE-RAGE signaling pathway was activated and expression of proinflammatory cytokines was upregulated in the hippocampus of mice exposed to CORT, and that co-treatment with DHM counteracted these changes.
Earlier studies have demonstrated that DHM alleviates depression-like behaviors in the lipopolysaccharides (LPS)-induced acute depression mouse model and in the chronic
Discussion
In the current study, we demonstrated that DHM relieved chronic CORT-induced depression-like behaviors in mice, supporting the notion that DHM has therapeutic potential for patients with depressive disorders. Additionally, network pharmacology analyses predicted that DHM's efficacy against depression is possibly mediated by different targets and signaling pathways, of which the AGE-RAGE was further examined. We confirmed that the AGE-RAGE signaling pathway was activated and expression of proinflamma-tory cytokines was upregulated in the hippocampus of mice exposed to CORT, and that co-treatment with DHM counteracted these changes.
Earlier studies have demonstrated that DHM alleviates depression-like behaviors in the lipopolysaccharides (LPS)-induced acute depression mouse model and in the chronic unpredictable mild stress-induced depression model [28,29]. DHM has also demonstrated antidepression effects in a diabetic neuropathic pain and depression comorbidity model [7,30]. Consistently, our data showed that chronic CORT injection resulted in increased immobility time in the FST, while DHM administration exerted an anti-immobility effect. Considering the effect of the level of locomotor activity on immobility time in the FST, the distance traveled by the mice was measured to evaluate locomotor activity in the OFT. Our data show that all the groups traveled an equivalent distance, indicating that neither CORT nor DHM administration influenced the level of locomotor activity. These data further strengthen the anti-despair role of DHM. Additionally, our data also show that chronic CORT injection resulted in decreased sucrose preference behavior in SPT, while DHM treatment boosted sucrose preference in chronic CORT-exposed mice, suggesting an anti-anhedonia role of DHM.
Network pharmacology is a system biology approach that is used to identify potential drug candidates and drug targets and to predict their functional mechanisms [31]. Recently, it has been widely applied in the discovery of antidepressants from natural products including medicinal plants and traditional Chinese medicine [32,33]. Network pharmacology analyses showed that DHM had multiple targets associated with its anti-depression function. Among these core targets, SIRT1, HIF1A, ESR1, VEGFA, HSP90AA1 and PTGS2 are well documented to be associated with depression [26,28,[34][35][36][37]. Furthermore, the KEGG enrichment approach mapped DHM-mediated signaling pathways which are possibly involved in the pathogenesis of depression. One of these pathways, AGE-RAGE, has also been suggested by the network pharmacology approach to be involved in the efficacy of herb medicines against chronic pain-related depression [32,33]. Previous studies have also demonstrated that the alleviation of depressive-like behaviors in rodent models by natural products (p-coumaric acid and melatonin) is mediated by the AGE-RAGE signaling pathway [9,38].
Inflammation plays a critical role in the development and progression of depression [19]. DHM has demonstrated an anti-inflammation capacity in various types of diseases by inactivating the NLRP-3 and NF-κB pathways and decreasing the production of proinflammatory cytokines such as IL-1β, IL-6 and TNFα [39]. Wei et al. reported that DHM reversed depressive-like behavior in LPS-treated rodents by blocking the TLR4/Akt/HIF1a/NLRP3 pathway, inactivating the NF-κB pathway and decreasing the expression of proinflammatory factors [28]. In the chronic unpredicted mild stress-induced depression mouse model, DHM's action against depression is associated with the upregulation of brain-derived neurotrophic factor (BDNF) and inhibition of inflammation in the hippocampus [29]. Similarly, DHM treatment also upregulates BDNF expression and decreases the expression of proinflammatory cytokines in the hippocampus of rats with depression and diabetic neuropathic pain [30]. The AGE-RAGE signaling pathway also regulates inflammation via downstream activation of the NF-κB pathway and the promotion of proinflammatory cytokine generation [27]. In the current study, we found that DHM treatment decreased the levels of AGE, RAGE and proinflammatory factors in CORT-treated mouse hippocampus, suggesting that the AGE-RAGE signaling pathway was inactivated. Previous studies have reported that BDNF inhibits hippocampus inflammation in type 1 diabetic mice via the inactivation of the RAGE-NF-κB pathway [40]. Since DHM has been shown to enhance BDNF expression in two other depression rodent models [29,30], it is reasonable to conclude that DHM also stimulates BDNF expression, controls the AGE-RAGE-NF-κB axis, and inhibits inflammation in the hippocampus of CORT-exposed mice, which is worthy of further investigation.
Depression is also a common clinical feature presented in neurodegenerative disorders, including Alzheimer's disease (AD) and Parkinson's disease (PD) [41,42]. The depressive phenotypes in AD or PD patients are similar to that of depressive patients without AD or PD. Both AD and major depression have overlapping disease mechanisms [43,44]. DHM may therefore offer benefits to patients with neurodegenerative diseases.
In conclusion, our present study showed that DHM eased depression-like behaviors in chronic CORT-treated animals. The protection of DHM against depression may be associated with multiple targets and signaling pathways, among which the AGE-RAGE pathway possibly plays a central role. The data suggest that DHM has therapeutic potential for treating patients with depression.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cells11233730/s1; Figure S1: the image of the original Western blot for Figure 7A;
Institutional Review Board Statement:
The study was performed following the Guidance for the Care and Use of Laboratory Animals, the University of South China (protocol code SYXK2020-0002, approved on 10 January 2020).
|
v2
|
2022-11-25T16:44:09.744Z
|
2022-11-22T00:00:00.000Z
|
253871566
|
s2orc/train
|
Early Postoperative Low Compliance to Enhanced Recovery Pathway in Rectal Cancer Patients
Simple Summary This research investigates the adherence and compliance to the ERAS pathway in patients operated for rectal cancer; the results highlights the important role of early postoperative compliance to the postoperative pathway with the development of complications. Abstract Early postoperative low compliance to enhanced recovery protocols has been associated with morbidity following colon surgery. The purpose of this study is to evaluate the possible causes of early postoperative low compliance to the enhanced recovery pathway and its relationship with morbidity following rectal surgery for cancer. A total of 439 consecutive patients who underwent elective surgery for rectal cancer have been included in the study. Compliance to enhanced recovery protocol on postoperative day (POD) 2 was evaluated in all patients. Indicators of compliance were naso-gastric tube and urinary catheter removal, recovery of both oral feeding and mobilization, and the stopping of intravenous fluids. Low compliance on POD 2 was defined as non- adherence to two or more items. One-third of patients had low compliance on POD 2. Removal of urinary catheter, intravenous fluids stop, and mobilization were the items with lowest adherence. Advanced age, duration of surgery, open surgery and diverting stoma were predictive factors of low compliance at multivariate analysis. Overall morbidity and major complications were significantly higher (p < 0.001) in patients with low compliance on POD 2. At multivariate analysis, failure to remove urinary catheter on POD 2 (OR = 1.83) was significantly correlated with postoperative complications. Low compliance to enhanced recovery protocol on POD 2 was significantly associated with morbidity. Failure to remove the urinary catheter was the most predictive indicator. Advanced age, long procedure, open surgery and diverting stoma were independent predictive factors of low compliance.
Introduction
Enhanced recovery protocols have been associated with a significant improvement of outcome after major surgery for gastrointestinal cancer [1][2][3]. The elderly and patients with multiple comorbidities can be included in the enhanced recovery program, but often require a tailored protocol [4,5].
Early postoperative low compliance to an enhanced recovery protocol has been reported in about one third of patients following elective colonic resection [6,7]. Patients with early low compliance after colonic resection had significant higher morbidity and longer hospital stays [8]. Few data are currently available on both the rate and causes of early low compliance to enhanced recovery protocols following rectal surgery and its relationship with morbidity occurring afterwards.
The first experiences with enhanced recovery protocols were carried out more than 20 years ago. New items reducing perioperative stress and invasiveness of surgery have been subsequently proposed [9,10]. Promising preliminary results have been obtained with low-pressure pneumoperitoneum, multimodal analgesia including abdominal wall blocks, and inferior mesenteric artery preservation in upper rectal cancer surgery [11][12][13][14].
The purpose of this study is to assess which variables can be associated with low compliance to enhanced recovery pathways. The relationship between low compliance and overall postoperative morbidity has also been investigated.
Materials and Methods
The present study is performed in accordance with STROBE guidelines [15]. Consecutive patients who underwent elective surgery for rectal cancer in seven Italian hospitals have been included in the study. Patients with combined resections (rectal and other viscera) were excluded. All patients have been prospectively registered in the database of the PeriOperative Italian Society. Each hospital applied a comprehensive ERAS pathway according to the ERAS ® Society recommendation in colorectal surgery [16] and followed a pathway implementation program before starting the study [8].
In 35 consecutive patients who underwent surgery for upper rectal cancer at Monza Hospital, Monza, Italy, (Monza subgroup), three operative items have been added to the study protocol: TAP (transversus abdominis plane) block instead thoracic epidural catheter, low pneumoperitoneum (8 mmHg), and inferior mesenteric artery (IMA) sparing.
Demographics, perioperative variables, adherence to each item of the protocol, and short-term outcome parameters were prospectively collected in all patients. Indicators of postoperative compliance were naso-gastric tube and urinary catheter removal, recovery of oral feeding and mobilization out of bed, and the stopping of intravenous fluids. Removal of the naso-gastric tube was planned at the end of surgery; and patients were mobilized the day of surgery. The starting of oral feeding and removal of urinary catheter were planned on postoperative day 1. Intravenous fluid infusion was discontinued as early as possible in accordance with the recovery of oral feeding. Low compliance on postoperative day (POD) 2 was defined as non-adherence to two or more items [8].
Criteria to identify each postoperative complication were defined a priori [17] and the Clavien-Dindo classification has been used to grade their severity [18]. Complications graded as IIIb to V were considered as major. Discharge criteria and time to readiness for discharge were defined according to a previous study [19]. Any hospital readmission due to postoperative complications occurring within 30 days after discharge has been registered.
Statistical Analysis
Continuous variables were reported as median along with the interquartile range (IQR) and compared with a Mann-Whitney's U test, while categorical variables were reported as percentages and compared with the Chi square test. Variables predictive of complications were individuated with uni and multiple logistic regression methods. The analysis of factors associated with low compliance on POD2 was carried out in uneventful patients.
Results
The present analysis includes 439 consecutive cancer patients who underwent elective rectal resection. An (American Society of Anesthesiology) ASA score of 3-4 was found in 141 (32.1%) patients, neoadjuvant chemo-radiotherapy was carried out in 113 (25.7%) patients, and laparoscopic surgery was successfully performed in 373 (82.7%) patients (Table 1). The overall adherence to preoperative and operative items was 81.3%. No patient received oral antibiotics before surgery. Mechanical bowel preparation was carried out in 172 (39.3%) patients, while an abdominal drain was placed in 345 (78.8%). The naso-gastric tube was removed at the end of surgery in 398 (90.8%) patients. Table 2 shows that a low protocol compliance on POD 2 was found in one-third of patients. The items with the lowest adherence were removal of the urinary catheter, the stopping of intravenous fluids, and mobilization. Table 3 shows that advanced age, long surgical procedure, open surgery, and diverting stoma were significantly associated to low compliance on POD 2, whereas operative volemia monitoring was associated with high compliance (p = 0.06). Table 4 reports short-term outcomes. Postoperative morbidity occurred in 149 (32.6%) patients and major complications occurred in 27 (6.2%) patients. Twenty-four (5.5%) patients underwent reoperation. Median time to readiness for discharge and length of hospital stay were 5 (4-8) and 6 (5-8) days. The readmission rate was 3.0% (13 patients). Figure 1 shows that patients with low compliance on POD 2 had higher overall morbidity and major complications. At multivariate analysis, failure to remove the urinary catheter on POD 2 was significantly correlated with postoperative complications (Table 5). Figure 1 shows that patients with low compliance on POD 2 had higher overall morbidity and major complications. At multivariate analysis, failure to remove the urinary catheter on POD 2 was significantly correlated with postoperative complications (Table 5). Table 6 reports data on patients of the Monza subgroup who have a higher rate of ASA 3 compared to the overall series. The TAP block and IMA sparing technique were successfully performed in all patients, while low pneumoperitoneum failed in 5 (14.2%) patients who needed an increase up to 12 mmHg. Lymph-node collection and postoperative pain score were similar to the overall series, while early mobilization was observed in 32 (91.4%) patients. No anastomotic leak occurred.
Discussion
Low compliance to an enhanced recovery protocol was found in about one-third of patients after rectal surgery. Patients with low compliance on POD 2 had higher overall morbidity and major complications. Variables associated with early low compliance were advanced age, long procedure, open surgery, and diverting stoma. Upon multivariate analysis, failure to remove the urinary catheter on POD 2 was significantly correlated with postoperative complications.
Operative fluid overload and inadequate pain control can be determinants of postoperative low compliance to enhanced recovery protocol [20][21][22]; however, low compliance can also be considered an early sign for underlying complications. In a series of colon cancer patients, the failure to remove the urinary catheter and to stop intravenous fluids on POD 2 was a predictive indicator of morbidity [8]. To detect an early low compliance might yield to identify patients with higher risk to develop complications afterwards. These patients could benefit from proper diagnostics and the early treatment of complications. This is very important, especially in patients with advanced age and multiple comorbidities.
Previous studies found that minimally invasive colorectal surgery had an independent role to favor early postoperative recovery, to reduce overall morbidity, and to shorten the hospital stay [9,10,23,24]. In the present series, successful laparoscopic surgery was widely performed and conversion to laparotomy was necessary in only 2.7% of patients. A multivariate analysis showed that open surgery was the most important variable associated with low compliance to enhanced recovery protocol on POD 2. Our data also suggest that the elderly and patients who were given a long surgical procedure or diverting stoma had a lower compliance rate. Therefore, a tailored approach with a tight postoperative monitoring should be performed in these patients.
The rate of low compliance on POD 2 was similar to that reported following colonic surgery [8]. The lowest protocol adherence was found for removal of the urinary catheter and the stopping of intravenous fluids, whereas the highest adherence was found for nasogastric tube removal and oral feeding recovery. Early low compliance to postoperative protocol was significantly associated with overall morbidity and major complications. In particular, failure to remove the urinary catheter on POD 2 played an independent role to favor postoperative morbidity. A delayed removal of urinary catheter increases urinary tract infections, and reduces the patient's mobilization favoring respiratory complications [25].
The impact of the three additional operative items was positive. Both TAP block and IMA sparing were successfully performed in all patients, allowing good pain control, no anastomotic leak, and a lymph-node collection comparable to the overall series. A failure of low pneumoperitoneum was recorded in 14% of patients. These promising results and previous reports [20][21][22] should encourage the incorporation of a TAP block, low pneumoperitoneum, and IMA sparing in the enhanced recovery protocols.
A possible limitation of the present study is that participating hospitals could differ in the degree of enhanced recovery pathway implementation. However, the high adherence to preoperative and operative items indicates that the vast majority of patients followed a comprehensive protocol. The wide range of patients' age and ASA score suggests a small likelihood of selection bias.
Conclusions
In conclusion, early low compliance to postoperative enhanced recovery protocols was associated with overall morbidity and major complications following rectal surgery. Variables associated with early low compliance were advanced age, long procedure, open surgery and diverting stoma, suggesting a tailored and careful approach in these patients.
Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board (or Ethics Committee) of ASST Monza (protocol code 0012747 17 April 2020).
Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study; data were collected anonymously without any identifying information.
|
v2
|
2022-11-25T17:03:12.093Z
|
2022-11-22T00:00:00.000Z
|
253872840
|
s2orc/train
|
Identification and Validation of Glycosyltransferases Correlated with Cuproptosis as a Prognostic Model for Colon Adenocarcinoma
Cuproptosis is a newly defined programmed cell death pattern and is believed to play an important role in tumorigenesis and progression. In addition, many studies have shown that glycosylation modification is of vital importance in tumor progression. However, it remains unclear whether glycosyltransferases, the most critical enzymes involved in glycosylation modification, are associated with cuproptosis. In this study, we used bioinformatic methods to construct a signature of cuproptosis-related glycosyltransferases to predict the prognosis of colon adenocarcinoma patients. We found that cuproptosis was highly correlated with four glycosyltransferases in COAD, and our model predicted the prognosis of COAD patients. Further analysis of related functions revealed the possibility that cuproptosis-related glycosyltransferase Exostosin-like 2 (EXTL2) participated in tumor immunity.
Introduction
The incidence of colon adenocarcinoma (COAD) is increasing year by year. Colon adenocarcinoma accounts for about 10% of all cancers and is the second most common cause of cancer death [1]. It is the third most common cancer in men and the second most common cancer in women [2]. The prognosis of colon adenocarcinoma is of great concern, but many factors can affect the patient's prognosis, such as TNM stage, clinical factors, pathological factors, biological factors, and so on. Studies have shown that about 20% of colon adenocarcinoma patients have metastases at diagnosis, and up to half of the patients who initially have localized disease will have metastases [3]. The establishment of a prognosis model for colon adenocarcinoma is of great significance for the selection and optimization of treatment methods.
Cuproptosis is a novel type of cell death that differs from known cell death mechanisms such as apoptosis, autophagy, and ferroptosis [4]. It was first reported by Tsvetkov et al. that excessive accumulation of intracellular copper direct binds to lipoacylated components of the tricarboxylic acid cycle, which leads to aggregation of lipoacylated proteins and destabilizes Fe-S cluster proteins, eventually leading to proteotoxic stress and eventual cell death. Copper is essential for life processes such as energy metabolism, reactive oxygen species detoxification, iron uptake, and signaling in eukaryotes. Mitochondria collect copper for the assembly of copper enzymes [5]. A copper imbalance has been linked to some diseases, including anemia, neutropenia, thrombocytopenia, and tumor development [6][7][8]. Consumption of copper causes metabolic reprogramming from oxidative to glycolytic metabolism and lowers energy generation, which in turn increases tumor angiogenesis [9,10]. Abnormal copper accumulation in cancer cells can be a target for new chemotherapeutic agents, and drugs targeting copper have been utilized for clinical treatment in cancer [11,12]. Copper may play a role in the etiology and progression of cancer. Protein post-translational modifications (PTMs) can change the stability, activity, and cellular localization of proteins, which is an important way to regulate protein function.
The most common post-translational modification in the body is glycosylation, and it is estimated that more than half of all proteins are glycosylated [13]. Changes in the glycosylation pathway actively drive the malignant phenotype of cancer, and glycosylation changes can be detected in the tissues and biological fluids of cancer patients, where these changes have utility as disease biomarkers [14][15][16]. The glycosylation of proteins is inseparable from the action of various glycosyltransferases. Based on sequence similarity, glycosyltransferases have been classified into 78 families (GT family) [17]. Another important enzyme involved in the alteration of glycosylation is glycosidase. It has been shown that glycosidases are higher in tumor interstitial fluid and tumor-bearing animal and human serum. At present, the identification, evolution, and function of glycosyltransferases are still far from being understood and deserve further exploration [18].
However, cuproptosis-related glycosyltransferase in colorectal cancer patients has not been reported as a biomarker. In this study, based on the TCGA database, the prognostic model of COAD was constructed, and several related glycosyltransferases were identified as potential biomarkers. In addition, we conducted a comprehensive analysis of risk models for functional enrichment, drug resistance, immunotherapy, immune infiltration, and somatic mutations.
Data Source
In this study, we downloaded transcriptome data and clinical data from the TCGA database (https://portal.gdc.cancer.gov, accessed on 15 June 2022). The transcriptome data included 398 tumor tissues and 39 normal colon tissues. The clinical data contained information such as age, sex, survival time, survival status, pT stage, pN stage, and pM stage. We selected the clinical and sample information of 367 patients and randomly divided them into training and test groups using R software (version 4.1.3; creator Ross Ihaka and Robert Gentleman, Vienna, Austria).
Recognition of Cuproptosis-Related Glycosyltransferase
To seek cuprotosis-related glycosyltransferases, we retrieved 186 glycosyltransferases from a research article [19]. Nineteen cuprotosis-related genes were identified from a literature review [20]. To find a correlation between cuprotosis-related genes and glycosyltransferases, we used Pearson's correlational analysis to investigate.
Biometric Analysis
We first used univariate COX regression analysis to screen for cuprotosis-related glycosyltransferases associated with prognosis; then, we used LASSO regression to build prognostic models and prevent overfitting. Next, we established cuprotosis-related glycosyltransferases for final inclusion in the model by multivariate COX regression analysis. Principal component analysis (PCA) uses dimensionality reduction to discern potential differences between high-risk and low-risk groups.
Cuproptosis-Related Glycosyltransferase Functional Analysis
Functional analysis of differential genes in high-and low-risk groups was performed by using GO and KEGG enrichment analysis, mainly including the analysis of biological processes (BP), cell components (CC), molecular functions (MF), and biological pathway information. Next, we analyzed the immune function of the differential genes. We analyzed their potential function by using STRING: functional protein association networks (https://string-db.org, accessed on 10 July 2022) to predict the interacting proteins of four glycosyltransferases.
Search for Potential Drugs for the Disease
The drug sensitivity of patients with different risk groups was assessed using the R "pRRophetic" package, which predicts the 50% inhibitory concentrations of chemotherapy drugs common to colon cancer (IC50). We used the Wilcoxon Signature Level Test to evaluate the differences between groups.
Exploring Cuproptosis-Related Glycosyltransferases with Prognostic Value in Patients with COAD
In the study, we extracted 19 cuproptosis-related genes and 186 glycosyltransferases from The Cancer Genome Atlas (TCGA) database COAD array. Co-expression analysis of glycosyltransferase and cuproptosis-related genes was performed to explore the correlation between them. Sankey's diagram was used to demonstrate their correlation ( Figure 1). To further explore the relationship between cuproptosis-related glycosyltransferases and the survival of COAD patients, we identified 19 prognostic glycosyltransferases ( Figure 2A) using univariate COX regression analysis. Nine cuproptosis-related glycosyltransferase genes (MGAT4B, GALNT11, CHPF, CHPF2, CHSY1, CHSY3, EXTL2, POMT2, and DPY19L2) were selected using LASSO regression analysis ( Figure 2B,C). Multivariate COX regression analysis identified four cuproptosis-related glycosyltransferase genes (CHPF2, EXTL2, MGAT4B, and POMT2) ( Figure 2D). These results suggest that cuproptosis-related glycosyltransferases may be utilized to predict the prognosis of patients with COAD. analyzed their potential function by using STRING: functional protein association networks (https://string-db.org, accessed on 10 July 2022) to predict the interacting proteins of four glycosyltransferases.
Search for Potential Drugs for the Disease
The drug sensitivity of patients with different risk groups was assessed using the R "pRRophetic" package, which predicts the 50% inhibitory concentrations of chemotherapy drugs common to colon cancer (IC50). We used the Wilcoxon Signature Level Test to evaluate the differences between groups.
Exploring Cuproptosis-Related Glycosyltransferases with Prognostic Value in Patients with COAD
In the study, we extracted 19 cuproptosis-related genes and 186 glycosyltransferases from The Cancer Genome Atlas (TCGA) database COAD array. Co-expression analysis of glycosyltransferase and cuproptosis-related genes was performed to explore the correlation between them. Sankey's diagram was used to demonstrate their correlation ( Figure 1). To further explore the relationship between cuproptosis-related glycosyltransferases and the survival of COAD patients, we identified 19 prognostic glycosyltransferases ( Figure 2A) using univariate COX regression analysis. Nine cuproptosis-related glycosyltransferase genes (MGAT4B, GALNT11, CHPF, CHPF2, CHSY1, CHSY3, EXTL2, POMT2, and DPY19L2) were selected using LASSO regression analysis ( Figure 2B, C). Multivariate COX regression analysis identified four cuproptosisrelated glycosyltransferase genes (CHPF2, EXTL2, MGAT4B, and POMT2) ( Figure 2D). These results suggest that cuproptosis-related glycosyltransferases may be utilized to predict the prognosis of patients with COAD.
Evaluation and Validation of Prognostic Models
We established a prognostic model by using the results of the multivariate COX regression analysis to verify the prognostic ability of four cuproptosis-related glycosyltransferase genes (CHPF2, EXTL2, MGAT4B, and POMT2). We first calculated the risk score for all patients and divided them into high and low-risk groups according to the median risk score; then, we randomly assigned all patients in the high and low-risk groups to the training group and the test group (Figure 3A,D,G). We found that the high-risk group had a higher death rate than the low-risk group ( Figure 3B,E,H). Glycosyltransferases with risk characteristics were significantly different between the high-risk and low-risk groups. GnT-IVb (N-Acetylglucosaminyltransferase-IVb, encoded by MGAT4B) was expressed highly in the low-risk group, while CHPF2 (Chondroitin sulfate glucuronyl-Cells 2022, 11, 3728 4 of 12 transferase, encoded by CHPF2), EXTL2 (Exostosin-like 2, encoded by EXTL2), and POMT2 (O-mannosyltransferase 2, encoded by POMT2) were highly expressed in the high-risk group ( Figure 3C,F,I). The progression-free survival analysis of the high-and low-risk groups found that the survival rate of the high-risk group was lower than that of the low-risk group ( Figure 3J). In addition, the area under the ROC curve further confirmed the predictive power of the prognostic model, with AUC values of 0.609, 0.616, and 0.548 for 1, 3, and 5 years, respectively.
Evaluation and Validation of Prognostic Models
We established a prognostic model by using the results of the multivariate COX
Verification of the Prognostic Ability of Four Cuproptosis-Related Glycosyltransferases in COAD Patients
In order to investigate the predictive value of cuproptosis-related glycosyltransferases in COAD patients, univariate and multivariate COX regression analyses were utilized. The results revealed a substantial correlation between the risk score and survival rates of COAD patients ( Figure 4A,B), suggesting that the risk score can be utilized as an independent prognostic trait to forecast the prognosis traits of COAD patients. The 1, 3, and 5-year survival calibration curve for COAD patients also further demonstrated the accuracy of prognostic predictive models ( Figure 4C). The C-index curve also showed that the risk score could be used as an indicator of patient prognosis compared to other risk factors ( Figure 4D). Next, we examined the differences between the high-risk group and the lowrisk group using PCA. The findings revealed that there was no difference in all genes, cuproptosis-related genes, and glycosyltransferases ( Figure 4E-G). However, the COAD patients were split into two distinct groups according to the expression of the cuproptosisrelated glycosyltransferases ( Figure 4H). Therefore, the prognostic model can help predict the prognosis of COAD patients and classify patients with different clinical features into high-risk and low-risk groups.
Potential Function of Cuproptosis-Related Glycosyltransferases
We compared the gene expression in the high-risk and low-risk group COAD patients and obtained 299 differentially expressed genes. Then, we analyzed the differential genes using GO and KEGG functional enrichment assays. The GO analysis results showed that these differential genes were mainly enriched in collagen-containing extracellular matrix, extracellular matrix structural constituents, and endoplasmic reticulum lumen ( Figure 5A). The KEGG results showed that the different genes were highly enriched in extracellular matrix organizations, extracellular structure organizations, and external encapsulating structure organizations ( Figure 5B). The results showed that among the 15 genes with the highest mutation frequency, the gene mutation rate of the high-risk group, except APC, was higher than that of the low-risk group ( Figure 5C,D). To find differences in the immune function between the high and low-risk groups, we further analyzed the immune pathways, and the results showed that the highly enriched pathways mainly included the Type II IFN response, parainflammation, and CCR (Chemokine Receptor) ( Figure 5E). These results showed that cuproptosis may be associated with tumor immunity. Four cuproptosis-related glycosyltransferase interaction proteins, including CHPF2, EXTL2, GnT-IVb, and POMT2 were predicted using the STRING website, and the results revealed that most of these glycosyltransferase-interacting proteins were proteoglycans or other glycosyltransferases ( Figure 5F-I). However, the results showed an interaction between EXTL2 and XBP1 (X-Box Binding Protein 1), a key protein in the endoplasmic reticulum stress pathway, suggesting that EXTL2 may play an important role in endoplasmic reticulum stress. XBP1 is located at 22q12.1 on chromosome 12 and consists of six exons. It encodes a 376-amino acid protein, XBP1s, and a 261-amino acid protein, XBP1u [21]. It has been shown that XBP1 binds to cAMP response element (CRE) sites or CRE-like elements in the promoters of target genes [22], which are involved in metabolism, cell proliferation [23,24], and ER stress [25]. XBP1 expression is associated with the prognosis of several cancers, including breast cancer [26] and lung adenocarcinoma [27]. The above results suggest that cuproptosisrelated glycosyltransferases not only play a role in tumor immunity but may also be involved in endoplasmic reticulum stress, which is necessary for cell survival.
The Role of Risk Scores in Drug Therapy
To explore the significance of the prognostic models in drug therapy, we assessed patients with COAD with different risk scores and sensitivity to various anticancer drugs. The statistical results showed that there were significant differences in drug sensitivity between the high-risk and low-risk groups. Patients in the low-risk group were more sensitive to A-770041, rapamycin, and cyclopamine than those in the high-risk group (Figure 6A-C,G-I). However, patients in the low-risk group were less sensitive to sorafenib, pyrimethamine, and MS-275 than those in the high-risk group ( Figure 6D-F,J-L). These results imply that risk scores in prognostic models can assist in the treatment and drug selection of patients with COAD.
Cells 2022, 11, x FOR PEER REVIEW 11 group (C) and the high-risk group (D). Difference analysis of the immune function between the and low-risk groups (E). Prediction of the interacting proteins for cuproptosis-re glycosyltransferases CHPF2, EXTL2, GnT-IVb, and POMT2 by the STRING website (F-I).
The Role of Risk Scores in Drug Therapy
To explore the significance of the prognostic models in drug therapy, we ass patients with COAD with different risk scores and sensitivity to various anticancer drugs statistical results showed that there were significant differences in drug sensitivity bet the high-risk and low-risk groups. Patients in the low-risk group were more sensitive 770041, rapamycin, and cyclopamine than those in the high-risk group ( Figure 6A-C, However, patients in the low-risk group were less sensitive to sorafenib, pyrimethamine MS-275 than those in the high-risk group ( Figure 6D-F, J-L). These results imply tha scores in prognostic models can assist in the treatment and drug selection of patients COAD.
Discussion
In this study, we explored the correlation between cuproptosis and glycosyltransferases and constructed a prognostic model using cuproptosis-related glycosyltransferases, which had not been explored before. In addition, we further analyzed the functions of the four glycosyltransferases involved in the model construction and found that they may be involved in a variety of biological functions, including tumor immunity, endoplasmic reticulum stress, and so on.
Glycosylation of proteins alters their biophysical properties, function, distribution, and retention at the plasma membrane and modulates cell behavior, cell interactions, specific ligand-receptor interactions, and immune recognition [28][29][30][31]. Moreover, studies have shown that glycosyltransferase and glycosylation levels are altered during inflammatory conditions, tumorigenesis, and metastasis [32]. EXTL2, one of three Ext-like genes homologous to EXT1 and EXT2 in the human genome, encodes an N-acetylhexosaminyltransferase. It also plays an important role in the biosynthesis of glycosaminoglycans [33]. Exostosin-like 2 (EXTL2) has dual catalytic activity in vitro. Enzymatic analysis showed that the enzyme could be used as both -GlcNAc and -GalNAc glycosyltransferase synthesis linker. GnT-IVb (encoded by MGAT4B) is an enzyme that catalyzes the formation of the β1,4-GlcNAc branch in N-glycans. The isoenzymes GnT-IVa (N-Acetylglucosaminyltransferase-IVa) and GnT-IVb (MGAT4B) catalyze the synthesis of the β1,4-GlcNAc branch in N-glycans both in vitro and in cells. GnT-IVa and GnT-IVb prefer different glycoproteins. Notably, GnT-IVb acted more efficiently on glycoproteins bearing an N-glycan premodified by GnT-IV [34]. Because the expression levels of each N-glycan branch on specific glycoproteins are highly correlated with various diseases, such as cancer, diabetes, and Alzheimer's disease [35], GnT-IVb may be a potential drug target to combat these diseases. They may act by participating in matrix interactions, glycosphingolipid metabolic pathways, and immune responses.
Furthermore, POMT2, encoding O-mannoyltransferase 2 (POMT2), is one of the pathogenic genes of α-DGP (Alpha-dystroglycanopathy). POMT2 catalyzes the first step in the biosynthesis of α-DG O-mannosylated glycans by transferring mannose to serine or threonine residues [36]. Mutations in POMT2 cause Walker-Warburg syndrome [37]. In recent decades, the activation of programmed cell death has become an intense research focus in cancer research. [38]. Cuproptosis is a new type of cell death, different from any previously known cell death type, which also suggests that cuproptosis may lead to new solutions for cancer treatment. According to a previous study, glycosylation of ATP7A, one of the key enzymes of cuproptosis, promotes the plasma membrane localization and copper sensitive trafficking pattern [39], which implied the possibility that glycosyltransferases regulate cell death by modifying cuproptosis-related genes.
Our study also had some limitations. First of all, our model was created using the TCGA database without clinical sample verification. In addition, although four glycosyltransferase genes (CHPF2, EXTL2, MGAT4B, and POMT2) showed strong prognostic ability in this study, they were limited to a single database. A large number of different data points are lacking for further verification. In conclusion, we screened four cuproptosis-related glycosyltransferases associated with the prognosis of COAD patients and further evaluated their ability to predict survival and prognosis. This study establishes the correlation between cuproptosis and glycosyltransferase in COAD, which may provide a new perspective for the study of cuproptosis and a new strategy for the treatment of COAD patients.
|
v2
|
2022-12-14T16:01:53.661Z
|
2022-11-22T00:00:00.000Z
|
254611581
|
s2ag/train
|
Dynamics of changes in expression of VEGF neoangiogenic factor in tumor tissue bioptates in patients with squamous cell carcinoma of oral mucosa receiving cetuximab treatment and chemotherapy
Purpose of the study. An analysis of changes in the expression of the VEGF neoangiogenic factor in the tumor tissue of patients with squamous cell carcinoma of the oral mucosa receiving targeted therapy with cetuximab and chemotherapy.Patients and methods. We performed an immunohistochemical study of tumor samples obtained from 60 patients with squamous cell carcinoma of the oral mucosa T3-4N0-1M0. The main group comprised 30 patients who received therapy with cisplatin and fluoruracil plus cetuximab. The control group included 30 patients receiving standard chemotherapy without targeted therapy. Each group was divided into two subgroups with different treatment efficacy: patients sensitive to treatment (n = 17 in the group with cetuximab and n = 12 in the group without cetuximab) and resistant to treatment (n = 13 in the group with targeted therapy and n = 18 in the group with standard chemotherapy).Results. Quantification of the VEGF expression demonstrated minimal numbers of vessels stained positively for this marker in the field of view in patients of the main group sensitive to chemotherapy and cetuximab. The value was 5.3 times lower than initial values, and 4.3 times lower than in the subgroup of patients resistant to the treatment (the data were statistically significant, р = 0.0132 and р = 0.0455, respectively). In the control group, patients who were sensitive to the treatment showed 1.4 times lower values than initially (р = 0.921), and patients who were resistant to the treatment had 1.1 times lower values than initial values (р = 0.936). The data were not statistically significant.Conclusions. The study showed that the number of microvessels in patients resistant to chemotherapy and cetuximab was 4.3 times higher than in patients with effective targeted therapy (р = 0.0455). The differences in the control group were not statistically significant.
|
v2
|
2022-11-23T14:36:08.463Z
|
2022-11-23T00:00:00.000Z
|
253797860
|
s2orc/train
|
Development and validation of a chromatin regulator prognostic signature in colon adenocarcinoma
Aberrant expression of chromatin regulators (CRs) could lead to the development of various diseases including cancer. However, the biological function and prognosis role of CRs in colon adenocarcinoma (COAD) remains unclear. We performed the clustering analyses for expression profiling of COAD downloaded from The Cancer Genome Atlas. We developed a chromatin regulator prognostic model, which was validated in an independent cohort data. Time-intendent receiver operating characteristics curve was used to evaluate predict ability of model. Univariate and multivariate cox regression were used to assess independence of risk score. Nomogram was established to assess individual risk. Gene ontology, and Kyoto Encyclopedia of genes and genomes, gene set variation analysis and gene set enrichment analysis were performed to explore the function of CRs. Immune infiltration and drug sensitivity were also performed to assess effect of CRs on treatment in COAD. COAD can be separated into two subtypes with different clinical characteristics and prognosis. The C2 had elevated immune infiltration levels and low tumor purity. Using 12 chromatin regulators, we developed and validated a prognostic model that can predict the overall survival of COAD patients. We built a risk score that can be an independent prognosis predictor of COAD. The nomogram score system achieved the best predict ability and were also confirmed by decision curve analysis. There were significantly different function and pathway enrichment, immune infiltration levels, and tumor mutation burden between high-risk and low-risk group. The external validation data also indicated that high-risk group had higher stable disease/progressive disease response rate and poorer prognosis than low-risk group. Besides, the signature genes included in the model could cause chemotherapy sensitivity to some small molecular compounds. Our integrative analyses for chromatin regulators could provide new insights for the risk management and individualized treatment in COAD.
Introduction
In recent years, the morbidity and mortality of colon cancer have been increasing year by year, becoming one of the main causes of tumor-related death worldwide, which has caused a serious burden on people's health and quality of life (Orangio, 2018;Ahmed, 2020). Metastasis and recurrence are the leading causes of death in most colon cancer patients (Labianca et al., 2010). At present, the main treatment for colon cancer is surgery, preoperative neoadjuvant chemoradiotherapy and postoperative chemoradiotherapy are the routine programs for comprehensive diagnosis and treatment of colon cancer (Wu, 2018). However, due to the insidious onset and asymptomatic progression of colon cancer, some patients with colon cancer are already in the middle and advanced stages when they are diagnosed, and conventional treatment cannot prolong the survival time of these patients (Freeman, 2013). Clinicians mainly assessed the prognosis of colon cancer patients by the disease process and tumor stage at the time of diagnosis (Pacal et al., 2020;Cerrito and Grassilli, 2021). However, traditional methods are insufficient to accurately assess the prognosis of colon cancer patients. Therefore, identifying biological markers related to colon cancer prognosis and survival is of great significance for patients with colon cancer.
Chromatin regulators (CRs) are a class of enzymes with specialized functional domains capable of recognizing, forming, and maintaining epigenetic states in a cellular contextdependent manner (Frye and Benitah, 2012). CRs are indispensable upstream regulators of epigenetics (Lam et al., 2005). According to their regulatory roles in epigenetics, CRs are generally classified into three major categories: deoxyribonucleic acid (DNA) methylation, histone modifications, and chromatin remodelers. Aberrant expression of CRs is associated with various biological processes such as inflammation, apoptosis, autophagy, and proliferation, suggesting that dysregulation of CRs may lead to the development of various diseases including cancer (Begolli et al., 2019;Smits et al., 2020;Lee and Kim, 2021). Therefore, CRs are expected to become new targets for the treatment of various diseases. However, the biological function and prognosis role of CRs in COAD remains unclear.
Many studies have shown that differences in tumor microenvironment, targets, and genes enhance the effects of traditional treatments and supplement the deficiencies of previous studies (Ansari et al., 2020;Lin et al., 2020). In the process of tumor progression, diagnosis, treatment and prognosis, bioinformatics has gradually played an important role with the continuous in-depth research of next-generation sequencing and big data centers (Jacoby et al., 2015). Through the analysis and comparative study of big data gene chip information, to calculate differential genes and immuneinfiltrating cell screening in colon cancer to provide important biological prediction data for tumorigenesis mechanism and prognosis . In our current research, we first explored the landscapes of chromatin regulators including differentially expressed genes, regulation network, correlations, and gene alterations in colon adenocarcinoma (COAD). Next, we performed the clustering analysis and identified the molecular subtypes and explored the characteristics of different subtypes. Then, we developed a prognostic model based on chromatin regulators in COAD, and validated the utility of this model in an independent cohort dataset, followed by the identification of an independent prognosis factors of risk score calculated by the chromatin regulators. Subsequently, we constructed a nomograph scoring tool for predicting the individual prognosis outcomes. Finally, we explored the pathways enrichment, immune filtration in different risk setting, evaluated the effect of chromatin regulators on immunotherapy in a cohort dataset, and identified the potential small molecular compounds associated with chemotherapy sensitivity. Our study highlights important role of chromatin regulators and provides new insights for individualized treatment in COAD.
Data source
We downloaded the sequencing expression data of COAD from The Cancer Genome Atlas (TCGA: https://portal.gdc. cancer.gov/) including 473 tumor samples and 41 normal samples. We exclude these samples with mean absolute deviation (MAD)<0.1. The clinical information was also extracted, including age, gender, stage, TNM classification. The other independent dataset was also downed from the Gene Expression Omnibus database (https://www.ncbi.nlm. nih.gov/geo/) (GSE103479: 156 patients with colon carcinoma). The gene alterations and copy number variations were also obtained. We obtained the 870 chromatin regulators from the previous studies (Lu et al., 2018).
Differential expression and gene alterations analysis
Using "limma" package, we identified the differentially expressed genes (DEGs) with the |log fold change|>1 and FDR p < 0.05. We built the protein-protein interaction network (PPI) using these DEGs in the STRING database (http://string-db.org), and these data were entered into Cytoscape Version 3.8 and generate the PPI network. We explored the correlation among these regulators using the Pearson correlation. Using "maftools" package, we analyzed the gene alterations and copy number variation in COAD.
Frontiers in Genetics frontiersin.org
Identification of molecular subtypes
We first identified risk and favorable factors of the CRs using univariate cox regression. We performed the clustering analyses using "ConsensusClusterPlus" package, and identified the optimal the number of K using consensus matrix and consensus cumulative distribution function plot (Wilkerson and Hayes, 2010). Principle component analysis was used to validate the subtypes distributions. The Kaplan-Meier analysis was used to compare the survival curve between different subtypes. We explored the correlations of molecular subtypes with clinical characteristics using the Chi-square test.
To explore the differences in different subtypes, we calculated the enrichment score of each sample using this dataset: c2. cp.kegg.v7.4. symbols and performed the get set variation analysis (GSVA) using GSVA package (Hanzelmann et al., 2013). We also compared the immune status between two subtypes including estimate score, stromal score, immune score, and tumor purity. The infiltration levels of immune levels were also evaluated.
Development and validation of prognostic model based on chromatin regulators
We first performed a univariate cox regression and identified the prognosis-related chromatin regulators (p < 0.001) in the TCGA training cohort. The least absolute shrinkage and selection operator (LASSO) regression was used to the identify the best genes number, followed by the multivariate cox regression to achieve the regression coefficient of the included genes in the model. We calculated the risk score of each sample according to the following formula: risk score = coef 1 *gene 1 expression+. . .+coef n *gene n expression. The COAD patients were divided into high-risk group and low-risk group according to the median of risk score. The Kaplan-Meier survival curves of different risk groups were plotted. We validated this established model using an independent cohort data (GSE103479). The time-intendent receiver operating characteristics curve (ROC) was plotted to calculate the area under the curve (AUC) at 1-year, 2-year, and 3-year in both TCGA training cohort and GEO validation cohort. PCA was also performed to identify the risk groups.
Clinical characteristics and independent analysis
To investigate the correlations of risk groups with clinical characteristics, we compared the risk scores among different age (age>=60 vs. < 60), gender (male vs. female), stage (I-II vs. III-IV), T (T1-2 vs. T3-4), N (N0 vs. N1-2), M (M0 vs. M1) classification. We also showed the clinical characteristics and identified genes expression level between high-risk and low-risk groups. We further performed the univariate and multivariate cox regression to detect whether risk score could be an independent prognosis predictor of overall survival in COAD.
Nomogram establishment and assessment
To estimate the individual's prognosis risk, we built a nomogram score tool based on the following clinical characteristics: risk score, age, gender, stage, TNM classification. Using this nomogram tool, we can easily calculate the 1-year, 3-year, and 5-year overall survival (OS) rate. We plotted the calibration fitting line between observed OS and nomogram-predicted OS at 1-year, 3-year, and 5-year, which can assess the accuracy of nomogram.
Then, we calculated the AUCs of all clinical parameters, risk score and nomogram tool, and identified the predictive ability of nomogram tool. Decision curve analysis was used to determine the clinical practicability of nomograms based on the net benefit according to different threshold probabilities in COAD patients.
Function enrichment and immune infiltration
To explore the biological function of different risk groups, we performed the gene setting enrichment analysis (GSEA) in highrisk group and low-risk group, respectively. We identified the top 5 signaling pathways of high-risk group and low-risk group. Then, we explored the immune infiltration status of high-risk and low-risk groups. We explored the correlations of risk score with immune cells by calculating the correlation coefficient. The tumor mutation burden level was also evaluated.
Immunotherapy and chemotherapy sensitivity
To explore the effect of chromatin regulators on treatment, we first calculated the tumor immune dysfunction and exclusion level (Chen et al., 2022). Based on tumor pre-treatment expression profiles, this tumor immune dysfunction exclusion (TIDE) module can estimate multiple published transcriptomic biomarkers to predict patient response to immunotherapy. We also used the IMvigor210 cohort for validating the effect of CR regulators on immunotherapy, and the immunotherapy cohort data for urothelial carcinoma (Mariathasan et al., 2018). Using the CellMiner database, we further explored the chemotherapy sensitivity by calculating the Pearson correlation coefficients (Reinhold et al., 2012). |R|>0.25 and p < 0.05 were considered significantly correlated.
Landscapes of chromatin regulator in colon adenocarcinoma
We depicted the landscapes of chromatin regulators in COAD using TCGA dataset. The flow of data processing was presented in Figure 1. We first performed differential expression analyses between tumor and normal samples using limma with | logFC|>1 and false discovery rate p < 0.05, and obtained 124 DGEs including 105 up-regulated genes and 19 downregulated genes. The volcano plot presented the distributions of DGEs between tumor and normal samples ( Figure 2A). Next, we built protein-protein interaction network and identified top 10 hub genes using normalized cross correlation methods (CHEK1, CDK1, TOP2A, CDC6, BUB1, AURK, TTK, RAD54L, PBK, UHRF1, Figure 2B). Among these chromatin regulators, we further identified 50 genes related to overall survival, including 6 favorable genes and 44 risky genes ( Figure 2C). Then, we explored the correlations among these chromatin regulators, and found ZNF592-BAHD1 and ARID3B, PHF21A-ZBTB4 and ZNF532, BRD3-PHF2 and BRD2, APOBEC3F-APOBEC3C and SP140 showed strong positive correlations (r > 0.5) while BCL10 showed negative associations with other genes ( Figure 2D). Finally, we analyzed the gene alterations of chromatin regulators in COAD. Our results indicated that the gene alterations ranged from 9% to 0% and top gene alterations of chromatin regulators in COAD were CHD4 (9%), CHD3 (7%), PPARGC1A (5%), PNK1 (5%), and PHF2 (4%) ( Figure 2E). The C > T variations accounted for most of single nucleotide polymorphism in COAD. Figure 2F show the locations of gene mutations in chromosome.
Identification of molecular subtypes
Using the chromatin regulators related to prognosis, we performed the consensus analysis. The consensus matrix showed that the optimal number is 2 ( Figure 3A). The consensus CDF achieved the best values when the number of clustering was 2 ( Figure 3B). The COAD can be divided into two subtypes (C1 = 187, C2 = 260). Then, the Kaplan-Meier analyses indicated that the C2 group had poorer prognosis than C1 group (p < 0.003, Figure 3C). The PCA also showed that COAD patients presented two distinguished two components. The Cluster 2 tend to be T III-IV stage (p < 0.01). There were no significant differences in age, gender, stage, N, M classification The flow chart of integrative analysis.
Frontiers in Genetics frontiersin.org ( Figures 3D,E). Some chromatin regulators were significantly down-regulated such as ORC1, MAPKAPK3, ELP3, TDRD7, and PPARGC1A. Furthermore, the GSVA indicated that some signaling pathways were significantly positive enriched in C2 such as Notch signaling pathways, GNRH signaling pathway, BASAL cell carcinoma, glycosaminoglycan biosynthesis chondroitin sulfate, ECM receptor interaction, focal adhesion, and MAPK signaling pathways. The glutathione and pyruvate metabolism, oxidative phosphorylation, peroxisome, terpenoid backbone biosynthesis, and citrate cycle tricarboxylic acid cycle were upregulated in C1 group ( Figure 4A).
Finally, we explored the immune infiltration status of two subtypes. The C2 had higher estimate, stromal and immune scores than C1 (Figures 4B-D). However, the tumor purity of C2 group was lower than C1 group ( Figure 4E). The C2 group also have higher B cells naïve, NK cells activated, and macrophages M0 infiltration levels while the plasma cells, Tcells CD4 memory activated, dendritic cells activated, mast cells activated, eosinophils and neutrophils level of C1 group were significantly elevated ( Figure 4F). We then considered the C2 group as "hot tumor" and C1 group as "cold tumor."
Development and validation of prognostic model based on chromatin regulators
We first developed the prognostic model in TCGA training cohort. Using the FDR p < 0.01, we identified the 18 genes related to prognosis in COAD including two favorable genes (PPARGC1A and MAPKAPK3) and 16 risky genes ( Figure 5A). We next performed the LASSO regression and identified the genes and number included in the prognostic model ( Figures 5B,C). Twelve genes were included in the final model, and we established the following formula for calculating the risk score of each sample: risk score = EXPAPOBEC3F*0.142 + EXPSMARCD3 * 0.376 -PPARGC1A * 0.223 + BRD9*0.370 + JDP2*0.592 + NEK9 * 0.028 + BAHD1 * 0.366 + PHF2 * 0.063 + PHF1*0.158 + PYGO2*0.435 -MAPKAPK3 * 0.577 + GADD45B * 0.007. We divided the COAD patients into high-risk group (n = Frontiers in Genetics frontiersin.org 223) and low-risk group (n = 224). The Kaplan-Meier analysis indicated that the high-risk group had worse overall survival than low-risk group (p < 0.001, Figures 5D,E). PCA also indicated two different risk groups ( Figure 5F). Subsequently, we validated this model in an independent cohort data. Our results showed that the established model was well validated in this cohort ( Figures 5G-I). The 1-year, 2-year, and 3-year AUCs were 0.735, 0.756, and 0.721 in the training cohort ( Figure 5J). The AUCs were 0.592, 0.585, 0.606 at 1-year, 2-year, and 3-year, respectively ( Figure 5K).
Clinical correlations and independent analysis
We further analyzed the correlations of risk score with clinical characteristics. The results indicated that age and gender were not associated with risk score (Figures 6A,B), while the patients with Stage III-IV, T3-4, N1-N2 and M1 had elevated risk score (Figures 6C-F). The high-risk group tend to be advanced clinical stage ( Figure 6G). APOBEC3F, SMARCD3, BRD9, JDP2, NEK9, NAHD1, Frontiers in Genetics frontiersin.org PHF2, PHF1, PYGO2, and GADD45B were significantly highexpressed in high-risk group. The univariate indicated that elevated risk score was significantly with poor overall survival (HR:3.34, 95%CI: 2.394-4.658, p < 0.001, Figure 7A), and the multivariate cox regression risk score is an independent prognosis predictor for COAD patients (HR:2.770, 95%CI: 1.960-3.915, p < 0.001, Figure 7B). Besides, Age, M1, and N1-2 classification were also risk factors for overall survival in COAD. Using clinical parameters and risk score, we built the nomogram score system ( Figure 7C). We estimated the 1-year, 3-year and 5-year OS were 0.94, 0.853, and 0.765 for an 85- year male patient with T3, N1, low-risk, and Stage II. The calibrations plots of 1-year, 3-year and 5-year showed the nomogram-predicted OS and observed OS can be fitted well. Furthermore, the nomogram achieved the best predict ability (AUC = 0.801) followed by risk score (AUC = 0.740, Figure 7D). The decision curve analysis also indicated the nomogram can be well applied in the clinical practice because the nomogram has the best net benefit ( Figure 7E).
Function enrichment and immune infiltration
The GSEA indicated the top 5 enrichments were cell adhesion molecules cams, cytokine receptor interaction, extracellular matrix receptor interaction, focal adhesion, and hematopoiesis cell lineage in high-risk group ( Figure 8A), while the top 5 enrichments were oxidative phosphorylation, Parkinson's disease, proteasome, ribosome, and systemic lupus Frontiers in Genetics frontiersin.org erythematosus in low-risk group ( Figure 8B). The risk score was positively associated with APOBEC3F, SMARCD3, BFRD9, JDP2, NEK9, BAHD1, PHF1, PHF2, PYGO2, and GADD45B. the PPARGC1A and MAPKAPK3 were negatively associated with risk score ( Figure 8C).
The immune infiltration analyses indicated that risk score was positively associated with T cells, CD8T cell, cytotoxic lymphocytes, B lineage, monocytic lineage, myeloid dendritic cells, endothelial cells, and fibroblasts ( Figure 8D). The risk score was also positively related to tumor mutation burden (TMB) level ( Figure 8E).
Immunotherapy and chemotherapy sensitivity
We also explored the effect of chromatin regulators on immunotherapy. We first evaluated the tumor immune dysfunction and exclusion level (TIDE). Our results indicated that the high-risk group had higher TIDE, exclusion, and dysfunction levels except MSI ( Figures 8F-I), which means the high-risk group had poor response to immunotherapy. The IMvigor data confirmed our results. The stable/ progression disease group had higher risk score than the com complete/part remission ( Figure 9A). Furthermore, the high-risk had poorer overall survival than low-risk group ( Figure 9B, p < 0.001). We then evaluated the effects of signature genes on chemotherapy sensitivity. We found that GADD45B can cause chemotherapy resistance to Bafetinib, Vmurafenib, Selumetinib, Dabrafenib, Cobimetinib, Hypothemycir, Trametinib, and Nilotinib. MAPKAPK3, BRD9, JDP2, PPARGC1A can enhanced the sensitivity of some small molecular compounds, including Fludarabine, Cladribine, 5-fluoro deoxy urine uracil, Acetalax, and Dabrafenib ( Figures 9C-R).
Discussion
The present study has the following several findings (Ahmed, 2020): The COAD can be separated into two subtypes with clinical characteristics, prognosis outcomes, and biological function enrichment. The C2 group had elevated immune infiltration levels and low tumor purity, which can be considered as "hot tumor" and the C1 group had low level Frontiers in Genetics frontiersin.org 12 immune status considered as "cold tumor" (Orangio, 2018). Using 12 chromatin regulators, we developed a prognostic model that can predict the overall survival and risk classifications among COAD patients. This model was well validated in an independent external cohort data (Labianca et al., 2010). We built a risk score that can be an independent prognosis predictor of COAD. The high-risk group based on risk score tended to have risky clinical characteristics (Wu, 2018). Using clinical parameters and risk score, we built the nomogram score system that can achieve the best predict ability and were also confirmed by decision curve analysis about its clinical application (Freeman, 2013). There were significantly different function and pathway enrichment, immune infiltration levels, and TMB level between high-risk and low-risk group (Pacal et al., 2020). The high-risk group had poor response to immunotherapy. The external validation data also indicated that high-risk group had higher SD/PD response rate and poorer prognosis than low-risk group. Besides, the signature genes included in the model could cause chemotherapy sensitivity to some small molecular compounds. Our integrative analyses for chromatin regulators could provide new insights for the risk management and individualized treatment in COAD.
Epigenetic changes, considered to be one of the most important markers of tumors, are driven by chromatin regulator (Dey, 2011;Florea and Karaoulani, 2018). The chromatin regulators dynamically regulate chromatin structure and epigenetic regulation of gene expression in response to endogenous and exogenous signaling cues (Weaver and Bartolomei, 2014). Somatic changes or misexpression of CR may reprogram the epigenetic map of chromatin, leading to a wide range of common diseases, especially cancer (Shu et al., 2012). Currently, the function role of chromatin regulators in COAD is still unclear. We first explored the relevance in prognosis and treatment for COAD. We identified two molecular subtypes using prognosis-related chromatin regulators. Two subtypes had different expression profiling of chromatin regulators and clinical characteristics. The cluster 2 showed elevated stromal and immune activation and was mainly enriched in some important tumor-related signaling pathways such as Notch, Gnrh, and MAPK signaling pathways, which had been suggested to be closely associated with tumor occurrences (Kranenburg, 2015;Lajko et al., 2019;Tang et al., 2021). ECM receptor interaction and focal adhesion were also highly enriched in cluster 2. On the contrary, the cluster 1 had low immune infiltration level and was mainly enriched in some metabolism-related pathways and functions such as glutathione, pyruvate, TCA cycle and oxidative phosphorylation. Thus, the cluster 2 can be regarded as "hot tumor," and the cluster 1 was called "cold tumor." Whether the tumor is hot or cold affects whether immunotherapy, represented by PD-1 inhibitors, is effective. This is because tumor cells overexpress PD-L1 protein and induce high expression of PD-1 on immune cells such as T lymphocytes. When the two are combined, they inhibit the function of T lymphocytes, allowing tumors to escape immune attack (Reschke and Olson, 2022).
Using these chromatin regulators, we established a prognostic model with twelve chromatin regulators. Previous studies also established prognostic models using other gene sets. Zhou et al. developed an autophagy-related lncRNA model for COAD and the 3-year predictive AUC was 0.790, which was close to our model (Zhou et al., 2020). Using 44 ferroptosis-related lncRNAs, Li developed a prognostic model with AUC of 0.860 that was slightly higher than our AUC . Li also built a prognostic model using immune-related genes, and the predictive ability was 0.792. Broadly speaking, all these model had similar predictive abilities, which suggested that our model was effective (Miao et al., 2020).
In this model, PPARGC1A and MAPKAPK3 were favorable genes in this model. Previous study had reported that the expression of PPARGC1A was negatively associated with some immune cells, which means that PPARGC1A may be responsible for regulating the immune components of tumor microenvironment (Ma et al., 2021). As a member of the Ser/Thr protein kinase family. MAPKAPK3 functions as a mitogenactivated protein kinase (MAP kinase)-activated protein kinase. Previous studies reported that ERK, p38 MAP kinase and Jun N-terminal kinase were all able to phosphorylate and activate this kinase, which suggested the role of this kinase as an integrative element of signaling in both mitogen and stress responses (Wagner and Nebreda, 2009;Sun et al., 2015). It was reported that MAPKAPK3 can promote autophagy via some phosphorylation pathway in vivo and vitro, which may explain its favorable role in COAD (Wei et al., 2015). The other 10 gene were oncogenes in the model. Such as APOBEC3F that could be a new treatment target in multiple cancers including COAD (Svoboda et al., 2016). SMARCD3 (Jiang et al., 2020), BRD9 (Sabnis, 2021), JDP2 (Mansour et al., 2018), were also reported to be a oncogene role in some cancer. We calculated the risk score for each sample based on the established prognostic model and divided COAD patients into high-risk and low-risk groups. The high-risk group and low-risk group had different overall survival. The time-independent ROC indicated that the prognostic signature with 12 chromatin regulators had accurate and reliable predictive ability. The established model was effectively validated in an independent cohort data. The univariate and multivariate cox regression also demonstrated that risk score was an independent risk factor for poor overall survival. Based on the risk score and clinical parameters, we constructed a nomogram scoring tool for individual' survival outcomes. The calibration, ROC and decision curve analysis had excellent predictive ability.
The risk score was found to be positively associated with many immune cells including T cells, CD8 T cells, monocytic lineage, endothelia cells and fibroblasts. We also found that the high-risk group and low-risk group had different immune Frontiers in Genetics frontiersin.org infiltration levels. Immune cell infiltration in tumor microenvironment affects the prognosis of tumor therapy (Bader et al., 2020;Lei et al., 2020). To explored the effect of chromatin regulators on immunotherapy, we further evaluated the TIDE levels of different risk groups. We found that the highrisk group had relatively high immune status including TIDE, exclusion, and dysfunction, which means the high-risk group may have poor prognosis when receiving immunotherapy. The data from an immunotherapy cohort data (IMvigor210) confirmed these assumptions that patients with high-risk score and immune infiltration had poor prognosis (Vander et al., 2019). Recently, several clinical trials had been performed to explore the efficacy of immunotherapy (Bao et al., 2020;Mlecnik et al., 2020). Our results provided some references for these researches. Finally, we evaluated the effect of chromatin regulators on chemotherapy sensitivity, and found GADD45B can cause chemotherapy resistance to Bafetinib, Vmurafenib, Selumetinib, Dabrafenib, Cobimetinib, Hypothemycir, Trametinib, and Nilotinib. MAPKAPK3, BRD9, JDP2, PPARGC1A can enhanced the sensitivity of some small molecular compounds, including Fludarabine, Cladribine, 5-fluoro deoxy urine uracil, Acetalax, and Dabrafenib. These findings will help clinical treatment for COAD patients.
The present study had several limitations. First, the sample size of validation cohort was small, and study with larger sample size were required. Based on suggestions from professional filed, at least two independent cohorts were required for the present prognostic model. Second, the biological function, molecular mechanism and the effect of chromatin regulators were not validated through experiments in vivo and vitro. Data from experimental research will further refine the present findings. Although we evaluated the effect of chromatin regulators on immunotherapy using two different methods, the immunotherapy was carried out in the other tumor types. Studies performed in COAD will be more persuasive.
In conclusion, we obtained two molecular subtypes in COAD using chromatin regulators, which had different clinical characteristics and immune landscapes. We further established and validated a chromatin-related prognostic model that can be capable of predicting overall survival of COAD patients. More important, we also found that chromatin regulators could affect the immunotherapy and chemotherapy sensitivity in COAD patients. Our study will provide new risk management and individualized treatment strategies for COAD that could bring more benefits for patients.
Author contributions
WY designed this study and contributed substantially to the design of the search strategy. WL and CL searched and selected the trials and extracted data. WY performed the analysis and interpreted the data. WY wrote the manuscript. CL and SC critically reviewed the manuscript. All authors read and approved the final manuscript.
|
v2
|
2022-11-24T06:17:19.385Z
|
2022-11-23T00:00:00.000Z
|
253800647
|
s2orc/train
|
Serum amyloid A and other clinicopathological variables in cats with intermediate- and large-cell lymphoma
Objectives Serum amyloid A (SAA) concentrations are increased in cats with lymphoma vs healthy cats; however, the association between SAA concentrations and prognosis in cats with lymphoma is unclear. The aim of this study was to evaluate if SAA concentrations were different in cats with nasal vs non-nasal lymphoma, if SAA concentrations are prognostic in patients treated with high-dose chemotherapy and if SAA concentrations are correlated with other clinicopathological variables. Methods Cats diagnosed with intermediate- or large-cell lymphoma between 2012 and 2022 with SAA concentration data available were included. Associations between tumour site (nasal vs non-nasal), stage, response to treatment and SAA concentration were evaluated using non-parametric statistics. Associations between SAA concentrations and stage with survival time were evaluated using Cox regression analysis. Patients with nasal tumours and those not receiving high-dose chemotherapy were excluded from the survival analyses. Results Thirty-nine cats were included. Median SAA concentrations were significantly higher in non-nasal compared with nasal lymphoma (42 µg/ml [range <0.3–797] vs <0.3 µg/ml [range <0.3–0.9]; P = 0.026). SAA concentrations did not correlate with tumour stage. Median survival time for patients with non-nasal tumour and undergoing chemotherapy was 49 days (range 2–1726). Responders had a better median survival time than non-responders (273 days [range 43–1728] vs 39 days [range 2–169]; P <0.001), whereas SAA concentrations were not associated with survival time. Lower haematocrit at presentation was associated with a reduced median survival time (P = 0.007). Conclusions and relevance In the population examined, no correlation between serum concentration of SAA and prognosis in patients with lymphoma was identified, while low haematocrit and lack of response to treatment were both found to be associated with survival time. SAA concentrations were elevated in patients with non-nasal lymphoma vs patients with tumours confined to the nasal cavity.
literature has suggested that a lomustine-based protocol is a viable alternative as a first-line and rescue treatment for cats with this malignancy. [9][10][11] Median survival times in cats with lymphoma range from 27 to 955 days, with a response rate to chemotherapy ranging from 22% to 95%. 1, 6,7,12 Compared with canine lymphoma, in which several prognostic factors such as the immunophenotype and the stage of the disease have been identified, in cats the response to treatment remains the most significant factor associated with survival time. 1, 6,7,[13][14][15] Other previously reported negative prognostic factors in cats include central nervous system (CNS) or bone marrow involvement, clinical substage b, granular cell morphology and anaemia. 1,4,7,16 In contrast, disease confined to the nasal cavity is considered to have a better prognosis. 6,12,17 Serum amyloid A (SAA) is a major acute phase protein (APP) in humans and cats. 18,19 The protein is mainly produced by the liver under the influence of inflammatory cytokines such as interleukin (IL)-1 and IL-6, and tumour necrosis factor-alpha in response to inflammation or tissue damage. 20,21 The role of SAA in the inflammatory response is not completely understood; however, previous literature suggests that the protein acts as an inflammatory mediator, particularly increasing cytokine production by monocytes and macrophages. [21][22][23] SAA concentrations were found to be increased in several inflammatory and neoplastic diseases in humans and cats, and were recognised as a marker of distant metastatic disease in people with different neoplastic diseases. [24][25][26] Additionally, in cats, SAA concentrations are an independent prognostic marker in patients with different diseases, including different types of neoplasia. 27 SAA concentration is higher in cats with lymphoma vs a healthy population and decreased in patients achieving remission after undergoing chemotherapy. 28 However, to our knowledge, no previous studies have studied the prognostic value of SAA concentrations in cats with lymphoma before undergoing chemotherapy.
The primary aim of the study was to correlate the SAA concentration with lymphoma in different tumour locations (nasal vs non-nasal), with tumour stage and with survival time. A secondary aim was to correlate SAA concentrations with other clinical and pathological variables, which were previously advocated as potential prognostic factors in other studies, in the same cohort of patients.
Materials and methods
The medical records of a university referral hospital were reviewed between January 2012 and March 2022 for cats diagnosed with lymphoma. Patients were included if they had a cytological or histological (± immunohistochemistry) diagnosis of intermediate-to-large-cell lymphoma. Patients with a diagnosis of small-cell or large granular-cell lymphoma were excluded from the study. SAA concentration data were retrieved from the previous medical history or measured using archived frozen (-80°C) samples obtained at the time of presentation. SAA concentrations were measured on an Olympus AU400 or AU480 analyser using a human immunoturbidimetric assay previously validated for use in cats. 29 The laboratory reference interval (determined as part of an internal validation study) for SAA concentration was <0.5 µg/ml and the limit blank of the assay was 0.3 µg/ml. Neutrophil count, haematocrit (Hct) and serum albumin at presentation were also recorded. Patients were excluded if they received any treatment with chemotherapy, steroids, radiation therapy or surgery before SAA measurement. Follow-up information was retrieved from the medical records of the hospital or by contacting the referring veterinary surgeons. Only patients that underwent full staging, including thoracic (thoracic radiographs or thoracic CT), abdominal (abdominal ultrasound [US]) imaging and fine-needle aspiration of the spleen and liver were included when evaluating the association between tumour stage and SAA concentrations. Based on these findings, patients were retrospectively staged according to a previously described staging system (Table 1). 30 The number of sites affected was also recorded. The time from treatment initiation to restaging was based on clinician preference. Patients were divided into two groups based on clinical response: responders (partial or complete remission) and non-responders (stable disease, progressive disease). Response to treatment was classified as complete remission (CR) if a complete disappearance of the visible disease was noted; partial remission (PR) if the lesion was reduced by at least 30% of its initial size but had not completely disappeared; stable disease (SD) if the disease had either reduced by less than 30% or increased up to 20%; or progressive disease (PD) if the disease had increased by more than 20%. Chemotherapy adverse events and their grade were classified according to the Veterinary Cooperative Oncology Group criteria. 31 Median survival time was measured from the time of diagnosis to the time of death for any cause. Patients that did not receive high-dose chemotherapy, patients diagnosed in 2022 (due to short follow-up time) and patients with nasal lymphoma were excluded from the responseto-treatment and survival analysis. Exclusion criteria for the single groups analysed are summarised in Table 2.
Statistical analysis was performed using commercially available statistical software (SPSS version 25.0; IBM). For statistical purposes, a SAA concentration below the limit of blank of the assay (0.3 g/l) was assigned an arbitrary value of 0.15 mg/l. The Kruskal-Wallis or Mann-Whitney U-test was used to compare SAA concentrations between groups (stage and site). Spearman's correlation coefficient was used to evaluate associations between continuous variables. Kaplan-Meier survival curves were constructed to calculate the median survival times of different groups, with survival between different groups compared using the log-rank test. Univariable Cox regression analysis was used to evaluate the association between SAA concentration, stage, site (nasal vs non-nasal) and response to treatment (responder vs non-responder) with overall survival time (all-cause mortality). Data are presented as median (range), unless otherwise specified, and a P value <0.05 was considered to be statistically significant.
Diagnosis
Diagnosis was achieved by cytology alone in 21 cases and by histology alone in six. Diagnosis was confirmed Table 1 Clinical staging system for patients with lymphoma (based on Mooney and Hayes) 30 Staging system for feline lymphoma Stage 1 • A single tumour (extranodal) or single anatomical area (nodal), including primary intrathoracic tumours Stage 2 • A single tumour (extranodal) with regional lymph node involvement • Two or more nodal areas on the same side of the diaphragm • Two single (extranodal) tumours with or without regional lymph node involvement on the same side of the diaphragm • A resectable primary gastrointestinal tract tumour, usually in the ileocecal area, with or without involvement of associated mesenteric nodes only Stage 3 • Two single tumours (extranodal) on opposite sides of the diaphragm • Two or more nodal areas above and below the diaphragm Stage 4 • Stages 1-3 with liver and/or spleen involvement Stage 5 • Stages 1-4 with initial involvement of the central nervous system or bone marrow, or both Patients with small-cell or large granular-cell lymphoma and patients that received steroids before the diagnosis were excluded from the initial population examined SAA = serum amyloid A; Hct = haematocrit by cytology and histology in 12 cats. The lymphoma was classified as intermediate cell in seven cases and as large cell in the remaining 32. The immunophenotype was available in 11 cases, and was assessed by immunohistochemistry in nine cases and by flow cytometry in two. Immunophenotype was characterised as a B-cell neoplasia in five cases, T-cell neoplasia in four cases and as showing an aberrant, suspected NK immunophenotype in two. In the first case, the diagnosis of an NK immunophenotype was suspected as, on flow cytometry, the majority of gated cells lacked expression of any of the tested markers (CD3, CD4, CD5, CD8, CD14, CD21) other than CD18. In the second case, on immunohistochemistry, the neoplastic cells were negative for CD3 and CD79a (these were the only markers tested).
Thoracic imaging was performed in 26 cases; by CT scan in five cases and by thoracic radiographs in the remaining 21. Abdominal US was performed on 32 patients. Sixteen of these patients underwent US-guided fine-needle aspiration of the spleen and liver. Patients were classified as stage 1 in two cases, stage 2 in three, stage 3 in two, stage 4 in six and stage 5 in three.
Treatment and survival
Of the entire population examined, 32 patients received chemotherapy alone, three patients radiation therapy alone, three were treated with palliative prednisolone only, and one received radiation therapy and chemotherapy. Of the patients treated with chemotherapy alone, 26 underwent a high-dose COP protocol and six received lomustine and prednisolone. One patient (with nasal lymphoma) that received radiation therapy and chemotherapy was treated with lomustine. The median number of chemotherapy doses administered in patients undergoing a COP protocol was five (range 1-12), while it was two (range 1-5) in the patients receiving lomustine. Lomustine (Bova Specials UK) was given at 10 mg/cat every 4-5 weeks in all cases. Data regarding the treatment received are summarised in Table 3. Chemotherapy-correlated adverse events were reported in 15 patients. Of these, nine developed neutropenia, which was graded as grade 1 in four patients, grade 2 in three, grade 3 in two and grade 4 in one. Gastrointestinal side effects were reported in seven cats and included three cats with vomiting, graded 1 and 2, three cats with diarrhoea, which was graded as 2 in two cats and 1 in the other, one case of constipation and two cats with grade 3 anorexia. Radiation therapy was administered to four patients in total. All these patients were diagnosed with nasal tumours. In all patients, a hypofractionated (three fractions), manually planned protocol was performed. The median dose administered was 24 Gy (range 11.5-24). Response to treatment information was available for 29 patients undergoing chemotherapy alone. In these, CR was achieved in 11 cases, PR in four and PD in 14. The overall response rate for the population receiving chemotherapy alone (PR and CR; all patients receiving chemotherapy) was 52%. The date of disease progression was recorded in 23/34 cats. The median time to first progression was 36 days (range 2-887). Rescue protocols were used in 14 cases. Lomustine was the most common rescue agent (10 cases), followed by COP (one case), L-asparaginase (one case), and vincristine and cytarabine (one case). Radiotherapy was used in one case. The median time to second progression was 20 days (range 2-481). Only one patient received a second rescue protocol with cytarabine and prednisolone and was euthanased 24 days later owing to progressive disease. At the time of writing, 33 patients were deceased. Of these, 31 died as a result of tumour-related causes, one died after trauma on day 750 and one died from heart failure due to a previously diagnosed restrictive cardiomyopathy at day 32.
When patients with nasal lymphoma and patients that did not receive chemotherapy were excluded (n = 25), the median survival time was 49 days (range 2-1726). Median survival time for cats with nasal lymphoma was 250 days (range . Median Hct was 26% (range 14-47%), and the median neutrophil count was 10 × 10 9 /l (range 2-27). Median serum albumin level was 25 g/l (range .
Association between SAA and nasal vs non-nasal lymphoma, stage, number of sites and haematobiochemical variables
When compared with other clinicopathological values, SAA concentration showed a weak positive correlation with neutrophil count (r s = 0.443; P = 0.005) and a weak negative correlation with serum albumin concentration (r s = −0.317; P = 0.049). There was no statistically significant correlation between SAA concentrations and Hct (r s = −0.109; P = 0.511).
After excluding patients that did not undergo full staging, 16 were available for analysis. SAA concentration of cats with different stages of lymphoma are shown in Figure 1. Owing to the low number of cases per stage group, statistical comparison was not possible; however, no obvious association between SAA concentrations and tumour stage was observed.
Clinicopathological factors associated with survival
After the exclusion of patients with nasal lymphoma, patients that did not receive any treatment, and patients diagnosed in 2022, 25 patients were included in the survival analysis. Of these, 22 were treated with a COP protocol and three with a lomustine protocol.
In the univariable analysis, response to initial treatment was the only factor that was statistically significantly associated with survival time, with responders living significantly longer than non-responders (273 days In the multivariable analysis, increasing Hct (HR 0.034, 95% CI 0.827-0.970; P = 0.007) and response to treatment (HR 0.34, 95% CI 0.007-0.177; P <0.001) were found to be independently associated with increased survival time.
When the three cats treated with lomustine were excluded and only cats treated with COP were included in the analysis, SAA concentration was not statistically significantly associated with survival time (HR 0.998, 95% CI 0.991-1.005; P = 0.576), while increasing Hct (HR 0.896, 95% CI 0.827-0.970; P = 0.007) and response to treatment (HR 0.34, 95% CI 0.007-177; P <0.001) remained statistically significantly associated with increased survival time.
Discussion
In this study, we evaluated the utility of SAA concentrations and other clinical and pathological variables as prognostic markers in a population of cats with intermediate-to-large-cell lymphoma. Feline nasal lymphoma tends to show a more localised behaviour than other forms of lymphoma in small animals. In most patients, the disease is confined within the nasal cavity, with only a minority showing systemic involvement. 12,32 In our population, all cats apart from one showed a disease localised to the nasal cavity. Owing to this peculiar clinical behaviour, some authors have advocated the use of more localised radiotherapy treatment as an alternative to chemotherapy for the management of this neoplasia. So far, no studies have found any substantial difference in response between the use of local radiation therapy and systemic chemotherapy; however, in approximately 10-17% of patients treated with local therapy, distant progression of the disease was noted. 6,12,17,32 For this reason, in our opinion, it is extremely important to have full knowledge of the systemic extent of the disease before choosing between the two treatments. In a study by Winkel et al, 28 SAA concentrations were increased in cats with lymphoma, although no difference was found between different anatomical locations. 28 In our population, the SAA concentration was increased in patients with nonnasal lymphoma vs those with nasal involvement. In our opinion, a low SAA value could reflect a less biologically aggressive disease, such as nasal forms of the tumour.
Previous studies in human medicine have proposed SAA concentrations as a marker of the extent of the disease in localised forms of haematological and solid neoplasia. 33,34 Additionally, in the same study, an increase in SAA concentrations was found to be significantly correlated with a reduction in the median time of tumour progression. 33,34 In our population, no correlation between the serum concentration of SAA and disease stage was observed. Larger studies regarding the value of SAA concentrations as a marker of tumour stage in cats are lacking. In a previous study in dogs, in which SAA was analysed in different neoplastic diseases, the serum concentration of the protein was not different between patients with lymphoma and leukaemia. 35 It is therefore plausible that SAA may be correlated with specific locations of the disease associated with different biological behaviour, such as nasal forms of the tumour, but does not give any additional information regarding disease stage. However, the number of patients in our staging group was small, with a low number of patients divided into the different groups; therefore, larger studies are necessary to confirm this finding. Additionally, it would be interesting to see if an increase in SAA would differentiate between a localised form of nasal lymphoma and systemic forms of the disease. However, our population of cats with nasal lymphoma was too small to perform this analysis.
Furthermore, in our population, SAA concentration was not associated with prognosis. To our knowledge, no previous studies have investigated the role of SAA as a prognostic marker in feline haematological malignancies. However, a previous study by Correa et al investigated the association of serum alpha 1-glycoprotein concentrations -another APP -with prognosis in cats with lymphoma, although no correlation was found. 36 This is in contrast with previous literature on human solid neoplasia, where the serum concentration of SAA at presentation was correlated with a reduced time to progression and survival time. 27,37 In the population examined, a low Hct and lack of response to treatment were the only factors statistically significantly associated with survival. Anaemia has previously been associated with a reduced time to progression and survival time in cats with lymphoma. 6,38 In patients with lymphoma, anaemia could be secondary to bone marrow infiltration, local disease or chronic gastrointestinal bleeding, or it could represent anaemia due to the inflammatory disease. Instead, the response to treatment has previously been reported as the most reliable prognostic factor in patients with lymphoma in multiple studies. 4,7,12,39 This study had several limitations, mainly due to its small size and retrospective nature. First, patient staging was not standardised; to include patients with complete staging, there was a marked reduction in the number of cases that could be included. Second, the number of rechecks and reassessments was not standardised in the examined population, which may have caused an alteration in the calculated time to progression and in the timing to assess the response to the treatment in this study. Additionally, owing to its retrospective nature, there may have been inconsistency in the reporting of the adverse events that occurred in the population. The final limitation is that SAA concentrations were measured at the time of diagnosis in only some patients and on the frozen sample in around 50% of cases.
To our knowledge, there are no published studies regarding the stability of SAA in frozen samples in cats.
A recent study in dogs did not show a significant difference in SAA concentration in frozen samples after 3 and 6 months. 39 A small pilot study undertaken in our laboratory has not identified any significant reduction in SAA concentrations in samples stored for between 1 and 5 years (n = 15, data not shown). Therefore, we do not believe that the use of frozen samples would have significantly confounded our data.
Conclusions
SAA concentration does not appear to be of value as a prognostic marker in cats undergoing chemotherapy for non-nasal, intermediate-to-large-cell lymphoma. However, the SAA concentration was elevated in patients with non-nasal lymphoma vs patients with a nasal form of the tumour. Further prospective studies with a larger population are required to confirm these findings.
Author note Preliminary study data were presented at the European Society of Veterinary Oncology (ESVONC), 25-28 September 2022.
Conflict of interest
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Ethical approval The work described in this manuscript
involved the use of non-experimental (owned or unowned) animals. Established internationally recognised high standards ('best practice') of veterinary clinical care for the individual patient were always followed and/or this work involved the use of cadavers. Ethical approval from a committee was therefore not specifically required for publication in JFMS. Although not required, where ethical approval was still obtained, it is stated in the manuscript.
Informed consent Informed consent (verbal or written)
was obtained from the owner or legal custodian of all animal(s) described in this work (experimental or non-experimental animals, including cadavers) for all procedure(s) undertaken (prospective or retrospective studies). No animals or people are identifiable within this publication, and therefore additional informed consent for publication was not required.
|
v2
|
2022-11-24T06:17:19.430Z
|
2022-11-23T00:00:00.000Z
|
253799734
|
s2ag/train
|
Blockage of MDM2-mediated p53 ubiquitination by yuanhuacine restrains the carcinogenesis of prostate carcinoma cells by suppressing LncRNA LINC00665.
Prostate cancer (PCa) is a challenging issue for men's health worldwide due to its uncontrolled proliferation and high metastatic potential. Increasing evidence has supported plant extracts and natural plant derivatives as promising antitumor therapy with less toxic side effects. Yuanhuacine is an active component isolated from Daphne genkwa and can effectively suppress the tumorigenesis of several cancers. However, its role in PCa remains unclear. In this study, yuanhuacine dose-dependently inhibited the proliferation and induced apoptosis of PCa cells. Moreover, yuanhuacine also restrained the invasion and migration of PCa cells. Mechanically, yuanhuacine decreased the ubiquitination and degradation of p53 protein, and ultimately increased p53 levels, which was regulated by inhibiting the phosphorylation and total protein levels of mouse double minute 2 (MDM2). Moreover, elevation of MDM2 reversed the suppressive efficacy of yuanhuacine in PCa cell viability, invasion, and migration. The network pharmacologic and bioinformatics analysis confirmed that MDM2 might be a common target of D. genkwa and LINC00665. Furthermore, yuanhuacine inhibited LINC00665 expression. Upregulation of LINC00665 reversed yuanhuacine-mediated inhibition in MDM2 protein expression and suppressed p53 levels by enhancing its ubiquitination in yuanhuacine-treated cells. Importantly, the inhibitory effects of yuanhuacine on cell viability and metastatic potential were offset after LINC00665 elevation. Together, the current findings highlight that yuanhuacine may possess tumor-suppressive efficacy by inhibiting LINC00665-mediated MDM2/p53 ubiquitination signaling. Therefore, this study indicates that yuanhuacine may be a promising candidate for the treatment of PCa.
|
v2
|
2022-11-24T06:17:19.635Z
|
2022-11-23T00:00:00.000Z
|
253800636
|
s2ag/train
|
Risk Factors Associated With First and Second Primary Melanomas in a High-Incidence Population.
Importance
An increasing number of people develop more than 1 primary melanoma, yet to date, no population-based prospective cohort studies have reported on risk factors for developing first vs second primary melanomas.
Objective
To compare the clinical characteristics of first and second melanomas and then to estimate the relative risks of developing 1 vs multiple melanomas associated with demographic, phenotypic, sun exposure, and genetic factors.
Design, Setting, and Participants
This population-based prospective cohort study included men and women aged 40 to 69 years recruited in 2011 and followed up until December 2018 in Queensland, Australia. Data analysis was performed from February to July 2022.
Exposures
Self-reported information about demographic, phenotypic, and sun exposure measures captured using a survey completed at baseline, and polygenic risk score for melanoma.
Main Outcomes and Measures
Incident first or second primary melanoma diagnosis, and histologic and clinical characteristics thereof. The Wei-Lin-Weissfeld model for recurrent events was used to estimate the association of each factor with the risks of first and second primary melanoma.
Results
A total of 38 845 patients (mean [SD] age at baseline, 56.1 [8.2] years; 17 775 men and 21 070 women) were included in the study. During a median follow-up period of 7.4 years, 1212 (3.1%) participants had a single primary melanoma diagnosis, and 245 (0.6%) had a second primary melanoma diagnosis. Second melanomas were more likely than first melanomas to be in situ; for invasive tumors, second melanomas were more likely to be thin (ie, ≤1 mm) than first melanomas. Having many moles at age 21 years (self-reported using visual scoring tool) was more strongly associated with second (hazard ratio [HR], 6.36; 95% CI, 3.77-10.75) than first primary melanoma (HR, 3.46; 95% CI, 2.72-4.40) (P value for difference between the HRs = .01). A high genetic predisposition (ie, polygenic risk score in tertile 3) was also more strongly associated with second (HR, 3.28; 95% CI, 2.06-5.23) than first melanoma (HR, 2.06; 95% CI, 1.71-2.49; P = .03). Second melanomas were more strongly associated with a history of multiple skin cancer excisions (HR, 2.63; 95% CI, 1.80-3.83) than first melanomas (HR, 1.86; 95% CI, 1.61-2.16; P = .05). For all other phenotypic characteristics and sun exposure measures, similarly elevated associations with first vs second melanomas were observed.
Conclusions and Relevance
Findings of this cohort study suggest that within the general population, the presence of many nevi and having a high genetic predisposition to melanoma were associated with the highest risks of developing second primary melanomas.
|
v2
|
2022-11-24T06:17:19.672Z
|
2022-11-23T00:00:00.000Z
|
253799579
|
s2ag/train
|
Association of Household Income at Diagnosis With Financial Toxicity, Health Utility, and Survival in Patients With Head and Neck Cancer.
Importance
While several studies have documented a link between socioeconomic status and survival in head and neck cancer, nearly all have used ecologic, community-based measures. Studies using more granular patient-level data are lacking.
Objective
To determine the association of baseline annual household income with financial toxicity, health utility, and survival.
Design, Setting, and Participants
This was a prospective cohort of adult patients with head and neck cancer treated at a tertiary cancer center in Toronto, Ontario, between September 17, 2015, and December 19, 2019. Data analysis was performed from April to December 2021.
Exposures
Annual household income at time of diagnosis.
Main Outcome and Measures
The primary outcome of interest was disease-free survival. Secondary outcomes included subjective financial toxicity, measured using the Financial Index of Toxicity (FIT) tool, and health utility, measured using the Health Utilities Index Mark 3. Cox proportional hazards models were used to estimate the association between household income and survival. Income was regressed onto log-transformed FIT scores using linear models. The association between income and health utility was explored using generalized linear models. Generalized estimating equations were used to account for patient-level clustering.
Results
There were 555 patients (mean [SD] age, 62.7 [10.7] years; 109 [20%] women and 446 [80%] men) included in this cohort. Two-year disease-free survival was worse for patients in the bottom income quartile (<$30 000: 67%; 95% CI, 58%-78%) compared with the top quartile (≥$90 000: 88%; 95% CI, 83%-93%). In risk-adjusted models, patients in the bottom income quartile had inferior disease-free survival (adjusted hazard ratio, 2.13; 95% CI, 1.22-3.71) and overall survival (adjusted hazard ratio, 2.01; 95% CI, 0.94-4.29), when compared with patients in the highest quartile. The average FIT score was 22.6 in the lowest income quartile vs 11.7 in the highest quartile. In adjusted analysis, low-income patients had 12-month FIT scores that were, on average, 134% higher (worse) (95% CI, 16%-253%) than high-income patients. Similarly, health utility scores were, on average, 0.104 points lower (95% CI, 0.026-0.182) for low-income patients in adjusted analysis.
Conclusions and Relevance
In this cohort study, patients with head and neck cancer with a household income less than CAD$30 000 experienced worse financial toxicity, health status, and disease-free survival. Significant disparities exist for Ontario's patients with head and neck cancer.
|
v2
|
2022-11-24T14:06:59.463Z
|
2022-11-23T00:00:00.000Z
|
253805584
|
s2orc/train
|
Efficacy of resistive exercise on skeletal muscle-related outcomes in cancer survivors: a systematic review protocol
Background Symptom burden and adverse treatment effects can negatively impact physical function, health-related outcomes, and quality of life in cancer survivors. Resistive exercise that improves skeletal muscle function can ameliorate these complications, but the central role of the skeletal muscle in mediating improvements in patient-related outcomes has not been explored. This protocol describes the rationale and methods for a systematic review that aims to determine the effects of resistive exercise on the skeletal muscle hypertrophy, muscle performance, and muscle-related biomarkers in cancer survivors. Methods A systematic review will be conducted on peer-reviewed randomized controlled trials (RCTs) that employ resistive exercise interventions for cancer survivors. The following electronic databases will be searched: AMED, CENTRAL, CINAHL, CIRRIE, EMBASE, MEDLINE, PEDro, REHABDATA, Scopus, and SPORTDiscus. Studies will be considered for inclusion if they present quantitative data in adult cancer survivors on skeletal muscle characteristics (e.g., muscle mass), muscle performance (e.g., strength), or skeletal muscle-related biomarkers (e.g., myocellular satellite cells). Secondary outcomes will be physical function (e.g., stair climb) and patient-reported outcomes (e.g., fatigue). Data will be reported through a narrative that describes study design, participants, interventions, and outcome characteristics. Discussion This systematic review will help clarify the influence of resistive exercise on factors relating to the skeletal muscle in adult cancer survivors. Findings may provide insight into optimal exercise selection for evidence-based practice. Systematic review registration PROSPERO: #277791 [under review] Supplementary Information The online version contains supplementary material available at 10.1186/s13643-022-02130-z.
Background
Cancer incidence and mortality are growing rapidly worldwide, with approximately 19.3 million new cancer cases and nearly 10 million cancer deaths in 2020 according to global cancer estimates (GLOBOCAN) [1]. Although therapeutic advances have significantly improved cancer survival, treatment toxicity and symptom burden persist and can negatively impact physical function, disease-related symptoms, and quality of life [2]. Rehabilitation, especially exercise, is recognized as a key strategy in improving these cancer-related health outcomes and compelling evidence from a 2019 American College of Sports Medicine International Multidisciplinary Roundtable strongly supports the role of exercise Open Access *Correspondence: [email protected] in mitigating the adverse effects of treatment across multiple cancer types [3].
The wide therapeutic efficacy of exercise is likely due to its influence on numerous physiological processes, including neuromuscular, cardiovascular, metabolic, and inflammatory systems [4,5]. Central to mediating the diverse effects of exercise is the skeletal muscle. Decades of research on the role of the skeletal muscle in performance, health, and disease have elucidated functions far beyond force generation and movement [6]. The skeletal muscle is now viewed as an integral component of the complex cross-talk between hormonal, metabolic, and inflammatory pathways in addition to performing its traditional role in the neuromuscular system [7]. Indeed, the influence of the skeletal muscle across multiple systems is apparent in diseases of muscle loss, such as cachexia, and to a lesser extent, sarcopenia and dynapenia, where dysregulation of these diverse pathways can lead to significant morbidity and mortality [7,8]. Thus, not only is the skeletal muscle essential for physical function, but it is also important for overall health. Furthermore, recent evidence supports the importance of the skeletal muscle in the oncology care continuum as low skeletal muscle mass is associated with higher surgical and postoperative complications, longer length of hospital stay, higher chemotherapy-related toxicity, lower physical function, poorer quality of life, and shorter survival [9,10].
Previous systematic reviews in the exercise oncology literature involving the skeletal muscle tend to follow a disease-specific (e.g., breast cancer), treatment-specific (e.g., during androgen deprivation therapy), or impairment-specific (e.g., strength loss) approach. For example, systematic reviews by Bourke et al. [11] and Stephensen et al. [12] examined the general benefits of exercise for men with prostate cancer [11] and adults with abdominal cancer [12], while Chen et al. [13] and Hasenoehrl et al. [14,15] focused specifically on the effect of resistance exercise on physical performance in prostate [13,14] and breast [15] cancer survivors. On the other hand, some systematic reviews have included all cancer types, but have focused on the type or timing of treatment when evaluating the effects of exercise on physical function [16][17][18]. One recent review has broadly considered the role of exercise across multiple cancer types, treatments, and impairments to identify key features of exercise interventions that improve physical function and other cancer-related outcomes [19]. This systematic review synthesized data from other exercise oncology systematic reviews with the goal of supporting exercise and rehabilitation intervention decision-making across all cancers. While the aggregate findings identified common features of exercise programs in the general cancer population, a wide range of outcomes were reported.
Hence, to the best of our knowledge, no systematic review has summarized the effects of resistive exercise specific to various skeletal muscle-related outcomes in cancer survivors. Given the significance of the skeletal muscle in physical function and health throughout the oncology care continuum, there is a need to better understand the contribution of the skeletal muscle to physiologic and patientreported outcomes in cancer survivorship. Therefore, the aim of this systematic review is to comprehensively evaluate the evidence on the effect of resistive exercise on factors relating to the skeletal muscle in adult cancer survivors. The skeletal muscle-related outcomes of interest are primarily focused on muscle mass, performance, and muscle-related biomarkers. The following research questions were formulated to assess this evidence: 1. Which skeletal muscle-related outcomes in cancer survivors are improved by resistive exercise? 2. What are the features of resistive exercise interventions conducted in cancer survivors that demonstrate improvements in skeletal muscle-related outcomes?
Design
This systematic review protocol has been developed according to the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) [20] (Additional File 1). The protocol is registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number: 277791).
Eligibility criteria
We will include studies that meet the following criteria.
Study type
Randomized controlled trials (RCTs) or randomized experimental trials written in English that assess the effect of resistive-type exercises on skeletal musclerelated outcomes in adult cancer survivors will be included. Data from reviews, meta-analyses, case reports, or editorials will be excluded.
Participants
Adults aged 18 years or older who are diagnosed with cancer regardless of tumor type, stage, or treatment will be included.
Intervention
Interventions using any form of resistive exercise are included. Resistive exercise refers to activity involving dynamic (concentric and/or eccentric) or static (isometric) muscle contractions opposed by a force that targets the development or performance of the skeletal muscle. Examples of resistive exercise modalities include free weights, body weight, elastic resistance, plyometrics, constant or variable resistance machines, isokinetic machines, and isometric training, including yoga, Tai Chi, or Pilates. Interventions may use any level of supervision, method of delivery, frequency, duration, or intensity. Multimodal interventions incorporating resistive exercise and other types of intervention, such as aerobic exercise, flexibility training, or diet interventions are not excluded.
Comparators
The comparison groups may be a non-resistive exercise group (e.g., aerobic exercise), non-exercise group (e.g., control or waiting list), or standard of care.
Outcomes
Studies that report absolute values and/or change from baseline to follow-up on at least one of the following primary outcomes will be considered: (1) hypertrophic characteristics of the skeletal muscle (i.e., muscle mass, cross-sectional area), (2) muscle performance (i.e., strength, muscular endurance, range of motion), or (3) muscle-related biomarkers (i.e., satellite cells, protein synthesis, regulatory gene expression, circulating markers released by muscle). In addition, secondary outcomes of interest will be examined if present alongside a primary outcome. These secondary outcomes include (1) physical function (i.e., stair climb, handgrip strength),
Search terms and keywords
The specific search keywords were developed collaboratively between all authors and included terms from 3 primary domains: (1) cancer (i.e., cancer, tumor, oncology), (2) resistive exercise (i.e., weight lifting, strength training), and (3) muscle outcomes (i.e., hypertrophy, strength, protein synthesis). The MEDLINE search strategy (Table 1) includes a combination of relevant subject headings and keywords. The search strategy for other databases will be adapted from the MEDLINE strategy. Searches will be run with no date limits but within a given time period to ensure consistent data retrieval. The searches will be independently conducted by three review authors (JKD, CDC, MN). Search results will be exported to the systematic review manager Covidence (Melbourne, AU), with duplicate articles removed.
Data management Identification and selection of studies
Initial screening of the first 100 articles will be performed by two authors (JKD, CDC) such that each author will examine every record. Any inconsistencies will be discussed until consensus is obtained. These measures are performed for training purposes to ensure the reliability of decision-making by more than one independent reviewer [21]. The two authors who performed the training (JKD, CDC) will then each lead a team of reviewers to independently screen the remainder of the articles.
Potentially relevant studies will be identified from titles or abstracts and marked in Covidence as eligible for fulltext review. Articles that cannot be safely excluded without reviewing the full text will be included. A second screening will be performed by the two teams of reviewers to assess the full text of eligible articles against the defined criteria for inclusion (Additional File 2). Articles that appear to meet the eligibility criteria will be recorded onto a Google data collection form. One author (JKD) will review all retrieved articles for final inclusion with any disagreements discussed until consensus is reached. Per PRISMA guidelines [22], a flow diagram will be used to describe the process of study selection with reasons for exclusion recorded.
Data extraction
Data from each included study will be extracted using predefined criteria that are entered into Covidence to create a data extraction template (Additional File 3). The extracted data include (1) study characteristics: authors, publication year, journal name, country, sample size, and funding source; (2) participant characteristics: inclusion/exclusion criteria, number of participants, age, cancer diagnosis, disease stage, treatment, physical characteristics, and demographics; (3) methods: study design, randomization, blinding, recruitment, retention, and adherence; (4) intervention characteristics: resistive exercise modality, frequency, intensity, session duration, and intervention duration; number of follow-up visits, supervision, progression of intensity, duration, or frequency; (5) primary outcome measures: skeletal muscle characteristics, muscle performance, and myocellular markers; and (6) secondary outcome measures: physical function, patient-reported outcomes, and adverse events. A team of review authors (DK, OB, RW, CC, JY, GC, LM, PM, CH) will independently perform data extraction on all included articles within Covidence. One review author who did not perform the extraction will check the extracted results (JKD).
Assessment of risk of bias
Methodological quality will be assessed using the risk of bias tool from the Cochrane Handbook for Systematic Reviews of Interventions [23]. The tool has seven domains, including sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other sources of bias. Each domain will be rated as either high risk, unclear risk, or low risk. One team of authors (DK, OB, RW, CC, JY, GC, LM, PM, CH) will evaluate the risk of bias independently, and differences will be resolved by discussion with an author who did not assess the risk of bias (JKD). Should further clarification of study methods be necessary to assess the risk of bias, we will contact the authors of the study for additional information. The Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) approach will be used to assess the quality of evidence across studies [24].
Data synthesis
A narrative synthesis of data will be conducted using the Guidance on the Conduct of Narrative Synthesis in Systematic Reviews [25]. The extracted findings will be examined through tables and a descriptive narrative, where data will be grouped according to characteristics to explore patterns between studies. Study quality, strengths, and limitations will be reported. We chose not to pool the data in a meta-analytical approach because of the significant heterogeneity in the reporting of exercise parameters across the different cancer types that would also make it challenging to calculate a valid effect size. Given that our review is the first systematic review on this broad topic, we intend for our descriptive findings to discern patterns that can be targeted in future metaanalytical reviews.
Discussion
This is the first systematic review to examine the effect of resistive exercise in cancer survivors on skeletal muscle-related outcomes that span multiple domains, including physical performance, muscle characteristics, and tissue level changes. While numerous systematic reviews have been conducted within the exercise oncology literature, prior reviews have focused on specific disease types, such as prostate cancer only, or certain aspects of muscle function, such as strength. Although the consensus is that resistance exercise is beneficial to cancer survivors, limited evidence has been synthesized from RCTs to provide a more holistic view of how the skeletal muscle may respond to resistance exercise to contribute to multiple patient-related outcomes. This review employs a comprehensive, reproducible approach to searching, screening, and extracting data based on published guidelines and validated methods. Limitations of the review include reporting bias in review authors and significant heterogeneity between studies that may limit the ability to synthesize data. We expect findings from this review to support clinical decision-making by oncology care providers, where evidence from this review may inform exercise selection to improve health-related quality of life in adult cancer survivors.
|
v2
|
2022-11-24T15:00:07.894Z
|
2022-11-23T00:00:00.000Z
|
253805180
|
s2orc/train
|
Strategies to enhance CAR-T persistence
Chimeric antigen receptor T (CAR-T) cell therapy has significantly improved the life expectancy for patients with refractory or relapse B cell lymphoma. As for B cell acute lymphoblastic leukemia (B-ALL), although the primary response rate is promising, the high incidence of early relapse has caused modest long-term survival with CAR-T cell alone. One of the main challenges is the limited persistence of CAR-T cells. To further optimize the clinical effects of CAR-T cells, many studies have focused on modifying the CAR structure and regulating CAR-T cell differentiation. In this review, we focus on CAR-T cell persistence and summarize the latest progress and strategies adopted during the in vitro culture stage to optimize CAR-T immunotherapy by improving long-term persistence. Such strategies include choosing a suitable cell source, improving culture conditions, combining CAR-T cells with conventional drugs, and applying genetic manipulations, all of which may improve the survival of patients with hematologic malignancies by reducing the probability of recurrence after CAR-T cell infusion and provide clues for solid tumor CAR-T cell therapy development. Supplementary Information The online version contains supplementary material available at 10.1186/s40364-022-00434-9.
Background
Chimeric antigen receptor T (CAR-T) cell immunotherapy has rapidly impacted the malignant tumor field and has achieved remarkable effects in recent years as the latest promising adoptive cell therapy. The CAR concept was first proposed in 1989, when G Gross et al. designed a chimeric T-cell receptor (TCR) gene consisting of the TCR constant domain and the antibody variable domain and transfected it into cytotoxic T cells, conferring these T cells with antibody-like specificity [1]. Moreover, the chimeric TCR was able to signal effectively upon activation and to execute its effector function [1]. Subsequently, with the emergence and development of transgenic technology, there are new options for adoptive T-cell therapy [2][3][4][5][6]. Up to 2011, CAR-T cells were used to treat patients with lymphoblastic leukemia, with striking results [7][8][9]. Based on multiple clinical trials with laudable results, the FDA approved the first CAR-T cell therapeutic product (Kymriah, CTL-019 from Novartis) for patients with relapsed/refractory B-cell acute lymphoblastic leukemia (r/r B-ALL) in 2017, and multiple similar products have since been approved [10][11][12][13]. Overall, the complete remission (CR) rate in patients treated with CD19 CAR-T is 30-70%, and in some trials, the rate was over 90% [10,12,14]. The fact that these products were generated using diverse technical schemes indicates the marked efficiency of CAR-T cells for treatment of lymphoblastic leukemia [15][16][17][18][19]. CAR-T cells have also been used for treating solid tumors, but curative effects are lacking compared to those for lymphoblastic leukemia [20][21][22][23][24]. Hence, there is an urgent need to further optimize CAR-T therapy to resolve existing issues, especially the limited persistence, high recurrence rate, and insufficient cytotoxicity, with regard to solid tumors.
The conventional CAR-T cell production process includes five steps: (1) T-cell collection and activation; (2) CAR structure preparation and transduction; (3) CAR-T cells expansion in vitro; (4) Verification of CAR-T cell phenotype and function; (5) CAR-T cells cryopreservation and storage until administration to the patient [25]. The five steps require approximately one to two weeks, at which point qualified CAR-T cells are ready for delivery to patients. Based on this process, two main areas of focus for optimizing CAR-T therapy have been developed: the design of CAR structure and the intervention during CAR-T cell expansion stage [26]. First, the CAR structure, including the binding domain, extracellular spacer, transmembrane domain, and intracellular signaling domain, is regarded as the core entity of the CAR-T cells [26,27]. This structure has undergone constant remodeling, encompassing five generations of evolution and optimization since the origin of CAR-T cells [28,29]. These five generations have been comprehensively studied and applied and thus are not introduced in detail in this review [29][30][31]. The second-generation CAR is the most widely adopted structure, which includes one costimulatory domain within the intracellular domain that is most commonly derived from CD28 or 4-1BB and can enhance TCR signaling [3,[32][33][34][35][36].
In addition to structural design, the intervention during CAR-T cell expansion stage could also contribute to the persistence after infusion. Sophisticated engineering strategies are applied to improve antitumor activity by regulating CAR-T cell development and differentiation. Overall, CAR-T cells exhibit strong proliferation and differentiation abilities in the culture stage [37]. These activities are accompanied by competition between effector function and long persistence, which determines the final quality of the cell therapy after infusion into patients. In this review, we describe the different approaches that have been recently adopted in vitro to improve the persistence and immunophenotype of CAR-T cells, including the choice of suitable cells, the improvements in the in vitro culture conditions, the application of conventional drugs with CAR-T cells, and the use of genetic manipulations. These approaches may improve the persistence of CAR-T cells which have superior immunotherapeutic effects.
AutoCAR-T cells or alloCAR-T cells
First, the quality and characteristics of T cells are crucial for the therapeutic efficacy of CAR-T cells [38]. CAR-T cells are classified as autologous or allogeneic (autoCAR-T or alloCAR-T) according to the source of T cells, and both have been evaluated in clinical trials [39][40][41]. Autoand alloCAR-T cells have unique benefits and challenges that need to be addressed. For autoCAR-T, the T cells are isolated from patients, and thus, immunological rejection does not occur [42]. However, long-term infiltration in the tumor microenvironment (TME) decreases cytotoxicity and induces exhaustion of T cells, which might lead to insufficient therapeutic efficacy of autoCAR-T cells for tumors [40]. Moreover, the whole stage, from cell collection to cell infusion, is time-consuming and is required for each patient to undergo autoCAR-T cell therapy, and this process is overly lengthy for some critically ill patients. Furthermore, the cost of CAR-T cell therapy is too high to wildly applied. AlloCAR-T cells potentially overcome these problems. They can be collected and prepared in advance and then supplied to patients without delay. Healthy donor T cells possess greater cytotoxicity when compared to T cells from patients. Moreover, multiple and systematic production of these cells may reduce the average cost and result in a process more easily carried out. The greatest obstacles for the application of alloCAR-T cells are host versus graft disease (HvGD) and graft versus host disease (GvHD), which need to be overcome by genetic manipulation of TCR and human leukocyte antigen (HLA), and it can be accomplished via the TALEN and CRISPR systems [43].
Beyond the principle and technical factors, recently published clinical results show that autoCAR-T cells have better therapeutic efficacy than alloCAR-T cells [44]. For example, the initial overall response rate for traditional autoCAR-T(CTL-019) was reported to be approximately 90%, with a 5-year response rate reaching 58% [45]. These patients with diffuse large B-cell lymphoma and follicular lymphoma were followed for a median of 60.7 months [45,46]. For patients with B-ALL who were treated with alloCAR-T, the initial overall response rate was 67%, the 6-month response rate was 55%, and the median survival time was 4.1 months (NCT02808442 and NCT02746952). The quantity and persistence of CAR-T cells are important factors for tumor recurrence. In one study, the CAR transgene was continuously detectable beyond 1 year in approximately 50% of patients who achieved complete remission after autoCAR-T treatment, but only 1 patient treated with alloCAR-T had a detectable CAR transgene beyond 120 days [47]. Furthermore, cytokine-release syndrome (CRS) occurs in 57% of patients administered autoCAR-T cells and in 91% of those administered alloCAR-T cells, with a similar neurotoxicity rate (39% vs. 38%) [44]. In comparison with alloCAR-T cells, autoCAR-T cells demonstrated better performance in a recent clinical trial in almost all aspects, especially in persistence [44]. Regardless, the core advantage of alloCAR-T cells, namely, their "off-the-shelf " nature, cannot be ignored and represents a significant benefit for large-scale systematic production [44]. In addition, the CRISPR genomic editing tool allows for easy manipulation of alloCAR-T cells, which is worth exploring in future tumor immunotherapy studies [48].
Moreover, T-cell-derived induced pluripotent stem cells have been verified as an ideal source of autoCAR-T cells that do not cause GvHD, which may facilitate the largescale development of potent autoCAR-T cells for a broad range of immunotherapies [49].
Subgroups of T cells
In addition to the source, cell subgroup also affects CAR-T cell persistence and therapeutic efficacy. For conventional CAR-T cell therapies, T cells are collected and isolated using the marker CD3, mainly resulting in two subgroups, CD4 + and CD8 + T cells [38]. The CD4/ CD8 ratio is not constant among patients or donors, which has attracted some attention. Several studies have focused on the relationship between the CD4/CD8 ratio and therapeutic efficacy in clinical trials [50]. First, as the bases of T cells, both CD4 + and CD8 + T cell subsets are capable of efficiently killing tumor cells [51]. Compared to the use of individual T cell subsets for CAR-T cells, the combination of CD4 + and CD8 + subsets exhibits synergetic antitumor effects both in vitro and in vivo [51]. Moreover, more uniform potency was obtained with defined T cell subsets compared with unselected T cells. Subsequently, a clinical trial involving 29 adult B-ALL patients with defined CD4 + and CD8 + T subsets (half and half ) for CD19 CAR-T cells observed a marked effect (NCT01865617). This result highlights the advantage of consistent and standardized CAR-T cells and response evaluation [52,53]. Nonetheless, a conflicting result has been reported, whereby CD4 + CAR-T cells showed superior antitumor activity in glioblastoma compared to CD8 + CAR-T cells, especially regarding the long-term antitumor response [54]. Additionally, a recent study that followed two patients for 10 years after CAR-T cell treatment showed a proportion of CD4 + CAR-T cells over 99%, whereas that of CD8 + CAR-T cells was less than 1% [15]. These findings suggest exaggerated persistence of CD4 + T cells in vivo [15]. Such a discrepant result might be caused by differences in tumor cell characteristics, the TME, the receptor, tumor antigens, and even the transduction and expansion methods. In summary, these studies consistently highlight the importance of defined CD4 + and CD8 + CAR-T cell compartments for effective antitumor activity. Due to the enormous success of CAR-T immunotherapy, interest in engineering innovative CARs using other types of immune cells has emerged in recent years. Although CAR-NK cells are well-known immune cell therapy that are undergoing comprehensive, meticulous preclinical and clinical trials, they are not reviewed here [55][56][57][58][59][60]. Natural killer T cells (NKTs), γδTs, regulatory T cells (Tregs), and a series of infrequent T cell subsets are also potential candidate for immune cell therapy. NKT cells are a group of T cells with both T cell receptors and NK cell markers and thus possess features of both cell types. These features include the nonspecific killing function of NK cells and the specific killing function of T cells, endowing NKT cells with powerful surveillance and killing capabilities [61]. Since their discovery and exploration, NKT cells have been used for cellular immunotherapy beginning in 2010 for patients with liver carcinoma, lung carcinoma, and gastric carcinoma [62]. In 2019, NKT cells were loaded with a GD2 CAR structure to specifically recognize and kill neuroblastoma cells, and clinical trials have shown a positive therapeutic effect [63]. More CAR-NKT cells are advancing in clinical trials [63,64].
T cells can be defined as two main subsets according to the difference in TCR molecules: αβT and γδT. The proportion of αβT cells exceeds 60%, and that of γδT cells is less than 5% [65]. Regarding anticancer efficacy, γδT cells can kill target cells through the perforin-granzyme-B pathway, Fas-FasL pathway, and antibodydependent cell-mediated cytotoxicity (ADCC) pathway [66,67]. Such multiple modes of action suggest a predominant anticancer effect [68]. In addition, γδT cells can recognize targets directly without major histocompatibility complex (MHC) restriction, which is their core advantage compared to αβT cells for use as offthe-shelf alloCAR-T cells without the complication of GvHD [65]. Based on these characteristics, several preclinical trials have demonstrated positive results using CAR-γδT cells, with favorable persistence in vivo, and the relevant clinical products have been described [69,70]. A recent preclinical study in 2022 reported that CD20 CAR-γδT cells have powerful tumor inhibition and low sensitivity between grafts and hosts, highlighting the potential advantages of CAR-γδT cells [69,71].
As a type of immunosuppressive cell, Tregs are critical for immune system tolerance and hyperimmune response avoidance [72,73]. Therefore, unlike other CAR cells for anticancer treatment, CAR-Tregs maintain the Treg phenotype and function. These cells are guided to the target tissue by the CAR structure and show enhanced suppression of autoimmune disease in susceptible organisms [74]. Furthermore, CAR-Tregs avoid MHC restriction, and compared to polyclonal Tregs, fewer cells are required. Preclinical studies demonstrate that CAR-Tregs protect vascularized grafts in fully immunocompetent recipients [75]. These cells have been modified to treat GvHD, type 1 diabetes, and Alzheimer's disease, and it is worth noting that CAR-Tregs markedly expand the applications of CAR engineering [76]. However, CAR-Tregs are only in the initial stages of development and application, and further optimization is urgently needed to improve their efficiency and stability for clinical application.
In summary, as the main component of CAR-T cells, the source and characteristics of T cells significantly determine the persistence and efficiency. Therefore, choosing a suitable source is the first and important step for successful CAR-T cell therapy (Table 1).
Differentiation of T cells influences CAR-T cell efficiency
In addition to the original subsets and proportion of T cells, their differentiation plays a significant role in regulating CAR-T cell anticancer activity. Similar to the components, the differentiation state of T cells also varies among individuals and culture platforms. T cells can be subdivided into five principal and typical subgroups according to the expression of cell surface markers, as shown in Fig. 1 [77]. For every T cell subpopulation, these different phenotypes suggest functional characteristics and diversity, which are vital to developing and understanding new strategies for CAR-T cell therapy [78,79].
Naïve T cells Before interacting with their cognate antigen or pathogen, immature T cells are considered naïve T (T N ) cells. These cells exhibit high expression of the CD45RA isoform, L-selectin (CD62L), and CXC chemokine receptor 7 (CCR7). The latter two representative markers indicate the ability of these cells to home towards lymph nodes [80]. T N cells possess high reproductive and survival abilities, but a sharp decline in the T N proportion during the expansion stage of CD3/CD28 antibody stimulation and activation cannot be avoided [37]. After stimulation and activation by CD3/CD28 antibodies and cytokines, T N cells experience rapid differentiation into memory T cells and effector T cells [37]. markers, including CD62L, CCR7, CD45RA and CD95 [79,80]. It is worth noting that the gene expression pattern of T SCM cells indicates that they are at an earlier differentiated stage than memory T cells. Studies have shown that T SCM cells can quickly respond to antigenic stimulation and maintain strong self-renewal capacity [80,81]. These findings suggest a tremendous potential of these cells for adoptive T cell therapy, despite their extremely low cell proportion [79,82]. Central memory T cells Similar to T SCM cells, central memory T (T CM ) cells express CD62L, CCR7, and CD95, but they express the CD45RO isoform instead of the CD45RA isoform [83]. As T CM cells have been previously exposed to an antigen, a second instance of antigen exposure causes rapid proliferation and an effective immune response. Compared to effector T cells, T CM cells are less cytotoxic, but their characteristics of rapid proliferation and long persistence enable superior performance [84,85]. Both T SCM and T CM cells maintain the ability to home towards lymph nodes, which is important for long persistence [85]. Thus, the ratio of T CM and T SCM cells has become a key performance indicator for evaluating CAR-T therapeutic effects (NCT01318317 and NCT01815749) [86]. Further comparison of the two memory T subsets reveals that T SCM cells can retain a higher proportion of naïve phenotypes than T CM cells, indicating stronger self-renewal ability [83,87]. Effector memory T cells Another type of memory T cell is the effector memory T (T EM ) cell, which lacks CCR7 and CD62L expression, in contrast to T SCM and T CM cells. This lack of expression prevents these cells from homing towards lymph nodes, but they can access target tissues. Once the cell encounters antigens, effector functions rapidly ensue as well as secretion of a series of cytokines, such as TNF-α, IFN-γ, and perforin [88,89]. Terminal effector T cells Continuous antigen exposure and stimulation induce T EM cells to further differentiate into terminal effector T (T EFF ) cells, the terminus of effector T cells. These cells do not express CD62L, CCR7, or CD45RO, but they do express CD45RA. T EFF cells are the primary T cell subset of anticancer cells. However, T EFF cells have a short lifespan and hardly any self-renewal ability. After addressing a chronic infection or killing cancer cells, T EFF cells inevitably lose their effector function and become exhausted [79,89]. These five differentiated T cell subsets have diverse features in terms of proliferation ability and antitumor effect that substantially contribute to the clinical efficacy of CAR-T cells [79]. Several preclinical and clinical trials have revealed that compared to traditional CAR-T cells, CD19 CAR-T cells derived from T SCM cells or from a higher ratio of memory T cells show long-term persistence and responses [90,91]. Therefore, a consensus has been reached that T SCM and T CM cells have better persistence and antitumor activity in vivo than T EM and T EFF cells. Accordingly, researchers have focused on strategies to alter T cell differentiation to induce more T SCM and T CM cells in vitro [37,92].
Culture conditions influence CAR-T cell differentiation and function
In addition to differences in surface markers and functions, metabolic requirements among the five subsets of T cells are diverse [93,94]. It has gradually become clear that T cell metabolic reprogramming occurs at the same time as T cell activation and is another critical component for adoptive therapy [94]. For T N cells, the main metabolic pathway and energy acquisition methods are oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO) in mitochondria [37]. In response to a CD3/CD28 antigen stimulation signal, T N cells accelerate metabolism to satisfy the increased biosynthetic demands [37]. The PI3K-AKT-mTOR pathway is activated and subsequently promotes aerobic glycolysis (Warburg effect) [95]. This metabolic reprogramming also results in the transformation of T N cells to T EFF cells [80]. For T SCM and T EM cells, the metabolic pattern is similar to that of T N cells and mainly depends on OXPHOS and low-level glycolysis [93].
To mediate T cell differentiation and improve the T SCM and T CM ratio during the CAR-T cell expansion phase, which impact the persistence and clinical performance, culture conditions and operating steps have been explored in depth. Combinations of multiple cytokines IL-2 is a widely used cytokine that not only induces rapid T cell proliferation in vitro but also evokes a switch from OXPHOS to glycolysis. It induces formation of more effector T cells and reduces memory T cell population [96]. Other cytokines, including IL-7, IL-15, and IL-21, mediate metabolic adaptation by boosting OXPHOS and inhibiting glycolysis [97,98]. The combination of IL-7 and IL-15 has been reported to intervene CAR-T cells with higher proportions of T SCM and T CM cells and superior anticancer activity in vivo compared with CAR-T cells cultured with only IL-2 [99]. This strategy has been applied for clinical trials in lymphoma, sarcoma, and osteosarcoma [9,100]. Metabolite adjustments In addition to cytokines, glucose is the direct target mediating glycolysis [101,102]. Many studies have explored and validated the abilities of several drugs and molecules to restrict or interfere with glycolysis metabolism. For example, 2-deoxy-D-glucose (2-DG), a synthetic glucose analogue, enters cells through glucose transporters (GLUTs) and is subsequently phosphorylated by hexokinase, acting as a competitive glucose inhibitor [101]. In the presence of 2-DG during expansion in vitro, CD8 + T cells alter their differentiation trajectory for more memory cells, leading to persistent antitumor functionality [101]. Metabolites have also been verified to influence T cell metabolic adaptability and differentiation fate. Metabolic analysis has shown that intracellular L-arginine is markedly depleted after T N cells become activated [101,103]. Additional L-arginine supplementation in the culture medium regulates T cell metabolic alternation from glycolysis to OXPHOS and increases the ratio of T SCM cells, leading to greater antitumor activity [103]. Glutamine is a substrate of the TCA cycle, and glutamine metabolism is augmented once T N cells are activated [104]. Glutamine is required for effector T differentiation and function, and its depletion increases the percentage of T CM cells [104,105]. Furthermore, glutamine-deficient condition caused by the antagonist 6-diazo-5-oxo-L-norleucine (DON) in the culture medium reportedly increases OXPHOS and reduces glycolysis in CD19 CAR-T cells [106]. Ultimately, these changes induce stronger lysis in vitro and more robust elimination of tumor cells in vivo, indicating a promising approach to optimize CAR-T immunotherapy [106]. Moreover, another study found that carnosine limits extracellular acidification to shift the metabolic profile from an acidic, stressed state towards an oxidative, energetic state [107]. Therefore, addition of carnosine enhances lentiviral gene delivery in activated T cells [107]. This finding provides the potential to improve the overall quality of the cell culture medium for CAR-T cell therapies. In addition to these specific metabolites, several growth factors originating from the immune microenvironment might participate in T cell proliferation, differentiation, or function. Recent studies have shown that adding human platelet lysate to the CAR-T cell culture medium enriches the T CM cell subset and ultimately improvs the antitumor effect in a mouse xenograft model [107].
Hence, it can be concluded that the core mechanism by which these strategies optimize culture conditions is related to the regulation of T cell metabolism through the increase in OXPHOS and appropriate inhibition of glycolysis in vitro. These processes lead to a high percentage of T SCM and T CM cells, ultimately resulting in superior longevity and antitumor potential for CAR-T immunotherapy.
Strategies for novel application of conventional drugs in the CAR-T cell expansion phase
In addition to improvements in the culture medium, several activators and inhibitors that target metabolic pathways and even conventional drugs used to treat other diseases have been used with CAR-T cells to achieve superior cancer immunotherapy. In this section, we focus on the application of these clinical molecules and drugs that potentially modulate metabolism and T cell differentiation as novel CAR-T adjuvants ( Table 2).
PI3K-AKT-mTOR inhibitors
Hyperactivation of PI3K has been verified to induce tumorigenesis in hematologic malignancies and in many solid tumors, such as melanoma, lung cancer, and breast cancer, which indicates its potential as a therapeutic target [118]. However, development of pan-PI3K inhibitors has advanced slowly for many years because of serious toxicity and low tolerance [119]. Several PI3K subunit inhibitors have shown remarkable antitumor effects in lymphoid cancer and have been approved by the FDA based on the results of clinical trials [120]. Moreover, in normal T cells, activation of PI3K family members, especially PI3K-δ and -γ, is pivotal for transmitting signals from the TCR/CD3 complex to activate downstream transcription factors (e.g., HIF-1α and c-Myc), promoting glycolysis and differentiation [121]. The effects of several PI3K inhibitors on T/CAR-T cells have recently been studied. In 2016, Rasha Abu Eid et al. demonstrated that the pan-PI3K inhibitor GDC-0941 (GDC) enhances the proliferative ability and survival of CD8 + T cells by delaying terminal differentiation and preserving the memory phenotype, which significantly slows tumor growth in B16 tumor bearing mice [108,122]. Subsequently, pretreatment with a series of PI3K inhibitors (e.g., Idelalisib, Duvelisib, and LY294002) during the expansion phase improved the CD19/CD33 CAR-T cells, as indicated by a higher frequency of CCR7 + CD62L + T cells and better antitumor therapeutic capacity in vivo [109,110,112,123]. Inhibitors of components downstream of the PI3K pathway, such as AKT and mTOR, may have a similar effect. For example, as a traditional AKT inhibitor for treatment of chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), Ibrutinib supplementation during CAR-T cell expansion phase might enrich the less-differentiated T cells and reduce expression of exhaustion markers [111,124,125]. Thus, this is an option to further improve the clinical outcomes of CLL patients [111]. Rapamycin (RAPA) has traditionally been administered to patients undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT) and kidney transplant to attenuate GvHD in clinical trials [126]. It has also been reported that pretreatment with RAPA enhances the infiltration capacity of EpCAM CAR-T cells into the bone marrow and increases the elimination of AML in mice by inhibiting mTOR activity [113].
Epigenetic drugs
Studies have gradually clarified that epigenetic factors, including DNA methylation, histone modification, and noncoding RNA, play vital roles not only in tumorigenesis and progression but also in T cell differentiation, metabolism, and function [127,128]. Knowledge of epigenetic remodeling suggests that epigenetic-targeted drugs may serve as an adjuvant for CAR-T cell immunotherapy. DNA methylation inhibitors DNA methyltransferases are gradually become activated during T cell differentiation [129]. Additionally, de novo DNA methylation was shown to promote T cell exhaustion and limit immunotherapy [129]. Therefore, inhibition of DNA methylation is capable of inducing T cell rejuvenation and restricting exhaustion. A recent study revealed the effect of Decitabine, a clinical DNA methylation inhibitor, on CAR-T cells [37]. Addition of this drug to CAR-T cell medium increased the expression of memory-related genes, strengthened the proliferative potential, increased cytokine production, and strengthened tumor cell lytic capacity, even at a very low effector/target ratio, compared to untreated CAR-T cells [37]. Therefore, adding demethylating drugs represents a convenient and economical option to improve the persistence and anticancer properties of CAR-T cells. HDAC inhibitors As significant posttranscriptional modifications of histone proteins, acetylation and deacetylation have been analyzed in detail [130]. Histone acetyltransferases (HATs) use acetyl-CoA, a critical intermediate metabolite and important signal transducer, as the primary substrate for histone acetylation, and acetyl group addition to a histone reduces the positive charge, leading to a relaxed DNA state suitable for active transcription. In contrast, histone deacetylases (HDACs) perform the opposite function [131]. HATs and HDACs have been considered important targets for various diseases [132]. Several HDAC inhibitors have received FDA approval for tumor therapy and are particularly utilized in clinical studies in combination with anti-PD-L1 mAb to restore the host immune response [133]. A recent study demonstrated that SAHA, a clinical HDAC inhibitor, reduced the expression of immunosuppressive molecules (e.g.,
CTLA-4 and TET2) in B7-H3 CAR-T cells and sharply
increased therapeutic efficacy at a low dose [114]. These findings support an unexpected application and potential clinical translation of HDAC inhibitors. BET bromodomain inhibitors Through defined comprehensive epigenetic target screening of chemical probes, JQ1, an inhibitor of bromodomain and extraterminal motif (BET) proteins, was selected because it is able to maintain the functional properties of T SCM cells [115]. Mechanistically, JQ1 inhibits the BET protein BRD4 and directly regulates the expression of the transcription factor BATF in CD8 + T cells, with an improved memory phenotype [115]. CAR-T cells pretreated with JQ1 exhibit enhanced persistence and antitumor effects in mice [115].
Several other conventional drugs with potential applications in CAR-T cell therapy
Metformin For more than 100 years, metformin has been known to reduce glucose levels, and it has been approved by the FDA since 1995 as a first-time treatment for type II diabetes [134]. Many studies have shown that metformin also has many other functions, such as reversing ageing and inhibiting tumor growth [135]. Recently, immunoregulatory potential has been revealed for metformin. Specifically, metformin reprograms T cell differentiation and maintains the phenotype and function of T SCM and T CM cells through the AMPK-miRNA-EOMES-PD1 pathway [92]. These activities lead to increased cytotoxicity in vivo, which may benefit cancer patients [92]. Moreover, a recent identification of targets and mechanisms may lead to a wider and a more accurate application of metformin [136]. Sulforaphane Sulforaphane (SFN) is a naturally occurring antioxidant enriched in cruciferous vegetables that has been regarded as one of the most promising treatments for autism spectrum disorder [137]. SFN can arrest the cell cycle and inhibit tumor progression, moreover, SFN has also been found to modulate immune cell differentiation and function [137].
Based on this, Chunyi Shen et al. evaluated the effect of SFN on CAR-T cells and reported that SFN pretreatment downregulates PD-1 expression in meso CAR-T cells by
inhibiting the PI3K-AKT pathway and improves antitumor efficiency both in vivo and in vitro [116]. Auranofin Auranofin is a gold-containing phosphine compound that has been approved for treating patients with rheumatoid arthritis [138]. It can significantly reduce accumulation of intracellular ROS, and pretreatment of CD19 CAR-T cells or TILs with auranofin increases the elimination of CD19 + tumor cells and autologous tumor spheroids [117]. These data suggest a potential strategy involving use of Auranofin to improve adoptive cellular immunotherapy [117].
In addition to single-drug studies, integrated drug screening and profiling have been implemented to identify small-molecule drugs that modulate CAR T-cell performance among a library of more than 500 approved or investigational compounds [139]. Taken together, these studies show that combining CAR-T cells with traditional drugs can elicit unexpected effects, indicating potential new uses of these drugs as immunomodulatory agents. This offers a new approach for CAR-T cells optimization.
Strategies for enhancing CAR-T cell function by genetic manipulation
Genetic manipulation has been extensively applied across the bioscience and medical field. In particular, the clustered regularly interspaced short palindromic repeat (CRISPR)-associated protein 9 (CRISPR/Cas9) system is considered the next generation of genomic editing technology and it has been widely applied [140]. This system represents a revolutionary innovation for research on the functional genome of various diseases, including monogenic disorders, polygenic disorders, and cancer, which can be used in immunotherapy [141,142]. In this section, we provide an overview of recent approaches and applications of CRISPR/Cas9 to enhance CAR-T cell functions.
Generation of alloCAR-T cells by CRISPR/Cas9
As discussed above, alloCAR-T cells have enormous advantages for immunotherapy, including a short production cycle, low cost, and consistent curative effects. However, given differences in MHC and TCR on the T cell surface between patients and donors, the issues of GvHD and HvGD must be completely resolved. Before the CRISPR system was developed, the first two generations of genomic editing tools, zinc finger nucleases (ZFNs) and transcription activator-like effector nucleases (TALENs), were applied to disrupt endogenous TCRs in alloCAR-T cell [29]. Compared to ZFN and TALEN, the CRISPR system has the highest editing effectiveness and simplest flexibility. Therefore, it has been widely adopted for alloCAR-T cells. Justin Eyquem et al. first targeted a CAR to the TRAC locus with CRISPR/Cas9 and enhanced its tumor elimination ability compared to traditional CAR-T [143,144]. Subsequently, Xiaojuan Liu et al. verified that CRISPR/Cas9 mediated editing of TRAC and B2M is readily applicable to alloCAR-T cells [145]. A series of relevant clinical trials have also been initiated [43,146]. And it has been verified that a one-step CRISPR process can yield alloCAR-T cells with up to four gene deficiencies, highlighting the enormous potential for simultaneous CAR-T cell remodeling [145,147]. Nonetheless, it is worth noting that coexpression of endogenous TCR plus CAR leads to superior persistence of T cells and significantly prolongs leukemia control in vivo compared to TCR knockout CAR-T cells, which suggests the significance of the TCR structure for T cells [148].
Improving CAR-T cell efficiency by CRISPR/Cas9
Compared to its significant curative effects in hematological malignancies, there are multiple obstacles for CAR-T immunotherapy application in solid tumors, primarily due to immunosuppressive TMEs, which disable and exhaust T cells. Knockout of several typical cell-surface immune checkpoint molecules, including PD-1, TIM-3, LAG-3 and CTLA-4, has been implemented in CAR-T cells and corresponding clinical trials are ongoing [149][150][151][152][153]. In addition to these coinhibitory molecules, a series of intracellular negative regulators of TCR signaling, such as tyrosine phosphatase 1B (PTP1B) and Cbl Proto-Oncogene B (CBLB), have been disrupted to enhance CAR-T cell function [154,155]. PTP1B attenuates cytokine induced JAK/STAT signaling by dephosphorylating and deactivating JAK2 and TYK2 [154]. PTP1B is upregulated in tumor infiltrating T cells and it limits T cell expansion and cytotoxicity. PTP1B-specific deletion in CD8 + T cells enhances antigen-induced expansion and cytotoxicity to suppress tumor growth by activating JAK/STAT5 signaling [154]. Furthermore, the pharmacologic inhibition or genetic deletion of PTP1B enhances the efficacy of CAR-T cells against solid tumors [154]. CBLB is an E3 ubiquitin ligase that plays a crucial role in TIL dysfunction [155]. Inhibition or deletion of CBLB could restore the effector function of TILs and reduce PD-1 and TIM-3 levels. In CAR-T cells, this deletion has a similar effect on preventing CAR-T cell exhaustion with superior antitumor capacity [155]. Considering the immunosuppressive effect of several chemokines in TME especially in solid tumors, their receptors are gradually drawing more attention. Disruption of TGF-β receptor II (TGFBRII) in CAR-T cells could decrease exhaustion and improve the killing efficiency to solid tumors [156,157]. As a key immunosuppressive metabolite in TME, adenosine could impair CAR-T cell function and induce sluggishness. Additionally, knockout of its receptor, A2AR, increases the capacity to inhibit solid tumor growth and development, indicating a potential target to promote CAR T-cell therapy [158,159]. Collectively, these data indicated that targeting several immunosuppressive factors in TME and their receptors in CAR-T cells was a favorable option that could significantly improve the use of CAR-T cell therapy in solid tumors.
Improving CAR-T cell efficiency via activator coexpression
The above studies show the remarkable results of strengthening CAR-T abilities by disrupting negative immune regulators with the CRISPR system. Correspondingly, several attempts to improve CAR-T therapy by coexpression of positive immune factors have acquired distinguished effects. Interleukins Addition of IL-2, IL-7, and IL-15 to culture medium stimulates T cell proliferation and preserves the T SCM and T CM proportions (reviewed in Sect. 2). Furthermore, the coexpression of interleukin-encoding genes with the CAR framework has led to promising results in preclinical and clinical studies. IL-7 or IL-15 coexpression in CAR-T cells has been assessed in clinical trials (NCT04381741, NCT03932565, and NCT03721068). Furthermore, recent research has shown the superior antitumor activity of IL-7 receptor (IL7R) coexpression in CAR-T cells targeting neuroblastoma and glioblastoma [160]. Related clinical trials have begun to assess the persistence and efficacy of this approach [160]. Further studies have revealed that coexpression of IL-12, IL-18, IL-21, or IL-23 in CAR-T cells significantly enhances the therapeutic effect against different malignancies, suggesting the significance of interleukins for CAR-T cell optimization [161]. Therefore, IL-12 not only enhances the ability to directly kill tumor cells, but also recruits host immune cells to enhance the anticancer immune response [162]. IL-18-secreting CAR-T cells were shown to modulate the TME and to evoke an effective endogenous antitumor immune response [163]. IL-21 was verified to enhance expansion of CAR-T cells after antigenic stimulation and reduce the apoptosis rate [164]. IL-23 overexpression could regulate CAR-T cells with increased granzyme B expression and decreased PD-1 expression [165]. Chemokine Receptors In the context of solid tumors, the first challenge is the weak infiltration of CAR-T cells into tumors due to the lack of sufficient and appropriate migratory signals. Chemokine systems offer promise as an approach to overcome this obstacle, and such systems have been confirmed to modulate the migration and function of immune cells through interactions between chemokines and specific chemokine receptors (CCRs). Further studies have revealed a significant correlation between chemokine genes and T cell migration. Hence, some studies have focused on rescuing the chemokine and CCR axes to improve the activity of CAR-T cells [166]. Linchun Jin et al. demonstrated that the coexpression of CXCR1 or CXCR2, the CXCL8 receptor, increased the migration and persistence of CAR-T cells in the TME. These increases led to superior tumor regression and long-term immunologic memory in several aggressive tumors, such as glioblastoma, ovarian cancer, and pancreatic cancer [167]. Synergy with CXCR2 coexpression was reported in an additional study [168]. CCL2-CCR2 signaling is another signaling pathway that has been confirmed to enhance the intratumor infiltration of immune cells. In the context of malignant pleural mesotheliomas that secrete high levels of CCL2, CCR2 transduction into meso CAR-T cells increases CCL2-induced calcium flux, cell migration and cell death [168]. Compared with control meso CAR T cells, those with CCR2 coexpression showed a 12.5-fold increase in tumor infiltration in a mouse xenograft tumor model [168]. Moreover, expression levels of proinflammatory cytokines, including IL-2, IFN-γ, and TNF-α, were increased by CCR2 in the treatment of non-small cell lung carcinoma [169]. Another study showed that adoptive T cell intratumoral trafficking is improved by CXCL16-CXCR6 (significantly higher IFN-γ production compared to conventional CAR-T) [170], CXCL12-CXCR4 (enhanced CAR-T cells recruitment into CXCL12-rich bone marrow in an acute myeloid leukemia mouse model) [171], and CCL22-CCR4 (enhanced antitumor efficacy against a subcutaneous xenograft model of human Hodgkin's lymphoma) [166,172]. Taken together, the results from abundant studies have highlighted the potential of combining the chemokine and receptor systems with CAR-T cells to markedly enhance immunotherapy, especially against solid tumors. Given that the concentration and variety of chemokines differ among tumors, it is important to select the optimum chemokine and receptor pathway for combination with CAR-T cells to optimize immunotherapy. Other Active Regulators Several active regulators of TCR signaling or downstream pathways have been gradually found to enhance CAR-T cell efficacy, such as CD80, 41BBL, OX40, and CD40L [173][174][175][176]. Moreover, IL-7 induces T cell polyfunctionality by activating signal transducer and activator of transcription 5 (STAT5) [177,178]. Beyond IL-7 supplementation in the culture medium during CAR-T cell expansion phase, persistent STAT5 activation by genetic manipulation has shown great results. Ectopic expression of a constitutively active form of STAT5 (CASTAT5) modulates CD4 + T cells with robust proliferation and vigorous infiltration abilities and enhances the antitumor response [179]. Further analysis suggested that CASTAT5 enables the remodeling of the genome-wide chromatin structure in CD4 + T cells and establishes a distinct epigenetic and transcriptional landscape [179]. When CASTAT5 was coexpressed in CD19 CAR-T cells, an optimal therapeutic outcome was achieved in a B-cell lymphoma model [179]. Activator protein 1 (AP1) is an important transcriptional regulator that participates in several cellular processes, including immune regulation [180]. Importantly, AP1 dysregulation in CD8 + T cells induces CAR-T cell exhaustion, involving downregulation of cytokines, reduction in expansion, increase in checkpoints, and exaggerated effector differentiation [181]. Further research has shown that c-Jun, a member of the AP1 family, is sufficiently dominant and its overexpression enabled an increase in the proportion of T SCM and T CM cells, long-term proliferation, cytokine secretion and T cells exhaustion [181]. In the context of c-Jun coexpression, CAR-T cells show superior selfrenewal ability, resistance to exhaustion, and antitumor ability in lymphoma and solid tumors [182]. C3aR, the receptor of complement fragment C3a, has been verified to enhance T cell responses and its cooperation in CAR-T cells tended to induce memory T cell phenotype with superior therapeutic potential in extramedullary leukemia [183]. Moreover, the introduction of Toll/interleukin-1 receptor domain of Toll-like receptor 2 in CD19 CAR-T cells showed improved expansion, persistency, and effector function and it has been adopted for relapse or refractory B-ALL patients [184,185]. Therefore, overexpression of several core positive cytokines and regulators might address the major barriers (weak infiltration and poor persistence) to CAR-T efficiency, especially for solid tumors [182].
Newly discovered targets for CAR-T cells with longer lifetimes and better efficiency
Numerous studies have focused on improving CAR-T immunotherapy using CRISPR mediated knockout or ectopic overexpression of genes. The majority of these genes have been verified to perform some function in the antitumor immune response. Prominent results have been achieved with this strategy, and there are several ongoing related clinical trials. To further optimize adoptive T cell therapy, especially for solid tumors, exploration of novel and undetected targets is urgently needed. The CRISPR mediated functional genome-wide screening platform has become an excellent tool for meticulously discovering new gene functions and molecular mechanisms [186,187]. Indeed, this screening strategy has been utilized to identify potential regulators of tumor-immune interactions as new immunotherapy targets in cancer treatment with impacts on processes such as T cell activation, effector function, exhaustion, and cytokine secretion and signaling (Table S1).
Screening for T cell activation To comprehensively search for genes that regulate T cell activation, Wanjing Shang et al. first developed a single-cell-based readout of T cell activation by measuring the expression of CD69, a well-defined early marker of T cell activation, using flow cytometry in Jurkat cells derived from T cells [188]. An unbiased genome-wide screening using a lentiviralinfected Jurkat cell library was performed. Cell was collected and analyzed according to the expression level of CD69, and the screening result confirmed the abundance of well-known regulators involved in T cell activation and identified several previously unexplored genes. For example, family with sequence similarity 49-member B (FAM49B), which was highly ranked in the CD69 high subset, is considered a negative regulator. Further study validated that FAM49B regulates cytoskeletal remodeling via the Rac-PAK axis during T cell activation and that the negative regulatory effect was reversed by FAM49B deficiency [188].
Screening for T cell effector function In addition to high-throughput loss-of-function screening, gain-offunction screening using a modified CRISPR system has been applied to identify functional boosters in primary T cells. Lupeng Ye et al. developed a dead-guide RNAbased genome-wide gain-of-function CRISPR activation screening system [189]. During the screening process, CD107a was chosen as the marker to reflect the ability of cytotoxic CD8 + T lymphocytes to kill tumor cells [189]. CD8 + T cells expressing CD107a at high levels (top 5%) were sorted and sequenced. Further analysis revealed several uncharacterized genes enriched in CD107a high cells, among which proline dehydrogenase 2 (PRODH2) was verified to reprogram proline metabolism and promote proliferation in CD8 + T cells [189]. Moreover, PRODH2 engineering by genomic knockin or lentiviral overexpression increases CD22 CAR-T cytotoxicity towards cancer cells [189].
Screening for T cell exhaustion For a more comprehensive understanding of the molecular events regulating CAR T-cell exhaustion, Dongrui Wang et al. developed a robust method for whole genome CRISPR-KO screening of human IL13Rα2 CAR-T cells targeting glioblastoma cells [190]. After transduction with sgRNA library lentivirus, the CAR-T cells were recursively incubated with excess GBM stem cells. The CAR-T cells were then sorted from the coculture system and grouped based on expression of the inhibitory receptor PD-1, a typical marker of T cell exhaustion [190]. A potential target for repressing CAR-T antitumor activity was detected in the PD-1 low subset. Knockout of the top hits (TLE4, IKZF2, TMEM184B, and EIF5A) repressed PD-1 expression, finally inhibiting exhaustion and improving CAR-T cell cytotoxicity in vitro [190].
Screening for T cell persistence CD62L was selected as a pivotal marker to further improve persistence of CAR-T cells [81]. Devikala Gurusamy et al. performed CRISPR screening of CD8 + T cells in mice [191]. Primary CD8 + T cells were stimulated twice and then classified by CD62, ROS, γH2AX, and proliferation signaling. The sgRNA distributions showed that p38 kinase can regulate the desired phenotypes of T cells. Further genetic and pharmacological studies suggested that p38 kinase blockade improves T cell fitness via metabolic and transcriptional alterations. Finally, p38 inhibition improved the persistence and antitumor efficacy of CAR-based adoptive immunotherapies [191]. Similar screening studies have also been related to T cell proliferation, cytokine regulation, and tumor cell resistance, but these studies differed in biosample types, library sources, sgRNA numbers, and screening protocols (Table S1) [192][193][194].
Taken together, increased application of whole-genome screening in immuno-oncology networks based on the CRISPR system has gradually revealed the astonishing ability to identify potential immune regulators. Thus, such studies may ultimately uncover novel opportunities and strategies for improving the efficiency of current immunotherapy agents.
Conclusion
CAR-T cell immunotherapy has emerged and rapidly developed into a promising treatment option for various types of cancer, especially hematologic malignancies. However, several barriers exist that need to be overcome to achieve superior clinical results. In recent years, numerous studies have concentrated on identifying strategies to optimize CAR-T cell efficiency. The aims are to regulate the T cell phenotype and drive T cell differentiation for superior effector function and longer persistence [93]. Therefore, we systematically examined the optimizating strategies that aim to improve the longterm persistence and antitumor performance of CAR-T cells (Fig. 2).
As the T cell source and components influence the applicable targets of CAR-T cells, they should be prudently chosen. According to existing clinical data, CD4 + T cell-based CAR-T cells show better persistence than CD8 + T cells, and autoCAR-T cells exist in vivo longer than alloCAR-T cells [15,44]. Culture conditions also influence CAR-T cell proliferation and differentiation by regulating cellular metabolism. Several metabolites, such as glutamine inhibitors, L-arginine, and 2-DG, have been found to alter the T cell development trajectory by enhancing OXPHOS and restricting glycolysis [101,103,106]. Interestingly, some conventional drugs improve the antitumor performance of CAR-T cells, relying on their ability to regulate metabolism and differentiation and thus prolong persistence. These drugs might have clinical applications as CAR-T adjuvants. In fact, Idelalisib, Rapamycin, Duvelisib, Decitabine, and Ibrutinib have been deemed safe in clinical trials (Table 2). Moreover, genetic manipulation has been quickly applied to CAR-T cells, especially with the discovery and development of the CRISPR system [195]. This approach not only targets traditional checkpoints but also several newfound targets. It is expected that the CRISPR system will increase CAR-T cell applications and lead to improvements. It is worth noting that combining these methods might result in superior in vivo persistence of CAR-T cells, particularly in the context of solid tumors surrounded by a hostile microenvironment [195]. To take advantage of these strategies for CAR-T cells to treat cancer, there is an urgent need to better understand the underlying mechanisms and to quickly carry out preclinical and clinical trials. It's worth noting that several factors in specific clinical setting also influenced the CAR-T persistence, such as the choice and effect of previous lymphodepleting chemotherapy which could influence the state of TME, the distribution and density of antigen target which induces CAR-T cells exhaustion [196][197][198][199][200]. Based on this, the clinical setting should not be ignored for improving CAR-T cells persistence in vivo.
Notably, several recent studies have revolutionized the in vitro CAR-T cell expansion process by shortening Fig. 2 Summary of strategies to improve CAR-T persistence. Optimization approaches to improve CAR-T cell persistence and antitumor performance, including choosing a suitable cell source, improving culture conditions with metabolites, combining CAR-T cells with conventional drugs, and applying genetic manipulations the time required [201][202][203]. Compared to the traditional 1-to 2-week waiting period, the new process only requires 1 to 3 days [201]. The CAR-T cells generated using this shorter in vitro culture process exhibited better proliferation, longer persistence, and better tumor killing ability than conventional CAR-T cells. A recent clinical study in which r/r B-ALL patients were treated with FasTCAR-T cells reported a complete CR at Day 28 and a CR rate of 83.3% at the 3-month assessment [204]. Further attention should be given to comparison between conventional and short-term CAR-T cells. Interestingly, several other strategies about vaccination and viruses application to improve CAR-T cells persistence and clinical efficacy have successively emerged. For example, when CAR-T cells were generated by donor Epstein-Barr virus-specific T-cells, they showed superior expansion/persistence ability and limited CRS, neurotoxicity, and GvHD in clinical trials [205]. Another virus, oncolytic virus, was capable to enter into tumor cells and delivers target protein which was recognized and killed by CAR-T cells, and this combination strategy showed promising results for multiple solid tumors [206]. Similarly, a nanoparticulate RNA vaccine has also been applied to deliver a protein into solid tumors as CAR-T target, which could selectively enhance the expansion of CAR-T cells in vivo [207]. These remarkable studies broadened the application and efficacy especially for solid tumors.
In this review, we summarize several strategies to regulate CAR-T cell metabolism and differentiation to improve CAR-T persistence, which aims at providing methods or clues for recurrence and refractory problem after CAR-T cell therapy. Also, the strategies could be applied in expanding immune cell therapy to more solid tumor, auto-immune diseases and cost reduction to benefit more patients.
Additional file 1: Table S1. CRISPR system-based screenings for CAR-T cell optimization.
|
v2
|
2022-11-24T15:00:52.966Z
|
2022-11-23T00:00:00.000Z
|
253805646
|
s2orc/train
|
Comprehensive analysis of the differences between left- and right-side colorectal cancer and respective prognostic prediction
Background Previous studies have reported that the tumor heterogeneity and complex oncogenic mechanisms of proximal and distal colon cancer (CRC) are divergent. Therefore, we aim to analyze the differences between left-sided CRC (L_cancer) and right-sided CRC (R_cancer), as well as constructing respective nomograms. Methods We enrolled 335 colon cancer patients (146 L_cancer patients and 189 R_cancer patients) from The Cancer Genome Atlas (TCGA) data sets, and 102 pairs of color cancer tissue and adjacent normal tissue (51 L_cancer patients and 51 R_cancer patients) from our hospital. Firstly, we analyzed the differences between the L_cancer patients and R_cancer patients, and then established the L_cancer and R_cancer prognostic models using LASSO Cox. Results R_cancer patients had lower survival than L_cancer patients. R_cancer patients had higher ESTIMATE and immune scores and lower tumor purity. These patterns of expression of immune checkpoint-related genes and TMB level were higher in R_cancer than in L_cancer patients. Finally, we using Lasso Cox regression analyses established a prognostic model for L_cancer patients and a prognostic model for R_cancer patients. The AUC values of the risk score for OS in L_cancer were 0.862 in the training set and 0.914 in the testing set, while those in R_cancer were 0.835 in the training set and 0.857 in the testing set. The AUC values in fivefold cross-validation were between 0.727 and 0.978, proving that the two prognostic models have great stability. The nomogram of L_cancer included prognostic genes, age, pathological M, pathological stage, and gender, the AUC values of which were 0.800 in the training set and 0.905 in the testing set. Meanwhile, the nomogram of R_cancer comprised prognostic genes, pathological N, pathological T, and age, the AUC values of which were 0.836 in the training set and 0.850 in the testing set. In the R_cancer patients, high-risk patients had a lower proportion of ‘B cells memory’, ‘Dendritic cells resting’, immune score, ESTIMATE score, immune checkpoint-related genes, and HLA-family genes, and a higher proportion of ‘T cells follicular helper’, ‘Dendritic cells activated’, and ‘Mast cells activated’. Conclusions We found significant differences between L_cancer and R_cancer patients and established a clinical predictive nomogram for L_cancer patients and a nomogram for R_cancer patients. Additionally, R_cancer patients in low-risk groups may be more beneficial from immunotherapy. Supplementary Information The online version contains supplementary material available at 10.1186/s12876-022-02585-3.
Introduction
Colon cancer (CRC) is one of the most common cancers and cause of cancer death globally, seriously endangering the health of patients [1]. In recent years, there has been a growing body of evidence demonstrating that the primary tumor location of CRC is an important prognostic factor, owing to distinct biological features [2][3][4]. Despite the fact that the primary tumor site is not generally considered in CRC management, left-sided colon cancers (L_cancer) and right-sided colon cancers (R_cancer) exhibit different clinical and biological characteristics [5]. A meta-analysis of 66 studies with more than 1.4 million patients with a median follow-up of 65 months revealed that the tumor side had a significant prognostic impact on overall survival, with a 20% percent longer life expectancy, independent of stage, race, adjuvant chemotherapy, year of study, number of participants, and quality of included studies. [6]. The differences in colon cancer by its location have been identified through extensive research, including survival, tumor microenvironment, methylation profile, microbiota, gene expression, and epigenetic changes. [2,3,[6][7][8]. In addition, the tumor location also influences the outcome of adjuvant chemotherapy, palliative therapy, or targeted therapy. Therefore, it is of special significance to classify CRC by its location.
Nomograms are widely used for prognosis in CRC patients. However, few previous studies have separately built predictive models to predict patient prognosis with respect to location. In this study, we separately build predictive models for L_cancer and R_cancer, identifying potential prognostic biomarkers for left and right CRC. Age, sex, histological classification, and so forth, are also important factors that can influence clinical outcomes and can improve the accuracy of models. Therefore, we also aimed to analyze the differences between L_cancer and R_cancer and construct respective nomograms for L_cancer and R_cancer, containing prognostic gene signatures and clinical prognostic factors, which are expected to allow for more accurate predictions in the prognosis of CRC, facilitating accurate diagnosis and treatment.
Data sets
The transcriptome data, somatic mutation data, and clinical information of CRC patients were downloaded from The Cancer Genome Atlas (TCGA, https:// portal. gdc. cancer. gov/), which includes transcriptome data for 332 CRC patients (146 L_cancer patients and 189 R_cancer patients) and somatic mutation data for 329 CRC patients (142 L_cancer patients and 187 R_cancer patients).
L_cancer patients were divided into L_cancer training and L_cancer internal validation sets at a ratio of 7:3. The L_cancer external validation set contained those who operated in our hospital, including 51 L_cancer patients.
R_cancer patients were also divided into R_cancer training set and R_cancer internal validation sets at a ratio of 7:3. The R_cancer external validation set contained those who operated in our hospital, including 51 R_cancer patients.
A total of 102 pairs of colon cancer and adjacent normal control samples were stored at − 80 °C. Patients were followed up by telephone interviews. As of the final data cutoff, December 30, 2021, the median duration of follow-up in the study was 4.5 years and the criterion to proceed with the final OS analysis was met.
The term "R_cancer" refers to any (histologically confirmed) adenocarcinoma arising from the caecum, ascending colon, or hepatic flexure. Any tumor that arises in the splenic flexure, descending colon or sigmoid colon was referred to as L_cancer.
Survival analysis
Using Kaplan-Meier survival analysis, we evaluated the differences in survival between patients with different clinicopathological characteristics, between highrisk and low-risk groups and between the L_cancer and R_cancer groups in the data sets mentioned above. The 'survival' package in R was used to perform a two-sided log-rank test and univariate and multivariate Cox regression analyses [9].
Differential gene analysis and functional annotation
By using the "edgeR" package in R, we identified differentially expressed genes (DEGs) between L_cancer and R_cancer, L_cancer and L_normal, R_cancer and R_normal based on differential expression analysis. To screen for DEGs, |log2 FC (fold-change)|> 1 and P < 0.05 were set as thresholds. To investigate the possible biological Conclusions: We found significant differences between L_cancer and R_cancer patients and established a clinical predictive nomogram for L_cancer patients and a nomogram for R_cancer patients. Additionally, R_cancer patients in low-risk groups may be more beneficial from immunotherapy. Keywords: Left-sided colon cancer, Right-sided colon cancer, Biomarkers, Nomogram, Immune microenvironment, Tumor mutation burden, Immune checkpoint processes, cellular components, and molecular functions of DEGs, GO enrichment and KEGG pathway analyses were performed by using the R software package "cluster-Profiler" [10][11][12].
Gene set variation analysis (GSVA)
By using the "GSVA" package in R, we evaluated the t-scores and assigned pathway activity conditions to L_ cancer and R_cancer patients to reveal pathway enrichment. The "limma" package in R was also used to show differences in pathway activation between L_cancer and R_cancer patients [13][14][15].
The proportion of immune cell infiltration and the calculation of tumor purity
In each cancer sample, the relative proportions of 22 immune cell types were calculated using the CIBER-SORT software [16]. A file called "LM22.txt", containing 547 gene signatures (https:// ciber sort. stanf ord. edu/ downl oad. php), is also needed in R. ESTIMATE was used to calculate immune, stromal, and ESTIMATE scores, as well as tumor purity, based on Yoshihara et al. [17].
Profiles of tumor mutation burden (TMB) and correlation analysis
The TMB was defined as: TMB = (total count of variants)/(the whole length of exons). In a waterfall plot, the mutation profiles of two groups were compared using the maftools package [18]. Afterward, the difference in mutation frequencies between the two groups was measured with the chi-square test. TMB was derived for each patient, calculated using Pearson correlation analysis with estimated P-values.
LASSO cox regression analysis
LASSO Cox regression analysis with the R package glmnet was then used to identify hub genes associated with the prognosis of L_cancer or R_cancer, and a Risk Score was calculated for each sample using the screened hub genes following the following formula [19]: where N represents the number of signature genes, Expi is the gene expression levels, and Coef is the estimated regression coefficient value from the Cox proportionalhazards analysis. Based on this optimal cutoff value, the R survival package "survminer" was used to divide patient groups into Low-and High-Risk groups. Moreover, model predictive power was evaluated by calculating the AUC of 1-, 3-, 5-, 7-year, and all time-dependent ROC curves, using the "survivalROC" package.
Building and validating a predictive nomogram
To construct the nomograms, we used univariate and multivariate Cox regression analyses. Forest plots were used to display the P-value, HR, and 95% CI for each variable, using R's 'forest plot' package. Based on independent prognostic factors, the nomograms were generated in R using the rms, nomogramEx, and ggDCA packages. In the next step, Using calibration curves, we determined whether the predicted survival outcome matched the actual outcome. Moreover, training set decision curve analysis (DCA) and internal validation set DCA, which is a statistical method for assessing and comparing predictive models, was used to determine the clinical suitability of our established nomograms.
RNA isolation and quantitative reverse transcription PCR assay
For total RNA isolation, the TRIzol reagent by Invitrogen was used, and for complementary DNA synthesis, the PrimeScript RT reagent kit by Takara was used. RT-PCR was carried out using SYBR Premix Ex Taq I. GAPDH served as an internal control. Relative RNA abundances were calculated by using the standard 2-ΔCt method.
Statistical analysis
A two-sided significance level of 0.05 was used to determine statistical significance in all analyses using R software (version 3.6.3). All significance levels were two-sided.
Differences between L_cancer and R_cancer patients Differences in demographic characteristics between L_cancer and R_cancer patients
An overview of the steps is presented as a flow chart in Fig. 1. The demographic characteristics of patients are summarized in Table 1. The L_cancer patients found a significant difference between R_cancer patients regarding age, stage N, and survival rate (P < 0.05). It is noteworthy that we observed lower survival after R_cancer versus L_cancer ( Fig. 2A).
Moreover, there is no difference between the training set and the verification set except T stage. The difference in the T stage may due to the poor stage of patients from our hospital, but it does not affect the internal validation.
Differential expressed genes and functional annotation between L_cancer and R_cancer patients
By comparing the transcriptome data, we identified 540 significantly up-regulated DEGs in the L_cancer group and 1507 significantly up-regulated DEGs in the R_cancer group (Fig. 2B). The heatmap was shown the top 40 DEGs with the greatest variation (Fig. 2C).
Differential immune microenvironment between L_cancer and R_cancer patients
By comparing the immune microenvironments between L_cancer and R_cancer patients, significant differences were observed between the two groups with regard to immune infiltration components.
Comparing the Stromal score, ESTIMATE score, immune score, and tumor purity of L_cancer and R_cancer patients, we found that the R_cancer patients had a lower tumor purity and higher ESTIMATE and immune scores (Wilcoxon test, P < 0.05; Fig. 3B) than L_cancer patients.
We also analyzed the immune checkpoint-related genes (PD-1, PD-L1, CTLA4, CD86, LAG3, HAVCR2, TIGIT) and HLA family-related genes levels, which are considered biomarkers for predicting the efficacy of immunotherapy, between L_cancer and R_cancer patients and found that the expression levels of immune checkpoint-related genes and HLA family-related genes were significantly higher in R_cancer patients (Wilcoxon test, all P < 0.05; Fig. 3C, D).
Identifying DEGs and functional annotation in tumor and normal patients
By comparing the transcriptome data of L_cancer and L_normal groups, we identified 4788 up-regulated DEGs and 4062 down-regulated DEGs (Fig. 5A). The top 20 up-regulated and down-regulated genes were displayed by heatmap (Fig. 5C). Further, we analyzed these DEGs between L_cancer and L_normal groups for functional enrichment analysis. This evaluation revealed the enrichment of 1139 GO terms and 65 KEGG pathways (FDR < 0.05). We chose to show the top 10 GO terms and 15 KEGG pathways in Fig. 5E, G.
Likewise, the DEGs between R_cancer and R_normal identified 6261 up-regulated DEGs and 4501 down-regulated DEGs (Fig. 5B). The top 20 up-regulated and downregulated genes were displayed by heatmap (Fig. 5D). These DEGs between R_cancer and R_normal groups be analyzed for functional enrichment analysis. A total of 1072 GO terms and 61 KEGG pathways had been enriched (FDR < 0.05). We chose to show the top 10 GO terms and 15 KEGG pathways in Fig. 5F, H.
Construction of prognostic gene model
To identify prognosis-related genes, we first screened genes using the Kaplan-Meier method in DEGs with P < 0.05, in order to screen survival-related DEGs as candidate genes affecting prognosis. Then, to avoid model overfitting, we performed a multivariate Cox regression analysis with the LASSO penalty algorithm to solve the multi-collinearity problem. Finally, we obtained 10 genes associated with the prognosis of L_cancer patients and 10 genes associated with the prognosis of R_cancer patients. These genes have a significant impact on the survival of patients (Additional file 1: Fig. S1).
The L_cancer patient prognosis features and risk score were calculated as: (Fig. 6A, B). The cutoff of risk score is 7.801, which had a great impact on OS (Fig. 6C). Scores lower than 7.801 have been defined as low-risk L_cancer patients, while scores higher than 7.801 have been defined as high-risk L_cancer patients. The AUC values of the risk score in the training set for 1-year, 3-year, 5-year, 7-year, and all-time OS were 0.554, 0.582, 0.593, 0.597, and 0.862, respectively (Fig. 6D). (Fig. 7A, B). The cutoff of risk score is 11.981, which had a great impact on OS (Fig. 7C). Scores lower than 1.981 have been defined as low-risk R_cancer patients, while scores higher than 1.981 have been defined as high-risk R_cancer patients. The AUC values of the risk score in the training set for 1-year, 3-year, 5-year, 7-year, and all-time OS were 0.557, 0.610, 0.626, 0.692, and 0.835, respectively (Fig. 7D).
Internal validation of the prognosis genes model and stratified analysis by clinical factors
The efficacy of the prognostic signature was validated using a testing set of TCGA patients. Five-fold cross-validation was used to assess the stability of the model.
Among the L_cancer patients, the area under the curve (AUC) values of risk scores predicted in the testing set for 1-year, 3-year, 5-year, 7-year, and all-time OS were 0.597, 0.696, 0.722, 0.723, and 0.914, respectively (Fig. 6E). The AUC values of fivefold cross-validation were 0.860, 0.792, 0.908, 0.854, and 0.978, respectively, and the integrated AUC value was 0.863 (Fig. 6F). The results revealed that the AUC values of fivefold cross-validation were high and similar, indicating that the model had good predictability and stability. Based on the obtained sample clinical characteristics, patients were stratified into age < 65 years and age ≥ 65 years sub-groups (Fig. 6G, H), female and male sub-groups (Fig. 6I, J), and pathological tumor Stage I/II and Stage III/IV sub-groups (Fig. 6K, L). The overall survival analysis was performed in each sub-group, based on the level of risk score, and all results showed statistical differences.
Likewise, in R_cancer patients, the AUC values of risk scores predicted in the test set for 1-year, 3-year, 5-year, 7-year, and all-time OS were 0.679, 0.725, 0.771, 0.801, and 0.857, respectively (Fig. 7E). The AUC values of fivefold cross-validation were 0.838, 0.727, 0.796, 0.793, and 0.826, respectively, and the integrated AUC value was 0.792 (Fig. 7F). The results revealed that the AUC values of fivefold cross-validation were high and similar, indicating the model had good predictability and stability. Patients were also stratified into age < 65 years and age ≥ 65 years sub-groups (Fig. 7G, H), female and male sub-groups (Fig. 7I, J), and pathological tumor Stage I/II and Stage III/IV sub-groups (Fig. 7K, L). Overall survival analysis was also performed in each sub-group, based on the level of risk score, and all the results showed statistical differences.
Incorporating clinical factors to develop individualized nomograms
Clinical characteristics, including Age, Gender, T, N, M, Stage, and risk score, were utilized to perform univariate analyses in the training sets of L_cancer (Fig. 8A) and R_cancer (Fig. 9A), respectively. After statistical adjustment for other variables with multivariate Cox regression analysis, we found that the Risk, pathological M, pathological stage, gender, and age were the only six independent prognostic factors that could be used to predict the survival rate in L_cancer (Fig. 8B), while the Risk, pathological N, pathological T, and age were the only four independent prognostic factors that could be used to predict the survival rate in R_cancer. (Fig. 9B). L_cancer patients' nomogram (Fig. 8C) and R_cancer patients' nomogram ( Fig. 9C) were developed using the above prognostic features, with the total points calculated by adding the points of individual prognostic features.
Predictive performance of the established nomogram
Among L_cancer patients, the calibration curve and decision curve analysis for predicting median survival time OS in the training and testing sets indicated that the nomogram-predicted survival similarly corresponded with actual survival outcomes (Fig. 8D, E). The AUC of the nomogram was 0.8 in the training set and 0.905 in the testing set (Fig. 8F, G).
In R_cancer patients, the calibration curve and decision curve analysis for predicting median survival time OS in the training and testing sets indicated that the nomogram-predicted survival similarly corresponded with actual survival outcomes (Fig. 9D, E). The AUC of the nomogram was 0.836 in the training set and 0.850 in the testing set. (Fig. 9F, G).
External validation of the prognosis signature by qRT-PCR
The obtained results were further validated by qRT-PCR, as shown in Fig. 10.
In 51 pairs of L_cancer patients, compared with adjacent cancer tissues, the expression of DAND5, SMPD1, KNG1, NKPD1, and CYP11A1 were found to be downregulated in cancer tissues (two-tailed paired t-test; all P < 0.05, Fig. 10A-E).
Moreover, in 51 pairs of R_cancer patients, compared with adjacent cancer tissues, the expression of LPO, METTL11B, and PTGS2 were found to be up-regulated, and ZC3H12C and MOCS1 were down-regulated in cancer tissues (two-tailed paired t-test; all P < 0.05, Fig. 10F-J).
Differences in the immune microenvironment, TMB landscape, immune checkpoint-related genes, and HLA-family genes level between high-and low-risk patients
Based on the difference in the immune microenvironment and TMB landscape between left and right CRC, we next analyzed the difference in these aspects between high-and low-risk patients based on prognostic gene models.
In the R_cancer patients, high-risk patients had a lower proportion of 'B cells memory' , 'Dendritic cells resting' , immune score, ESTIMATE score, immune checkpoint-related genes, and HLA-family genes, and a higher proportion of 'T cells follicular helper' , 'Dendritic cells activated' , and 'Mast cells activated' (Wilcoxon test, P < 0.05; Fig. 11A-E). These results indicate that R_cancer patients in high-and low-risk groups may have different responses to immunotherapy, and immunotherapy in R_cancer low-risk patients may be more beneficial.
In the L_ancer patients, there was no difference in these indicators between high-and low-risk patients (Additional file 2: Fig. S2A-E).
Correlation of hub gene and risk score with immune-related score and genes
Correlation analyses were carried out for risk scores and hub genes with immune-related scores and genes. As we can see, in R_cancer patients, R_cancer risk score was strongly correlated with immune-related scores and genes (Fig. 12). In particular, it has a significant negative correlation with immune checkpoint-related genes, Stromal score, immune score, and ESTIMATE score and a positive correlation with tumor purity. These results prove that R_cancer patients with R_cancer low-risk score may benefit more from immunotherapy. In addition, the R_cancer risk score was positively associated with the content of 'B cells memory' , 'T cells CD4 naïve' , 'T cells regulatory Tregs' , 'Macrophages M0' , and 'Dendritic cells resting' and negatively associated with the content of 'T cells follicular helper' , 'Dendritic cells activated' , 'Mast cells activated' and 'Neutrophils' .
In L_cancer patients, L_cancer risk score was no correlation with immune-related scores and genes (Additional file 3: Fig. S3).
Discussion
CRC has a heterogeneous tumor composition and complex oncogenic mechanisms. The development of individualized treatment strategies and the evaluation of patient prognoses based on tumor location are crucial. This study is the first to separately build predictive models for L_cancer and R_cancer, to the best of our knowledge. We presented two nomograms for CRC classified with Fig. 11 A The comparison of immune infiltration levels between high-risk and low-risk groups in R_cancer patients, based on CIBERSORT. B The Stromal Score difference, Immune Score difference, ESTIMATE Score difference, and tumor purity difference between high-risk and low-risk groups in R_cancer patients. C The immune checkpoint-related gene expression levels in high-risk and low-risk groups in R_cancer patients. D The tumor mutation burden difference between high-risk and low-risk groups in R_cancer patients. E HLA-related gene expression level between high-risk and low-risk groups in R_cancer patients. Notes: ns P > 0.05, * P < 0.05, ** P < 0.01, *** P < 0. Numerous studies have confirmed that the right-and left-sided colons are distinct due to their embryological origins. The right-side colon originate from the midgut, whereas the left-side colon originate from the hindgut. In this study, we confirmed that there exist significant differences in the TMB and immune microenvironment between right-and left-sided CRC patients. Furthermore, right-sided CRC tend to have worse prognosis than left-sided CRC patients. The difference between rightand left-sided CRC patients' survival rates is might be caused by the higher frequency of mutations in addition to changes in the tumor microenvironment associated with tumor purity. According to recent research, mutation prevalence differs depending on side and location. RAS mutations declined from 70% in patients with rightsided CRC to 43% in those with left-sided CRC, while the number of BRAFV600 mutations increased from 10 to 22% between the same locations. Sigmoid and rectal tumors with left-sided mutations were more likely to harbor TP53 mutations than PIK3CA, BRAF, or CTNNB1 mutations [3]. Consistent with our results, in left-sided tumors, TP53 (L_cancer: 68%, R_cancer: 48%) showed a higher mutation rate; meanwhile, in right-sided tumors, PIK3CA (L_cancer: 18%, R_cancer: 33%) and KRAS (L_cancer: 36%, R_cancer: 46%) showed higher yield mutation rates. The results in our study align well with a recent report by Marshall et.al., who also demonstrated significant differences between L_cancer and R_cancer in mutation patterns.
The tumor microenvironment (TME) refers to the physical environment around a tumor, including the immune cells, neurons, blood vessels, extracellular matrix, and other cellular functions related to tumor progression and therapy effects. We also confirmed that the immune microenvironment affects the prognosis of patients with CRC. Aggressively growing tumors create a highly immunosuppressive TME that depletes antitumor responses and promotes tumor progression [19,20].
Based on the Estimation of STromal and Immune cells in MAlignant Tumor tissues using Expression data approach, immune score and tumor purity can reveal information about the tumor's immune status. Low immune scores and high tumor purity have been associated with better prognoses in several studies [21][22][23]. Based on this, we examined the differences in tumor immune microenvironment between right-and left-sided CRC patients. In our study, L_cancer patients not only had poor prognosis but also had high ESTIMATE and immune scores, as well as low tumor purity. Thus, we further analyzed the effect of high-or low-risk on immune infiltration in patients in both L_cancer and R_cancer models. We found that, in the R_cancer model, high-risk patients had lower immune and ESTIMATE scores and higher tumor purity than low-risk patients. However, there was no difference between high-and low-risk in the L_cancer model with respect to immune infiltration. Besides, in the R_cancer model, high-risk patients were significantly different from low-risk patients in terms of immune infiltrating cell types, such as memory B-cells, dendritic, T follicular helper cells and mast cell activation. Nevertheless, in the L_cancer model, the high-and low-risk patients showed no difference. These results may be related to our different models for L_cancer and R_cancer. The findings of some studies were in line with our study, where low tumor purity result in poor prognosis in glioma and CRC [21,22]. Additionally, the proportions of CD8 T-cells and T follicular helper cells were significantly higher in the R_cancer group, while M0 macrophages had higher infiltration in L_cancer groups. A recent single-cell RNA-Seq study between right-and left-sided CRC patients discussed the difference in single-cell transcriptomes between the two groups, which was in line with our findings. In summary, there has been increasing awareness of the body's ability to fight tumors through various types of cells cytokines, and chemokines. Immune cells, especially, play a critical role in this. Immunotherapy has become increasingly popular as a treatment option for cancer patients with refractory malignant tumors, which can benefit significantly from immune checkpoint inhibitors. To determine whether immunotherapy is effective, TMB, TME, and immune checkpoint levels are considered as biomarkers [23][24][25]. A previous study has demonstrated that, in CRC patients, the prognostic impact of PD-L1 and PD-1 expression varies according to the primary tumor site. Moreover, the presence level of PD-L1 is an independent prognostic factor for right-side tumors [26]. This finding was in line with our study, which demonstrated that there were significant differences in PD-1, PD-L1, and CTLA4 expression between right-and left-sided CRC patients.
Given this, this study independently assessed the effect of the tumor microenvironment in L_cancer and R_cancer of high-and low-risk patients from two aspects (TMB and immune microenvironment), leading us to speculate that R_cancer-especially low-risk R_cancer-patients may benefit more from immunotherapy [27,28]. Validation is needed, but these results could be clinically significant as they indicate that tumor location is important to consider in therapeutic decisions, including eligibility for immunotherapy.
The hub genes in the signature have previously been shown to be potential biomarkers. Relevant research has reported that PTGS2-driven inflammatory responses can induce tumor expression of microRNA-21, which can increase the level of the inflammatory mediator prostaglandin E2 (PGE2) by down-regulating PGE2-metabolizing enzymes, contributing to colorectal cancer development [28][29][30][31][32]. PLEKHA8P1 expression has been associated with the development and progression of many malignancies in humans, such as CRC and renal cancer [33]; moreover, research has shown that its dysregulated expression affects 5-Fluorouracil-induced chemoresistance in the human hepatocellular carcinoma cell line FT3-7 [34]. Prior studies found ZC3H12A has links with immune homeostasis and post-transcriptional regulation which can stimulate tumor progression in lung and colon cancer [35][36][37]. LPO can collaborate with activated Wnt signaling to induce intestinal neoplasia [38]. METTL11B expression has been associated with poor prognosis in colorectal cancer and is higher in cancer tissues than in neighboring normal tissues [39]. NKPD1 has been predicted to be linked with the de novo synthesis of sphingolipids [40]. Increased DAND5 level is an independent risk factor for both colorectal and breast cancers and the prediction of poor prognoses [41,42]. SMPD1 encodes lysosomal acid sphingomyelinase, which converts sphingomyelin to ceramide. Prior studies have found that the functional inhibition of acid sphingomyelinase contributes to tumor cell death by overactivation of hypoxia stressresponse pathways [43]. Another study has shown that down-regulation of SMPD1 is linked with resistance to chemotherapy regimens including 5-Fluorouracil [44]. Studies have shown CYP11A1, which can hydroxylate the side-chain of vitamin D3 at carbons 17, 20, 22, and 23, are related to susceptibility to breast cancer [45,46]. KNG1 can regulate the expressions of VEGF, cyclinD1, ki67, and caspase-3/9, exerting anti-angiogenic properties and inhibiting the proliferation of endothelial cells. Over-expression of KNG1 can inhibit the activity of PI3K/Akt, decrease tumor growth, and promote apoptosis [47]. On the contrary, other researchers have found that KNG1 expression was significantly increased in colorectal cancer lesions [48]. At present, there has been no reported association between MOSC1, RP11-278A23.1, RP11-452K12.7, RP11-742B18.1, RP11-626H12.2 RP11-59D5_B.2, CTD-2184C24.2, RP11-680F8.3, RP11-51F16.9, CTD-2012K14.8, and cancer. In the end, RT-qPCR was performed to verify the results from the bioinformatic analyses of LCC and RCC. We revealed that the prognostic gene expression results were consistent with the outcomes of our survival analysis, indicating that our results are reproducible and reliable. In addition, this further confirmed that these key genes are related to the occurrence and development of colon cancer.
This study had some limitations. The signatures and nomograms constructed in this study using vast datasets from TCGA and our patient database were robust, but the study was still a retrospective one. Second, we explored the TMB and immune microenvironment landscape between right-and left-sided CRC patients and between patients in different risk groups; however, the study lacked experimental verification. Third, as previously noted, obtaining risk scores requires knowledge of ten genes expressed in tumor tissues, thereby increasing the difficulty of applying the nomograms. It appears that many molecular diagnostic or prognostic models have the same problem. Researchers and clinicians need to figure out how to simplify the application of these models in clinical settings. In the future, molecular detection technology may solve this dilemma. The constructed nomograms may be used routinely.
Conclusions
We found significant differences between L_cancer and R_cancer patients, including clinical features, transcriptome, TMB, immune microenvironment landscape, suggesting that colon cancer can be classified and analyzed into different clinical types with respect to their differences in anatomical location and gene expression, thus aiding in the early diagnosis and prognosis of colon cancer. We established two clinical predictive nomograms in combination with clinical features to provide a basis for the personalized and precise treatment of L_cancer and R_cancer. These hub genes may become promising biomarkers for the diagnosis, treatment, and prognosis of colon cancer. Moreover, The findings support previous studies suggesting that proximal and distal CRC can be classified differently in terms of epidemiology, pathology, and genetics.
|
v2
|
2022-11-24T15:00:54.361Z
|
2022-11-23T00:00:00.000Z
|
253805166
|
s2orc/train
|
Introducing new plan evaluation indices for prostate dose painting IMRT plans based on apparent diffusion coefficient images
Background Dose painting planning would be more complicated due to different levels of prescribed doses and more complex evaluation with conventional plan quality indices considering uniform dose prescription. Therefore, we tried to introduce new indices for evaluating the dose distribution conformity and homogeneity of treatment volumes based on the tumoral cell density and relative volumes of each lesion in prostate IMRT. Methods CT and MRI scans of 20 male patients having local prostate cancer were used for IMRT DP planning. Apparent diffusion coefficient (ADC) images were imported to a MATLAB program to identify lesion regions based on ADC values automatically. Regions with ADC values lower than 750 mm2/s and regions with ADC values higher than 750 and less than 1500 mm2/s were considered CTV70Gy (clinical tumor volume with 70 Gy prescribed dose), and CTV60Gy, respectively. Other regions of the prostate were considered as CTV53Gy. New plan evaluation indices based on evaluating the homogeneity (IOE(H)), and conformity (IOE(C)) were introduced, considering the relative volume of each lesion and cellular density obtained from ADC images. These indices were compared with conventional homogeneity and conformity indices and IOEs without considering cellular density. Furthermore, tumor control probability (TCP) was calculated for each patient, and the relationship of the assessed indices were evaluated with TCP values. Results IOE (H) and IOE (C) with considering cellular density had significantly lower values compared to conventional indices and IOEs without considering cellular density. (P < 0.05). TCP values had a stronger relationship with IOE(H) considering cell density (R2 = -0.415), and IOE(C) without considering cell density (R2 = 0.624). Conclusion IOE plan evaluation indices proposed in this study can be used for evaluating prostate IMRT dose painting plans. We suggested to consider cell densities in the IOE(H) calculation formula and it’s appropriate to calculate IOE(C) without considering cell density values. Supplementary Information The online version contains supplementary material available at 10.1186/s13014-022-02163-7.
One of the strategies to improve the efficiency of radiotherapy is increasing the prescribed dose in regions with higher cancer cell densities [8]. Higher doses ranging from 74 to 80 Gy has been reported to have improvements in the outcome of prostate cancer treatment [4][5][6][7][9][10][11][12][13][14]. However, delivering these high doses is impossible using conventional radiotherapy without a significant probability of occurring severe radiation toxicities [14]. Modulated techniques (IMRT, VMAT, Tomotherapy) can reduce these toxicities by optimizing radiation conformation [15,16].
The usual clinical protocol in prostate radiotherapy is to deliver a uniform dose to a defined planning target volume (PTV) [17]. Dose painting (DP) was introduced to increase the tumor local control rates by delivering higher dose levels to the regions with higher cellular densities or radioresistance tissues while sparing healthy tissue. It is designed to give additional doses to subvolumes with high radioresistance due to hypoxia or other reasons as quantified by functional imaging [18]. Higher dose levels in tumor nodules or dominant intraprostatic lesions can improve local control without increasing complication rates [4-7, 14, 15].
The application of DP radiation therapy is increased in recent years. However, the aspects of plan evaluation remain controversial now [30]. For example, conventional indices, such as the conformity index (CI) and the homogeneity index (HI), commonly used in routine clinical practice for plan evaluation, are formulated based on the paradigm of uniform dose prescription. Therefore, these indices need to be modified for DP planning. There are few studies that modified these indices for use in DP plan evaluation [31][32][33]. In a study by Park et al. [31], they introduced a new plan quality index, named "index of achievement (IOA)". Their introduced index assesses how close the planned dose distribution is to the prescribed one considering the differences between the prescribed and the delivered dose for each voxel multiplied by the relative volume of the voxel in the target volume. Their index did not account for the importance of different lesions with different cellular densities or radioresistance properties and also need to be calculated by a computer program. We think that the plan evaluation index in DP planning must be easy to calculate without a computer program and accounts importance of different lesions inside target volume. Therefore, in this study we tried to propose new dose painting plan evaluation indices with simpler calculation methods incorporating cellular density as an index of lesion importance obtained from MRI diffusion images for prostate cancer IMRT.
Methods
This single-center retrospective study was performed in accordance with the national ethical guidelines and regulations. The national Ethics Committee has approved the methods of this study. MRI and CT images of patients were used in this study without any intervention in the diagnostic or treatment procedures. In addition, gathering the informed consent was waived because of the retrospective nature of the study.
Imaging data (CT and MRI scans) of 20 male patients having local prostate cancer who had no previous surgery, hormone therapy (AST or ADT), and prostate radiation therapy with at least one non-high risk intraprostatic lesion (IL) in stages of T1 to T3a, were used in this study. Patients' ages ranged from 54 to 85 years, with a mean age of 69.4.
The CT (Matrix size: 512*512; Slice thickness: 3 to 5 mm), T2w-MRI (fast spin echo pulse sequence with TE: 80 ms and TR: 7800 ms), diffusion-weighted MRI (echo planar imaging with TE: 88 ms, and TR: 4600 ms) and apparent diffusion coefficient (fast spin echo pulse sequence with TE: 100 ms and TR: 3000 ms) images were taken using a Siemens 16-slice Emotion CT and a 1.5 Tesla Avanto MRI machine (Siemens Healthcare GmbH, Germany). The patients were placed in supine positions for both imaging procedures. Diffusion-weighted images (DWI) were gathered with three signals per image with a scattering-sensitive gradient in three orthogonal planes and b-values of 0, 250, 500, and 1000 s per square. DW-MRI images have a resolution of 1.64 × 1.64 × 3 mm and a FOV of 210 × 210 mm, a matrix size of 128 × 128 pixels, and a NEX (number of excitation) parameter equal to four. Apparent diffusion coefficient (ADC) maps (images) were automatically calculated from DW-MRI images. ADC is a measure of the magnitude of water molecules diffusion within tissue [34], and can show the cellular density in some tumors like prostate [35].
CT and MRI images were combined using a rigid registration algorithm in the treatment planning software (TPS) based on bone landmarks, gold markers implanted in the prostate and skin surfaces, and then verified by a specialized physician. The MRI and CT registration were used to contour the lesion volumes inside the patients' prostate and radiation-sensitive organs. It also allows for more precise target volume delineation in prostate cancer patients [36].
An in-house MATLAB program was developed to automatically identify lesion regions on ADC images based on ADC values. The MATLAB code is available in the "Additional file 1" section. Two types of predominant lesions were considered in the prostate, one related to lesions with ADC values lower than 750 mm 2 /s (PTV-1), and the other was related to lesions with ADC values higher than 750 and less than 1500 mm 2 /s (PTV-2). The upper limit of ADC values in tumor tissues varies in different studies but usually was considered more than 1300 mm 2 /s [37]. In this study, the apparent diffusion coefficient threshold for distinguishing tumor tissue from normal prostate tissue was 1500 mm 2 /s. The highest measurable value of ADC in MATLAB software was 5000 mm 2 /s. The ADC images were imported to the MATLAB program, and the voxels in a specific range of ADC values were determined. The related voxels for determining areas must at least have a minimum number (400 voxels) located next to each other so that the software can identify those areas separately. Considering the relationship between Gleason score (GS) and ADC cut-off value based on a study by Pepe et al. [38], different target volumes were identified within the prostate. Each patient's output DICOM RS file was then transferred to the treatment planning system to contour these new structures (prostatic lesions) on the CT images. Contouring of other organs at risk (OARs) was performed by an experienced radiation oncologist in the treatment planning system.
Prescription dose levels, except in dominant intraprostatic lesions (DILs), were taken from the Jereczek-Fossa et al. study [39]. The upper limit for the prescribed dose for DIL was considered 70 Gy in 27 fractions for high risk DILs. This hypofractionated dose escalation was used and evaluated in many studies [40][41][42][43][44][45][46][47]. According to a study by Onjukka et al. [48], this hypofractionated dose was equivalent to 86 Gy in 37 sessions (used in the study of Uzan et al. [49]. A prescribed dose for DILs with lower risks was considered 66 Gy. The clinical target volume of the base of the seminal vesicles was considered CTV 53Gy . The planning target volume for this target (PTV 53Gy ) was formed by adding eight millimeter isotropic margins to CTV 53Gy in order to account for patient and equipment placement errors. The whole prostate volume (except the DILs) was considered CTV60; similarly, PTV 60Gy was formed by adding 5 mm margins to the CTV 60Gy . The margin was reduced to zero in the posterior region, where the target volume overlaps with the rectum. Two millimeters margins were added to CTV 66Gy and CTV 70Gy to create PTV 66Gy and PTV 70Gy without extending beyond CTV 60Gy or overlapping with the rectum, bladder, and urethra due to the uncertainty in defining the DILs [49]. Furthermore, planning at risk volumes (PRVs) were created for high-risk organs, including the rectum, urethra, and bladder, with margins of two millimeters. The dose escalated DIL regions with the whole prostate were presented in Fig. 1. Furthermore, the procedure used to automatically contour the intraprostatic DILs is illustrated in Figs. 1-a and 1-b.
The IMRT plans were designed with Eclipse software (version 11, Varian Corporation, USA) for each patient. IMRT plan with nine coplanar fields in gantry angles of 0, 30, 60, 105, 140, 220, 260, 300, and 330 was designed to irradiate PTVs with prescribed doses. All the plans were interactively optimized based on our institutional planning protocol derived from a previous study by Pollak et al. [50]. The optimization algorithm was Dose Volume Optimizer (DVO) which is enclosed in Eclipse treatment planning software and clinically approved by previous studies [51]. The planning optimization objectives are presented in Table 1. An experienced physicist evaluated all the treatment plans to ensure compliance with reported dose constraints [52].
After treatment planning optimization, final dose calculations were performed by the anisotropic analytical algorithm (AAA) in the Eclipse software. The calculation accuracy of this algorithm was previously approved in several studies [53][54][55]. Dose volume histograms (DVHs) of the CTVs for each patient were entered in BioSuite software [56]. The tumor control probability (TCP) values were calculated using the Poisson model [57], based on radiobiological model parameters proposed by Deb and Fielding [58].
We introduced two indices of effectiveness (IOE) for evaluating IMRT dose painting plan dose distribution. One IOE can evaluate the conformity of CTVs, IOE(C), and another IOE can assess the overall dose distribution homogeneity of target volumes, IOE(H). The previous equation proposed by Park et al. [31] has relative volume coefficients, and these coefficients were included in our IOE equations accounting for the effect of each target (DIL) on the overall value of IOE. Furthermore, cell density values obtained from ADC maps were used in the IOE equations. The cell density is a measure of the clonogenicity level for each of the tumor volumes. The equations of IOE(H) and IOE(C) are as follows: Fig. 1 The procedure used to automatically contour the intraprostatic dose escalated DILs. a Delineating the whole prostate manually on ADC images. b Automatic contouring of different intraprostatic lesions based on ADC values. c Importing the contours from ADC images on registered CT images. d Determining DILS and margins to create different DILs with different dose levels on CT images. e Schematic of intraprostatic DILs and their PTVs with different dose levels HI 1 , HI 2 , and HI 3 are the homogeneity indices for CTV 70Gy , CTV 66Gy , and CTV 60Gy respectively. Similarly, the CI 1 , CI 2 , and CI 3 are the conformity indices for CTV 70Gy , CTV 66Gy , and CTV 60Gy respectively. V 1 , V 2 , and V 3 are the volumes, and CD 1 , CD 2 , and CD 3 are the cell density of these CTVs. In addition, V T and CD P are the total volume and mean cell density value of the whole CTV (sum of all CTVs).
Furthermore, the IOE(C) and IOE(H) were calculated without considering the cell density, and they were compared with IOE indices considering cell density (our proposed indices) and also the mean of the conventional HI and CI values. Equations related to IOE indices without considering cell densities and the mean of conventional indices are presented in Eqs. 3-6. (1)
Statistical analysis
Kolmogorov-Smirnov (K-S) test was used to evaluate the variables' normality distribution. The results showed that the distributions of the assessed parameters for 20 patients studied in this study were not normal. Therefore, the Wilcoxon statistical analysis was used to evaluate the differences in the relevant indices with considering cell density, without considering cell density and the mean of conventional indices. The significance level in these tests was considered equal to 5%, and the P-values less than 0.05 were considered significant differences. Spearman test was also used to investigate the relationship between IOE indices and radiological parameter (TCP of different targets) values. All of the statistical tests were performed in the SPSS software package Version 22 (SPSS Inc., Chicago, IL, USA).
(3) Mean and standard deviation values of IOE (C) with and without considering cell density were 0.392 ± 0.124, and 0.993 ± 0.004, respectively (Fig. 3). Furthermore, the mean and standard deviation of conventional CI was 0.994 ± 0.004. The IOE (C) with considering cell density had a significantly lower value compared to conventional CI and IOE (C) without considering cell density (P-value ≤ 0.02).
Correlations between the IOE and tumor control probability values
The mean ± standard deviation values of tumor control probability for CTV 70Gy was 94.72 ± 0.65. These values for CTV 66Gy , and CTV 60Gy were 91.58 ± 1.34 and 77.48 ± 3.28 respectively. TCP values greater than 0.7 (70%) have been reported to be "appropriate for total tumor control" in a study by Casares et al. [59].
The Spearman non-parametric correlation was used to investigate the correlation between tumor control probability (TCP), the IOE indices introduced in this study, and conventional HI and CI. The Correlation coefficients between TCP and IOE indices with and without considering cell density were presented in Table 2. Furthermore, the correlation between the TCP and conventional HI and CI values are provided in this table.
The correlation coefficient between TCP and IOE (HI) with considering cell density showed a moderate and negative correlation. IOE (HI) without cell density coefficient was not correlated with tumor control probability. In addition, the correlation coefficient between TCP and IOE (CI) with considering cell density has a moderate and negative correlation. In contrast, the IOE (CI) without cell density coefficient had a strong and positive correlation with TCP. The correlation of TCP with conventional HI mean and CI means were not significant.
Discussion
Dose painting radiotherapy is a technique that can produce more targeted dose delivery to tumor-rich regions while saving organs at risk and critical normal tissues. Dose painting planning would be more complicated due to different levels of the prescribed dose levels and harder to evaluate with conventional plan quality indices considering uniform dose prescription. Therefore we tried to introduce new indices for evaluating the conformity and homogeneity based on the tumoral cell density and relative volumes of each lesion in prostate IMRT.
There is a recent study [31] tried to introduce new indices to evaluate the plan quality of inhomogeneous irradiated targets and indicate "achievement" in DP plans based on an introduced IOA (index of achievement) as an alternative to the conventional homogeneity index [60]. Their introduced indices may not necessarily be correlated with the biological effect. The proposed indices in this study can be easily modified to incorporate such an effect. We incorporated the tumor cell density obtained from the DWI and ADC images in the plan evaluation indices, and therefore our introduced indices contain the biological effects. However, the cell density can be obtained from other imaging modalities such as "positron emission tomography" and can be evaluated in future studies.
Applying single or just several indices in plan evaluation is an easy and clinically acceptable method. However, using one or several indexes for plan evaluation can suffer from lacking detailed information. Therefore, it must be mentioned that the introduced indices can not replace the standard tools, such as isodose lines and DVH curves evaluation for treatment plan assessment. Although, they can present additional information. There is an alternative to the standard DVH, delta-volume histogram (ΔVH), which was introduced by Witte et al. [30]. They mainly addressed cumulative ΔVH. But in another study by Park et al. [31], an IOA was introduced, which could be supported by differential ΔVH (dΔVH). It was assumed that each target voxel has an equal amount of impact on the calculation of plan evaluation metrics. However, the impact of each voxel can be different. For example, under-dose regions in a PTV with a higher prescription dose may have clinically higher risks compared to the under-dose regions in a lower dose PTV. This issue can be resolved by adding voxel-specific or region-specific weighting factors to the previous plan evaluation indices. We used cell density obtained from the ADC map for different PTVs as region-specific weighting factors. We observed that the IOE values based on cellular densities could be completely different when applying the weighting factor. If the biological importance of hotness and coldness becomes much more apparent, a more accurate weighting factor system can be found in the future.
We evaluate the relationship of our proposed indices with TCP. Analytical radiobiological parameters (such as TCP and normal tissue complication probability (NTCP)) have been widely used for evaluating the quality of treatment plans [61][62][63]. In particular, several studies proposed TCP models for inhomogeneously irradiated tumors or planning target volumes [18,64]. Our results showed that IOE(H) with considering cell density and IOE(C) without considering cell density had a stronger relationship with TCP. Therefore it may be concluded that considering cell density values in calculating IOE(C) was not an appropriate idea. However, cell densities must be included in the IOE(H) calculation formula.
The cell density values used in our proposed formula of IOE indices were calculated based on the ADC map. Furthermore, the dose escalation was also based on the regions extracted based on these ADC images. A number of previous studies have advocated the strategy of dose escalation to the imaging-defined targets and dose de-escalation to the rest of the prostate. High dose areas in DP plans can increase the delivery uncertainty due to the limited capability of the treatment planning optimization algorithm to deliver high doses to small and isolated areas. Therefore, these high-dose areas must be defined accurately, and the method of determining these areas must have high repeatability. Comparable findings on the repeatability of ADC features in MRI prostate imaging are reported in the literature. Toivonen et al. [35] reported an ICC of 0.89 for ADC intensity in prostate cancer using MRI, although performed on an ROI basis. Koh et al. also reported high repeatability for ADC measurements in a two-center phase I clinical trial [65].
Van Lin et al. [66] performed a dose panting planning study on five patients with standard whole-prostate RT conventional plan to 78 Gy and a plan with DIL dose escalation to 90 Gy based on dynamic contrast-enhanced and 1H-spectroscopic MRI, and the remainder of the prostate dose de-escalation to 70 Gy. They reported that both plans had similar TCPs; however, the dose painting had lower NTCPs. In another study by Seppala et al. [21], a planning study of 12 patients was performed with DILs defined based on 11 C acetate PET scans. Six different dose escalation plans were performed and compared for each patient, including a whole-prostate RT plan to 77.9 Gy, and DIL dose escalations to 77.9 Gy, 81 Gy, 84 Gy, 87 Gy and 90 Gy, with remaining prostate dose deescalations to 72 Gy. They reported that the dose painting plans had higher TCP values compared with the standard whole-prostate plan and that the highest probability of tumor control without complication was related to a plan with an average dose of 82.1 Gy to the DIL. In a study by Chang et al. [23] the technical feasibility of IMRT dose painting using 11 C-choline PET scans were evaluated in eight patients with localized prostate cancer. Two DILs were defined including 60% and 70% of the maximum standardized uptake values (SUV 60% and SUV 70% ). Three IMRT plans were designed including: PLAN 78 (wholeprostate irradiation with 78 Gy); PLAN 78-90 (whole-prostate RT to 78 Gy, a boost to the SUV 60% and SUV 70% to 84 Gy, and 90 Gy, respectively); and PLAN 72-90 (wholeprostate RT to 72 Gy, a boost to the SUV 60% and SUV 70% to 84 Gy, and 90 Gy, respectively). TCP based on PET scan-defined volumes (TCP PET ) and on prostatectomydefined volumes (TCP path ), and rectal NTCP were compared between the plans. They reported that both dose painting plans (PLAN 78-90 and PLAN 72-90 ) had significantly higher TCP PET and TCP path values than conventional IMRT plan (PLAN 78 ), without significant differences in TCP PET or TCP path between dose painting plans. Furthermore, There were no significant differences in rectal NTCPs between the 3 plans.
We used rigid image registration for fusing CT and MRI images. Deformable registration can also be used for this purpose. Rigid registration is very effective in cases when no anatomic change is expected [67]. In our study, patients underwent MR imaging after CT imaging in one day with a maximum delay of one hour. Their positioning was similar in both imaging and therefore, we don't expect significant anatomical changes between the imaging techniques. If there is a big time gap between CT and MR imaging, patients might experience anatomical changes due to tumor shrinkage/growth, weight loss, or physiological organ shape variations. In these cases deformable registration can manage the distortion between two image sets and provide superior results [68][69][70]. In comparison to rigid registration, deformable registration has a significantly greater degrees of freedom [67], and can deform the image and structures with different algorithms (such as intensity-based approaches, landmark-based thin-plate spline, or biophysical and finite element modelling-based registration) (67).
This study has some limitations, and several factors must be addressed before clinically adopting this strategy. First, deformable registration might be superior in cases with the significant time intervals between CT and MR imaging because this could deal with changes in the prostate shape and discrepancies in the prostate size between imaging modalities more adequately than was possible using rigid registration. Second, the proposed plan evaluation indices in this study, including IOE(H) and IOE(c) with and without considering cell densities, must be assessed for a bigger group of patients and also in other cancer sites.
Conclusions
New IOE dose painting plan evaluation indices proposed in this study have simple calculation methods and incorporate cellular density as an index of lesion importance obtained from MRI ADC images for prostate cancer IMRT. These indices can be used for evaluating prostate IMRT dose painting plans. Cell densities must be considered in the IOE(H) (calculation formula, and it's more appropriate to calculate IOE(C) without considering cell density.
|
v2
|
2022-11-24T15:00:57.122Z
|
2022-11-23T00:00:00.000Z
|
253805167
|
s2orc/train
|
Whole-exome sequencing of BRCA-negative breast cancer patients and case–control analyses identify variants associated with breast cancer susceptibility
Background For the majority of individuals with early-onset or familial breast cancer referred for genetic testing, the genetic basis of their familial breast cancer remains unexplained. To identify novel germline variants associated with breast cancer predisposition, whole-exome sequencing (WES) was performed. Methods WES on 290 BRCA1/BRCA2-negative Singaporeans with early-onset breast cancer and/or a family history of breast cancer was done. Case–control analysis against the East-Asian subpopulation (EAS) from the Genome Aggregation Database (gnomAD) identified variants enriched in cases, which were further selected by occurrence in cancer gene databases. Variants were further evaluated in repeated case–control analyses using a second case cohort from the database of Genotypes and Phenotypes (dbGaP) comprising 466 early-onset breast cancer patients from the United States, and a Singapore SG10K_Health control cohort. Results Forty-nine breast cancer-associated germline pathogenic variants in 37 genes were identified in Singapore cases versus gnomAD (EAS). Compared against SG10K_Health controls, 13 of 49 variants remain significantly enriched (False Discovery Rate (FDR)-adjusted p < 0.05). Comparing these 49 variants in dbGaP cases against gnomAD (EAS) and SG10K_Health controls revealed 23 concordant variants that were significantly enriched (FDR-adjusted p < 0.05). Fourteen variants were consistently enriched in breast cancer cases across all comparisons (FDR-adjusted p < 0.05). Seven variants in GPRIN2, NRG1, MYO5A, CLIP1, CUX1, GNAS and MGA were confirmed by Sanger sequencing. Conclusions In conclusion, we have identified pathogenic variants in genes associated with breast cancer predisposition. Importantly, many of these variants were significant in a second case cohort from dbGaP, suggesting that the strategy of using case–control analysis to select variants could potentially be utilized for identifying variants associated with cancer susceptibility. Supplementary Information The online version contains supplementary material available at 10.1186/s40246-022-00435-7.
Introduction
Breast cancer (BC) is the most common malignancy and the leading cause of cancer-associated mortality among women worldwide [1]. It accounts for one in four cancer cases among women and one in six cancer deaths, ranking first in the vast majority of countries for incidence [1]. Approximately, 10-20% of all BC patients have a family history of cancer with multiple family members affected across generations [2]. Germline mutations in specific genes such as BRCA1, BRCA2, CDH1, PALB2, PTEN and TP53 confer an increased risk of developing BC [3].
Recent advances in next-generation sequencing have led to reduced costs for multigene panel testing of cancer predisposition genes for individuals referred for genetic testing, resulting in a higher uptake of testing. However, it is estimated that pathogenic variants in known cancer predisposition genes only account for around 25% of hereditary BC cases [4,5].
Whole-exome sequencing (WES) is revolutionizing our ability to identify novel genetic variants associated with cancer predisposition. To date, multiple candidate BC predisposition genes have been identified by WES, predominantly from studies on women of European ancestry [6,7].
Here, we aimed to identify novel candidate BC predisposition genes and variants by performing WES on germline DNA from Asian BC patients referred for cancer genetic risk assessment but who were BRCA1/2-negative. Pathogenic variants identified from WES were filtered and prioritized using in silico bioinformatic tools, followed by case-control analysis and only significant variants in known cancer genes were selected for further analysis. Notably, we have identified pathogenic variants in our cases that had a statistically significant difference in frequency as compared to the Genome Aggregation Database (gnomAD) East-Asian (EAS) controls and Singaporean controls [8].
Demographics and clinical information on the study population
Information on the demographics, age at diagnosis, ethnicity, family history, and clinicopathological characteristics of the 290 BC cases are provided in Table 1. The study population consisted of only females, and a large proportion were Chinese (69.3%). The age of first cancer diagnosis ranged from 19 to 75 years, with a mean and median age of 37.5 and 37 years, respectively. Of 290 patients, 65 (22.4%) presented with a family history (including first-degree, second-degree, and third-degree relatives) of BC, 23 (7.9%) with a family history of other cancers and 218 (75.2%) with no family history of breast or any other cancers (Table 1, Additional file 1: Fig. S1). Of the 290 BC cases, 225 patients (77.6%) had early-onset breast cancer (≤ 40 years).
Filtering of candidate variants
Whole exome sequencing of 290 BC patients revealed 1,196,466 variants before filtering. Among these, 1,101,796 (92.1%) passed Dynamic Read Analysis for GENomics (DRAGEN) quality-control checks. Further filtering to retain functional variants with gnomAD (EAS) minor allele frequency (MAF) less than 1%, predicted pathogenic variants with scaled Combined Annotation-Dependent Depletion (CADD) score greater than 20, and variants in the known or predicted cancer gene lists in the Network of Cancer Genes (NCG) database, left only 2,496 variants (0.2% of the total; Fig. 1).
The genes of our shortlisted variants were further prioritized using cancer genes databases such as Catalogue of Somatic Mutations in Cancer (COSMIC), cancer driver genes based on nucleotide context, and computationally discovered and experimentally validated cancer driver genes [9] (Additional file 4: Table S1). Finally, we shortlist only variants that were present in three or more patients. All variants were checked with IGV (Additional files 1, 2: Figs. S1, S2).
All 42 nonsynonymous SNVs had CADD scores greater than 20. The remaining 7 variants which were not nonsynonymous SNVs also had CADD scores greater than 20, except for a frameshift deletion variant in HLA-A. Thirty variants were classified as variants of uncertain significance (VUS) (61.2%), two stop-gain mutations in KMT2C were considered pathogenic (4.1%), and the remaining variants were benign (14 variants, or 28.6%) or likely benign (3 variants, or 6.1%) ( Table 2).
Case-control analysis of the Singapore cases
Case-control analysis was performed for 49 selected variants for our Singaporean cases against the gnomAD (EAS) and SG10K_Health control cohorts (Table 3). Apart from the two variants in BRD7 and NBEA that were not reported in gnomAD (EAS), all of our remaining 47 variants were significantly enriched in our cohort as compared to gnomAD (EAS). In the SG10K_ Health control cohort, seven of our 49 selected variants were absent, including the aforementioned variants in BRD7 and NBEA; and additional variants in KMT2C, GPRIN2, H3F3A, and MAF. Of the remaining 42 variants which could be found in SG10K_Health, 13 were significantly enriched at α = 0.05 in our cohort versus SG10K_Health (Table 3).
Case-control analysis using a breast cancer case cohort from dbGaP
Case-control analysis for the 49 germline variants identified from our Singapore breast cancer cohort was repeated using a case cohort from dbGaP (phs000822. v1.p1) against the same control cohorts ( (Table 3). Altogether, 14 variants were significantly enriched in cases, or missing in the control cohorts, across all four sets of case-control comparisons. These variants were found in 89 out of 290 breast cancer patients (30.7%) where 24 of the 89 cases had more than one pathogenic variant (Additional file 4: Table S3).
Discussion
Here, we report the largest WES study on germline DNA from Asian breast cancer patients who had undergone cancer risk assessment and were BRCA1 and BRCA2 mutation-negative. The approach that was taken was to select only pathogenic variants that showed a statistically significant difference against gnomAD East-Asian controls and Singapore controls. This was followed by an additional prioritization step of selecting only variants occurring in well documented cancer genes such as those listed in COSMIC, NCG and cancer driver gene databases [9][10][11].
In total, we have identified 49 rare pathogenic germline variants in 37 genes which were significantly enriched in breast cancer patients. These were all predicted to be pathogenic using in silico tools and all had a minor allele frequency of less than 1% or were unreported in gnomAD (EAS). We further validated these results with an independent United States-based case cohort obtained from dbGaP, of 466 early-onset breast cancer patients. Across four sets of comparisons involving two case and two control cohorts, 14 variants were consistently enriched in breast cancer cases (Table 3).
Of these 14 variants, seven variants in GPRIN2, NRG1, MYO5A, CLIP1, CUX1, GNAS, and MGA were confirmed by Sanger sequencing. To the best of our knowledge, these specific germline variants identified here have not been reported in any cancer-related studies thus far. However, their respective gene functions have been implicated in many cancer types [12][13][14][15][16][17]. The NRG1 nonsynonymous SNV (rs113317778) lies in an immunoglobulin-like domain, while other affected residues in GPRIN2 (rs4445576), CUX1 (rs782176246), GNAS (rs563844600), and MGA (rs61736074) are located within a protein disordered region, where it lacks a stable tertiary structure and adopts different structural conformations [18][19][20]. Interestingly, a computational study has predicted the mutation in GPRIN2 (p.S328C) to generate new microstructural elements in the disordered region and may disrupt protein functions or protein-protein interactions [20]. Other exome sequencing studies have also identified a damaging germline mutation in GPRIN2 (p.A233S) in Iranian patients with familial esophageal squamous cell carcinoma (ESCC) [21] as well as somatic mutations in melanoma samples [22].
Additionally, a frameshift deletion variant in TPTE2 (c.483delT) and two nonsynonymous SNVs in NBEA (c.C2317A) and BRD7 (c.A44C) could not be confirmed by Sanger sequencing. NBEA has segmental duplications on chr15, while BRD7 is mapped to segmentally duplicated regions on chr3 and chr6. Furthermore, the TPTE2 variant is within a short 8-nucleotides homopolymer, and it has two segmental duplications on chrY and chr21 [23]. Due to high sequence similarities, sequenced reads which arise from segmental duplications may be wrongly aligned and result in false-positive variant calls. Seven nonsynonymous SNVs in RNF43, HLA-B, ERBB3, NTRK1, TET2, and DCC identified here, have previously been implicated in various cancer types Additional file 4: Table S4. For example, the HLA-B c.A161G variant, which was detected in 9 patients (3.1%) here, was also found to be associated with high-grade cervical preinvasive lesions and invasive cervical cancer in a recent genome-wide association study [24]. A different study reported that the ERBB3 c.A3355T variant was significantly associated with poor survival in ER-positive cases [25]. Nonetheless, none of these variants were significantly enriched in our case-control analyses.
Of our 49 variants, 4.1% (2/49) were classified as pathogenic and 61.2% (30/49) as VUS by InterVar, respectively. This high VUS rate is consistent with our previous study and that of others on Asian populations [26,27]. In a large US study on germline genetic testing, Asian patients had approximately two-fold more VUS compared to non-Hispanic White patients, at a VUS rate above 40% [27]. These substantially higher VUS rates in Asians may reflect the underlying lack of variant data from Asian control populations available for variant reclassification.
Besides the variants identified in this current study, WES has been performed to detect candidate variants in BRCA -negative patients from other populations. In a study on 7 families from France, Italy, Netherlands, Australia and Spain, investigators found 12 variants in genes involved in DNA repair, cell proliferation and survival, or cell cycle regulation [28]. Sequencing of 52 individuals from 17 Greek families with HBOC and further validation in additional cohorts from Canada, TCGA and the UK Biobank, led to the prioritization of missense variants in the SETBP1 and c7orf34 genes [29]. In another European study, 54 BRCA -negative families from Belgium underwent WES and 44% harbored variants in known cancer predisposition genes. In particular, it was observed that nonsense variants in cancer-associated genes involved in DNA repair were enriched in breast cancer patients as compared to controls [30]. From 113 families from Tunisia, eight BRCA -negative unrelated patients were selected for WES. Of 24 genes that were prioritized from WES data, five were selected based on their significant association with survival, as determined from analysis using TCGA data [31]. Notably, the strategies for the prioritization and filtering of genes/variants differ between studies with differing variants identified. It is possible that these variants could be population-specific or low penetrance variants.
Our study has limitations. We had used an independent breast cancer cohort of US patients with early-onset breast cancer [35 years or younger] from dbGaP to validate the frequency of the 49 variants discovered in our cohort that were found to be associated with breast cancer. However, 17 of the 49 variants were not present in this dbGaP case cohort, possibly due to differences in genetic ancestry between the populations. Hence, further studies in additional Asian as well as European populations are necessary to validate the variants described in this current study. Secondly, DNA samples from family members of our cases were not available for segregation analysis. Thirdly, due to limited access to the SG10K_Health cohort, we had used the gnomAD (EAS) population for variant filtering. The gnomAD (EAS) cohort is comprised of individuals of Korean, Japanese and Chinese descent, whereas our study population were South-East Asians, mainly of Chinese, Malay and Indian ethnicity. Nonetheless, the gnomAD (EAS) was the most suitable publicly available control population available, and thus was selected.
Conclusions
In summary, the current study has identified 49 pathogenic variants in 37 genes associated with breast cancer predisposition, many of which have not been previously documented. Our study provides new insights into the genetic susceptibility to BC, and it is imperative that further studies in additional populations of diverse ethnic background be undertaken to determine the frequency of these variants, and to confirm their association with BC risk.
Study participants
Two hundred and ninety breast cancer patients who fulfilled one or more of the following criteria were selected for WES: 1. having a family history of breast cancer in first-and/or second-degree relatives; 2. having bilateral breast cancer; and, 3. having early-onset breast cancer at the age of 40 years or below (Additional file 1: Fig. S1) [26]. Written informed consent was obtained from all participants and the study was approved by the Sin-gHealth Centralised Institutional Review Board (CIRB Ref: 2018/2147).
Whole-exome sequencing
Genomic DNA was isolated from peripheral blood samples, collected from breast cancer patients as described previously [32,33]. Samples for sequencing and libraries were prepared according to Agilent SureSelect Human All Exon V6 kit (Agilent Technologies, CA, USA) and the library preparation and enrichment were carried out according to Agilent SureSelect protocols. Enriched samples with paired-end sequencing (2X150 bp) were performed on the Illumina NovaSeq 6000 platform. Variants were aligned and called with Illumina DRAGEN version 3.5.7 on the BaseSpace Sequence Hub cloud platform [34], with median 80 × coverage per base.
Prioritization and filtering of variants
The variants were annotated for their transcript effects, CADD v1.3 scaled score [35], and gnomAD minor allele frequencies using ANNOVAR [36]. CADD v1.3 indel scores were filled in manually using the CADD web server. The American College of Medical Genetics and the Association of Molecular Pathology (ACMG-AMP) classifications were obtained using InterVar [37]. We removed variants which did not pass DRAGEN's default quality control checks, variants with gnomAD (EAS) MAF greater than 1%, and variants found in only two or fewer patients. Frameshift indels, stop-gains; and nonsynonymous SNVs with scaled CADD v1.3 score greater than 20 were chosen for further analysis. A CADD score of 20 and above represents the top 1% of pathogenic variants as scored by CADD.
Prioritization of candidate genes
From the genes of our prioritized variants, we selected only known or candidate cancer genes as listed by the NCG [9]. These genes were then further curated for those that were strongly implicated in cancer in at least one other cancer gene database: the COSMIC database [10], cancer driver genes based on nucleotide context [11], and computationally discovered and experimentally validated cancer driver genes [38] (Additional file 4: Table S1).
Manual checking with IGV
All prioritized variants were manually checked with Integrative Genomics Viewer (IGV) [39], except those in highly repetitive regions in MUC4 or KMT2C, or highly polymorphic genes HLA-A or HLA-DRB1, as their alignments were too complex (Additional file 2: Fig. S2). Variants suspected to be false positives were excluded (Additional file 3: Fig. S3).
Case-control analysis
Case-control analysis for the variants was performed for two breast cancer cohorts (cases described in this study and the phs000822.v1.p1 dataset from dbGaP) and two control cohorts (gnomAD (EAS) and SG10K_Health). The dataset from dbGaP is a breast cancer dataset of 466 patients with early-onset breast cancer (diagnosed on or before the age of 35) from the United States of America. The gnomAD (EAS) cohort (gnomAD v2.1.1) comprises 9,977 individuals of East Asian descent while the SG10K_ Health cohort consists of whole genomes from 9,770 healthy Chinese, Indian, and Malay volunteers from Singapore [8].
Polymerase chain reaction and Sanger sequencing
Variants that were significant by case-control analysis were validated by polymerase chain reaction (PCR) and Sanger sequencing. PCR primer sets were designed using Primer-BLAST [40]. DNA amplification by PCR was performed using HotStartTaq (Qiagen, Venlo, Netherlands) or Q5 High-Fidelity (New England Biolabs, Ipswich, MA, USA) DNA polymerase, as described in the manufacturer's protocol. Primer sequences and their respective cycling conditions are listed in Additional file 4: Table S5. The PCR products were then analyzed by 2% agarose gel electrophoresis and purified with ExoSAP-IT Express (Thermo Scientific, USA) prior to sequencing. Cycle sequencing reactions were performed using BigDye Terminator v3.1 kit (Applied Biosystems, Foster City, CA) and the sequencing products were analyzed on a Genetic Analyzer. DNA sequences were visualized and aligned using Geneious Prime version 2022.1.
Statistical analysis
For case-control analyses, a two-sided Fisher's exact test was used and p values were adjusted for multiple testing using the Benjamini-Hochberg method [41].
the frameshift deletions H HLA-A NM_001242758.1:c.268delA and I HLA-DRB1 NM_002124.3:c.118_122del are not associated with any obvious gaps in read alignments; nor is the frameshift insertion J HLA-DRB1 NM_002124.3:c.126_127insTTA AGT TT represented by insertions in its read alignments.
Additional file 3: Fig. S3. Representative IGV screenshots of alignments supporting two likely-false positive frameshift insertions. Panel A shows the alignment for PABPC1 NM_002568.4:c.1336_1337insACC TCA TC and B for CIC NM_015125.4:c.4778_4779insGG. Red boxes indicate where the insertion would have been expected to appear, red arrows point to the soft-clipped alignments which support the existence these frameshift insertions. C Reads supporting the PABPC1 insertion map partially to both PABPC1 and PABPC3 (reverse complement) genes on reference genome loci NC_000008. 10
|
v2
|
2022-11-24T15:00:57.866Z
|
2022-11-23T00:00:00.000Z
|
253805804
|
s2orc/train
|
KTN1-AS1, a SOX2-mediated lncRNA, activates epithelial–mesenchymal transition process in esophageal squamous cell carcinoma
Kinectin 1 antisense RNA 1 (KTN1-AS1), a long non-coding RNA (lncRNA), has been proved to have tumor-promoting properties and its expression is enhanced in several human tumors. However, the role of KTN1-AS1 in the pathogenesis of esophageal squamous cell carcinoma (ESCC) remains unknown. This study aimed to investigate the expression status, functional roles, and molecular mechanisms of KTN1-AS1 in the development of ESCC. Considerable upregulation of KTN1-AS1 was confirmed in esophageal cancer cells and ESCC tissues and its expression was associated with TNM stage, pathological differentiation, and lymph node metastasis. SOX2 directly activated transcription of KTN1-AS1, and overexpression of KTN1-AS1 facilitated ESCC cells proliferation and invasion in vitro and in vivo. Furthermore, KTN1-AS1 could bind to retinoblastoma binding protein 4 (RBBP4) in the nucleus and enhanced its binding with histone deacetylase 1 (HDAC1), thereby activating the epithelial–mesenchymal transition (EMT) process through downregulating E-cadherin expression at the epigenetic level. In conclusion, KTN1-AS1, induced by SOX2, acts as a tumor-promoting gene and may serve as a potential therapeutic and prognostic biomarker for ESCC.
SOX2 induces KTN1-AS1 expression in ESCC cells. Given the high expression of KTN1-AS1 in ESCC, the mechanism leading to its upregulation attracted our attention. To study the potential transcription factors regulating KTN1-AS1, we searched two online databases, hTFtarget and animalTFDB3, and predicted that there were 127 transcription factors ( Supplementary Fig. S1A). We further analyzed the correlation between the 127 transcription factors and the expression of KTN1-AS1 through the GEPIA database, and found that there were 36 transcription factors with an R value greater than 0.3, including SOX2 ( Supplementary Fig. S1B). A positive correlation was found between SOX2 and KTN1-AS1 ( Fig. 2A,B) and the mRNA expression of SOX2 was also significantly upregulated in ESCC tissues and cell lines (Fig. 2C,D). After transfection of pcDNA3.1-SOX2 and si-SOX2 into Kyse150 and Kyse170 cells (Fig. 2E), the expression level of KTN1-AS1 was substantially increased in SOX2 overexpressing cells and remarkably decreased in SOX2 knockdown cells (Fig. 2F). So we chose SOX2 for the following study.
As shown in Fig. 2G, according to the location of the three possible binding sites (site 1: − 137 bp to − 129 bp, site 2: − 408 bp to − 400 bp, and site 3: − 1113 to − 1105 bp), we constructed a luciferase reporter gene plasmid containing the − 1239 bp to + 56 bp region of KTN1-AS1 promoter, and co-transfected it with pcDNA3.1-SOX2 in Kyse150 and Kyse170 cells, respectively. The luciferase activity was found to be substantially increased. To further clarify the specific site of action, a truncated plasmid containing the fragment of − 594 bp to + 56 bp was constructed, and its luciferase activity was also considerably upregulated. The elevated luciferase activity was significantly decreased when site 2 was mutated, while no obvious change was observed accompanied with site 1 or site 3 mutation, suggesting the key role of site 2. The binding effect of SOX2 on site 2 of the KTN1-AS1 promoter was further confirmed by ChIP assay (Fig. 2H). These results collectively suggest that the elevated expression of KTN1-AS1 may be regulated by SOX2 in ESCC.
KTN1-AS1 facilitates proliferation, migration, and invasion of ESCC cells.
To investigate the biological function of KTN1-AS1 in ESCC cells, the overexpression plasmid of KTN1-AS1 was transfected into Kyse150 and Eca109 cells, and a striking upregulation of KTN1-AS1 was detected in the transfected cells (Fig. 3A). Furthermore, si-KTN1-AS1 was used to knockdown the expression of KTN1-AS1 in Kyse170 and Kyse150 cells (
KTN1-AS1 interacts with RBBP4 in the nucleus. The subcellular localization analysis in ESCC cells
showed that KTN1-AS1 was distributed in both nucleus and cytoplasm (Fig. 4A). RNA pull-down assay followed by mass spectrometry analysis was then performed to detect the RNA binding proteins, and RBBP4 was proved to be one of the differentially expressed proteins (Fig. 4B). The interaction between KTN1-AS1 and RBBP4 was further verified by Western blot and RIP assay (Fig. 4C,D). However, there were no significant differences in RBBP4 mRNA and protein expression levels in KTN1-AS1 overexpressed or knockdown cells (Fig. 4E), which was consistent with the poor correlation predicted by the GEPIA database (Fig. 4F). The expression level of RBBP4 was upregulated in esophageal carcinoma according to GEPIA database (Fig. 4G). Our results also showed an upregulation of RBBP4 in ESCC tissues and cell lines (Fig. 4H,I).
KTN1-AS1
relates to the epithelial-to-mesenchymal transition (EMT) process by inhibiting the expression of E-cadherin at the epigenetic level. Considering that SOX2 could influence the migration and invasion capability of esophageal cancer cells and up-regulate the expression of KTN1-AS1 at transcriptional level, we then detected the influence of KTN1-AS1 on EMT related markers. As shown in Fig. 5A,B, overexpression of KTN1-AS1 in Kyse150 cells considerably decreased the mRNA and protein expression levels of E-cadherin, and increased those of N-cadherin, vimentin, and MMP2; while downregulation of KTN1-AS1 in Kyse170 cells demonstrated the opposite tendency, suggesting the possible role of KTN1-AS1 in EMT process. Subsequently, we explored the expression changes of some EMT related markers, including E-cadherin, N-cadherin, Vimentin, MMP2, Snail1, and Twist1, after knocking down RBBP4. With the downregulation of RBBP4 in Kyse170 cells, the mRNA expression level of MMP2 was correspondingly downregulated, while the expression of E-cadherin was upregulated ( Supplementary Fig. S3).
As previous studies have demonstrated the involvement of HDAC1 in the transcriptional regulation of E-cadherin 17,18 , and RBBP4 and HDAC1/2 have been proved to form a core deacetylase complex in both the NuRD and Sin3 complex 19 www.nature.com/scientificreports/ of E-cadherin due to overexpression of KTN1-AS1 could be alleviated by the inhibition of RBBP4 (Fig. 5C). The interaction between RBBP4 and HDAC1 was noticeably strengthened in KTN1-AS1 overexpressed cells, while impaired in KTN1-AS1 knockdown cells detected by Co-IP assay (Fig. 5D), furthermore, KTN1-AS1 was also verified to bind with HDAC1 ( Fig. 5E), suggesting that KTN1-AS1 could combine with RBBP4 and HDAC1 to form a complex and simultaneously enhanced the binding effect of RBBP4 and HDAC1. In addition, a rescue experiment was conducted using the HDAC1 inhibitor TSA. As shown in Fig. 5F, TSA treatment alleviated the downregulation of E-cadherin caused by overexpression of KTN1-AS1. Moreover, as shown in Fig. 5G, overexpression of KTN1-AS1 weakened the enrichment of acetylation of histone H3 (ac-H3) at the promoter region of E-cadherin, whereas knockdown of KTN1-AS1 enhanced the enrichment of ac-H3, indicating that the binding action of KTN1-AS1 with RBBP4 and HDAC1 could finally influence the expression of E-cadherin via regulating the level of histone acetylation.
KTN1-AS1 promotes ESCC cell growth in vivo. To further verify the carcinogenic effect of KTN1-AS1
on ESCC, the tumor xenograft experiments were conducted to investigate the effects of KTN1-AS1 on tumor growth in vivo. Compared with control group, the tumor size, tumor volume and weight in the KTN1-AS1 upregulated group were significantly increased (Fig. 7A,B). Furthermore, the expression of KTN1-AS1 in the xenograft tissues of the KTN1-AS1 overexpression group was increased, accompanied by the increased mRNA expression level of N-cadherin, Vimentin, and MMP2, while the expression of E-cadherin was decreased (Fig. 7C). All these findings indicated that KTN1-AS1 could promote ESCC tumor growth in vivo.
Discussion
It is well known that lncRNAs take up a significant proportion acting as either tumor suppressors or oncogenes in the tumor carcinogenesis. In this study, KTN1-AS1 was proved to be upregulated in esophageal cancer tissues and cells, and exhibited as an oncogene in facilitating ESCC cells proliferation, migration, and invasion. The expression of KTN1-AS1 maybe induced by SOX2, and has a relationship with the EMT process. In non-small cell lung cancer (NSCLC), signal transducer and activator of transcription 1 (STAT1) was proved to bind to the promoter region of KTN1-AS1 and activated its transcription 11 . Since the promoter region of KTN1-AS1 is riched in transcriptional regulatory elements, we speculated that there should be other transcriptional factors involved in the regulation of abnormal KTN1-AS1 expression in ESCC. Amplification of SOX2 is one of the gene characteristics of ESCC and the expression of SOX2 is specifically highly in ESCC 21 . In the present study SOX2 was proved to activate transcription of KTN1-AS1. In addition, SOX2 acts as a potential EMT-inducing transcriptional factor in promoting cancer cells invasion and metastasis, and SOX2 induced lncRNAs have been demonstrated to play pivotal roles in tumorigenesis [22][23][24] , so there may be potential promoting role for KTN1-AS1 on ESCC cells progression.
Nuclear-localized lncRNAs can recruit chromatin modification and remodeling complexes to specific genomic sites to change the chromosome structure and modification status, and DNA/RNA methylation status, and further control the related genes expression 25 . Accumulating evidence has demonstrated that lncRNAs participate in these processes by binding to specific proteins 26 . Our present study found that KTN1-AS1 could bind with RBBP4 in the nucleus. As a chaperone protein, RBBP4 exerts its oncogenic function by participating in the formation of gene regulatory complexes, such as polycomb repressive complex 2 (PRC2) 27 , nucleosome remodeling factor (NURF) 28 , NuRD 29 , Sin3 complex 19 , and the deacetylase module (HDAC1/2, MTA1/2/3, RBBP4/7) complex 30,31 . Previous studies of KTN1-AS1 have primarily focused on the mechanism of competing endogenous RNA (ceRNA). In this study, we first discussed the underlying mechanism between KTN1-AS1 and RNA-binding protein in ESCC.
EMT is a cellular program known to be critical for malignant progression of tumors 32 . Since KTN1-AS1 could regulate the expression of EMT-related genes, we paid more attention to whether KTN1-AS1 regulates the expression of EMT-related genes through RBBP4 and HDAC1. E-cadherin, a typical epithelial cell marker, is a Ca 2+ dependent transmembrane glycoprotein closely related to intercellular adhesion. The dysregulation of E-cadherin expression that leads to carcinogenesis happens mostly at the epigenetic level 33,34 . HDACs are enzymes mediating the removal of acetyl from lysine residues in either histones or other proteins, causing the repression of gene transcription and subsequent changes in signaling events 35 , and HDAC1 was demonstrated to be involved in the transcriptional regulation of E-cadherin expression 17,18 . Our subsequent studies found that KTN1-AS1 could also bind to HDAC1 and affect the binding ability of RBBP4 to HDAC1. KTN1-AS1 may silence the expression of E-cadherin by forming a complex with RBBP4 and HDAC1 and enhancing its deacetylation effect on the promoter of E-cadherin in the nucleus, thereby promoting the EMT process in ESCC.
Conclusions
In summary, this study identifies the novel oncogenic role of lncRNA KTN1-AS1 in ESCC. Transcriptionally activated KTN1-AS1 by SOX2 may silence the expression of E-cadherin at epigenetic level by binding with RBBP4 and HDAC1 in the nucleus. KTN1-AS1 may act as a potential therapeutic and prognostic biomarker for ESCC.
Methods
Clinical specimens. One hundred and eleven ESCC patients were included in this study. The patients didn't undergo any radiotherapy or chemotherapy before operation between the years of 2013 to 2015 in the Fourth Hospital of Hebei Medical University. The clinical information was collected from the hospital records. The study was reported in accordance with ARRIVE guidelines and informed consent was obtained from all patients. Cell proliferation assay. The cellular proliferation capability was detected using MTS and clone formation assays. For MTS assay, 1 × 10 3 cells after 24 h of transfection were seeded into 96-well plate. The CellTiter 96 ® AQ ueous One Solution Reagent (Promega) was added after incubation of 0 h, 24 h, 48 h, 72 h, and 96 h. Then the optical density for each well was measured after incubation for 2 h. As to clone formation assay, 3 × 10 3 /5 × 10 3 cells after 24 h of transfection were seeded into a 6-well plate and were routine cultured for 1 week. The 4% paraformaldehyde was used to fix the cells and stained with crystal violet solution.
Cell migration and invasion assay. Wound healing and transwell assays were performed to detect the migration and invasion ability. For wound healing assay, transfected cells were inoculated in a 6-well plate. A straight scratch was made in each well when the cell density was close to overgrown, and pictures were taken at the same position at 0 h, 12 h, and 24 h using a microscope. For transwell assay, 1 × 10 5 cells after 24 h of transfection were seeded onto the upper compartment of the matrigel-coated chamber (Corning Costar, Corning, NY, USA) with 200 μL of serum-free RPMI-1640 medium; in the lower chamber, 600 μL of the medium containing 10% fetal bovine serum were added. Cells on the upper surface were wiped off after 24 h of incubation; the 4% paraformaldehyde was used to fix the invasive cells on the lower surface of the chamber and stained with crystal violet solution.
Western blot assay. The RIPA lysis buffer and PMSF (Solarbio) were used to extract proteins from cells. Luciferase reporter assay. In Kyse150 and Kyse170 cells, the promoter-reporter gene plasmids were separately co-transfected with pcDNA3.1-SOX2 or pcDNA3.1 empty plasmid. After 48 h of transfection, the luciferase activity was detected using the Dual-Luciferase Reporter Assay System (Promega) and normalized to Renilla activity.
Chromatin immunoprecipitation (ChIP) assay. ChIP assay was carried out using the EZ-ChIP™ kit (Millipore). After sonicating the cross-linked chromatin DNA, the antibodies against SOX2 (ZenBioScience, Cat# 864316) and acetyl histones H3 (Active motif, Ca# 61937) were used for precipitating the DNA fragments overnight at 4 °C, and protein A/G agarose beads were added to collect the precipitated complexes. The precipitated DNA fragment was purified and subjected to PCR detection. Primers used for the ChIP sequence were listed in Supplementary Table S4. RNA immunoprecipitation (RIP) assay. Cells were re-suspended in NP-40 lysis buffer. RNA was immunoprecipitated with antibodies against RBBP4 (ZenBioScience, Cat# 385565) and HDAC1 (Proteintech, Cat# 10197-1-AP). The qRT-PCR method was performed to detect the expression of KTN1-AS1.
Co-immunoprecipitation (Co-IP) assay. Cells were lysed with NP-40 lysis buffer and the lysates were precleared with protein A/G agarose beads (Santa Cruz Biotechnology, Dallas, Texas, USA). The supernatant was incubated with an antibody against RBBP4 overnight at 4 ℃ and protein A/G agarose beads were then added for further incubation. The precipitates were washed with lysis buffer and then suspended in 5 × SDS-PAGE sample loading buffer. After boiling for 10 min, the samples were analyzed by western blot and detected by the relevant antibodies.
Tumor xenograft model. G-418 bioreagent (Merck, Rahway, NJ, USA) was used to generate KTN1-AS1 stable expression Kyse150 cells. A total of 5 × 10 6 cells were subcutaneously injected into one side of male BALB/ c-nude mice purchased from Beijing HFK Bioscience CO., Ltd. Tumor volume was measured and calculated every 4 days. The mice were dissected on the 28 days, and tumors weight were measured. The animal experiments were conducted at the Experimental Animal Center of the Fourth Hospital of Hebei Medical University under the guidelines of NIH, and this study was approved by the Committee on the Ethics of Animal Experiments of the Fourth Hospital of Hebei Medical University.
Statistical analysis. SPSS 22.0 software package and GraphPad Prism 8.0 were used to perform data analysis and graphing. The statistical differences analysis between the two groups was conducted using Student's t-test. For overall survival analysis, Kaplan-Meier method and log-rank test were used. Univariate and multivariate Cox regression analysis was used to investigate the independent prognostic parameters. All experiments data came from three independent experiments performed in duplicate and presented as mean ± SD. P < 0.05 was considered statistically significant.
Ethics statement. The study involving the usage of patients tissues was performed in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of the Fourth Hospital of Hebei Medical University. The study was reported in accordance with ARRIVE guidelines.
Data availability
The datasets analyzed during the current study are available from the corresponding author on reasonable request.
|
v2
|
2022-11-25T06:17:29.486Z
|
2022-11-23T00:00:00.000Z
|
253838470
|
s2ag/train
|
Piperazine derivatives with potent drug moiety as efficient acetylcholinesterase, butyrylcholinesterase, and glutathione S-transferase inhibitors.
Cholinesterases catalyze the breakdown of the neurotransmitter acetylcholine (ACh), a naturally occurring neurotransmitter, into choline and acetic acid, allowing the nervous system to function properly. In the human body, cholinesterases come in two types, including acetylcholinesterase (AChE; E.C.3.1.1.7) and butyrylcholinesterase (BChE; E.C.3.1.1.8). Both cholinergic enzyme inhibitors are essential in the biochemical processes of the human body, notably in the brain. On the other hand, GSTs are found all across nature and are the principal Phase II detoxifying enzymes in eukaryotes and prokaryotes. Specific isozymes are identified as therapeutic targets because they are overexpressed in various malignancies and may have a role in the genesis of other diseases such as neurological disorders, multiple sclerosis, asthma, and especially cancer cell. Piperazine chemicals have a role in many biological processes and have fascinating pharmacological properties. As a result, therapeutically effective piperazine research is becoming more prominent. Half maximal inhibition concentrations (IC50 ) of piperazine derivatives were found in ranging of 4.59-6.48 µM for AChE, 4.85-8.35 µM for BChE, and 3.94-8.66 µM for GST. Also, piperazine derivatives exhibited Ki values of 8.04 ± 5.73-61.94 ± 54.56, 0.24 ± 0.03-32.14 ± 16.20, and 7.73 ± 1.13-22.97 ± 9.10 µM toward AChE, BChE, and GST, respectively. Consequently, the inhibitory properties of the AChE/BChE and GST enzymes have been compared to Tacrine (for AChE and BChE) and Etacrynic acid (for GST).
|
v2
|
2022-11-25T14:13:42.767Z
|
2022-11-23T00:00:00.000Z
|
253841562
|
s2orc/train
|
A vasculogenic mimicry prognostic signature associated with immune signature in human gastric cancer
Background Gastric cancer (GC) is one of the most lethal malignant tumors worldwide with poor outcomes. Vascular mimicry (VM) is an alternative blood supply to tumors that is independent of endothelial cells or angiogenesis. Previous studies have shown that VM was associated with poor prognosis in patients with GC, but the underlying mechanisms and the relationship between VM and immune infiltration of GC have not been well studied. Methods In this study, expression profiles from VM-related genes were retrieved from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Cox regression was performed to identify key VM-related genes for survival. Subsequently, a novel risk score model in GC named VM index and a nomogram was constructed. In addition, the expression of one key VM-related gene (serpin family F member 1, SERPINF1) was validated in 33 GC tissues and 23 paracancer tissues using immunohistochemistry staining. Results Univariate and multivariate Cox regression suggested that SERPINF1 and tissue factor pathway inhibitor 2 (TFPI2) were independent risk factors for the prognosis of patients with GC. The AUC (> 0.7) indicated the satisfactory discriminative ability of the nomogram. SsGESA and ESTIMATE showed that higher expression of SERPINF1 and TFPI2 is associated with immune infiltration of GC. Immunohistochemistry staining confirmed that the expression of SERPINF1 protein was significantly higher in GC tissues than that in paracancer tissues. Conclusion A VM index and a nomogram were constructed and showed satisfactory predictive performance. In addition, VM was confirmed to be widely involved in immune infiltration, suggesting that VM could be a promising target in guiding immunotherapy. Taken together, we identified SERPINF1 and TFPI2 as immunologic and prognostic biomarkers related to VM in GC.
Introduction
Gastric cancer (GC) is one of the most common gastrointestinal malignant tumors worldwide. It accounts 7.7% for of cancer-related deaths in 2020 and is recognized as the fourth leading cause of cancer-related mortality (1). Surgery accompanied by systemic chemotherapy is currently the main treatment modality for GC. Nevertheless, recurrence and metastasis are still often occurred especially in patients in the advanced stage (2,3). Unfortunately, other treatment options such as target therapy and immunotherapy are restricted because of drug resistance (4). Therefore, it is of great necessity to understand the underlying mechanisms involved in the recurrence and resistance of GC.
Malignant solid neoplasms depend on blood and oxygen supply to maintain growth and promote metastasis. Endothelium-dependent tumor angiogenesis has long been thought to be the sole pattern of blood supply (5). However, the clinical efficacy of anti-angiogenic targeted therapy for GC is still unsatisfactory, which could be explained by that the tumor hypoxic microenvironment further exacerbates the genetic instability of gastric tumor cells and activates the tumor driver gene, causing GC to be resistant to chemotherapy as well as antiangiogenic target therapy (6). Recent studies have revealed that highly aggressive tumor cells can form a vascular-like channel through their deformation and extracellular matrix remodeling to meet their energy demand, which is called vasculogenic mimicry (VM) (7, 8). VM, an epithelial-independent tumor microcirculation pattern, can promote tumor growth and facilitate metastasis by providing blood perfusion and promoting the secretion of protein hydrolases by tumor cells to degrade the basement membrane and extracellular matrix (9). It was found that VM is mostly present in highly malignant tumor tissues and is closely associated with tumor metastasis, recurrence and patient prognosis (10,11). In recent studies, VM formation was reported to be closely associated with poor prognosis in tumors such as glioblastoma (12), breast cancer (13), lung cancer (14), colorectal cancer (15), gallbladder cancer (16), and so on. Various VM-related genes, including VEGF, cadherin 5 (CDH5), tissue factor pathway inhibitor 2 (TFPI2), c-myc, hypoxia-inducible factor (HIF)-1alpha, Nodal, Twist, serpin family F member 1 (SERPINF1), and mutant fibronectin ED-B, were reported to be involved in VM process (10,11). Xu et al. reported that VEGF could induce VM formation by the PI3K signal transduction pathway (17). CDH5 is highly expressed in many aggressive cancer cells and its knockdown prevented VM (10,11). TFPI-2 has been reported to produce some of the phenotypic changes associated with aggressive, vasculogenic melanoma cells, thus contributing to VM plasticity (18). However, to date no study has reported the synergistic effect of these VM-related genes.
This study collected VM-related genes from literature research in PubMed (18-31) and investigated VM in GC using databases including The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO). Through Cox regression, VM-related genes were selected, and based on that, a nomogram was constructed to predict the prognosis of GC. In addition, the correlation of VM-related genes with immune checkpoints and immune infiltration was evaluated. This study highlights a functional role for the VM-related gene signature and uncover a potential prognostic biomarker for GC, providing novel insights into potential therapeutic targets and strategies for the treatment of GC.
Acquisition of VM-related genes and dataset preparation
We extracted 24 VM-related genes from the earlier reviews, then obtained the RNA-Seq data (HTSeq-FPKM), clinical data and survival data for patients with digestive system malignancy, including cholangiocarcinoma (CHOL), colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), liver hepatocellular carcinoma (LIHC), pancreatic adenocarcinoma (PAAD), rectum adenocarcinoma (READ) and stomach adenocarcinoma (STAD) from The Cancer Genome Atlas (TCGA) database (http://portal. gdc.cancer.gov/). The Gene Expression Profiling Interactive Analysis (GEPIA) (http://gepia.cancer-pku.cn/) (32) is an interactive web that includes 9,736 tumors and 8,587 normal samples from TCGA and the GTEx projects. We used GEPIA to detect the outcome with VM-related genes and generate survival curves. The R package "pheatmap" was used to draw heatmaps. In addition, 431 GC samples were also obtained from the Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo) with the accession number of GSE84437.
GO and KEGG enrichment analysis
Gene Ontology (GO) (33) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis (34) were performed by the "clusterProfiler" (an R package). GO has three independent branches: molecular function (MF), biological process (BP), and cellular component (CC). The KEGG database facilitates the systematic analysis of the intracellular metabolic pathways and functions of the gene.
Construction and validation of VM index in GC
Univariate Cox regression analysis was used to screen for VM-related genes in GC that were significantly associated with prognosis by "survival" R package. Moreover, Multivariate factors Cox regression analysis identified the VM-related genes in GC. KM-plotter (https://kmplot.com/analysis/) (35) was used to validate the prognostic effect of hub genes in GC. Multiplying the gene expression by its corresponding Cox regression coefficient and adding up was the risk scores which named VM index. To validate the effectiveness of the VM index, we detected the relationship of clinical information like T stage, N stage, tumor stage and overall survival from the TCGA database with the VM index and plotted by the "ggplot2" package in R. Based on the analysis of VM index, we divided patients into high and low-risk by the median VM index value. Then the VM index model was validated in the TCGA cohort and GEO cohort. Kaplan-Meier (KM) survival curves and risk score distribution model were plotted respectively by the "survival" R package. R packages "pheatmap" and "glmnet" were used to show the relationship between the expression of VM-related genes and VM index.
Differentially expressed genes analysis based on VM index
To identify the putative biological pathways and functions of DEGs, the "limma" R package was performed between high and low-risk VM index groups in GC. P-value ≤0.05 was selected as the threshold value.
Gene set enrichment analysis
The selected KEGG gene set was downloaded from the Molecular Signatures Database (MSigDB), and GSEA (version 4.0.3) (36,37) was performed to explore the potential molecular mechanisms in the high and low VM index groups and to acquire pathways for up and down regulation. A false discovery rate (FDR) of <0.05 was considered statistically significant.
Evaluation of microenvironment and immune cells infiltration
To evaluate the microenvironment of GC, the "ESTIMATE" R package (38) was used to acquire the tumor mutation burden (TMB), immune score, stromal score and estimate score. Then we used "single sample GESA (SsGESA)" and "pheatmap" packages to estimate differences in the infiltration of 29 immune cell types between high and low VM index groups. The Stemness index of the tissue samples containing DNA methylation-based stemness scores (DNAss) and mRNA expression-based stemness scores (RNAss) was also assessed (39). We also detected the relationship between immune checkpoints and the VM index from the data of the TCGA database.
Development and validation of the nomogram model
W e i n c o r p o r a t e d a l l s t a t i s t i c a l l y s i g n i fi c a n t clinicopathological parameters identified via multivariate Cox analysis and further established a visualized nomogram model including T stage, N stage and VM index through "rms" and "survival" R package, thus predicting the 3-, 5-year overall survival (OS) probability of patients. The Receiver Operating Characteristic (ROC) Curve of the nomogram was plotted to estimate the nomogram's predictive abilities with respect to GC patients' prognosis.
Immunohistochemical staining
Immunohistochemical (IHC) staining was performed in paraffin-embedded tissue from patients with GC at the Second Affiliated Hospital of Soochow University (Suzhou, China) between April 2017 to July 2018 to measure SERPINF1 protein level using an anti-SERPINF1 antibody (1:200, Abcepta, Suzhou, China). There were 33 cases of GC and 23 samples of paracancer tissues. All tissues were collected from patients who had not received chemotherapy or radiotherapy prior to surgery in the Second Affiliated Hospital of Soochow University and the informed consents were signed by all patients. Detailed clinicopathologic characteristics of the patients were listed in Supplementary Table 1. Each GC sample was evaluated based on the staining intensity and the percentage of cells with. The H-score was calculated as previously reported (40). The H-score value ranges between 0 and 300. Unpaired t-test was used to compare the SERPINF1 expression in GC tissues and the paracancer tissues.
Statistical analysis
Statistical analyses were carried out using R software (version 4.0.5; http://www.Rproject.org). Correlation with survival was evaluated by means of Kaplan-Meier plots, logrank test, and univariate and multivariate analyses based on the Cox proportional hazards method. Student's t test was applied for categorical variables. The correlation between two variables was assessed using Spearman's correlation test. In this study, P < 0.05 was identified as statistically significant.
Results
The expression level of VM-related genes and functional enrichment analysis To detect the expression level of VM-related genes in digestive system cancers and normal tissues, we draw a heatmap by the data from TCGA and found out that VM genes were highly expressed in varied digestive system cancers, including CHOL, COAD, ESCA, LIHC, PAAD, READ, STAD and so on ( Figure 1A). Enrichment analyses including GO and KEGG was performed to investigate the functions and involved pathways of VM-related genes and their potential relationship with GC development. KEGG pathways analysis was carried out to investigate comprehending functions of VM-related genes and the results showed that VM genes were involved in TNF signaling pathways, HIF-1 signaling pathway, VEGF signaling pathway, and focal adhesion ( Figure 1B). GO analysis revealed that the top five enriched BP terms were "aortic valve morphogenesis", "positive regulation of cell motility", "positive regulation of locomotion", "regulation of cellular component movement" and "aortic valve development". The top five enriched MF terms were "protein kinase activity", "E-box binding", "endopeptidase inhibitor activity", "peptidase inhibitor activity" and "endopeptidase regulator activity". The top five enriched CC terms were "extracellular matrix", "caveola", "plasma membrane raft", "collagen-containing extracellular matrix" and "secretory granule lumen" ( Figure 1C). Additionally, we found out that high expression levels of VM-related genes bring worse outcomes through GEPIA analysis ( Figure 1D).
Identification of VM-related hub genes in GC
Our univariate Cox regression analysis showed that 5 VMrelated genes were correlated with the prognosis of patients with GC (Figure 2A), including SERPINF1, TFPI2, CDH5, prostaglandin-endoperoxide synthase 2 (PTGS2) and snail family transcriptional repressor 2 (SNAI2). Furthermore, we conducted the multivariate Cox regression on the genes obtained in the previous steps, and we retained SERPINF1 and TFPI2 as the hub genes of VM in GC (Figure 2A). We also detected the relationship between expression levels of these two VM-related genes with OS status (Figures 2B, C). Also, expression levels of SERPINF1 and TFPI2 were associated with worse OS and progression-free survival (PFS) (Figures S1A-D). These results suggested that SERPINF1 and TFPI2 are risk factors in GC.
Construction and validation of a risk prediction model named VM index
VM index was constructed based on the two VM hub genes (SERPINF1 and TFPI2), which means VM index is the sum of proportional expression of SERPINF1 and TFPI2. We found that higher T stage, N stage, Tumor stage and poorer overall survival promoted the increase of VM index ( Figures 2D-G), which indicated that higher VM index represented poor prognosis, revealing the potential of VM index in prognosis prediction. To verify the effectiveness of the VM index prognosis prediction, Kaplan-Meier curves for the training cohorts in TCGA and validation cohorts in the GEO database were performed and the results were shown in Figures 3A-C. VM index was negatively correlated with prognosis. We then divided patients into high and low-risk groups according to the best cut-off value of the VM index. The distribution of the survival data and VM index for each patient, as well as the heatmaps of SERPINF1 and TFPI2, are shown in Figures 3B-D, in which patients with higher VM index usually had shorter survival time. These results verified that SERPINF1 and TFPI2 are important risk factors of GC.
Correlations of VM index with DEGs in GC and tumor microenvironment
By comparing the gene expression levels of high and low VM index groups, we screened DEGs and the most five highly expressed genes were HAND2 antisense RNA 1 (HAND2-AS1), stimulator of chondrogenesis 1 (SCRG1), heart and neural crest derivatives expressed 2 (HAND2), cholinergic receptor muscarinic 2 (CHRM2) and myocilin (MYOC) ( Figure 4A). To further explore the difference in enrichment pathways between high and low VM index groups, GESA analysis was performed. In the high VM index group, pathways mainly related to tumor metastasis and angiogenesis included basal cell carcinoma, TGF-b-signaling pathway, MAPKsignaling pathway, VEGF-signaling pathway, pathways in cancer, WNT-signaling pathway and JAK-STAT-signaling pathway. In contrast, the low VM index group was mainly related to gene repair such as base excision repair, nucleotide excision repair, RNA degradation, mismatch repairs and DNA replication ( Figure 4B).
Then, ESTIMATE was performed to compare TMB, immune and stromal scores in two groups. As shown in the diagrams, a high VM index was negatively associated with TMB, while positively correlated with the immune score, stomal score and ESTIMATE score ( Figures 4C-F). Also, SsGESA and ESIMATE were conducted to examine the relationship between the expression of SERPINF1 and TFPI2 with tumor microenvironment in GC.
Higher expression of SERPINF1 and TFPI2 was accompanied by lower cancer stemness while with a higher stromal score, immune score and ESTIMATE score ( Figure S2).
VM index and immune cell infiltration analysis
For further understanding of the relationships between immune cell infiltration and VM index, data from TCGA was used to quantify the activity or enrichment levels of immune cells, functions and pathways in GC by R package SsGESA. Heatmap and box plot of 29 kinds of immune cells indicated that the VM index was correlated to immune response ( Figure 5A). We further examined the relationships of immune cells and checkpoints with the VM index, and the results showed that the expression of immune checkpoints including CD28, CD86, BTLA, CD40LG, CD4 and CD8A was positively correlated with the VM index ( Figures 5B-E).
Construction of a nomogram for GC
To quantify the risk assessment and survival status for GC patients, a nomogram was built with VM index ( Figure S3) as well as other clinicopathological features including T stage, and N stage ( Figure 6A). In the nomogram model, the above parameters were assigned scores, and the score of each parameter was based upon plotting upward a straight line. For each GC patient, the survival probability of 3 and 5 years was estimated through the drop line from the total points line to the result line. To prove the accuracy of the nomogram, we made a comparison of the ROC curves between our nomogram model and additional clinicopathological variables (T stage, N stage and tumor stage) in GEO ( Figure 6B) and TCGA ( Figure 6C) cohorts. With the area under the curve (AUC) of the nomogram (combined model) larger than any single factor, our nomogram model, constituted by VM index, T stage, and N stage, was an optimal model to predict the long-term OS of GC patients.
Validation of the expression of SERPINF1 in GC tissues
To further identify the VM-related hub genes in the pathologenesis of GC, the expression of one hub gene (SERPINF1) was assessed in 33 GC tissues and 23 paracancer tissues using an IHC staining assay. Our IHC staining results on the GC tissue microarray demonstrated that SERPINF1 expression in GC tissues was obviously higher than that in adjacent non-tumor tissues (P<0.05, Figures 7A, B). Generally, the data indicated that SERPINF1 could be the candidate biomarkers for the VM process of GC.
Discussion
GC is one of the most common malignancies and acts as the leading cause of cancer-related death worldwide (1). Although surgery accompanied by systemic chemotherapy is recommended as the main treatment method for GC currently (4,41), recurrence and metastasis still happen in patients with resected GC and remains disturbing. Target therapies and immunotherapies have been applied as standard treatment for systemic therapy for unresectable locally advanced, recurrent or metastatic GC (1,(42)(43)(44), but unfortunately, the clinical efficacy of anti-angiogenic therapy is unsatisfactory due to drug resistance caused by hypoxic tumor microenvironment (45-47). Therefore, the identification of novel factors for prognosis prediction of patients with GC is of great necessity. VM, an alternative blood supply to tumors that is independent of endothelial cells or angiogenesis, has been demonstrated as a critical factor involved in the pathogenesis of solid tumors and is significantly associated with increased resistance to chemotherapy, low survival, and poor prognosis of patients with malignant tumors (7, 48, 49). It was reported that VM could promote tumor neovascularization to favor metastasis (50) and to drive resistance to antiangiogenic therapy (12,51). A meta-analysis showed that VM was related to the poor OS and DFS of patients with digestive cancer (52). Baeten et al. reported that VM formation could be used to predict prognosis in colorectal cancer (15). To date, only one review has reported that VM is associated with a poor prognosis in patients with GC in China (53). The underlying mechanisms and the relationship between VM and immune infiltration is still unclear. In this study, we demonstrated that VM-related genes as previously reported were upregulated in GC and involved in the HIF-1 signaling pathway (54) and VEGF signaling pathways (55). Two VM-related genes, SERPINF1 and TFPI2, were identified as independent risk factors for the prognosis of patients with GC through Cox regression analysis (18, 56). SERPINF1 was reported to be involved in the migration and invasion by extracellular matrix (ECM) remodeling in GC (19). TFPI2 may take part in promoting colorectal cancer by changing the DNA methylation status of colon epithelial cells (39,57,58). In our study, we constructed a model named VM index which could well evaluate the expression levels of VM in GC. Interestingly, for the first time, we found that the VM index was correlated with immune cells and immune checkpoints such as CD28, CD86, BTLA, CD40LG, CD4, and CD8A in GC, suggesting that VM may promote the pathogenesis and metastasis of GC through regulating immune cells and immune surveillance. We also demonstrated that the upregulation of VM is accompanied with lower cancer stemness indices including RNAss and DNAss, while with higher immune infiltration, including stromal score, immune score and ESTIMATE score, indicating that cancer stem cells and immune infiltration were of great necessity in the regulation of GC cells by VM. Tumor hypoxic microenvironment, as an indispensable part of cancer that has been exclusively focused on and increasingly acknowledged for decades, was reported to be inseparable from VM (59). Herein our study analyzed the relationship between VM, several immune infiltration criteria, and the tumor microenvironment. ESTIMATE and GSEA revealed that a high VM index was negatively associated with TMB, while positively correlated with the immune score, stomal score and B C A FIGURE 6 Construction of a nomogram based VM-index for prognosis prediction of GC. (A) The nomogram using T stage, N stage and VM index. For each patient, three lines are drawn upward to verify the points received from the three predictors of the nomogram. The sum of these points situates on the 'Total Points' axis. Then a line is drawn downward to assess the 3-, and 5-year overall survival of GC. (B, C) The ROC curve to evaluate the nomogram in GEO and TCGA database. Y-axis, Sensitivity; X-axis, Specificity. ESTIMATE score. Recently, a review reported that cancer stem cells have been identified to be involved in VM in gastrointestinal cancer (45). Consistent with that, our study revealed the correlation between the two VM key genes (SERPINF1 and TFPI2) and stemness. Besides, the VM index was identified to be positively associated with an immune score, stromal score and ESTIMATE score, which are important metrics of the tumor microenvironment (38). VM, an alternative mechanism of vasculatures, has been reported to be involved in resistance to anti-angiogenic therapies, whereas the combination of anti-angiogenesis and immune therapy could bring out better clinical efficacy (60)(61)(62)(63). Our study further investigated the regulatory role of VM in immune cells. There are several possible mechanisms by which the VM index is involved in immune infiltration. On the other hand, the VM index could upregulate immune checkpoints including CD28, CD86, BLTA and CD40LG to inhibit immune response, thus leading to tumor immune escape. The VM index could also directly upregulate immunotherapeutic genes including CD4 and CD8A. Both of these two mechanisms provide potential targets and novel insights into the treatment of VM by immunotherapy.
To explore the impact of VM on the prognosis of patients with GC, a nomogram based on VM index, T stage and N stage was constructed to visualize the effects of clinical features and VM index on patients' 3-and 5-year survival probabilities. Time-dependent AUC identified that our nomogram had high prediction efficiency and was better than the T stage and N stage, demonstrating the potential value of our nomogram in clinical practice. Furthermore, this study has allowed the development of strategies with therapeutic potential directed against VM formation. In clinic, the combination of traditional anti-angiogenic therapies with SERPINF1 and TFPI2, the two potential anti-VM targets may improve the outcomes of patients with GC. Besides, our nomogram may be a valuable tool for assessing the prognosis of patients with GC to date, and researchers and clinicians may conveniently access it.
Conclusion
In summary, our study constructed a novel risk score model in GC named VM index based on SERPINF1 and TFPI2. The VM index showed satisfactory predictive performance in both training and validation cohorts and could be applied to predict the prognosis and tumor microenvironment in patients with GC. Moreover, a precise and convenient nomogram based on VM index and clinical factors including the T stage and N stage was well constructed and validated by the ROC curve, in which the nomogram showed excellent performance in the prediction of GC compared with merely the T stage or N stage. In addition, we confirmed that VM is widely involved in the regulation of immune infiltration and immune checkpoints. Therefore, we have not only constructed a precise and convenient model for prognosis prediction but also have proposed that VM could be a promising molecular target in guiding immunotherapy.
Data availability statement
Publicly available datasets were analyzed in this study. This data can be found here: TCGA (http://portal.gdc.cancer.gov/) and Gene Expression Omnibus (https://www.ncbi.nlm.nih.gov/ geo) under the accession number GSE84437.
Ethics statement
The studies involving human participants were reviewed and approved by the Second Affiliated Hospital of Soochow University. The patients/participants provided their written informed consent to participate in this study.
|
v2
|
2022-11-25T16:58:18.291Z
|
2022-11-23T00:00:00.000Z
|
253876604
|
s2ag/train
|
Optimization modeling of anti breast cancer drug candidates based on data mining
In drug research and development, in order to save time and cost, the method of establishing compound activity prediction model is usually used to screen potential active compounds, In order to become a candidate drug, a compound not only needs to have good biological activity, but also needs to have good pharmacokinetic properties and safety in human body, which is collectively known as ADMET. This paper adopts data mining technology, Firstly, the use of random forest to find the main variables of modeling is studied, and its independence is verified by high correlation filtering. The 20 main operating variables selected are MDEC-23, maxHsOH etc; Secondly, a five layer BP neural network is used to establish a compound bioactivity prediction model, which can predict the IC50 value and the corresponding pIC50 value of the compound; Then the improved BP neural network model is used to establish the classification prediction model of compounds Caco-2, CYP3A4, ERG, hob and Mn. The algorithm verifies that the accuracy of CYP3A4 is 94.3%, and the accuracy of the five models is more than or close to 90%, which is more practical than the prediction value of the improved BP neural network; Finally, the main variables of genetic algorithm are used to make the compound pair inhibit er α The value range of biological activity is optimized, which has certain practical significance.
|
v2
|
2022-11-25T17:06:11.014Z
|
2022-11-23T00:00:00.000Z
|
253858734
|
s2ag/train
|
Recognition of urban pollution types based on BP neural network
There have been many studies on the survival analysis of patients with non-small cell lung cancer (NSCLC). However, most of the studies are based on the extraction of tumor radiomics features based on the tumour label outlined by the physician, followed by a combination of clinical and pre-treatment PET/CT image features of the patient for survival analysis. Survival analysis of patients with locally advanced NSCLC based on whether pre- and post-treatment FDG-PET can be performed without tumors label using a deep learning approach. The consistency index (C-index) of the convolutional neural network model was 0.67 when using pre- and post-treatment FDG-PET, suggesting that simultaneous reading with pre- and post-treatment PDG-PET can predict the probability of patient risk.
|
v2
|
2022-11-26T16:38:20.678Z
|
2022-11-23T00:00:00.000Z
|
253913229
|
s2orc/train
|
Differences in Cancer-Specific Mortality after Trimodal Therapy for T2N0M0 Bladder Cancer according to Histological Subtype
Simple Summary Trimodal therapy represents an accepted treatment option for non-metastatic muscle-invasive bladder cancer, which is an alternative to radical cystectomy. Evidence regarding trimodal therapy efficacy has predominantly, or even exclusively, been applied to urothelial carcinoma of the urinary bladder patients. To address this void, we tested for differences in cancer-specific mortality in trimodal therapy-treated bladder cancer patients, according to histological subtype, namely urothelial carcinoma vs. neuroendocrine carcinoma vs. squamous cell carcinoma vs. adenocarcinoma. Abstract We aimed at assessing the impact of non-urothelial variant histology (VH), relative to urothelial carcinoma of the urinary bladder (UCUB), on cancer-specific mortality (CSM) in T2N0M0 bladder cancer patients treated with trimodal therapy (TMT). TMT patients treated for T2N0M0 bladder cancer were identified within the Surveillance, Epidemiology, and End Results database (2000−2018). Patients who underwent TMT received trans-urethral resection of the bladder tumor, chemotherapy, and radiotherapy. CSM-FS rates were tested using Kaplan–Meier plots and multivariable Cox-regression (MCR) models according to histological subtype: UCUB vs. neuroendocrine carcinoma vs. squamous cell carcinoma vs. adenocarcinoma. A total of 3846 T2N0MO bladder cancer patients treated with TMT were identified. Of these, 3627 (94.3%) harbored UCUB, while 105 (2.7%), 85 (2.2%), and 29 (0.8%) harbored neuroendocrine carcinoma, squamous cell carcinoma, and adenocarcinoma, respectively. In Kaplan–Meier analyses, 3-yr CSM-FS rates were 57% for UCUB, 51% for neuroendocrine carcinoma, 35% for squamous cell carcinoma, and 60% for adenocarcinoma (p-value < 0.0001). In MCR models, only squamous cell carcinoma exhibited higher CSM than UCUB (HR 1.98, 95%CI 1.5–2.61, p-value < 0.001). Despite the small number of observations, squamous cell carcinoma distinguished itself from UCUB based on worse survival in T2N0M0 patients after TMT.
Introduction
Trimodal therapy (TMT) represents an accepted treatment strategy for non-metastatic muscle-invasive bladder cancer, which is an alternative to radical cystectomy (RC) [1]. It consists of maximal trans-urethral resection of the bladder tumor (TURBT), chemotherapy, and radiotherapy [2]. TMT might be considered for patients not eligible for RC, or for select, highly motivate patients interested in bladder-sparing regimens. To date, no randomized trials have compared the efficacy of TMT vs. RC. However, prospective studies and retrospective comparative analyses have shown encouraging oncological TMT results relative to RC [3][4][5]. In particular, large single-institution data has shown favorable oncological results for TMT-treated patients diagnosed with organ-confined disease [6]. However, these observations have predominantly, or even exclusively, been of UCUB patients. In consequence, cancer control outcomes are largely unknown for non-urothelial variant histology (VH) patients treated with TMT [7,8]. We addressed this knowledge gap in T2N0M0 bladder cancer patients treated with TMT who harbored histological subtypes other than UCUB and compared those individuals to their UCUB counterparts.
We hypothesized that no survival differences would be recorded according to bladder cancer histological subtypes. We addressed this objective using the Surveillance, Epidemiology, and End Results (SEER) database (2000-2018).
Study Population
Within the SEER database (2000-2018) [9], we focused on patients 18 years or older with histologically confirmed bladder cancer (International Classification of Disease for Oncology [ICD-O-3] site code C67.0-67.6 and C67.8-67.9). We only considered T2N0M0 patients treated with TMT. Patients who underwent TMT received TURBT, chemotherapy, and radiotherapy. The four histological subtypes of grades included were UCUB, squamous cell carcinoma, adenocarcinoma, and neuroendocrine carcinoma. Cancer-specific mortality (death from bladder cancer) was defined according to the SEER cause-specific death classification. Exclusion criteria consisted of unavailable information about grade and histology as well as all autopsy, death certificates, and missing follow-up data.
Statistical Analyses
Statistical analyses focused on cancer-specific mortality free-survival (CSM-FS) using Kaplan-Meier analyses and multivariable Cox-regression (MCR) models. Results were stratified according to histological subtypes: UCUB vs. neuroendocrine carcinoma vs. squamous cell carcinoma vs. adenocarcinoma. In MCR models, covariates consisted of age, grade (low vs. high), and sex. All statistical tests were two-sided with a level of significance set at p< 0.05. Analyses were performed using the R software environment for statistical computing and graphics (version 4.2.1; http://www.r-project.org/; accessed on 25 August 2022).
Multivariable Cox Regression Models Predicting CSM
In MCR models, relative to UCUB, only squamous cell carcinoma emerged as an independent predictor of higher CSM (HR 1.98, 95% CI 1.5-2.61, p-value < 0.001) after adjusting for all covariates ( Table 2).
Multivariable Cox Regression Models Predicting CSM
In MCR models, relative to UCUB, only squamous cell carcinoma emerged as an independent predictor of higher CSM (HR 1.98, 95% CI 1.5-2.61, p-value < 0.001) after adjusting for all covariates (Table 2).
Discussion
Few reports compared cancer control outcomes after TMT between UCUB and nonurothelial VH patients [10,11]. We addressed this void and hypothesized that no differences would distinguish VH from UCUB in TMT-treated T2N0M0 patients. Our analyses led to several noteworthy observations. First, of all 3846 patients, a very marginal proportion of non-urothelial VH patients was recorded. Specifically, we identified 105 (2.7%), 85 (2.2%), and 29 (0.8%) individuals harboring neuroendocrine carcinoma, squamous cell carcinoma, and adenocarcinoma, respectively. These observations indicate that only very large-scale epidemiological databases might allow for comparisons between VH and UCUB after TMT. Moreover, it also explains the lack of studies addressing differences in survival according to different histological subtypes after TMT, based on institutional databases. Additionally, the rarity of non-urothelial VH patients treated with TMT also indicates the limited confidence that the urological community places on bladder-sparing strategies in the context of non-UCUB histological subtypes. Indeed, data validating the efficacy of TMT, predominantly, or even exclusively, stem from UCUB patients. In consequence, the use of TMT in non-UCUB patients cannot be based on strong objective evidence. These observations motivated the conduct of the current study.
Second, in Kaplan-Meier analyses, squamous cell carcinoma exhibited lower 3-yr CSM-FS than UCUB (35% vs. 57%). Moreover, in MCR models, squamous cell carcinoma represented an independent predictor of higher CSM (HR 1.98, 95% CI 1.5-2.61, p-value < 0.001), relative to UCUB. In consequence, our observations suggest that the use of TMT in T2N0M0 squamous cell carcinoma patients is associated with worse cancer control outcomes vs. their T2N0M0 UCUB counterparts. However, it should be emphasized that our findings are based on a relatively low number (n = 85) of squamous cell carcinoma patients. This limitation also applies to other non-UCUB histological subtypes. Specifically, only 105 neuroendocrine carcinoma and 29 adenocarcinoma patients were identified. Those histological subtypes did not reach independent predictor status for CSM, relative to UCUB. In consequence, our observations indicate worse cancer control outcomes in T2N0M0 squamous cell carcinoma patients treated with TMT, according to sufficient numbers of observations to allow valid statistical testing of potential CSM differences. However, comparisons of other histological subtypes to UCUB are based on marginal, if not insufficient, numbers of observations to justify valid conclusions.
We tested for cancer control differences in T2N0M0 patients treated with TMT, according to histological subtype, within a relatively large population-based cohort. Despite the large size of the current cohort, we encountered critical limitations due to sample size when non-UCUB histological subtypes other than squamous cell carcinoma were considered. Other analyses that addressed TMT in non-UCUB patients were affected by similar sample size limitations. For example, within an NCDB (2004-2013) analysis, Fischer-Valuck et al. [11] directly compared squamous cell carcinoma (n = 78) vs. UCUB (n = 3252) patients treated with TMT and detected worse median overall survival in squamous cell carcinoma patients (15.1 vs. 30.4 months; p-value = 0.013). Unfortunately, unlike in the current study, the authors did not include other non-UCUB histological subtypes to allow comparisons relative to UCUB. Therefore, our results regarding other histological subtypes cannot be directly compared to the Fischer-Valuck study [11]. However, the comparison of squamous cell carcinoma vs. UCUB patients after TMT revealed similar survival disadvantage, as recorded in the current study, even though a somewhat different endpoint was considered, which was the overall survival in NCDB vs. CSM in the current study. Unfortunately, NCDB only allows overall survival analyses and CSM cannot be distinguished from other-cause mortality. In a different study, which also focused on TMT-treated patients, Janopaul-Naylor et al. [10] (NCDB 2004-2015) tested for overall survival differences in RC-and TMT-treated patients according to different histological subtypes. The authors reported worse overall survival for squamous cell carcinoma TMT-treated patients (n = 94, HR 1.49, 95% CI 1.25-1.77, p < 0.001) as well as for adenocarcinoma TMT-treated patients (n = 38, HR 1.75, 95% CI 1.36-2.25, p < 0.001) vs. their RC counterparts harboring the same histological subtype. Again, the different study design, population, and endpoint do not allow direct comparisons with the current study. Furthermore, Krasnow et al. [12] compared outcomes of pure vs. variant UCUB patients after TMT. Specifically, 66 individuals (22%) harbored variant UCUB. Of these, 24 patients exhibited squamous differentiation (36%). In MCR models, variant UCUB was not associated with worse disease specific survival (p = 0.3). However, our findings are not comparable with Krasnow et al. [12] since the authors addressed differences in patients with UCUB according to variant histology differentiation. Moreover, patients exhibiting dominant histological subtypes different from UCUB (e.g., adenocarcinoma, squamous cell carcinoma) were excluded from the analyses. Unfortunately, to the best of our knowledge, no institutional or multi-institutional studies directly addressed the comparison made in the current study, namely CSM after TMT in T2N0M0 patients according to histological subtype.
Taken together, our study identified a very low number of T2N0M0 non-UCUB patients treated with TMT (n = 219). Of those, the majority harbored neuroendocrine carcinoma (n = 105), followed by squamous cell carcinoma (n = 85) and adenocarcinoma (n = 29) vs. 3627 UCUB patients. Squamous cell carcinoma patients distinguished themselves from UCUB patients based on significantly worse survival (3-yr CSM-FS 35% vs. 57%, MCR HR 1.98, p-value < 0.001). Previous reports showed marginal benefit from perioperative chemotherapy for squamous cell carcinoma patients treated with RC [13,14]. Thus, the radiosensitizing effect of chemotherapy might not be sufficient in squamous cell carcinoma TMT-treated patients to achieve adequate disease control. However, for other non-urothelial VH, the number of observations were insufficient to perform valid testing and interpretation of results. In consequence, it may be postulated that in TMT-treated T2N0M0 squamous cell carcinoma patients, significantly less favorable cancer control outcomes may be expected than when TMT is applied to their UCUB counterparts. In this context, novel therapy regimens (e.g., checkpoint inhibitors) and risk stratification based on genomic profiling might improve squamous cell carcinoma TMT treated patients' cancer control outcomes [15].
Despite the novelty of our findings, several limitations need to be acknowledged. The first and foremost limitation consists of patient origin. Specifically, our findings are applicable to individuals who were captured within the SEER database. Therefore, the observations made within the current study cannot be applied to patients originating from outside of the United States or even patients that are not comparable to those included in the SEER database. For example, patients treated at centers of excellence, such as the Memorial Sloan Kettering Cancer Center or MD Anderson Cancer Center, are not included in the SEER database. Therefore, institutional or multi-institutional data reflecting cancer control outcomes of such individuals should be used if available. Second, our analyses relied on limited numbers of observations. Sample size represented a critical limitation, even within the current, very large-scale database. Consequently, it is unlikely that smaller scale databases, except for NCDB, will provide larger sample size results. Third, unlike other studies, we only focused on T2N0M0 bladder cancer patients. Such consideration was based on the concept that TMT is of marginal or no value in the presence of extra-vesical bladder cancer. Fourth, within the SEER database, only the dominant histological subtype is reported; therefore, analyses adjusting for presence of mixed histology variants cannot be assessed. Fifth, the SEER database does not provide specific information about size and focality of bladder tumors or completeness of TURBT. Similarly, the type, dose, and timing of radiotherapy as well as chemotherapy, were not available. Sixth, our endpoint consisted of CSM-FS. However, in TMT studies, endpoints may consist of bladder preservation, local recurrence with cystectomy or distant recurrence with preserved bladder. These endpoints cannot be addressed using SEER but are provided in institutional and multi-institutional analyses [6]. Finally, our report represents a retrospective analysis with high potential for selection biases.
Conclusions
Despite the small number of observations, squamous cell carcinoma distinguished itself from UCUB based on worse survival in T2N0M0 patients after TMT. Institutional Review Board Statement: All analyses and their reporting followed the SEER reporting guidelines. Due to the anonymously coded design of the SEER database, study-specific Institutional Review Board ethics approval was not required.
|
v2
|
2022-11-26T16:58:58.021Z
|
2022-11-23T00:00:00.000Z
|
253913438
|
s2ag/train
|
Gut microbiota in human health: insights and discussion on the role of probiotics
Gut microbiota remains stable and individualized throughout life, but there are inter-species or intra-species variations that may be controlled by various environmental factors. Several diseases may be associated with dysbiosis like inflammatory bowel disease (IBD), obesity, diabetes mellitus, irritable bowel syndrome (IBS), gastric and colon cancer, and sometimes colorectal polyps, non-alcoholic steatohepatitis (NASH), and liver cirrhosis, because of the gut-derived neurotoxins. There is a 10% to 30% risk of development of post-infectious IBS despite the beneficial effects of a low FODMAP diet on IBS symptoms. This diet reduces the luminal concentration of one of the most common bacteria that is the Bifidobacterium. Therefore, probiotics help in the restoration of normal gut microbiota, are a valuable tool in the treatment of certain diseases and help in recovering microbial balance in the gut. Bifidobacterium W11 is a novel probiotic with certain special characteristics that can be of benefit in dysbiosis. This review evaluates gut microbiota dysbiosis, associated complications of dysbiosis, and benefits of treatment with probiotics based on focused group discussions of various experts from India, followed by guidance statements based on analysis of published literature. The beneficial effects of Bifidobacterium longum W11 (B. longum W11) in the management of IBS, IBD, and liver diseases have been elaborated. The proposed management strategy can effectively aid the management of gut dysbiosis in several gastrointestinal conditions and help in understanding the judicious use of probiotics.
|
v2
|
2022-11-26T17:11:50.406Z
|
2022-11-23T00:00:00.000Z
|
253920643
|
s2orc/train
|
The Prognostic Significance of HALP Index for Colon Cancer Patients in a Hispanic-Based Population
Background Survival and recurrence rates following locoregional colon cancer surgical resection are highly variable. Currently used tools to assess patient risk are still imperfect. In the present work, we evaluate, for the first time, the prognostic value of the recently developed HALP (hemoglobin, albumin, lymphocyte, and platelet) index in Hispanic colon cancer patients. Patients and Methods. We conducted a retrospective cohort study in Mexican patients with a nonmetastatic colon cancer diagnosis who underwent surgical resection. We determined the preoperative HALP score optimal cut-off value by using the X-tile software. We plotted survival curves using the Kaplan–Meier method and performed a multivariate Cox regression analysis to explore the association of preoperative HALP score with two primary endpoints: overall survival (OS) and disease-free survival (DFS). Results We included 640 patients (49.8% female). The optimal HALP cut-off value was 15.0. A low HALP index was statistically significantly associated with a higher TNM stage. Low HALP score was statistically significantly associated with shorter median OS in the Kaplan–Meier analysis (73.5 vs. 84.8 months) and in the multivariate Cox regression analysis (HR = 1.942, 95% CI = 1.647–2.875). There was no significant association between the HALP score and DFS. Conclusions Our findings show that the HALP index is an independent factor associated with survival in Hispanic patients, despite recurrence. It seems to reflect both the anatomical extent of the disease and traditionally unaccounted nutritional and inflammatory factors that are significant for prognosis.
Introduction
Colon cancer is a public health concern. At a global level, colorectal carcinoma is currently the third most incidence of cancer and the second most common cause of cancer death. In Mexico, during 2020 alone, this cancer accounted for 14,901 new cases and 7,755 deaths [1].
Surgery is the mainstay of treatment in locoregional colon cancer and continues to be the only curative modality. However, surgery results in a cure only in approximately 50% of cases, as recurrences following surgical resection remain a major problem and are a frequent cause of eventual death [2,3].
Colon cancer is characterized by being heterogeneous. Whether recurrence occurs or not is possibly dependent on a myriad of factors that result in diferent individual risks of recurrence [4]. Some high-risk patients beneft from the addition of adjuvant chemotherapy and strict surveillance, while others do not. Te latter group of patients could be spared from signifcant chemotherapy side efects or bothersome procedures, while the former greatly beneft from these interventions, improving survival rates [4,5]. Terefore, elucidating prognostic factors is an essential task.
Te tumor-node-metastasis (TNM) staging system by the American Joint Committee on Cancer (AJCC) is currently the sturdiest and most frequently used tool for assessing patients' prognosis to guide management in colon cancer [6,7]. However, this tool still has limitations, as individuals within the same stage have highly variable survival rates, and patients' prognoses across diferent stages sometimes overlap [7,8].
Cancer progression and metastasis are not solely dependent on tumor characteristics and anatomic extent [9]. It is well established that both systemic infammation and nutrition play an important role in prognosis [10]. Tese two factors are linked to many tumor characteristics, including proliferation, invasion, metastasis, and recurrence [11]. Various infammatory and nutritional markers have been individually associated with survival outcomes in several types of cancer. Moreover, studies have used a combination of these markers to successfully predict prognosis with the use of a single index; examples include the prognostic nutritional index (PNI), the neutrophil-tolymphocyte ratio (NLR), and the platelet-to-lymphocyte ratio (PLR) [12].
In 2015, Chen et al. developed the novel HALP index, which utilized preoperative hemoglobin, albumin, lymphocyte, and platelet levels to assess gastric cancer patient's prognosis [13]. Shortly after, Jiang et al. studied the prognostic value of this index in colorectal cancer, fnding superb results. Teir study showed that patients with lower HALP scores had an increased risk of death and cancerrelated death, with lower overall survival and cancerspecifc survival. All the aforementioned associations were statistically signifcant and independent of other factors in the multivariate analysis, both in the training and validation sets [14].
Given that the HALP index is an easily reproducible tool, which utilizes widely-available routinary lab values, its adoption in clinical practice could signifcantly beneft developing countries, such as those in Latin America. Unfortunately, so far, this index has not been validated for Hispanic populations, where colon cancer epidemiological and prognostic characteristics are distinct [18]. A formal validation is needed to implement the HALP index in clinical practice.
In the present work, we aim to evaluate the prognostic value of the HALP index for Hispanic colon cancer patients in one of Mexico's largest oncological centers.
Study Design.
We conducted a retrospective cohort study in Mexican colon cancer patients without metastatic disease. We explored the association of the HALP index with our two primary endpoints: overall survival (OS) and disease-free survival (DFS). Tis study was approved by the Institutional Review Board (IRB) at Mexico's National Cancer Institute (No. 2021/143). Written informed consent from patients was waived due to the retrospective nature of this study.
Patients.
All patients with a nonmetastatic colon cancer diagnosis who underwent surgical resection at Mexico's National Institute of Cancer, between January 2008 and January 2020, were retrospectively screened for inclusion. We included patients with histologically confrmed stage I-III primary colonic adenocarcinoma who were subjected to radical surgery for the primary tumor. We excluded patients who had been previously treated for cancer, had recurrent or metastatic disease, synchronous malignancies, hereditary colon cancer, underwent other types of surgeries (local excision or palliative surgery), and/or had missing information needed to calculate the HALP index.
Data Collection.
Clinical and pathological variables of all included patients were collected from electronic medical records. Tese included age, gender, body mass index (BMI), date of diagnosis, tumor characteristics (location, diferentiation grade, presence of lymphovascular/perineural invasion, and staging characteristics), surgical margin status, and information regarding complications. Also, to calculate the HALP index, preoperative serum albumin, hemoglobin, lymphocyte, and platelet values were collected for all patients. All patients were staged according to the 7th edition AJCC TNM staging system [6]. Information regarding patients' vital status and disease recurrence was collected in order to assess our primary endpoints (OS and DFS). Patients were followed-up regularly (every two weeks while receiving chemotherapy and every two months otherwise). Follow-up care included clinical examinations, laboratory testing, and imaging (TAC, PET-CT), according to international guidelines. Follow-up for each patient began at the date of diagnosis and continued until (1) loss of followup, (2) death, or (3) last visit before the cut-of date (January 2020).
Statistical Analysis.
HALP index was calculated with the following formula: hemoglobin level (g/L) × albumin level (g/L) × lymphocyte (/L)/platelet count (/L) [13]. Te optimal cut-of value for HALP was determined by using the X-tile software (Version 3.6.1, Yale University, USA) [19]. Te cutof with the lowest p-value calculated from the chi-squared test for OS was selected and patients were classifed as having either low or high HALP. Te optimal cut-of value was frst determined for the entire cohort, and later it was identifed independently within two subgroups (patients who underwent emergency surgery vs. patients who underwent elective surgery), in order to validate our obtained value.
For the analysis, quantitative variables were converted into categories (groups). Age was grouped according to three categories (younger, middle-aged, and senior adults) and BMI was grouped according to nutritional status categories. Frequencies and proportions were reported for all collected variables. Chisquared test was used to detect associations between the collected variables and the dichotomized HALP index.
Te Kaplan-Meier method was used to calculate the median OS and DFS for each variable and plot survival curves. A log-rank test was used to compare survival between groups. Additionally, multivariate Cox regression analysis was performed to identify associations between each variable and survival; hazard ratios (HRs) and the respective 95% confdence intervals (CIs) were calculated. Only statistically signifcant variables in the univariate analysis were included in the subsequent multivariate analyses. All statistical analyses were performed with SPSS software (SPSS 19.0, IBM, Chicago, IL, USA). A two-tailed p value of <0.05 was considered statistically signifcant.
Patient
Characteristics. 893 patient records were screened for inclusion. Ultimately, 640 patients were deemed eligible and included in our fnal sample.
Cut-Of Value Determination.
Te optimal cut-of value in the entire cohort for HALP was set at 15.0, and with this dividing point, patients were classifed as having either a low or high HALP index. Te optimal cut-of points for the two evaluated subgroups, patients who underwent elective surgery vs. those who had emergency surgery, were 15.0 and 14.7, respectively. Given their similarity to the cut-of value obtained for the whole cohort, this value (15.0) was concluded to be a valid cut-of point. X-Tile results and plots are displayed in Figure 1.
3.3.
HALP Association with Clinicopathological Characteristics. HALP index was statistically signifcantly associated with pTNM stage, colon cancer location, BMI, tumor diferentiation degree, and achievement of adequate lymphadenectomy (12 or more resected lymph nodes).
HALP and Clinicopathological Characteristics Association
with Survival. Te median follow-up time, OS, and DFS were 46.39, 51.9, and 47.9 months, respectively. In the (univariate) Kaplan-Meier analysis, low HALP score was signifcantly associated with shorter median OS (73.5 vs. 84.8 months; log-rank test p � 0.013; Figure 2), as were advanced pTNM stage, poorly diferentiated tumors, lymphovascular invasion, perineural invasion, presence of cancer obstruction, N positive disease, and positive surgical margins. Multivariate Cox regression analysis established that a low HALP score is an independent factor associated with shorter OS (HR � 1.942, 95% CI � 1.647-2.875; p � 0.031). All variables' association with OS can be reviewed in Table 2.
Tere was no statistically signifcant association, in the univariate ( Figure 3) or multivariate analyses, between HALP score and DFS. Te only statistically signifcant variable associated with DFS in the multivariate analysis was cancer perforation (HR � 1.692, 95% CI � 0.986-2.902; p � 0.046). All variables' association with DFS can be reviewed in Table 3.
Biological and Clinical Signifcance of the HALP Index.
Te HALP index has been found to be a reliable marker for prognosis in cancer patients. It utilizes individuallyvalidated preoperative markers and computes them into a single value, able to identify patients with a higher risk, utilizing the following formula: hemoglobin (g/L) × albumin (g/L) × lymphocytes (/L)/platelets (/L) [13]. Patients with a low HALP score have been shown to have worse outcomes, both in previously published Asian-based studies and in the present, Hispanic-based, study.
Hemoglobin has been widely validated as a prognostic factor for disease progression and survival in a variety of human cancers. Cancer-related anemia (CRA) is present in over 30% of patients at diagnosis. CRA is a consequence of chronic infammation and poor nutritional status. Both tumor cells and tumor-reacting immune cells release proinfammatory cytokines that directly alter the hematopoietic environment and deteriorate nutritional status, further worsening anemia.
Patients with advanced cancer have a more intense and prolonged infammatory and nutritiously impoverished state. Consequently, hemoglobin levels serve as an indicator of the degree of disease progression [20]. Moreover, anemiagenerated hypoxia in the tumor microenvironment increases the tumor's proliferative and metastatic potential. Anemia has been directly linked to treatment resistance and aggressive disease [21].
Since anemia is an independent risk factor for perioperative morbidity and mortality, hemoglobin levels are a commonly assessed value in surgical patients [22]. In the context of colorectal tumor surgical resection, anemia has been well-validated as a prognostic factor associated with decreased OS and DFS [21]. Serum albumin is one of the most widely used markers to assess nutritional status [23]. Malnutrition in cancer patients derives from cancer's catabolic state and infammationinduced anorexia. It results in metabolic alterations and muscle wasting, which are directly linked to an increased risk of chemotherapy toxicity, postoperative complications, and death [24]. Concomitantly, albumin is a main negative acute phase reactant. Similarly to hemoglobin, its decay refects the systemic infammation produced by cancer [25]. Albumin is also known to stabilize cell growth, promote DNA repair and have antioxidant properties [26]. In vitro, albumin has been shown to suppress tumor proliferation [27]. However, this efect may vary in vivo depending on the stage the tumor is at, so further studies are needed to elucidate the direct efect of albumin on tumor progression. Clinically, albumin levels are associated with the development of surgical complications and have been well-defned as a long-term survival predictor [26]. Particularly, hypoalbuminemia is associated with wound healing and anastomotic complications [28].
Lymphocytes play an important role in the immune response toward malignant cells. Trough both humoral and cellular immune responses, lymphocytes are able to limit tumor growth and metastatic potential. Lymphocyte count and tumor-infltrating lymphocytes are thought to be indicators of the host's ability to mount an efective immune response to cancer [29]. In colorectal cancer, lymphopenia has been found to be associated with decreased chemotherapy efectiveness and lower overall survival [30,31].
In contrast with the aforementioned factors, low platelet count has been consistently related to a better prognosis. A cohort that included 112,231 cancer patients found that a high platelet count was associated with an increased rate of cancerspecifc death [32]. Platelets can release a variety of cytokines, growth factors, and proangiogenic molecules which directly induce tumorigenesis, cancer proliferation, and metastasis. In addition, platelets are able to engulf tumor cells, aiding in cancer immune evasion and further promoting uncontrollable growth and dissemination [33]. Colon cancer patients with elevated platelet counts have shorter OS and DFS [34].
Te HALP index efciently compiles the aforementioned marker values into a joint score, to bolster their independent prognostic efect. Tis way, a single and practical value is put forth with the aim of reaching the required predictive value for implementation in clinical practice.
Key
Results. In our study, the HALP index was found to be independently associated with overall survival. Having a low HALP score nearly doubled the hazard of death (HR � 1.942, 95% CI � 1.647-2.875; p � 0.031). Te HALP index represented the third strongest variable associated with OS, right after the well-defned and widely implemented variables: TNM stage (Stage II vs. Stage III) and surgical margin status (R0 vs. R1/R2).
We did not fnd a statistically signifcant association between the HALP index and DFS. Tis could be attributable to the fact that DFS is largely dependent on other variables that were not included in our analysis, such as receiving adequate adjuvant chemotherapy. It should be noted that a statistically signifcant association was not found between the TNM stage and DFS either, which gives us further reason to believe that other factors, which were not considered in our analysis, played an important role in determining DFS.
It should be borne in mind that the association identifed here between a high HALP score and longer OS suggests that the HALP index is able to predict long-term survival, despite recurrence.
Our sample size was fairly large and the included patients had clinical characteristics similar to those included in previous colon cancer studies undertaken in Mexico and Journal of Oncology 7 other Latin-American countries [35][36][37]. Surgical quality indicators, such as successful lymph node resection and microscopically negative surgical margins, were most often achieved. All this supports the generalizability of results to other Hispanic colon cancer populations where quality standards of treatment are met. Te optimal HALP index cut-of point for our population was lower than that reported in previous studies. Using the X-tile software, we determined that 15.0 was the optimal cut-of value, while previous study cut-of points ranged from 22.2 to 56.8 [13,15]. It should be noted that those studies were undertaken in Asian populations. Te diference in our cut-of point may be purely incidental, but it could also stem from biological and clinical diferences between Asian and Hispanic populations.
In order to validate our cut-of point of 15.0, we independently calculated the optimal cut-of point in two subgroups (patients who underwent elective surgery vs. those who underwent emergency surgery). In these calculations, we obtained nearly identical cut-of values (15.0 and 14.7, respectively), which confrms the validity of this cut-of point in our overall population. Moreover, even though the nutritional and infammatory status of patients undergoing elective and emergency surgery may difer signifcantly, our results support the adoption of the same cut-of point for both populations.
Previous studies have reported female gender [12,13,15,17] and older age [11][12][13]15] to be associated with lower HALP scores. Reasons for this association include the fact that female patients tend to have lower hemoglobin levels, while older patients tend to have reduced levels of both hemoglobin and albumin [13,38,39]. Nevertheless, this association was not seen in our population. In contrast, left colon cancers in our cohort were associated with lower HALP scores, an association that has not been reported elsewhere. Just as in previous reports, our study found the TNM stage to be associated with the HALP score [12,13,[15][16][17]. Tis is consistent with the belief that computed values for HALP calculation are correlated with the degree of disease progression.
Study Limitations.
Our study has some potential limitations. It was a single-center study undertaken in a thirdlevel cancer hospital in Mexico City, which might restrict, to some degree, the generalizability of results to other countries and populations. Additionally, the retrospective nature of this study increases the risk of selection bias. Further assessment of the HALP index with our cut-of point in diferent Hispanic-based populations should be undertaken.
Future Remarks.
Since the HALP index is constructed through modifable lab values, it is reasonable to believe that preoperative interventions could be undertaken in patients with a low HALP to improve outcomes. Eforts to improve some of the individual HALP-included values have already been undertaken [20,26]. Knowledge gathered on the prognostic value of the HALP index in our population might indicate that potential preoperative interventions could
Conclusion
Overall, our fndings suggest that the HALP index is a viable independent preoperative predictor of survival for Hispanic patients with TNM stages I-III who are undergoing primary tumor surgical resection. Tese study results give us reason to believe that the HALP index refects both the anatomical extent of the disease, measured through the TNM staging system, and traditionally unaccounted nutritional and infammatory factors that play a signifcant role in prognosis. Te wide availability of the routinary lab values needed to compute this score, along with the very practical nature of its implementation, make its adoption in clinical practice feasible. We provide a cut-of value that can be used in similar populations to assess prognosis and guide management.
Data Availability
All datasets used for the development of this work are available from the corresponding author upon reasonable request.
|
v2
|
2022-11-26T17:18:35.675Z
|
2022-11-23T00:00:00.000Z
|
253926115
|
s2orc/train
|
Prostate Multiparametric MRI: Evaluation of Recurrence and Post-treatment Changes
This article reviews all the most common therapeutic strategies of prostate cancer, systemic or local, and all the following morpho-structural alterations, with the aim of helping the radiologist to recognize the signs of recurrence by using mp-MRI. According to the most recent evidences, prostate mp-MRI has now become a strong, non-invasive, and valid tool to evaluate all patient treated for prostatic carcinoma across the time, especially in the suspicion of biochemical recurrence. The minimal signs of focal recurrence can put a strain on radiologists, especially if they are novice with multi-parametric prostate MRI. Familiarizing themselves with the outcomes of treatment, local or systemic, and its characteristics to MR imaging is indispensable to avoid diagnostic pitfalls and, subsequently, unnecessary reinterventions.
Introduction
In the vast field of cancer diseases that most frequently affect the male sex of Western Europe, prostate cancer (PCa) is undoubtedly one of the most numbered, representing one of the main causes of neoplastic death along with skin and lung cancers [1 •• ]. Although significant progress has been made in early detection as well as in the treatment of prostate cancer, the greatest challenge at the moment is the management and follow-up of patients following treatment, whether it be focal or radical. Specifically, the first-line therapeutic choice with curative intervention is represented by radical prostatectomy (RP), followed by radiotherapy treatment (RT), with external beam radiotherapy (EBRT) or with brachytherapy, as a viable alternative in patients with medium and low-risk PCa (below T3a in the TNM staging system) and, in addition, in patients not eligible for surgery by age or comorbidity [2]. In addition, in recent years, further therapeutic proposals focused mainly on focal ablation techniques (FA) such as laser therapy, cryotherapy or highintensity focused ultrasound (HIFU) have been developed [3]. Regardless of treatment, biochemical recurrence (BR) is not uncommon in patients treated for PCa and strongly influences their management and subsequent therapeutic choices [4]. Especially in the category of high-risk patients, who are subjected to RP, the probability of a BR occurring within the next 15 years after treatment amounts to about 50%; so high percentage is explained by the frequent need to preserve the vascular beam-nervous and the urethral sphincter during the surgical procedure [5]; no less is the probability of risk of BR after RT treatment, that is around 67% [6].
BR, also called biochemical failure (BF), does not necessarily indicate a recurrence of disease confined to the prostate bed and surrounding tissues, but may also refer to metastases at a distance, that is why attention should be paid to which parameters to assess for a correct distinction between the two conditions [4].
The first approach in the diagnosis of recurrence of prostatic disease is represented by the serum dosage of prostatic specific antigen (PSA) and the evaluation of its kinetics over time, specifically the PSA-Doubling Time (PSA-DT) and the PSA-velocity (PSA-VE), parameters closely related to the treatment to which the patient has been subjected. As a general rule, in patients undergoing RP, one would expect to find an undetectable PSA or PSA with values very close to zero (0.1 ng/ml) after 21-30 days of surgery, considering any increment, even minimum, a possible warning sign of persistence or recurrence of disease [4]. In contrast, serum levels of PSA that are low but still measurable with certainty after RT or FA, given the persistence of normally functioning residual glandular tissue, should be considered physiological [7].
According to the most recent evidence and in agreement with the European Urology Association (UAE) two consecutive values of PSA [ 0.2 ng/ml after RP are strongly suggestive of BR, but is not sufficient to distinguish between locoregional and systemic relapse [8]. Therefore, in order to ensure an appropriate therapeutic choice, an evaluation of serum PSA levels and their evolution over time will be necessary. Broadly speaking, reference PSA values have been identified for one condition or another and are given in Table 1 [9].
On the contrary, estimating the risk of BR following RT is a much more complex undertaking and the PSA itself has a poorly defined role in pursuing this aim, since it may happen that after radiation treatment serum PSA levels have a very slow descent or never reach undetectable values. This phenomenon obviously occurs due to the presence of residual and functioning prostatic tissue, which causes the achievement of the so-called ''PSA nadir'' after RT even 3 or more years later than the RP. According to the 2005 American Society of Therapeutic Radiology and Oncology's ''Phoenix Criteria'', two consecutive PSA values above 0.2 ng/ml are highly suggestive of relapse of disease after RT [10].
As mentioned above, the need to identify and differentiate with certainty a local recidivism from a systemic one, aims to choose with accuracy an appropriate treatment as tailored as possible on each patient. In general, according to the UAE guidelines, in cases of locoregional relapse, which therefore presupposes the presence of residual glandular tissue in the prostate bed, a treatment with RT defined ''rescue'', before the PSA levels exceed the value of 1 ng/ml, is indicated, keeping in mind, as a general rule, that the treatment will be able to have better effects the lower the initial PSA is: in particular, recent evidences show that the best results are obtained specifically when the PSA level is lower than 0.5 ng/ml [11].
Alternative therapies to RT and equally useful in the treatment of local relapse can be represented by RP or FA, whose choice is guided by an overall assessment of the patient's health status, his comorbidities and, not least, to the location and extent of residual tumor tissue.
On the contrary, the certain diagnosis of systemic disease renders completely useless the only RT as a first- choice treatment, which is therefore supplanted in most cases by androgen deprivation therapy (ADT). Regardless of the serum levels of PSA which are obviously one of the first alarm bells of possible disease's recovery, the decisive step in the definition of locoregional or systemic BF is made by using imaging. In the field of PCa imaging, mp-MRI is the master technique thanks to its excellent potential to provide excellent anatomical spatial resolution (with T1w and T2w sequences) as well as good functional information with dynamic sequences (DCE) and diffusion (DWI) [12].
Although in the mp-MRI of the BF the same imaging sequences are used, as those suggested by the PIRADS v 2.1 for the primitive PCa, it is necessary to take into account any anatomical changes, functional or simply signal that is likely to incur in the assessment of MR imaging depending on the type of treatment the patient has been subjected to (e.g., anatomical changes or simply ferromagnetic artifacts for the presence of metal clips after RP) [13].
The aim of this article is to review all the possible changes that the prostate undergoes as a result of therapy, radical or focal, as well as describing the mp-MR imaging characteristics of locoregional relapse and/or potential pitfalls in its interpretation.
mp-MRI in the Active Surveillance and the Precise Score
As mentioned in the introductory paragraph, mp-MR has been used for several years in many centers as a choice imaging technique in the diagnosis of PCa. The combined use of functional and anatomical sequences, in fact, increased the sensitivity and specificity of mp-MRI in the detection of clinically significant PCa, thus pushing the European Society of Urology and the American College of Radiology to develop guidelines on standardized acquisition and interpretation protocols known as ''Prostate Imaging-Reporting and Data System (PI-RADS)'' (latest version PI-RADS v2.1 released in 2019) [14].
In general, T2-weighted hinge sequences (T2w) have been identified for the anatomical study of the gland and male pelvis as a whole, together with a functional study with diffusion sequences (DWI), dynamics (DCE), and spectroscopy. Moreover, among these, preferential and more significant tailored sequences were chosen on the zonal anatomy of the prostate: the T2w sequences for the transition zone (TZ) and the DWI sequences for the peripheral zone (PZ) [15 • ].
In view of the above, it is clear that, in addition to diagnosis, mp-MRI is useful in the management of patients with locally developed and/or low-risk PCa and in those who undergone treatment with RP, RT, or FT. The aforementioned categories of patients, in fact, are subjected to active surveillance (AS) which has as its ultimate aim to reduce overtreatment and to determine how much a treatment is deferable, always remaining in the time window of curability.
In 2014, the National Institute for Health and Care Excellence (NICE) suggested that mp-MRI also acquired a role in the management of PCa, alongside or replacing prostate biopsy for restaging in patients with suspicion of recurrence [16].
Mp-MRI could clearly be considered a viable alternative to prostate biopsy during active surveillance, as it would prevent patients from undergoing invasive procedures on a regular basis or at least limit their frequency.
In 2016, a team of experts in the field of urological oncology and radiology elaborated recommendations on the standardized management of patients in AS with the mp-MRI, today known as the Prostate Cancer Radiological Estimation of Change in Sequential Evaluation (PRECISE) system and built on a 5-point scale ( Table 2); its clinical usefulness, however, requires official validation due to a still small number of literature data [17].
MRI After Radical Prostatectomy
RP is the primary therapeutic choice for healthy patients younger than 70, with PCa confined to the prostate gland. Post-surgical recurrence is not an infrequent phenomenon and, for this reason, the early detection of a BR becomes, therefore, of vital importance to undertake as soon as possible a second-level therapeutic process (e.g., radiation therapy with external beams with or without androgen deprivation therapy). In a study conducted on 84 patients subjected to RP, Panebianco et al. [18] have shown that mp-MRI has a much higher diagnostic accuracy than the most common methods used (PET/TC in particular) in the diagnosis of BR, reporting a sensitivity of 92% (versus 62%) and a specificity of 75% (versus 50%). Compared to detection of in situ PCa, BR often appears either a hypointense or as a hyperintense nodular formation in T2w sequences, compared to pelvic muscles, located near the vesicourethral anastomosis (VUA), typically a low signal strength area related to post-surgical fibrosis. The dynamic sequences show a typical nodular-shape contrastenhancement (with an Is/T ''wash-in/wash-out'' curve of type III) near the VUA, very suggestive finding of local recidivism ( Fig. 1) and much more significant than the classic hyperintensity of restricted signal in DWI (with relative low intensity in the ADC-map), an mp-MRI sequence strongly susceptible to artifacts from metal clips [19].
Although RP is by definition the complete removal of the prostate and seminal vesicles, mp-MRI is useful in this sense because it can differentiate the local recurrence from possible glandular residues, as well as from inflammatory processes. The presence of a glandular residue may appear as a hyperintense nodule in the T2w anatomical sequences, making it difficult to distinguish with relapsing PCa which could appear as a slightly hyperintense nodule as well, using pelvic muscles as a benchmark for signal strength. Functional sequences are used for this purpose, since they allow to lean toward a radiological diagnosis of benignity in the presence of a lack of signal restriction in DWI/ADC and a focal, slow, and progressive contrast-enhancement (Is/T curve of type I-II) or a ''mute'' DCE sequence (Fig. 2).
Always within the scope of post-surgical residues, even seminal vesicles are to be considered a confusing element in the mp-MRI evaluation of the prostate bed; their tubular conformation, however, makes them more easily identifiable with T2w sequences yet, in which they show a typical hyperintense signal due to their fluid content [20,21]. Another considerable challenge is the possible presence of granulation tissue which, like fibrosis, typically occurs in the vicinity of VUA but, contrary to it, shows signal hyperintensity in T2w and a high perfusion given its hypervascularity, like a BR; on the contrary, its signal restriction characteristics (no DWI signal) are more typical of a benign finding. b Fig. 1 A 65-year-old man who presented two consecutive serum PSA levels above 4.3 ng/ml, 3 years after radical prostatectomy (RP) and candidate for mp-MRI evaluation in the suspicion of biochemical recurrence (BR). Axial T2w (A) shows a hypointense nodule (red arrow), which can be confused with post-surgical fibrosis located at the left side of vescico-urethral anastomosis (VUA). The restricted signal on DWI (B) and the corresponding ADC-map (C) increase the suspicion of a local recurrence of disease, which is largely confirmed by dynamic sequences after contrast medium injection (DCE), which show a hyperintense signal on the corresponding area (D) with type II-III Is/T curve (E), suggesting an hyper-vascular behavior (Color figure online) In general, in the light of the above, each mp-MRI sequence has its usefulness in post-RP imaging and it is equally clear that dynamic sequences, with intravenous contrast medium administration, are the key point to the discrimination of the various pathogenic noxae that can mimic a BR: a ''mute'' contrast-enhancement in the arterial phase and progressively intense during the acquisitions in the venous phase is defined as a finding of normality; on the contrary, a slight change or even a reversal of the contrast medium dynamics of the post-operative prostatic bed is a highly suggestive alarm bell of locoregional recurrence [3].
In order to support the aforementioned, several studies highlight the usefulness of DCE sequences after RP. Among these, an analysis of 46 patients conducted by Casciani et al. [22] which showed an increase in sensitivity from 48 to 88% and specificity from 52 to 100% given by the choice to add dynamic sequences to T2w alone. Or, moreover, Cyril et al. [23] which reported similar results on a cohort of 72 patients with specificity values of about 89.3% (versus 82.1%) and sensitivity of 84% (versus 61.4%).
MRI After Radiation Therapy
Radiation therapy (RT) is currently the second therapeutic option for medium-and low-risk PCa (stages I-III) after RP and, for obvious reasons, does not make the patient totally immune from recurrence of locoregional disease and with an average time of onset of distance metastasis of about 3 years [24].
Used in about 40% of patients over 65 and in about 25% of those under 65 years old, RT can be used both with external beams (EBRT), generally in the earliest stages of the disease, either in a more targeted form with brachytherapy.
Because of all the morpho-structural changes that the prostate face as a result of RT, regardless of whether it is EBRT or brachytherapy, the imaging method of choice for a more sensitive detection and localization of relapsing PCa is mp-MRI.
In the evaluation of pelvic MR imaging after RT, it is important not only to keep in mind the changes that directly affect the PCa (reduction of its own volume, reduction of capsular bulging and/or ECE), but also glandular atrophy and fibrotic changes mainly responsible for morphological/anatomical variations and signal strength [25].
The post-RT changes, moreover, affect the anatomical structures placed in close proximity of the prostate gland and can be visible both in the form of morpho-structural alterations, such as the wrinkled appearance of the seminal Fig. 3 A follow-up MRI in an 83-year-old man treated for PCa with radical prostatectomy (RP) and radiation therapy (RT), who presented a serum PSA level increase. This is a practical example of how the functional sequences of mp-MRI are of vital importance in distinguishing the effects of the RT from the signs of local recurrence: the restricted signal in DWI (hyperintense focus in A) and in the corresponding ADC-map (hypointense focus in B), in addition to the early contrast-enhancement, demonstrated by high signal intensity of the lesion after contrast media injection (C), confirm the presence of a recurrent focal disease vesicles, or the thickening of the peri-rectal fascia, bladder, and rectal wall, either as variations in signal intensity (e.g., adipose bone marrow replacement in skeletal segments of the pelvis, visible as increased signal in T1w and hypointensity in T2w) [26].
It is important to keep in mind that the variations, both anatomical and signal, which follow the RT are different depending on the mode of delivery of the radiation and, therefore, it is necessary to assess specifically the consequences of EBRT and brachytherapy separately.
External Beam Radiation Therapy (EBRT)
The imaging evaluation after EBRT represents a real challenge for the radiology, especially considering all the post-RT changes that involve the prostate gland in its entirety. The irradiated glandular tissue, in fact, undergoes fibrotic changes and a significant reduction in volume, which result on the one hand in an overall reduction of the signal intensity in the T2w sequences, on the other hand in a loss of a clear distinction of the zonal anatomy of the gland. The widespread hypo-intensity of the gland in T2w, in particular, diminishes the usefulness of this sequence in the identification of a recurrent PCa, making it difficult to differentiate between healthy and neoplastic tissue.
The recurrent PCa, similar to a radio-treated neoplastic nodule, in the T2w images is presented as a focal hypointensity nodular morphology, difficult to distinguish from the surrounding glandular tissue, but that may nevertheless show indirect signs related to its localization within the prostate, such as capsular bulging derived from rather rapid tumor growth. Knowledge of the previous location of the tumor is also a factor not to be underestimated in the search for cancer recurrence, since only a very low percentage of prostatic cancer, ranging between 4 and 9%, occurs in a different location from the primitive [27].
Despite this, it is clear that T2w sequences alone are not sufficient to express themselves about the presence or not of local recurrence. Westphalen et al. [28], in a study conducted on a cohort of 64 patients, evaluated the usefulness of T2w sequences in the certain identification of recurrence of PCa, obtaining an area under the curve (AUC) only of 67%. Almost similar results emerged from another analysis carried out by Sala et al. [29] on a cohort of 45 patients diagnosed with BR, where T2w sequences showed a sensitivity between 36 and 75% and a specificity between 65 and 81%. Therefore, these data largely highlight the limitations of the exclusive use of T2w sequences.
In this scenario come into play the functional sequences DCE and DWI that, however, are not entirely free from RT-induced alterations, linked to reduced vascularity and cellularity of the fibrotic gland. Although present, morphostructural and signal alterations of dynamic sequences do not appear to be as marked as in T2w, especially if we keep in mind that the continuous angiogenesis linked to the tumor hesitates in a perfusion activity visible in DCE sequences in contrast to the surrounding fibrotic tissue [30].
Remaining within the scope of the MR functional prostate imaging after EBRT, DWI sequences also deserve mention, since neoplastic relapse seems to have signal characteristics far similar to those of the primitive PCa, with evidence of diffusiveness restriction that results in a typical signal hyperintensity in DWI sequences with high b-values ([ 1000) and a corresponding hypo-intensity in the ADC-map (Fig. 3). Several studies analyzed the usefulness of DWI used in combination with T2w sequences, compared to the latter alone, demonstrating a patient-based sensitivity and specificity for DWI of 100%, with regionbased accuracy of 89% [32].
Apparently, in the evaluation of patients undergoing EBRT, there is no functional sequence that shows greater utility than another; however, there is evidence that the use of both DWI and DCE sequences greatly improves the diagnostic performance of MR imaging, especially in relation to the exclusive use of basic sequences such as T2w [33].
Brachytherapy
Although the alterations which the prostate undergoes after brachytherapy are almost analogous to those induced by EBRT, some peculiar aspects of this type of RT should not be underestimated, primarily related to the use of specific radioactive seeds contained in metal capsules which are responsible for some ferromagnetic artifacts, due to their intrinsic characteristics, that can distort the appearance of the gland and signal intensity especially in DWI sequences [34].
In the course of brachytherapy, as occurs in the EBRT, the prostate undergoes a progressive volumetric reduction in conjunction with the onset of fibrosis of the glandular parenchyma, resulting in widespread signal hypo-intensity in T2w sequences, in which context radioactive seeds appear as small hypointense ellipsoid bodies increasingly placed at the periphery of the atrophic gland [35].
In the presence of a confirmed BR, the nodule of PCa appears as a hypointense focus in the T2w sequences with characteristic intense and rapid contrast-enhancement in dynamic sequences and, limited to the presence of magnetic artifacts, restricted signal (hyperintense) in DWI at high b-values ([ 1000). This latter aspect is the only reason why DCE sequences should be considered as pivotal sequences in the evaluation of post-brachytherapy recurrence, rather than DCE and DWI together [34].
MRI After Focal Therapy
In the field of targeted locoregional treatments there is a new emerging method known as focal therapy (FT) that brings together a series of different procedures depending on the type of energy used to attack the neoplastic lesion: cryotherapy, laser ablation focal (FLA), and high-intensity focused ultrasound (HIFU) [2].
The mp-MRI, with morphological and functional sequences, is now an excellent diagnostic method both in the assessment of the response to the treatment itself, along with biopsy and serum PSA, and in the long-term follow-up. With regard to this last aspect, in fact, it is recommended to perform an mp-MRI in the first week immediately following treatment, in order to capture the early findings and compare them with those of imaging of the next 5 years, in accordance with surveillance protocols [36].
Regardless of the type of focal therapy that the patient with PCa undergoes, in order to not incur errors of interpretation in the assessment of MR imaging, it is good to know not only the effects of FT on the tumor itself and the prostate gland as a whole, but also the treatment location, pre-procedure imaging and, last but not least, the type of energy used for FT [37].
In general, all FT treatments induce atrophy, volume reduction and reduced or absent perfusion of the treated area, visible on MRI as a focal hypo-intensity in T2w sequences and low signal in DWI, with a variable signal in ADC and DCE [38]. The effects on the tumor and the consequent appearance of MR imaging depend, in addition to the time elapsed, mainly on the type of FT used (Table 3).
At present there is still small evidence of the certain appearance of recurrent PCa in an area of prostate treated with FT, which suggests, therefore, the importance of knowing the effects of different physical energies on cancer tissue and, at the same time, the utility of integrating as always morphological and functional imaging sequences (Fig. 4).
MRI Post-hormonal Therapy
In the therapeutic protocols currently in use, patients with PCa, especially if metastatic and/or symptomatic, are subjected to hormone therapy (HT) or androgenic deprivation (ADT) with the aim of inducing apoptosis of prostate cells.
ADT exerts its action either through a defined mechanism of ''castration'', by direct inhibition of testicular secretion of androgens, or properly through an ''anti-androgenic'' action by inhibition on peripheral receptors [39,40]. It is not unusual that patients treated with ADT, both for locally advanced PCa and metastatic, are subject to rapidly progressive BR with ECE and/or seminal vesicle invasion.
Like the treatments described in the previous paragraphs, HT is also responsible for morpho-structural alterations of the prostate and possible errors of diagnostic interpretation in MR imaging. Typically, the gland will experience an overall reduction of its volume, especially if Fig. 4 A 68-year-old patient who was evaluated at our institution after focal laser therapy (FLA). The use of mp-MRI, in particular functional sequences, has highlighted the absence of loco-regional recurrence due to the lack of signal restriction in DWI (A) and the corresponding ADC-map (B) both showing low intensity of the treated area (red arrow in B) (Color figure online) Table 3 The most used types of focal treatment (FT) for PCa and their principal effects on the treated area Cryotherapy Responsible for coagulative necrosis on a larger area than the focal neoplasia, which than appears changed in architecture and hypointense on T2w sequences, with a lower perfusion detected on DCE sequences HIFU It uses focused ultrasound in order to induce heating of the treatment area, resulting in a T2w hypo-intensity and still a slight peripheral perfusion on DCE sequences FLA It determines a heterogeneous T2 hypo-intensity, but with a restricted diffusion on DWI resulting in a high signal intensity the drug therapy is associated with RT, with evidence of reduction of the signal intensity in T2w sequences, mainly visible in the peripheral area (PZ), and increase of the ADC signal (Fig. 5). Functional sequences, on the other hand, will be indicative of reduced overall vascularity showing reduced perfusion of the gland in DCE sequences [41].
Conclusion
The role of MRI imaging in the study of prostate and PCa is not limited to detection and staging of the tumor itself, but also it extends to the periodic follow-up to patients following treatment. The percentage of recurrence of PCa, unfortunately, is not irrelevant in the spectrum of neoplastic prostate disease and, given the existence of a BR through clinical-laboratory data, it is important to differentiate the local recurrence from the systemic one and, to this end, mp-MRI plays a decisive role. Although the imaging acquisition and processing protocol does not differ at all from the classic one used for the routine study of the prostate, it is important for the radiologist to become familiar with all the changes that the gland and neighboring structures undergo, both as a result of focal and systemic therapies.
This work offers the basic tools for the radiologist not to fall into the trap of diagnostic pitfalls, describing the effects of each type of treatment on the prostate gland and listing the most frequent findings deriving from them.
Funding Open access funding provided by Università di Foggia within the CRUI-CARE Agreement. No funds, grants, or other support was received.
Declarations
Conflict of interest The authors have no relevant financial or nonfinancial interests to disclose.
Research Involving Human and Animal Participants This article does not contain any studies with human or animal subjects performed by any of the authors.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.
|
v2
|
2022-12-14T16:17:39.538Z
|
2022-11-23T00:00:00.000Z
|
254615961
|
s2ag/train
|
Synthesis and In-silico Study of Novel 1,3,4-Oxadiazole Derivatives : A Biologically active Scaffolds which induce Anti-tubercular activity by targeting Pteridine Reductase and Dihydrofolate Reductase
Heterocyclic compounds possess diverse biological properties that have led to intense study and research of these compounds. One of these compounds is Oxadiazole which has been found to exhibit various pharmacological activities. 1,3,4-oxadiazole having heterocyclic nucleus is a novel molecule which attract the chemist to search a new therapeutic molecule. Research on 1,3,4-oxadiazole and their synthetic analogues have revealed a variety of pharmacological activities including anti-microbial, anti-tubercular and insecticidal agents. Some of these compounds have also analgesic, anti-inflammatory, anti-cancer, anti-HIV agent, anti-parkinsonian and anti-proliferative agent. It was our interested to make novel derivatives of the titled compounds and evaluate the anti-tubercular activities. 1,3,4-oxadiazole and its derivatives (4a-4e) were obtained. The current study discusses the microwave irradiation synthesis of derivatives with the goal of generating new medications with high specificity for mycobacterium tuberculosis and low harm to the human.
|
v2
|
2022-11-24T14:06:32.879Z
|
2022-11-24T00:00:00.000Z
|
253841554
|
s2orc/train
|
Prevention and treatment of human papillomavirus in men benefits both men and women
Men should not be overlooked in research on human papillomavirus (HPV) and its associated genital diseases. This is because men infected with HPV are not only at higher risk of genital cancers, but also increase their partners’ risk of HPV infection and reinfection through sexual contact. Herein, we summarized the state of knowledge regarding the prevention and treatment of HPV infection in men as well as the possible effects of the prevention and treatment of HPV in men on their female partners. Condom use, smoking cessation, male circumcision, and HPV vaccination for men each play an important role in preventing HPV infection within heterosexual couples. Additionally, men could choose to test for certain types of HPV, such as the oncogenic HPV16 or HPV18 strains, as part of a routine screening program when their partner is positive for HPV. Although there is no recognized treatment for HPV infection as of yet, immunotherapy drugs, such as toll-like receptor agonists, therapeutic HPV vaccines, and immune checkpoint inhibitors, have shown promising results in clinical trials and in actual clinical practice. HPV infection in men also increases the risk of cervical cancer in their female partners. Because of the high partner concordance for HPV demonstrated in prior research, the prevention and treatment of HPV in men should be explored more comprehensively in future research.
Introduction
Human papillomavirus (HPV) is the most common reproductive tract virus. HPV belongs to the Papillomaviridae family. Every year, HPV causes 630,000 cancer cases in men and women (i.e., 4.5 percent of all cancer cases), thereby posing a serious threat to public health on a global scale. In 2012, HPV-related malignancies accounted for 8.6% of Abbreviations: HPV, human papillomavirus; hr-HPV, high-risk human papillomavirus; lr-HPV, low-risk human papillomavirus; MC, male circumcision; ICI, immune checkpoints inhibitor. (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. all cancer cases in women and 0.8% of all cancer cases in men (Serrano et al., 2018). The Martel study, based on the data from 2012, was updated to report that cervical cancer accounted for 83% of HPV-attributable cancers, while other HPV-attributable cancers were head and neck cancer, anus cancer (half for each sex), penile cancer, vaginal cancer, and vulvar cancer (de Martel et al., 2017). Moreover, HPV is one of the most common causes of sexually transmitted diseases in sexually active women and men, causing the proliferation of scaly epithelium on the mucous membranes of human skin (Burd, 2003). At present, more than 200 HPV subtypes have been identified, of which more than 85 types have been identified in the human body (Burd, 2003). Approximately, 40 species can be transmitted to the anogenital organs and surrounding skin through sexual activity .
HPV is divided into broad high-risk (hr-HPV) and low-risk types (lr-HPV) according to carcinogenicity. Infection with hr-HPV can cause head and neck cancer and oropharyngeal cancer in addition to anogenital cancer. In contrast, benign skin lesions, such as genital warts, are generally attributed to lr-HPV (Arbyn et al., 2012). Although most HPV infections and their related precancerous lesions can resolve spontaneously, HPV infection and its sequelae is still an important cause of cancers of the cervix, vulva, vagina, penis, prostate, and anus (Giuliano et al., 2015). HPV infections can also influence sperm status, potentially leading to infertility (Foresta et al., 2010). The prevention and treatment of HPV-related diseases remains a major focus in the field of medicine.
Recent studies suggest that men can take a series of measures to reduce the incidence of certain HPV types, for example, men should be vaccinated against HPV (this is especially true for adolescents) (Petrosky et al., 2015;Schmeler and Sturgis, 2016) and circumcision should be performed for males (Smith et al., 2021). However, current prevention and treatment protocols in regard to HPV infections focus only on femalespecific HPV-related diseases, especially cervical cancer (which ranks fourth among female-specific diseases worldwide). There is presently sparse literature and few clinical compasses that recommend prevention, screening, and treatment for HPV infections in men.
At present, the joint prevention and treatment of HPV in men and women (which is especially critical for sexual partners) is not at the core of the screening and treatment protocols that are currently in place. However, cross-infection between couples results in persistent HPV infections more easily than in other situations, thereby elevating the risk of developing high-grade cervical lesions and eventually cervical cancer (de Lima Rocha et al., 2012). Hence, the aim of this paper is to provide an overview of the current understanding of the prevention and treatment of HPV for men and new ideas for heterosexual couples' joint health.
Prevention of HPV Condom use
Sexual contact is the primary transmission route for HPV. Males act as both virus carriers and vectors and this is an important component of the epidemiological chain for HPV . Further, women are more likely to transmit HPV to their male partners than men are to transmit HPV to their female partners (Malagon et al., 2021).
The fact that cross-infection occurs between members of a couple should not be ignored, as this is one of the reasons for the poor control of HPV evident in the literature and in clinical practice. Condoms are effective in physically isolating HPV infection, and men who do not use condoms have higher rates of HPV infection (Vardas et al., 2011). According to finding by , the number of condoms used in the previous three months is linked to a lower prevalence of HPV. One cross-sectional analysis covering three countries suggested that consistent condom use is an important factor in the low detection rate of any HPV type, any oncogenic type, and multiple types (Repp et al., 2012). Another cross-sectional study of 393 men showed that regular condom use during sexual intercourse is correlated with a reduction in oncogenic and overall HPV risk, which is similar to the results of the former study (Baldwin et al., 2004). Therefore, the use of condoms during sexual intercourse is essential to preventing HPV transmission and infection.
Smoking cessation
Smoking is a known independent risk factor for HPV infection. Schabath and colleagues successively demonstrated that current smoking overall as well as current smoking with a history of greater than five pack-years of smoking was associated with a higher incidence of HPV infection (especially oncogenic infection) and a lower probability of infection clearance in men (Schabath et al., 2012;Schabath et al., 2014). Researchers have also found that smoking 10 or more cigarettes per day was associated with HPV infection in men . For women, the prevalence of HPV was found to be 40.8% for smokers versus 25.2% for nonsmokers; the corresponding values for men were 68.2% versus 63.2% (Kaderli et al., 2014). This suggests that men with a history of smoking are more likely to be infected with HPV than women with a history of smoking. Therefore, timely cessation of smoking in men may play an especially critical role in HPV prevention.
Male circumcision
Numerous studies have demonstrated that male circumcision (MC) is effective in reducing the incidence of multiple HPV infection strains in men (Castellsague et al., 2002;Svare et al., 2002;Gray et al., 2010;Smith et al., 2021), thereby also decreasing the incidence of HPV-related diseases. In Baldwin and colleagues' cross-sectional study, it was suggested that circumcision reduces the risk of overall HPV in addition to oncogenic and non-oncogenic HPV, respectively (Baldwin et al., 2004). Men who have been circumcised may be less likely to allow viral invasion through epithelial abrasions, subsequent viral shedding, and viral persistence. Thus, circumcision is also the most effective factor in reducing the clearance of oncogenic and any HPV infection (Lu et al., 2009). Similar to HPV vaccines, not only can MC help men avoid acquiring certain genital diseases, but also is beneficial to women's health. Men who undergo MC reduce the risk of HPV infection to their female sexual partners (Morris et al., 2019). Moreover, the presence of foreskin in a woman's sexual partner is considered a risk factor for cervical cancer (Agarwal et al., 1993). Evidence is emerging that MC can meaningfully reduce the prevalence of cervical cancer in female partners within heterosexual couples (Castellsague et al., 2002;Svare et al., 2002;Morris et al., 2019). We draw the conclusion that MC should be included as a primary preventive measure for cervical cancer, penile cancer, and other HPV-related cancers within updated medical guidelines.
HPV vaccines
To date, 107 countries have introduced HPV vaccination programs, among which developed countries (led by Australia) have high vaccine implementation coverage; in contrast, in middle and low-income developing countries, the scale of HPV vaccine introduction has not yet been satisfactory (Bruni et al., 2021). It is thought that the HPV vaccine plays an indispensable role in preventing cervical cancer in women. However, HPV vaccination is not just an issue for women. Men are also exposed to the possibility of developing various diseases as sequalae of HPV infection. Moreover, high HPV vaccine coverage in men strongly benefits women by reducing the risk of cervical cancer (Lehtinen et al., 2018). Therefore, the HPV vaccine is undoubtedly equally needed for men and women.
A growing number of studies on gender-neutral HPV vaccination have emerged, and investigators have suggested that HPV16 eradication in the general population is predicted when 75% coverage of early adolescents (both boys and girls) is achieved (Lehtinen et al., 2019;Vanska et al., 2020). We additionally note that the herd effect refers to the indirect protective effect of vaccination on the unvaccinated population by reducing infection transmission within the susceptible population. A community-randomized trial previously reported that gender-neutral vaccination of early adolescents produced a striking population effect, substantially increasing the protective impact of vaccination on women's health (Lehtinen et al., 2018).
Nevertheless, barriers still exist to achieving prevalent HPV vaccination in men, including but not limited to a lack of knowledge regarding HPV, prejudices against the vaccine, various sociodemographic and religious factors, fear of side effects, and concerns about cost (Grandahl and Neveus, 2021). Consequently, it is necessary to enrich knowledge regarding HPV-related diseases in the general population so that more people are willing to get vaccinated. Simultaneously, the existing healthcare system should be modified with respect to making HPV vaccines more accessible.
There are three types of HPV vaccines: a bivalent HPV vaccine, four-valent HPV vaccine, and nine-valent HPV vaccine. The bivalent vaccine protects against HPV16 and HPV18; the four-valent vaccine protects against HPV6, HPV11, HPV16, and HPV18; and the nine-valent vaccine protects against HPV6, HPV11, HPV16, HPV18, HPV31, HPV33, HPV45, HPV52, and HPV58. These three types of vaccines are all effective against HPV16 and HPV18, including HPV-related cancers, as it can be prevented effectively through the use of vaccination campaigns, as the majority of these cancers are caused by HPV16.
According to the recommendation of the Advisory Committee on Immunization Practices (Petrosky et al., 2015; Oshman and Davis, 2020), females aged 11 or 12 years should be routinely vaccinated with bivalent, four-valent, or nine-valent HPV vaccines, while males of the same age should be vaccinated with four-valent or bivalent HPV vaccine. The Advisory Committee on Immunization Practices also recommended vaccination for females and males aged 13-26 years who have not received the HPV vaccine or who have not completed the required three doses. We note that each type of vaccine is administered in a three-dose schedule. According to current recommendations, the second dose should be administered 1-2 months after the first dose and the third dose should be administered 6 months after the first dose.
Detection of HPV
As most HPV infections clear spontaneously without intervention, a positive result does not indicate the need for immediate treatment of the patient or his or her sexual partners. Nevertheless, asymptomatic HPV infection in men is thought to be an important cause of ongoing transmission to female partners, and HPV infection in men increases the risk of cervical cancer in women (Barrasso et al., 1987). HPV infection also poses a risk for genital warts, penile cancer, and anal cancer in men. Therefore, HPV screening is necessary for men. Currently, however, only standardized HPV screening for women is emphasized, and there are no routine HPV screening programs in place for men. To our knowledge, data on the most reliable sampling site, the standardized sampling method, and the quality of sampling for men are not currently available. Some researchers suggest that, in men, samples collected from the external genital region yield a higher detection rate for oncogenic HPV than samples collected in the anal region, and the penile shaft is recommended as the optimal anatomical site for HPV detection Giuliano et al., 2007). Moreover, testing for HPV DNA appears to be the best strategy for detecting HPV infection in males, as revealed by Nicolau and colleagues. Brush material obtained from the distal urethra as well as from the external surface of the penis tends to be the most effective approach to diagnosing HPV infection in men (Nicolau et al., 2005). Targeted screening for certain types of HPV may also be used as a testing tool for HPV detection; for example, E6 seropositivity for HPV16 has been used as a prognostic and surveillance tool for oral cancer (Holzinger et al., 2017).
HPV testing for men should remain a focus within future research and clinical endeavors. We recommend several models herein: 1) men could choose to test for HPV when their female partner is positive for HPV (especially for those whose partners are positive for hr-HPV); 2) HPV screening programs for men should be developed as part of an easy and routine program; and 3) certain types of HPV infection, such as HPV16 and HPV18, should be highlighted because these infection strains are linked to the majority of malignancies of the penis, anus, and head and neck as compared with other HPV strains.
Treatment of HPV-related lesions
Current treatment is focused on addressing individual HPVrelated lesions, such as cervical cancer, vulvovaginal cancer, and penile cancer. Surgery, radiotherapy, chemotherapy, and targeted therapy are widely used in clinical practice. There is presently no standardized treatment for HPV infection only.
However, numerous studies regarding immunotherapy for HPV infection, a treatment modality that aims to achieve therapeutic goals by restoring local immune cell function, have been emerging as of now. The specific mainstream immunotherapy approaches currently under evaluation are described later in this report.
Toll-like receptor agonists
It is well known that toll-like receptors activate innate immunity by activating downstream signaling pathways through the recognition of pathogen associated molecular p a t t e r n s , t h e r e b y s t i m u l a t i n g t h e p r o d u c t i o n o f proinflammatory cytokines and type I interferons (Mifsud et al., 2014;Owen et al., 2020). In summary, toll-like receptor agonists stimulate the body's innate immune system and enhance innate immune function to clear pathogens and protect the body from infection (Mifsud et al., 2014). Imiquimod, a typical toll-like receptor agonist, has been increasingly used in the treatment of HPV-associated intraepithelial neoplasia and squamous cell carcinoma in situ of the penis as an alternative treatment option to surgery, with fewer adverse events and tumor recurrence (Schroeder and Sengelmann, 2002;Tristram et al., 2014).
Therapeutic HPV vaccines
Among the HPV proteins, E6 and E7 (proteins involved in tumorigenesis and progression) are considered ideal targets for cervical cancer immunotherapy (Pal and Kundu, 2019). Therapeutic HPV vaccines targeting E6 and E7 proteins have therefore been proposed; these vaccines are capable of enhancing the T cell immune response (Garbuglia et al., 2020). Live-vectorbased, peptide-based, protein-based, dendritic cell-based, and genetic vaccines each have great advantages as well as demonstrated effectiveness in the treatment of HPVrelated diseases (Chandra et al., 2021). Although no vaccine has presently been approved for clinical use, the therapeutic HPV vaccine has a promising future as an effective treatment strategy.
Immune checkpoint inhibitors
Immune checkpoints and their ligands have been found to be constantly upregulated in the tumor microenvironment of diverse malignancies, representing a major obstacle in the initiation of the body's effective innate anti-tumor immune response (Toor et al., 2020). Immune checkpoints inhibitors (ICIs) have become a popular research target in the field of tumor immunotherapy and are currently the main therapeutic strategy employed in immunotherapy. ICIs induce the blockade of programmed cell death protein 1, programmed death-ligand 1, and cytotoxic T-lymphocyteassociated protein 4. Ipilimumab (targeting cytotoxic Tlymphocyte-associated protein 4), and nivolumab and pembrolizumab (both targeting programmed cell death protein 1) are presently licensed for marketing. Moreover, programmed cell death protein 1 inhibitors have entered clinical trials with respect to the treatment of advanced cervical cancer following HPV infection (Chung et al., 2019;Naumann et al., 2019). Still, the clinical efficacy of any particular ICI alone is limited; according to current findings, ICIs should instead be combined with other therapeutic modalities to improve treatment outcomes. In summary, more convincing clinical trials proving the efficacy of this treatment modality are needed to address the current difficulties in the application of immunotherapy in HPVrelated diseases.
Necessity of HPV treatment for men
Reciprocal transmission of HPV is prevalent between couples. In a previous study conducted by Burchell and colleagues that evaluated couples in which both partners were positive for any type of HPV, 87% were found to be concordant for at least one type of HPV strain (Burchell et al., 2010). Compelling evidence from a well-designed cross-sectional study conducted in 2014 showed that 68% of the evaluated couples in which both partners were positive for HPV had at least one genotype (i.e., strain) in common; moreover, if male partners were positive for at least one HPV genotype, this had a substantial impact on their female partners' positivity status (de Lima Rocha et al., 2012). Moreover, Bleeker et al. demonstrated a high concordance of partner HPV types, and presented findings that this concordance may be associated with an increased viral load (Bleeker et al., 2005).
Among heterosexual partners, female patients diagnosed with cervical intraepithelial neoplasia have been shown to increase the risk of HPV infection in their sexual partners (Martin-Ezquerra et al., 2012). Although the majority of infections are cleared through enhanced immune function, there is still a risk of progression to severe disease among those infected with HPV.
Discussion
Men are an important component of the cycle of the transmission of sexually transmitted diseases, and that HPVpositive men are also responsible for their female partners' reinfection status (Skoulakis et al., 2019). As carriers and vectors of hr-HPVs, male partners may also cause a significant impact on the development of cervical cancer in their female partners (Skegg et al., 1982;Barrasso et al., 1987;Bosch et al., 1996). This reminds us that HPV-specific treatment and preventive medicine efforts should not only be targeted to women, but also toward men with HPV infections.
Hence, in addition to various screening and preventive efforts, the co-treatment of HPV for male and female partners is extremely crucial to slow down HPV transmission. Moreover, partner co-therapy for HPV infection should gradually be included within clinical treatment practices and formal medical guidelines informing clinical decision-making.
Conclusion
HPV infection in men and the effects of HPV infection on their partners are increasingly being emphasized in the medical literature, in ongoing epidemiologic and clinical research efforts, and in clinical practice. Since sexual transmission is the main route of HPV infection, condoms can help prevent HPV transmission by physically isolating contact with the mucous membranes of the skin. Smoking has also been recognized as an independent factor contributing to HPV infection, and quitting smoking is one of the known measures for preventing HPV. Moreover, studies have successively proven that MC and male vaccination exert a crucial effect on preventing HPV in female sexual partners.
We also note that, on the one hand, HPV infection in men may be a potential risk factor for cervical cancer in women. On the other hand, the high concordance of HPV types in partners suggests that cross-infection is a barrier to HPV control. Therefore, in summary, the treatment of HPV infection in men should undoubtedly be addressed equally as that for women.
Early intervention in men can protect against the transmission of HPV infection, while also reducing the incidence of cervical cancer in female partners and facilitating the treatment of female patients with HPV. However, there is still much to learn about the prevention and treatment of HPV and its related malignant diseases. We strongly recommend that this disparity be investigated in male patients as well in female patients, and that this should become a key focus and hotspot for future research.
Author contributions
KZ contributed to the manuscript drafting and final approval. YH contributed to manuscript revising and critical discussion. ZL provided practical suggestions and critically revised the manuscript. All authors have read and approved the final manuscript. All authors contributed to the article and approved the submitted version.
Funding
This study was supported by the Medical Science and Technology Project of Sichuan Provincial Health Commission (No. 21PJ050).
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
|
v2
|
2022-11-24T14:16:11.350Z
|
2022-11-24T00:00:00.000Z
|
253840899
|
s2orc/train
|
Exposure to endosulfan can cause long term effects on general biology, including the reproductive system of mice
Increased infertility in humans is attributed to the increased use of environmental chemicals in the last several decades. Various studies have identified pesticides as one of the causes of reproductive toxicity. In a previous study, infertility was observed in male mice due to testicular atrophy and decreased sperm count when a sublethal dose of endosulfan (3 mg/kg) with a serum concentration of 23 μg/L was used. However, the serum concentration of endosulfan was much higher (up to 500 μg/L) in people living in endosulfan-exposed areas compared to the one used in the investigation. To mimic the situation in an experimental setup, mice were exposed to 5 mg/kg body weight of endosulfan, and reproductive toxicity and long-term impact on the general biology of animals were examined. HPLC analysis revealed a serum concentration of ∼50 μg/L of endosulfan after 24 h endosulfan exposure affected the normal physiology of mice. Histopathological studies suggest a persistent, severe effect on reproductive organs where vacuole degeneration of basal germinal epithelial cells and degradation of the interstitial matrix were observed in testes. Ovaries showed a reduction in the number of mature Graafian follicles. At the same time, mild vacuolation in liver hepatocytes and changes in the architecture of the lungs were observed. Endosulfan exposure induced DNA damage and mutations in germ cells at the molecular level. Interestingly, even after 8 months of endosulfan exposure, we observed increased DNA breaks in reproductive tissues. An increased DNA Ligase III expression was also observed, consistent with reported elevated levels of MMEJ-mediated repair. Further, we observed the generation of tumors in a few of the treated mice with time. Thus, the study not only explores the changes in the general biology of the mice upon exposure to endosulfan but also describes the molecular mechanism of its long-term effects.
Introduction
Pesticides are applied in the environment to increase agricultural yield by controlling pests. Currently,~2 million tonnes of pesticides are utilized around the globe each year, which has been estimated to increase to 3.5 million tonnes by 2020 (Eddleston et al., 2002;Zhang, 2018;Sharma et al., 2019). Organochlorine pesticides like DDT (Dichlorodiphenyltrichloroethane) and BHC (Benzene hexachloride) are one of the earliest chemical pesticides used around the world (Simonich and Hites, 1995;Pimentel, 1996;Vijgen et al., 2011). Due to their persistence and long-range transport, they were replaced with the less persistent organophosphate and carbamate pesticides (Goff and Giraudo, 2019). Nevertheless, the residue of organochlorine pesticides is still detected in the environment in many countries where it is not at all used. Moreover, developing countries, including several Asian countries, still use these pesticides (Iwata et al., 1993;Senior, 2009;Gill et al., 2020). Thus, investigating the health and environmental effect of these pesticides are still relevant and necessary.
Commercially, ES is a mixture of two isomeric forms, αand β-Endosulfan, in a ratio of 3:1 (Carrera et al., 2002;Singh et al., 2014). In mammalian systems, it is metabolized into the most persistent and toxic metabolite, endosulfan sulfate, and endosulfan diol, which is further metabolized into endosulfan ether, hydroxy ether, and lactone (Kataoka and Takagi, 2013). Endosulfan and its related isomers were listed as persistent organic pollutants due to their high toxicity, bioaccumulation, long residual period in soil, and long-distance transportation under the Stockholm Convention held in May 2011 (UNEP, 2011). It is persistent in the environment, taking decades to biodegrade. Endosulfan is frequently detected in river water in Africa, Europe, and Asian countries, including India (Dalvie et al., 2003;Hellar-Kihampa et al., 2013;Loha et al., 2020;Montuori et al., 2020). A study showed that a concentration of αand β-ES ranged from non-detectable limit to 35.21 μg/L and 37.56 μg/L, respectively, in Tapi river water in India (Sarker et al., 2021). The mean concentration of ES in the river Ganga in India was~36 ng/L (Khuman and Chakraborty, 2019). Some of its initial degradation products are structurally similar to parent endosulfan isomers and share the same toxic effects (Dorough et al., 1978).
A previous study from our laboratory has shown that exposure to ES in mice (3 mg/kg body weight) induces an increase in DNA damage, triggers compromised DNA damage response, and promoted the error-prone microhomologymediated end-joining (MMEJ) pathway of DNA repair in the male reproductive organ (Sebastian and Raghavan, 2016). ESinduced DNA damage was highest in the lungs and testes when analyzed through immunohistochemistry, immunofluorescence, and Western blot analysis (Sebastian and Raghavan, 2015a;Sebastian and Raghavan, 2015b;Sebastian and Raghavan, 2016;Sebastian and Raghavan, 2017). TUNEL assay revealed DNA damage in spermatogonia mother cells, Sertoli cells, and primary spermatocytes in the same study. The potential of ES to induce DNA damage and compromised DNA repair mechanisms in germ cells is critical for fertility and the stable propagation of species. In the same study, when a sublethal dose of ES was used with a serum concentration of 23 μg/L after 24 h, male mice (33%) infertility was observed due to testicular atrophy. It decreased sperm count and increased mortality (Sebastian and Raghavan, 2015b). However, the serum concentration of ES is much higher in people living in highly exposed areas, with a concentration range of 8.85-547.6 μg/L ES in the affected human population, which can be 700 μg/L after 2 h of exposure (Mnif et al., 2011;Sebastian and Raghavan, 2015b).
ES poisoning increases the incidence of neurological complications and congenital and reproductive abnormalities among the affected people (James and Emmanuel, 2021). However, the effect of ES on the female reproductive system was not investigated previously. Epidemiological studies show that farmers exposed to ES were more prone to developing tumors (Zahm and Ward, 1998) and multiple myeloma (Khuder and Mutgi, 1997). Similarly, people exposed to ES were reported to have an increased risk of developing various cancers in their life (Melangadi, 2017). Although the immediate effects of pesticides are studied extensively, their delayed actions are not explored at the molecular level. There is a knowledge gap between pesticide exposure and its long-term effects on humans. Therefore, to investigate the long-term impact of ES in an exposed human population, we have used a mice model system, considering the relevant concentrations of ES in the exposed areas. In the present study, we report that exposure to ES leads to significantly compromised fertility in both male and female mice. Further, we find a long-term persistent effect on the histopathology of reproductive tissues, liver and lung. Persistent elevated levels of DNA damage leading to compromised DNA repair were also observed. Further, we report an increase in the incidence of tumor development in the mice exposed to ES in their later life.
Materials and methods
Chemicals, reagents, and antibodies Endosulfan (pestanal) was purchased from Millipore-Sigma (Catalogue No. 32015). Other chemicals and reagents were obtained from Millipore-Sigma (St. Louis, MO, United States) or Sisco Research Laboratories Ltd. (SRL), India. Antibodies were procured from Santa Cruz Biotechnologies (United States) and BD Biosciences (United States).
Animals
All animal experiments were carried out with the approval of the animal ethical committee of the Indian Institute of Science (IISc), Bangalore, India (CAF/Ethics/228/2013; CAF/Ethics/793/ 2020). Balb/c mice of 4-6 weeks old weighing 20 g-25 g were purchased from Central Animal Facility (CAF), IISc. The animals were housed in polypropylene cages and provided a standard pellet diet (21% protein, 5% lipids, 4% crude fiber, 8% ash, 1% calcium, 0.6% phosphorous, 3.4% glucose, 2% vitamin, and 55% nitrogen-free extract (carbohydrates) and purified water ab libitum. The mice were maintained under controlled temperature and humidity conditions with a 12 h light/dark cycle.
Administration of endosulfan in mice and hematological analysis
ES was dissolved in 0.05% methylcellulose, and mice were orally fed using a gastric gavage (Sebastian and Raghavan, 2015b). The treatment consisted of a dose of 5 mg/kg on every alternative day for 10 doses.
After 31 days of ES exposure (5 mg/kg, 10 doses), two animals from the treated and control groups were sacrificed using CO 2 asphyxiation. Blood was collected in EDTA-coated vials by heart puncture. Analysis of blood samples was performed at Rohana Veterinary Diagnostic Lab, Bangalore, India. Red blood cells (RBC), white blood cells (WBC), platelet counts (PLT), packed cell volume (PCV), hemoglobin (HGB), mean cell volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), lymphocytes, neutrophils, eosinophils, and monocytes were analyzed as described before Sharma et al., 2021).
Impact of endosulfan exposure on fertility
Mating experiments in Balb/c mice were performed as described before (Chiruvella et al., 2012;Sebastian and Raghavan, 2015b). After 7 days post-completion of ES treatment, 1:2 (male: female) mating was set up in separate cages for 10 days or two oestrus cycles. At the end of the 10th day, males and females were separated, and the females were observed for pregnancy in the following days. A male was considered infertile if all two females in the cage failed to get impregnated. Four groups of animals were maintained for mating; 1) treated male and untreated female; 2) treated female and untreated male; 3) treated male and treated female; 4) control group comprising untreated male and untreated female (n = 5/group for male and n = 10/group for female). The experiments were repeated three independent times on separate occasions.
Histopathological evaluation
Mice organs (brain, lungs, liver, ovary, testes, intestine, and spleen) collected from the animals after 8 months of ES treatment were processed as per standard protocol (Srivastava et al., 2012;Sharma et al., 2013;Gopalakrishnan et al., 2021) and embedded in paraffin. Microtome sectioning was done using rotary microtome with a section thickness of 5 µm (Leica Biosystems, Germany) and stained with hematoxylin and eosin. Images were captured using a bright field microscope (Carl Zeiss, Oberkochen, Germany). Tissues from two mice were subjected to histopathological analysis a minimum of two independent times. The veterinarians analyzed histopathological results at Central Animal Facility, IISc, Bangalore.
Immunofluorescence analysis
Immunofluorescence studies were performed as described before (Chiruvella et al., 2012;Kumari et al., 2019). Paraformaldehyde fixed paraffin embedded tissues sectioned at Frontiers in Genetics frontiersin.org 03 5 µm thickness were de-paraffinized and rehydrated in a series of alcohol gradients. Antigen retrieval was done in 10 mM sodiumcitrate buffer (pH 6.0), followed by blocking in PBST containing 1% BSA and 10% FBS. Primary antibody [53BP1 (SC 22760) dilution 1:100 and Ligase III (BD 611876) dilution 1:500]; incubation was carried out overnight at 4°C. Slides were washed and incubated with appropriate Alexa Fluor conjugated secondary antibody for 3 h at room temperature. After washing the slide, it was mounted with DAPI-DABCO, and images were taken in a confocal microscope (Olympus, FLUOVIEW FV3000, Japan) and processed by Olympus software. The experiments were repeated independently with multiple technical repeats on separate occasions.
Pharmacokinetics of endosulfan
ES was administered orally to Balb/c mice (5 mg/kg, one dose). Animals were sacrificed at 4, 8, 10, 12, 24, and 48 h postadministration, deproteinization was done using acetonitrile, and the sera were used for HPLC analysis (Shimadzu, Kyoto, Japan) (Srivastava et al., 2012;Sebastian and Raghavan, 2015b;Sharma et al., 2021). Liquid Chromatography (LC) was carried out using a C18 analytical column (Shimadzu). The mobile phase consisted of an isocratic mixture of acetonitrile and water at 70:30 v/v. Spectra were acquired at 214 nm. The injection volume was 20 μl (automatic), and the flow rate used was 0.5 ml/min. The standard solution of ES was prepared by diluting the stock to 1, 5, and 10 μM of the compound in the serum. Standard calibration curves were plotted using peak area against the concentration and retention time of the compound, were also determined. Plasma concentrations of the treated groups were extrapolated from standard curves. Pharmacokinetic parameters were analyzed using LabSolutions software (Shimadzu, Japan). The values obtained were plotted with GraphPad Prism (ver5.1) software, where "C" is the predicted concentration, and "t" is time. Data were analyzed using nonlinear regression analysis. Maximum drug plasma concentration (C max ) and time to reach maximum concentration T max were determined by the area under the curve versus the time curve. The clearance rate (CL) was calculated by the rate of drug elimination in plasma (Vartak et al., 2016;Sharma et al., 2021).
Quantification and statistical analysis
All quantifications were performed using GraphPad Prism software, and statistical significance (mean ± SEM) was calculated. All statistical analyses were performed in GraphPad Prism software using one-tailed unpaired student's t-tests. Error bars have been shown depicting mean ± SEM (ns: not significant, *p < 0.05, **p < 0.005, ***p < 0.0001). Colocalization analysis was done using JaCoP in ImageJ software. The values were plotted in GraphPad Prism 5.0, and the significance was calculated using the same.
Result
The bioavailability of endosulfan increases with an increased dose The appropriate concentration of ES for analysis of physiomolecular changes in mice model system, which can be correlated with humans, is important for the investigation of its effect on human health as exposure to it is reported to cause abnormalities and congenital disabilities in areas of extensive use (Satheesh, 2017;James and Emmanuel, 2021). Since the people living in ES-exposed areas showed a serum concentration of up to 550 μg/L of the pesticide, our study selected a concentration that can mimic the situation. Mice were treated with 5 mg/kg body weight of endosulfan based on the previous investigation (Sebastian and Raghavan, 2015b) ( Figure 1A). HPLC analysis was performed to ascertain the serum concentration of ES in mice by collecting serum samples at different time points (4, 8, 10, 12, 24, and 48 h) after oral ingestion ( Figure 1B). For standard, serum samples from control mice were collected and spiked with different concentrations of ES. Our results show that the serum concentration of ES after 24 h was 50 μg/L compared to the previously observed 23 μg/L when 3 mg/kg of ES was used (Sebastian and Raghavan, 2015b) ( Figure 1B).
HPLC analysis post oral dose of ES revealed that the maximum concentration of ES (C max ) was 237 μg/L and the time to reach C max , i.e., T max , was 8 h ( Figure 1C). We observed the clearance rate (CL) of 5.14 ml/min/kg in mice serum based on the elimination rate in plasma ( Figure 1C). It was eliminated from the blood at the rate of 5.14 ml/min/kg. We did not detect ES in serum after 48 h of oral ingestion.
Exposure to endosulfan affects the normal physiology of mice Evaluation of biochemical changes in organs is essential to understand the effects of xenobiotics. Therefore, we were interested in evaluating the impact of exposure when 5 mg/kg (10 doses) of ES was used. To do this, we performed liver and kidney function tests after 31 days of ES treatment (Figure 2A). Results showed no significant difference between the control and treated mice in the level of SGPT, total protein, albumin, bilirubin, and ALP ( Figure 2B). However, analysis of kidney function (creatinine, phosphorous, uric acid, and BUN levels) showed a significant decrease in BUN levels indicating liver and kidney injury ( Figure 2C). This result suggests that ES affects the function of the kidney and liver.
Frontiers in Genetics frontiersin.org 04 Hematological parameters are routinely used to indicate the physiological or sublethal stress responses to endogenous and exogenous changes. To analyze any biochemical changes, we assessed different blood parameters in ES-exposed mice (5 mg/kg) after 31 days compared to the control mice ( Figure 2D). The quantitative analysis showed a significant increase in total leukocyte counts and neutrophils percentage, possibly due to inflammation but a significant decrease in lymphocyte percentage compared with the controls, indicating that the mice were under stress ( Figure 2D). Although an increase in the monocyte population was observed, it was not significant. Blood parameters like platelet counts, RBC, MCHC, MCV, Eosinophils, MCH, and PCV remained unchanged. These results indicate that ES alters the hematological and biochemical parameters in mice.
Endosulfan induces the reproductive toxicity
Mating experiments (1:2 ratio, male: female) were set up the post-one week of the last dose of ES treatment (5 mg/kg, 10 doses) in the mice. The mating was set up in four different groups, Group 1, where ES-exposed male mice were mated with normal female mice; Group 2, where ES-exposed female mice were mated with normal male mice; Group 3, where both male and female mice were exposed to ES and Group 4 was the untreated control group ( Figure 3A).
Male mice were considered infertile if they could not impregnate any female mice. In contrast, female mice were considered infertile if it was not able to get pregnant even after the mating period of 10 days. The percentage of infertility in Group 1 (only male treated) was 55%, while in Group 2 (only female treated), it was 62%. The most severe effect was observed when both parents were treated (67% infertility). However, no infertility was observed when untreated mice were subjected to mating ( Figure 3B). Therefore, a significant increase in infertility was observed in the present study compared to the previous report (Sebastian and Raghavan, 2015a).
Endosulfan affects the reproductive organs of mice even after several months of exposure Although the acute effect of pesticides is studied in detail, its long-term effects are not followed up systematically. When we evaluated the mice for long-term effects of pesticide exposure, we did not find any difference in food and water intake between Frontiers in Genetics frontiersin.org 05
FIGURE 2
Evaluation of physiological effects of ES in mice. (A) Schematic representation of ES-treatment in mice. Mice were exposed to 5 mg/kg body weight of ES through oral gavage every alternate day for 10 doses. (B,C) Liver and kidney function test of ES-treated mice. Bar graphs indicating enzymatic activities reflective of liver and kidney function at 31st day after ES treatment (5 mg/kg, n = 2). (D) Blood parameter analysis following oral administration of ES in mice after 31st day (n = 2). Analysis of red blood cells (RBC), white blood cells (WBC), haemoglobin (HGB), platelets, MCHC, MCV, MCH, PCV, neutrophils, monocytes, eosinophils, and lymphocytes counted among control and treated after 31 days of administration. Error bars denote mean ± SEM (ns: not significant, *p < 0.05).
FIGURE 3
Evaluation of the effect of ES on fertility in mice. (A) Schematic is showing mating groups following exposure to endosulfan (5 mg/kg body weight); Group 1 (male treated), Group 2 (female treated), Group 3 (male and female treated), and Group 4 (untreated control). (B) Bar graphs show the difference in fertility levels when ES was given only to males (n = 5), only to females (n = 10), or to both males and females (n = 15). Mating was in the ratio of 1:2 of male to female. Experiments were repeated 3 times. Error bars denote mean ± SEM.
Frontiers in Genetics frontiersin.org 06 control and ES-exposed mice. Besides, their social behaviour remained unchanged.
Histopathological analysis was performed to understand the direct and long-lasting effect of ES on the reproductive organs of exposed mice (Figure 4). After 8 months of ES treatment, analysis of the testes revealed vacuolar degeneration of basal epithelial cells and increased interstitial spaces due to the shrinkage of seminiferous tubules compared to control mice ( Figure 4A).
FIGURE 4
Histopathological examination of reproductive organs from ES-exposed mice. (A) Histopathology of testis of mice following ES administration. Control indicates testis tissue from mice with no treatment, and ES represents tissue from ES-treated mice (5 mg/kg, 10 doses; n = 2; Magnification: ×5, ×10, and ×20). To evaluate long-lasting effects of exposure to ES, tissue samples were collected after 8 months following exposure. Nuclear components, including heterochromatin and nucleoli were stained by haematoxylin as deep blue or purple while eosin-stained cytoplasmic components like collagen, elastic fibers, muscle fibres and red blood cells. The blue arrows indicate the spermatogonia cells in basal epithelial layer, black arrow in ES indicates the interstitial space between the seminiferous tubules. (B) Histopathology of the ovary of mice following ES administration (5 mg/kg, 10 doses) after 8 months of exposure. The experiments were repeated two independent time for each tissue. Control indicates ovary tissue from mice with no ES treatment and ES represents tissue from ES exposed mice (Magnification: ×5, ×10, and ×20; Scale bar: 100 µm).
Frontiers in Genetics frontiersin.org 07
Many seminiferous tubules were affected, which had depleted spermatogonia mother cells in ES-exposed animals. A visible difference was also observed in the sperm population in the lumen of seminiferous tubules in treated mice compared to controls ( Figure 4A).
Similar to the histopathology of mice testes, decreased follicular activity and an increased number of corpus luteum were observed in the ovary of treated female mice even after 8 months of ES exposure ( Figure 4B). An increase in infertility in females correlates with decreased follicular activity in ES-exposed mice ovary. Thus, our results reveal the persistent, long-lasting effect of ES exposure on the male and female reproductive system.
Endosulfan exposure leads to persistent DNA breaks in the reproductive organs of mice
Since the histological investigation of reproductive organs implicated the persistent effect of ES, we evaluated the reproductive tissues for long-term persistent DNA damage. DNA damage was investigated by immunofluorescence staining of 53BP1 (p53-binding protein 1) in both testes and ovaries in ES-exposed mice ( Figure 5). 53BP1 is a well-known DNA damage response factor, and it is recruited at the site of DNA damage (Gupta et al., 2014). 53BP1 staining on testis sections from ES-exposed mice showed significantly elevated expression in seminiferous tubules of testes when compared to controls, suggesting the generation of DNA double-strand breaks (DSBs) (Figures 5A,B). While the 53BP1 foci were restricted only to the spermatogonial mother cells surrounding the basal epithelial layer of the seminiferous tubules in the control mice testis, a significantly elevated number of 53BP1 foci were seen in spermatocytes and spermatids in the case of treated mice testis. These results implicated the persistent DNA damage in testicular cells when exposed to ES.
Further, a significant increase in 53BP1 foci was observed in the ovary of ES-exposed mice compared to controls ( Figures 5C,D). Specifically, 53BP1 foci were seen in primordial follicles and the granulosa cells of the primary follicles in treated mice. Considering that granulosa cells are involved in the maturation of the entire follicle and may play a crucial role in ovarian physiology, damage to these cells affects the proper maturation of the oocytes resulting in infertility (Amsterdam and Rotmensch, 1987).
FIGURE 5
Evaluation of ES induced DNA breaks in mice reproductive organs. To evaluate the long-term effects of ES, mice (n = 2) were exposed to ES (5 mg/kg, 10 doses), and tissue samples were collected after 8 months and examined for 53BP1 foci, a hallmark of DSBs. (A) Immunofluorescence for 53BP1 expression in testis of ES treated mice. Control indicates sections of testes from untreated control mice, while treated refers to sections from testes of ES-exposed mice. Frontiers in Genetics frontiersin.org 08 Exposure to endosulfan results in overexpression of DNA ligase III in reproductive organs Previously, we observed that ES induces error-prone DNA repair through MMEJ in which DNA ligase III is one of the most critical enzymes responsible for the final sealing of breaks (Wang et al., 2005;Sharma et al., 2015;Sebastian and Raghavan, 2016). Therefore, we examined the expression of DNA Ligase III in the reproductive tissues by immunofluorescence staining on the testis and ovary sections after 8 months of ES exposure ( Figure 6). In the ES-exposed mice testes, the expression of DNA Ligase III was significantly higher than that of the control (Figures 6A,B). A similar analysis of the ovary section revealed elevated expression of DNA Ligase III in ES-exposed mice ( Figures 6C,D). Interestingly, DNA Ligase III expression was observed highest in granulosa cells, as was the case of 53BP1 ( Figures 5C,D), which indicated the induction of the alternate error-prone DNA repair pathway. Thus, our results indicate that the persistent ES-induced DNA double-strand breaks in the reproductive organs of exposed mice led to the elevated expression of DNA Ligase III.
Endosulfan induces damage in the lungs and liver of mice but not in other organs
We were interested in testing the persistence and direct effect of ES on the vital organs of mice after 8 months of its exposure (5 mg/kg, 10 doses). The histopathological analysis of the lungs, liver, kidney, intestine, cerebellum, and spleen was performed on ES-exposed mice after 8 months of the treatment. Results showed changes in the architecture of alveoli, which was more congested in the lungs of treated mice than in control ( Figure 7A). Mild vacuolation in hepatocytes was observed in ES-treated mice liver, indicating irritant in the blood system ( Figure 7A). We did not find any remarkable difference in the histology of the spleen, kidney, intestine, and cerebellum of the treated mice when Mander's coefficient on ImageJ. In panels (B,D) at least 20 fields each were analyzed from two independent batches (ns: not significant, *p < 0.05, **p < 0.005, ***p < 0.0001).
Frontiers in Genetics frontiersin.org 09 compared with the control ( Figure 7B). Thus, histopathological analysis revealed that ES-exposure had a persistent effect on the liver, lungs, and reproductive organs.
Endosulfan induces tumorigenesis in mice
Pesticides are considered one of the major factors contributing to increased tumor incidence, although the direct link is still not well established (Anwar, 1997;Dich et al., 1997;Bharathi et al., 2013). In the present study, we observed the development of tumours in ES exposed (5 mg/kg, 10 doses) mice after 3 and 16 months of completion of the dose (Figures 8, 9). Abnormal growth was observed in the lungs of two ES-exposed male mice ( Figure 8A, top and bottom panel). Histopathology of the suspected tumour tissues revealed undifferentiated cells ( Figure 8B, top panel). Further, infiltrating cells in the bronchiole of the lung were also observed ( Figure 8B, bottom panel, marked with a red arrow). To confirm the cancerous nature of the tissues, immunohistochemical analysis was performed with anti Ki67, a cell proliferation marker, and p53, the tumour suppressor gene, in ES-exposed mice. Results revealed the expression of Ki67 and p53 in the lung, confirming that the abnormal growth was indeed cancerous ( Figure 8C). We also observed the tumour development in the neck region of one of the ES-exposed female mice 3 months after the dosing (5 mg/kg, 10 doses) ( Figure 9A). Further, histopathological analysis of these tissues revealed the presence of undifferentiated cells ( Figure 9B). Sheets of neoplastic cells which did not have well-defined borders were also visible. The cells were uniform in size and shape and were mononucleated. Thus, our observations show compelling evidence that ES has the potential to cause long-term harmful effects on organisms, including tumorigenesis.
Discussion
A preponderance of epidemiology studies shows that pesticides have a health burden on non-target species like humans due to their intrinsic toxicity and limited species selectivity. Widespread pesticide use contaminates the food chain and causes mutagenic activation or inactivation of the ingested chemicals (Pednekar et al., 1987;Bolognesi, 2003;Mostafalou and Abdollahi, 2013;Kapeleka et al., 2021). In the present study, we have used mice as a model system to study the chronic effects of pesticides once exposed when they are 4-8 weeks old. Therefore, the present study can link pesticide exposure and its long-term/persistent impact on humans and provide a molecular basis for its persistent effects.
FIGURE 7
Histopathological examination of vital organs of mice post 8 months of ES treatment. (A) Histopathology of liver and lung of mice following ES administration (5 mg/kg, 10 doses). Tissue samples were collected after 8 months of completion of the dose. Control indicates tissue from untreated mice and ES represents tissue from ES treated mice. Hepatocytes and central vein are marked using blue arrows in case of liver. Changes in the architecture of alveoli in lungs is also indicated using arrows. The magnification shown is ×20. (B) Histopathology of spleen, kidney, intestine, and cerebellum of mice following ES administration (5 mg/kg, 10 doses). Control indicates tissue sections from untreated mice and ES represents tissue from ES-treated mice. The experiments were repeated two independent time for each tissue. Magnification shown is of ×20. Scale bar: 100 µm.
Frontiers in Genetics frontiersin.org A distinct increase in infertility and increased reproductive issues have been associated with occupational or environmental exposure to pesticides during the last few decades (Oliva et al., 2001;Joffe, 2003). In the present study, a serum concentration of 50 μg/L (24 h) was used, as this could mimic relevant concentration in such ES-exposed areas. Interestingly, we observed a significant increase in infertility in both male (30%-55%) and female (20%-62%) mice compared to the previous study in which serum concentration was 23 μg/L (24 h) (Sebastian and Raghavan, 2015b). Several studies have also reported that ES affects the reproductive tissues in different animal models (Sinha et al., 1995;Hiremath and Kaliwal, 2002b;Rey et al., 2009;Milesi et al., 2015). The observed increase in infertility in male and female mice was consistent with the epidemiological studies from the affected region (Jayadevan et al., 2005;Kanthila and Shenoy, 2012;Satheesh, 2017;Badiger et al., 2019). It has been shown that ES affects testicular functions due to its effects on reproductive hormones, leading to abnormal spermatozoa and decreased sperm count and sperm motility (Ozmen, 2011;Sebastian and FIGURE 8 Histopathological examination of abnormal growth in lungs of mice post treatment with ES. (A) Abnormal growth was observed in the lungs following 16 months of ES exposure (5 mg/kg, 10 doses) in male mice. The white arrow indicates abnormal growth. (B) Histopathological analysis of abnormal growth in lungs from ES exposed male mice. The red arrow indicates the infiltrating cells (×5, ×10, and ×20). (C) Immunohistochemical analysis of abnormal growth in lungs of ES exposed mice. Immunohistochemical analysis of control mice lung with Ki67 (top panel) and treated mice lung with Ki67 and p53 (bottom panel) (×20). In panels (B,C) experiment was repeated three independent times from the tissues of same animal. Control indicates lung tissue from untreated mice and treated indicates ES-treated tissue from male mice. Scale bar: 100 μm.
Frontiers in Genetics frontiersin.org Raghavan, 2015a; Sengupta and Banerjee, 2013). Although there exists a limited amount of human data on the effect of ES, a cohort study by Saiyed et al. (2003) among school children (10-19 years old) showed delayed male sexual maturity and interference with male sex-hormone synthesis due to exposure to ES (Saiyed et al., 2003). ES is also shown to cause ovarian regression in females, alterations in hormone synthesis, follicular maturation, ovulation process, and ovarian cycle, which leads to an increase in infertility (Hiremath and Kaliwal, 2002a;Sharma R. K. et al., 2020). The effect of ES on the reproductive system is linked to its hormonal-disrupting function, and such pesticides are often called endocrine-disrupting chemicals (EDCs). These EDCs significantly impact the reproductive system; their activity can be due to direct binding with hormone receptors owing to their conformational similarity with receptor-binding portions of natural steroid hormones (Mrema et al., 2013). Numerous studies have shown that ES behaves as an anti-androgen (Singh and Pandey, 1990;Wilson and LeBlanc, 1998;Murono et al., 2001;Viswanath et al., 2010). We have also shown through docking studies and bioinformatic analysis that ES can bind to the ligand-binding site of the androgen receptor with considerable energy compared to its natural ligand dihydrotestosterone in the previous study (Sebastian and Raghavan, 2015a). The reproductive toxicity of ES is not limited to mammals but also extended to various other animals like zebrafish (Han et al., 2011), crocodiles (Tavalieri et al., 2020), C. elegans (Du et al., 2015) and Newt (Park et al., 2001) indicating that it could affect most forms of life in the ecosystem.
Besides reproductive organs, the present study revealed that exposure to ES affected the lungs and liver in mice, which was also consistent with our previous results (Sebastian and Raghavan, 2015b). Studies from other animal models have also reported the liver, lungs, brain, and kidney as the major target organ for ES-exposure (Gupta and Chandra, 1977;Seth et al., 1986;Paul and Balasubramaniam, 1997;Caglar et al., 2003;Mor and Ozmen, 2003;Jia and Misra, 2007).
Although acute toxicity studies of pesticides are reported, their long-term effects are not well investigated, particularly in humans. Previously, immediate effects of ES were observed in the testes, lungs, and liver when the histochemical analysis was performed (Sebastian and Raghavan, 2015b). In the present study, when the long-term effect on the vital organs of mice was evaluated, we noted that such effects were persistent. We observed vacuolar degeneration of basal epithelial cells and increased interstitial spaces due to shrinkage of seminiferous tubules in mice testes even after
FIGURE 9
Histopathological examination of abnormal growth in ES-exposed mice post treatment. (A) Abnormal growth in neck region observed after 3 months of ES exposure indicated by black arrow in ES-exposed female mice (5 mg/kg, 10 doses). (B) Histology of abnormal tissue from the neck region of the ES-exposed female mice (×20). Experiment was repeated three independent times from the tissues of same animal. Scale bar: 100 μm.
Frontiers in Genetics frontiersin.org 8 months of ES-exposure. A visible difference in the sperm population was also observed in the lumen of seminiferous tubules in the ES-exposed mice. A decrease in follicular activity in ES-exposed mice ovaries further suggests the pesticide's long-term effect on female reproductive tissues. A persistent effect was also seen in the lungs and liver of the ES-exposed mice. The architecture of the alveoli of the lungs was changed in ES-exposed mice. In the liver, mild vacuolation in hepatocytes was observed even after 8 months of ES exposure. The only available long-term study conducted in liver tissues of ES-treated mice, which was assessed after 6 months, demonstrated that exposure to ES caused liver tissue damage illustrated by hepatic/somatic index and increased levels of liver Lactate Dehydrogenase (Kurutaş et al., 2006). Further, our results, in conjunction with a previous study, revealed that ES-induced persistent DNA damage in testes and ovaries, altered DNA damage response, and possibly elevated levels of microhomologymediated end joining (Sebastian and Raghavan, 2016). A study by the International Agency for Cancer Research (IARC) has reported potential carcinogenicity of chemicals such as phenoxy acid herbicides, 2,4,5-trichlorophenoxyacetic acid (2,4,5-T), lindane, methoxychlor, toxaphene and several organophosphates in laboratory animals but epidemiological data on cancer risk in farmers are conflicting (Bolognesi, 2003). In vitro studies have reported the carcinogenic potential of ES (Flodström et al., 1988;Soto et al., 1994;Bulayeva and Watson, 2004;Zhu et al., 2008;Bedor et al., 2010). Besides the general public, farmers and industrial workers are exposed mainly to pesticides. Although there is controversy regarding the carcinogenicity of ES in the exposed region, we report the development of tumours in three different mice exposed to ES several months prior. Two male mice developed a lung tumour, while one female mouse developed a tumour near the neck region. Histopathology and immunohistochemistry studies indeed revealed the existence of cancer cells. Therefore, our study points toward the carcinogenic potential of ES. Further, this was consistent with our previous results, where ES induced an error-prone MMEJ in the reproductive and lung tissues leading to deletions and other genomic rearrangements (Sebastian and Raghavan, 2016). Thus, our study reveals the persistent effect of pesticides and indicates the need for a thorough investigation of the impact of pesticides on human health.
Although the present study is focused on one pesticide, these results can be extrapolated to other pesticides and may be helpful during the policymaking of the use and manufacture of such pesticides. Further, this study also underlines the need for detailed epidemiological and animal studies to understand the extent of pesticide poisoning to human health and the environment.
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Ethics statement
The animal study was reviewed and approved by the Animal ethical committee of the Indian Institute of Science (IISc), Bangalore, India (CAF/Ethics/228/2013; CAF/Ethics/ 793/2020).
|
v2
|
2022-11-24T14:17:06.199Z
|
2022-11-24T00:00:00.000Z
|
253841374
|
s2orc/train
|
Expression patterns of platinum resistance-related genes in lung adenocarcinoma and related clinical value models
The purpose of this study was to explore platinum resistance-related biomarkers and mechanisms in lung adenocarcinoma. Through the analysis of gene expression data of lung adenocarcinoma patients and normal patients from The Cancer Genome Atlas, Gene Expression Omnibus database, and A database of genes related to platinum resistance, platinum resistance genes in lung adenocarcinoma and platinum resistance-related differentially expressed genes were obtained. After screening by a statistical significance threshold, a total of 252 genes were defined as platinum resistance genes with significant differential expression, of which 161 were up-regulated and 91 were down-regulated. The enrichment results of up-regulated gene Gene Ontology (GO) showed that TOP3 entries related to biological processes (BP) were double-strand break repair, DNA recombination, DNA replication, the down-regulated gene GO enriches the TOP3 items about biological processes (BP) as a response to lipopolysaccharide, muscle cell proliferation, response to molecule of bacterial origin. Gene Set Enrichment Analysis showed that the top three were e2f targets, g2m checkpoint, and rgf beta signaling. A prognostic model based on non-negative matrix factorization classification showed the characteristics of high- and low-risk groups. The prognostic model established by least absolute shrinkage and selection operator regression and risk factor analysis showed that genes such as HOXB7, NT5E, and KRT18 were positively correlated with risk score. By analyzing the differences in m6A regulatory factors between high- and low-risk groups, it was found that FTO, GPM6A, METTL3, and YTHDC2 were higher in the low-risk group, while HNRNPA2B1, HNRNPC, TGF2BP1, IGF2BP2, IGF2BP3, and RBM15B were higher in the high-risk group. Immune infiltration and drug sensitivity analysis also showed the gene characteristics of the platinum-resistant population in lung adenocarcinoma. ceRNA analysis showed that has-miR-374a-5p and RP6-24A23.7 were lower in the tumor expression group, and that the survival of the low expression group was worse than that of the high expression group. In conclusion, the results of this study show that platinum resistance-related differentially expressed genes in lung adenocarcinoma are mainly concentrated in biological processes such as DNA recombination and response to lipopolysaccharide. The validation set proved that the high-risk group of our prognostic model had poor survival. M6A regulatory factor analysis, immune infiltration, and drug sensitivity analysis all showed differences between high and low-risk groups. ceRNA analysis showed that has-miR-374a-5p and RP6-24A23.7 could be protective factors. Further exploration of the potential impact of these genes on the risk and prognosis of drug-resistant patients with lung adenocarcinoma would provide theoretical support for future research.
Introduction
Globally, the mortality rate of lung cancer is the highest among all tumors (Xia et al., 2022). Non-small cell lung cancer (NSCLC), which accounts for 80% of all lung cancer cases (Miller et al., 2016), can be divided into three main pathological subtypes: adenocarcinoma (40%), squamous cell carcinoma (30%), and large cell carcinoma (15%) (Ruiz-Cordero and Devine, 2020; Travis et al., 2016). The standard first-line treatment is still platinum-based combined chemotherapy (Scagliotti et al., 2002;Schiller et al., 2002). Although chemotherapy can bring benefits to lung adenocarcinoma (LUAD) patients, the median progression-free survival time is only 5.5 months (West et al., 2019), and drug resistance is inevitable. Although many studies have explored the mechanism of platinum drug resistance, there is still no clear mechanism or targets of platinum drug resistance, and few research results can be used in clinical application. Therefore, we aimed to explore the potential biomarkers and mechanisms of platinum-based drug resistance genes in LUAD, establish a prognostic model, and conduct related research on clinical prognosis and risk.
The rapid development of immunotherapy over the last decade has led to the improvement of immune checkpoint inhibitors, which has improved the clinical outcomes of some patients with advanced cancer and changed the treatment status of NSCLC (Osmani et al., 2018;Queirolo and Spagnolo, 2017). Therefore, attention has been paid to immune cell infiltration, the role of immune infiltration in the occurrence, development, and prognosis of platinum-based LUAD, prognostic information, and predicting the efficacy of immunotherapy. Drug sensitivity analysis can also provide guidance for follow-up treatment of platinum-resistant patients with LUAD. Several studies have shown that mRNA methylation plays an important role in the occurrence and development of some cancers (e.g., glioblastoma, renal clear cell carcinoma, and pancreatic cancer) (Cui et al., 2017;Du et al., 2020;Geng et al., 2020;Lan et al., 2019;Wang et al., 2020). These studies indicated that the development of tumors may be related to the expression of key genes related to the function of the m6A regulator. However, there is no research on the m6A methylation regulatory factor with respect to platinum drug resistance in LUAD.
In this study, we aimed to identify the characteristics of platinum drug resistance genes in LUAD, and explore the characteristics of patients after drug resistance, to pave the way for further study of drug resistance mechanisms. We first explored the expression patterns of platinum-resistant genes related to LUAD in The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO), and A database of genes related to platinum resistance. Functional annotations and channel analysis of different platinum resistance genes were performed through Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Set Enrichment Analysis (GSEA). Subsequently, we evaluated the ability to transform into clinical applications through non-negative matrix factorization (NMF) cluster analysis and the establishment of a prognosis model. Immune infiltration and drug sensitivity analyses were used to evaluate the possible applicability of patients with LUAD platinum resistance to other clinical treatments. The results of this study provide guidance for the development of clinical drugs for platinum-resistant patients with LUAD. Workflow is shown in Figure 1.
Expression datasets and bioinformatic analysis 2.1.1 Data and preprocessing
We downloaded the gene expression profile, miRNA expression profile, and clinical information data of LUAD from the TCGA database (n = 585) (https://tcga-data.nci.nih. gov/tcga/). A total of 912 genes related to platinum resistance were downloaded from A database of genes related to platinum resistance (http://ptrc-ddr.cptac-data-view.org). Finally, expression data and clinical information of 116 LUAD samples (data number GSE26939) were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/query/acc. cgi?acc=GSE26939) for use as the verification dataset.
Differentially expressed gene (DEG) analysis
We analyzed the differences in gene expression between cancer and normal patients using the R package DESeq2 (v1. Frontiers in Genetics frontiersin.org 32.0) (Love et al., 2014). We set | log2fold change | (| log2fc |) ≥ 1 and adjust p value <0.05 as the threshold of differential genes; log2fc ≥ 1 and adjust p value <0.05 gene was used to identify upregulated genes, while log2fc ≤ −1 and adjust p value ≤0.05 were used to identify downregulated genes. Cisplatin-based chemotherapy is a common method to treat LUAD. However, after developing resistance to cisdiammine di chloroplatinum (CDDP), a considerable number of patients' tumors recurred. Therefore, screening patients with primary resistance to cisplatin in the LUAD population can maximize the clinical benefit. To further reveal the biological functions affected by the DEGs related to platinum resistance, the intersection of platinum resistance genes and DEGs was defined as PRR-DEGs. We used a volcano map and heatmap to visualize platinum resistance genes with significant differential expression. The volcano map was drawn using the R package ggplot2 (v3.3.5), the heat diagram was drawn using the R package pheatmap (v1.0.12) (Figure 2A). The detailed information is shown in Supplementary Table S1. The expression of different genes is interrelated, especially among genes that regulate the same biological process. Therefore, to reveal the relationships between the PRR-DEGs, a protein-protein interaction network (PPI) was constructed based on the platinum resistance-DEGs. Using the String database (https://www.string-db.org) (Mering, 2003), the above genes were used as input, and the default confidence threshold was 0.4 ( Figure 2B). The PPI network was constructed, and the visualization was carried out using the Cytoscape (v3.8.2) (Shannon et al., 2003) software. The expression correlation of PRR-DEGs was calculated and visualized using a nomogram.
Functional enrichment analysis
GO (Http://geneontology.org/) is a common method for largescale functional enrichment of genes in different dimensions and levels. It is generally carried out from three levels: biological process (BP), molecular function (MF), and cellular component (CC) (Harris et al., 2004). The R package cluster Profiler (v4.0.5) (Wu et al., 2021) was used for GO functional annotation analysis of all the genes with significantly different expression levels to identify the biological processes and pathways with significant enrichment. The enrichment results were visualized using the R package GOplot (v1.0.2) (Walter et al., 2015). The significance threshold of enrichment analysis was set as adjust p value ≤0.05 ( Figures 3A,C). KEGG (https://www.kegg.jp/) is a utility database resource for understanding advanced functions and biological systems (such as cells, organisms and ecosystems), genome sequencing and other high-throughput experimental techniques generated from molecular level information, especially large molecular data sets. It was established in 1995 by Kanehisa Laboratory of Bioinformatics Center of Kyoto University, Japan. The significance threshold of enrichment analysis was set as p adjust p value ≤0.05 ( Figures 3B,D).
GSEA is a calculation method used to determine whether a group of predefined genes shows a statistical difference between two biological states. It is usually used to estimate changes in pathways and biological process activities in expression dataset samples (Subramanian et al., 2005). To study the differences in biological processes between the two groups of patients, we used the gene expression profile dataset from the MSigDB database (Liberzon et al., 2015)
NMF cluster analysis
Nonnegative matrix factorization, referred to as NMF, is a matrix factorization method proposed by Lee and Seung in the journal Nature in 1999 (Lee and Seung, 1999). It makes all the decomposed components non-negative (requiring purely additive description), and at the same time realizes nonlinear dimension reduction. The samples were analyzed by NMF unsupervised cluster analysis, which was realized using the NMF package in R (v0.23.0) (Zushi, 2021). The correlation between the expression of PRR-DEGs and clinical features (including race, stage, age, and sex) based on NMF classification was visualized ( Figures 4A-C). Gene set variation analysis, referred to as GSVA, is a non-parametric and unsupervised algorithm. Unlike GSEA, GSVA does not need to group samples in advance and can calculate the enrichment scores of specific gene sets in each sample. In other words, GSVA transforms gene expression data from a single gene as a feature expression matrix to a specific gene set as a feature expression matrix. To further analyze differences between NMF classifications, GSVA analysis was implemented using the GSVA package (v1.40.1). Finally, the features of NMF classification were visualized using the FactoMineR (v2.4) and factoextra packages (v1.0.7) (Figures 4D-F).
Prognostic model construction
Owing to the importance of platinum resistance in the treatment of LUAD, different patients may have different platinum resistance states; as such, it is extremely feasible to construct a diagnostic model based on differentially expressed platinum resistance genes. Here, firstly, we used the least absolute shrinkage and selection operator (LASSO) regression method to screen differentially expressed platinum resistance genes; the R package glmnet (v4.1-2) was used to realize this method and select the best lambda symbol value. After regression, only genes with coefficients other than 0 were retained ( Figures 5A,B).
To analyze the relationship between the prognosis model and survival status, we used Kaplan-Meier survival analysis and risk factor analysis. Then, to verify the predictive efficiency of the diagnostic model, receiver operating characteristic (ROC) curves were drawn using the R package pROC (v1.18.0) (Robin et al., 2011). The area under the curve (AUC) of 1 year, 3 years, and 5 years were evaluated. To further prove the robustness of the model prediction, external data (GSE26939) were used for verification ( Figures 5C-F).
To further verify the efficacy of the prognostic model of PRR-DEGs, we incorporated clinical indicators into the model, evaluated the univariate and multifactorial prognostic models of clinical factors using the survival package (v3.2.11), and displayed them as forest maps. In addition, we took clinical factors into account and used the rms package (v6.2-0) to construct clinical prediction nomogram and corresponding calibrate correction charts ( Figures 5G,H).
We investigated whether clinical features are related to prognosis. Univariate Cox regression analysis showed that risk score, sample intermediate dimension, tumor stage, person Frontiers in Genetics frontiersin.org neoplasm cancer status were significantly correlated with OS ( Figure 6A). Finally, these univariate prognostic variables were used as covariates of multiple cox regression analysis. The results showed that risk score and tumor status were independent prognostic factors related to OS ( Figure 6B). In order to evaluate whether our model can effectively predict the prognosis of patients in the clinical environment, we incorporated OS-related factors into the model and constructed a nomogram ( Figure 6C) to predict the OS: 1, 3, and 5-year survival rates of patients. The nomogram model once again confirms the reliability and prospective clinical applicability of the risk model. At the same time, we Calibration the nomogram and found that the predicted results were highly correlated with the actual survival rate ( Figure 6D).
Immune infiltration analysis
The immune microenvironment mainly consists of immune cells, inflammatory cells, fibroblasts, interstitial tissues, and various cytokines and chemokines, and is a loaded comprehensive system. The infiltration analysis of immune cells in tissues plays an important guiding role in disease research and treatment prognosis prediction.
To further explore the relationship between differentially expressed prognostic platinum resistance genes and the infiltration level of immune cells, CIBERSORT (Steen et al., 2020) was used to evaluate the infiltration level of immune cells. The contents of 22 kinds of immune cells in each patient were calculated based on the LM22 background gene set provided by the CIBERSORT website (https://cibersort.stanford.edu/) to reflect the infiltration level. The results were visualized using box diagrams drawn by the R package ggplot2 (v3.3.5) (Figures 7A-D). Significant differences between high and low immune cell expression groups of patients may be related to the prognosis of PRR-DEGs. We used the R package ggExtra (v0.9) and a p-value < 0.01 to identify extremely significant differences in levels of immune cell infiltration, DEGs, and the prognosis of platinum resistance gene expression; the results were visualized using scatter plots and correlation curve fitting. At the same time, we checked the high-and low-risk groups of the immune checkpoints (CD274, CD47, HAVCR2, LAG3, IDO1, SIRPA, TNFRSF4, YTCN1, PDC D1, CTLA4, and TIGIT). The tumor immune dysfunction and exclusion (TIDE) score reflects the sensitivity to immune checkpoints, and the TIDE score calculated for each tumor sample could be used as a substitute biomarker to predict responses to immune checkpoint blocking (Figures 7E,F).
N-methyladenine (m6A) modification is the most common and abundant RNA modification in eukaryotes. To explore differences in m6A between high and low groups, we used m6A regulatory factor data from Yongsheng Li et al. , including 11 readers, 7 writers, and 2 erasers.
Drug sensitivity analysis
The LUAD cell line-drug action dataset was obtained from the Genomics of Drug Sensitivity in Cancer database (GDSC Frontiers in Genetics frontiersin.org www.cancerRxgene.org) (Yang et al., 2012). The drug sensitivity of the expression data in TCGA-LUAD was analyzed using the R package oncoPredict (v0.2) (Maeser et al., 2021), and the sensitivity differences of high-and low-risk groups to different drugs were compared ( Figure 9).
Statistical analysis
All data calculations and statistical analyses are carried out in the R language (v4.1.0). For comparison between the two groups, we used a variance test, p ≤ 0.05 was considered statistically significant.
DEG analysis
To reveal biological differences between LUAD patients and healthy people at the transcriptome level, DEG analysis was conducted between the two groups. After screening by a Frontiers in Genetics frontiersin.org statistical significance threshold, a total of 252 genes were defined as platinum resistance genes with significant differential expression (Figures 2A,B), of which 161 were upregulated and 91 were downregulated. The PPI network diagram ( Figure 2C) of PRR-DEGs was constructed using the String database. In addition, we calculated the correlation of differentially expressed platinum resistance genes, and found that the correlation between upregulated and downregulated genes was closer ( Figure 2D).
FIGURE 7
Results of immune infiltration analysis. (A,B) Correlation of immune infiltration in high-risk and low-risk groups. (C) Differences in immune infiltration between high and low risk groups. (D) Differences in immune checkpoints between high and low risk groups. Differences in (E) TIDE between high and low risk groups. Differences of (F) m6A regulon in high and low risk groups.
Functional enrichment analysis
In order to further reveal the biological functions and processes affected by the differential expression of platinum resistance genes, up-regulated and down-regulated genes were enriched by GO, KEGG, and GSEA and visualized in various forms. The enrichment results of up-regulated gene GO showed that TOP5 entries related to biological processes (BP) were doublestrand break repair, DNA reconstruction, DNA replication, intrinsic apoptotic pathway, and a response to oxidative stress ( Figure 3A), KEGG enriched to cell cycle, cellular senescence, IL-17 signaling pathway, p53 signaling pathway, platinum drug resistance etc. ( Figure 3B). The down-regulated gene GO enriches the TOP5 items about biological processes (BP) as a response to lipopolysaccharide, muscle cell proliferation, response to molecule of bacterial origin, response to mechanical stimulus, and regulation of cell-cell adhesion ( Figure 3C). Down-regulated gene KEGG enrichment is mainly related to the MAPK signaling pathway, IL-17 signaling pathway, EGFR tyrosine kinase inhibitor resistance, TNF signaling pathway, and other related pathways ( Figure 3D). In addition, GSEA analysis showed that TOP3 was E2F targets, G2M checkpoint, TGFβ signaling pathway ( Figures 3E-G).
NMF clustering analysis
We molecular-typed the data of TCGA-LUAD according to the PRR-DEGs. In the NMF method ( Figure 4A), the abscissa corresponding to the first sharp decline of the cophenetic graph is the optimal cluster number. The results ( Figure 4B) showed that the samples of LUAD could be divided into two categories. Then, we visualized the relationships between the PRR-DEGs and clinical information such as stage, race, age, and sex ( Figure 4C), and identified that PRR-DEGs can basically be classified according to NMF. In addition, we calculated the GSVA analysis scores of cancer hallmarks ( Figure 4D). WNT betacatenin signaling and hedgegcg signaling, among others, scored higher in the C1 category, g2 checkpoint and e2f targets, among others, scored higher in the C2 category. The survival analysis of the two classifications of NMF ( Figure 4E) showed that the survival rate of C1 was slightly higher than that of C2, and was statistically significant (p < 0.05). Principal component analysis (PCA) analysis suggested that most samples of the two categories of NMF can be separated ( Figure 4F).
Construction and evaluation of prognosis model
We constructed a prognosis model based on the PRR-DEGs in order to translate the research results into practical clinical application. First, 252 PRR-DEGs were screened using the LASSO regression method (Figures 5A,B). There were 17 genes with non-zero retention coefficients; therefore, we constructed a prognosis model with 17 genes, and the influence coefficient of each gene was the coefficient of the LASSO regression results. To verify the prognostic efficacy of the prognostic model, survival analysis ( Figures 5C,D) was carried out based on TCGA-LUAD data and the validation dataset (GSE26939). The survival of the high-risk group was poor. In addition, the predicted ROC curve was drawn and the AUC was calculated. The results show that the prediction model has excellent prediction efficiency for both sets of data, and the AUC was~0.7 (Figures 5D-F). Finally, in order to evaluate the correlation trend between each gene and risk score, risk factor analysis ( Figures 5G,H) was carried out. Genes such as HOXB7, NT5E, and KRT18 were positively correlated with risk score in the two groups of data.
We investigated whether clinical features are related to prognosis. Univariate Cox regression analysis showed that the prognosis model gene, KRAS mutation, stage, and so on, were significantly correlated with OS ( Figure 6A). Finally, these univariate prognostic variables were used as covariates of multivariate Cox regression analysis. The results showed that all of the prognostic model genes except CASP14 were independent prognostic factors related to OS ( Figure 6B To evaluate whether our model could effectively predict the prognosis of patients in clinical environments, we incorporated OS-related factors into the model and constructed a Nomogram ( Figure 6C) to predict OS 1-, 3-, and 5-year survival rates. The Nomogram model again confirmed the reliability and prospective clinical applicability of the risk model. At the same time, upon calibration to the Nomogram, the predicted results were highly correlated with the actual survival rate ( Figure 6D).
Immune infiltration analysis
To further explore the degree of immune cell infiltration in patients, the CIBERSORT method was used to calculate the degree of infiltration in all samples based on 22 kinds of background genes of immune cells. First, the correlation between the infiltration degree of immune cells in high-and Frontiers in Genetics frontiersin.org low-risk groups ( Figures 7A,B) was calculated. In the high-or low-risk group, the correlation between macrophage M, naïve B cells, and plasma cells was high. In addition, we examined the differences in immune cell infiltration for the different groups ( Figure 7C) and found significant differences in 8 of the 22 kinds of immune cells in both groups. Resting T cells CD4 memory, resting dendritic cells, resting mast cell, and monocytes have a higher degree of infiltration in the low-risk group, while T cells CD4 memory activated, macrophage M0, macrophage M1, and mast cell activated have a higher degree of infiltration in the highrisk group. In addition, we examined the differential expression of ten common immune checkpoints in the high-and low-risk groups ( Figure 7D) and found that only CD47 was highly expressed in low-risk groups. We further explored the differences in TIDE scores between the two groups ( Figure 7E) and found that the TIDE score of the low-risk group was higher. Then, we analyzed the differences in m6A regulatory factors between the high-and low-risk groups ( Figure 7F)and found that FTO, GPM6A, METTL3, and YTHDC2 were higher in the low-risk group, while HNRNPA2B1, HNRNPC, TGF2BP1, IGF2BP2, IGF2BP3, and RBM15B were higher in the high-risk group. Finally, to reveal the relationship between the expression of 17 prognostic platinum resistance genes and the infiltration level of immune cells more directly, a scatter plot was drawn and the correlation curve was fitted by taking the expression values of 8 significant differential immune cells and prognostic platinum resistance genes in the CIBERSORT results. The results showed that the average expression value of prognostic platinum resistance genes was positively correlated with macrophage M0, macrophage M1 and T cells CD4 memory activated ( Figures 8A-C), while mast cell resting, monocytes and T cells CD4 memory resting were negatively correlated ( Figures 8D-F). Then, to further identify drugs that may interact with the highrisk group, we identified the drug sensitivity of TCGA-LUAD patients ( Figure 9) according to the gene expression data of TCGA-LUAD patients and the GDSC database. The results showed that the high-risk group was sensitive to drugs ABT7371910, Axitinib1021, Afatinib1032, Afuresertib1912, Axitinib1021, Ipatasertib1924, and Ibrutinib1799, and had lower IC50 ( Figure 9A). In contrast, the high-risk group was nonsensitive to Afatinib1032, Bortezomib1191, Dasatinib1079, Docetaxel1007, Erlotinib1168, Gefitinib1010, Lapatinib1558, ZM4474391050, Paclitaxel1080, and Tozasertib1096 ( Figure 9B).
ceRNA network construction
In total, 17 PRR-DEGs related to prognosis were identified, 5 mi-RNA ( Figure 10A) interacting with PRR-DEGs were identified based on Tarbase and Targetscan databases, and 18 lncRNA ( Figure 10D) interacting with miRNA were found through lncBase predicted V.2 and the StarBase V2.0 databases, which constituted the ceRNA network. The difference in expression of lncRNA and miRNA between the acute tumor group and normal group ( Figure 10B, C), as well as the KM survival of high-and low-risk groups, were analyzed. The results showed that among the five mi-RNA, only has-miR-374a-5p had different expression and survival curves between the tumor group and the normal group. The expression of HAS-MIR-374a-5P in the tumor group was lower, but its survival in the high-expression group was worse ( Figure 10E). The expression of RP6-24A23.7 in lncRNA was also lower in the tumor group, and the prognosis was worse in the high expression group ( Figure 10F).
Discussion
Although the incidence of lung cancer is lower than that of female breast cancer, the mortality rate remains the highest worldwide (Xia et al., 2022). In China, the lung cancer death rate is the highest among all cancers (Miller et al., 2016). Until 1995, landmark meta-analysis confirmed that cisplatin-based chemotherapy could significantly prolong the survival of NSCLC patients compared with meta-supportive treatment (Listed, 1995). Subsequent studies further confirmed the importance of chemotherapy in the treatment of NSCLC (Socinski et al., 2013). Since then, targeted therapies and immunotherapies have been developed. However, many patients cannot use targeted drugs because they are resistant to the drugs or because they contain driving genes; moreover, immunotherapy cannot be used because of unqualified immune indexes. As such, chemotherapy remains the best option for these patients, despite its toxic nature and strong side effects. Cisplatinbased chemotherapy is still the main method for the treatment of many cancers, but patients treated with platinum drugs will inevitably develop drug resistance.
In this study, we explored the mechanisms of drug resistance, prolonged drug resistance time, and longer survival time for patients. M6A is a methylation modification of RNA adenine (A), which is one of the most abundant modifications in eukaryotic mRNA. It is mainly regulated by the m6A methylation regulator Liu et al., 2019). Previous studies have focused on the relationship between m6A and the occurrence and development of LUAD Ma et al., 2022;Qian et al., 2021), or the relationship between the m6A regulatory factor and chemotherapy resistance of small cell lung cancer (Zhang et al., 2021a;Zhang et al., 2021b). Some studies have also found that the m6A regulatory factor is closely related to LUAD resistance to erlotinib (Li K et al., 2021). As far as we know, our study is the first to explore the relationship between m6A and LUAD resistance to platinum. Some past studies have found that FTO promotes the growth of lung cancer cells , but in this study, the expression of FTO in the low-risk group was higher than that in the high-risk group, which suggests that the Frontiers in Genetics frontiersin.org m6A regulatory gene may have changed in platinum-resistant patients with LUAD. In this study, by comparing the genes of LUAD patients from the TCGA database with platinum resistance genes from A database of genes related to platinum resistance, 252 platinum resistance genes with significant differential expression were obtained, of which 161 were upregulated and 91 were downregulated. Among the significantly different drug resistance genes, LIN28B is the most upregulated gene. LIN28, a structurally conserved RNA-binding protein, is highly Frontiers in Genetics frontiersin.org expressed in embryonic hepatocytes. It can promote rapid cell proliferation and is highly expressed in a variety of tumor tissues and tumor cell lines. The high expression of the LIN28 gene can increase the ability of liver cancer cells to metastasize to distant places (McDaniel et al., 2016), moreover, LIN28 can increase the resistance of ovarian and breast cancer cells to chemotherapy drugs by regulating let-7i (Yang et al., 2008). High expression of LIN28 can increase the insensitivity of lung cancer cells and pancreatic cancer cells to radiotherapy (Oh et al., 2010). However, when LIN28 is inhibited, the growth of NSCLC is reduced . Our research indicates that the high expression of LIN28B may lead to an increase in the resistance of LUAD cells to platinum drugs. This is helpful for further verification in subsequent experiments. A PPI network diagram of upregulated and downregulated genes was constructed, and correlation analysis showed that the correlation between upregulated and downregulated genes was greater. The essence of oxidative stress is the imbalance of the oxidation-antioxidation system in vivo. However, the intracellular oxidation-reduction system of many tumor cells is out of balance, and so the drug resistance of LUAD cells is also closely related to oxidative stress (Winterbourn, 2008). There is DNA recombination in the GO pathway of gene enrichment, and the study shows that the occurrence and development of lung adenocarcinoma are closely related to it (Liang et al., 2022). We speculate that platinum resistance of lung adenocarcinoma is also closely related to it, and we will focus on it in the follow-up research. The research and development of lipopolysaccharide drugs have further improved drug utilization (Guo et al., 2019). In our down-regulated gene GO enrichment, top1 is the reaction pathway to lipopolysaccharide. We speculate that in patients with lung adenocarcinoma drug resistance, the metabolism of lipopolysaccharide substances decreases, which may lead to resistance to platinum drugs, which has a certain hint for us to improve platinum drugs. The top three GSEA enrichment sites were e2f targets, g2m checkpoint, and rgf beta signaling. E2f is a transcription factor gene family. Previous research found that E2F can regulate the expression of mitochondria-related genes, and the loss of this regulation leads to serious mitochondrial defects that affect cell metabolism and tumor cells Frontiers in Genetics frontiersin.org (Benevolenskaya and Frolov, 2015). Yao et al. (Yao et al., 2020) found that E2F, the most abundant pathway in our GSEA analysis, is a potential biomarker and therapeutic target of colon cancer, which indicates that LUAD resistance to platinum is also closely related to the E2f family. The G2m checkpoint pathway is an important part of the cell cycle and is related to the occurrence and development of many tumors. A High G2M score is always related to the overall survival rate of pancreatic cancer (Oshi et al., 2020). NMF, which can be used to solve the complex and excessive calculation issues caused by huge data, is a decomposition nonprobability algorithm using matrix decomposition, belonging to the linear algebra algorithm group (Egger, 2022). NMF processes the data after TFIDF conversion by decomposing a matrix into two low-level matrices (Obadimu, 2019). We used the NMF algorithm for the molecular typing of PRR-DEGs and found that LUAD samples could be divided into two categories. We then used heatmap visualization to identify associations between PRR-DEGs and clinical information such as stage, race, age, and sex. Using PCA, we also found that the two categories of samples could be distinguished easily. In GSVA analysis, 4dk wnt betacatenin signaling and hedgegcg signaling had the highest scores in C1. Wnt/β-catenin is a classic signal pathway, and the occurrence and development of many tumors are closely related to it, including colon cancer, hepatocellular carcinoma, desmoid tumor, pancreatic cancer, gastric cancer, melanoma, ovarian cancer, renal cancer (Guillen-Ahlers,2008), and prostate cancer (Robinson et al., 2008). Our analysis shows that the platinum resistance of LUAD is also related to Wnt/β. In contrast, g2 eckpoint and e2f scored higher in C2. Some studies have shown that the expression of abnormal cyclin G2 is the key factor leading to the pathological process of cancer, including glioma. Among the platinum resistance genes in LUAD, the related gene of cyclin G2 is also very important and warrants attention.
To better guide clinical work, we constructed a prognostic model to evaluate the PRR-DEGs, and screened 252 PRR-DEGs using the LASSO regression method. There were 17 genes with a non-zero retention coefficient, and so we constructed a prognostic model with 17 genes; the influence coefficient of each gene was the coefficient of the LASSO regression results. Based on TCGA-LUAD data and a validation set (GSE26939), survival analysis was carried out to verify the prognostic efficacy of the prediction model, which confirmed that the prognosis of the high-risk group was poor. Moreover, risk factor analysis was used to evaluate the correlation trend between each gene and risk score. We found that genes such as HOXB7, NT5E, and KRT18 were positively correlated with risk score. Studies (Yan et al., 2022) have shown that the HOXB gene cluster contributes to cancer development; increased expression of HOXB3, HOXB6, HOXB7, HOXB8, and HOXB9 in LUAD patients is linked with poor overall survival (OS). Our data mining also illustrated the close relationship between the Frontiers in Genetics frontiersin.org HOXB7 gene and LUAD platinum resistance. Past research (Dong et al., 2020) has also shown that NT5E levels are significantly higher in NSCLC tissues and cells. In our model, NT5E was also associated with the platinum resistance of LUAD, suggesting that the NT5E gene may be related to the development of lung cancer. The study has shown that the overexpression of ALDOA increases the migration and invasion of lung cancer cell lines in vitro and the formation of metastatic lung cancer in vivo (Chang et al., 2019). Our analysis suggests that ALDOA may also be associated with platinum drug resistance. Cystatin (CASPs) is an important regulator and executor of the apoptosis pathway. It has been found that the CASP family is closely related to the prognosis of non-small cell lung cancer (Lee et al., 2010). However, there is no study to explore the relationship between CASP and platinum resistance. Our analysis shows that CASP12, and CASP14 may be related to platinum resistance. This will guide the following research. The researchers have found that the deletion of FAT1 can accelerate the occurrence and malignant progression of tumors. In mouse and human squamous cell carcinomas, the loss of FAT1 function can promote tumorigenesis by inducing a mixed EMT state (Pastushenko et al., 2021). The other study has shown that FAT1 mutation is associated with better immunogenicity and ICI efficacy, which may be considered as a biomarker of patients who choose to receive immunotherapy . The results of the expression of the same gene will be different under different treatment regimens. No one has studied the relationship between FAT1 and platinum resistance.FEN1 is the main component of the basic excision and repair pathway of the DNA repair system. Studies have shown that the high expression of the FEN1 gene is essential for the rapid proliferation of lung cancer cells, and the FEN1 gene can also increase the resistance of lung cancer cells to cisplatin (He et al., 2017). Studies have shown that the over-expression of GDF15 significantly inhibits the proliferation of NSCLC in vitro and in vivo (Lu et al., 2018). Through our analysis and prediction, GDF15 may be also related to platinum resistance of lung adenocarcinoma, which needs further verification by subsequent experiments. Studies have shown that high TXNRD1 protein levels are associated with shorter disease-free survival and postoperative distal metastasis-free survival in patients with NSCLC, including some individuals receiving platinum adjuvant chemotherapy (Delgobo et al., 2021;Guo et al., 2021), indicating that TXNRD1 is an important predictor of poor prognosis, which is consistent with our conclusion. Some studies have shown that UBE2S promotes the metastasis of lung adenocarcinoma cells by activating NF-κ B signal transduction, while other studies have shown that UBE2S regulates Wnt/β-catenin signal and promotes the progression of non-small cell lung cancer (Ho et al., 2021;Qin et al., 2020).In our predictive model, UBE2S is also an important factor in platinum resistance in patients with lung adenocarcinoma. The clinical prediction model established by Luo Yu et al. also shows that WFDC2 is an important factor affecting the prognosis of lung adenocarcinoma. Whether WFDC2 is also an important factor affecting platinum resistance in lung adenocarcinoma needs further experimental verification (Luo et al., 2022). Univariate Cox regression analysis showed that prognostic model genes, KRAS mutations, stages, and so on, were significantly correlated with OS. These univariate prognostic variables were used as covariates of multivariate Cox regression analysis. Except for CASP14, the other 16 prognostic model genes were independent prognostic factors related to OS. The nomogram model once again confirmed the reliability and prospective clinical applicability of the risk model. When the nomogram was used for calibration, the predicted results were highly correlated with the actual survival rate. This shows that our risk model has good clinical application value. Focusing on the immune status of tumor patients will help us to explore the mechanisms of drug resistance and new therapeutic targets. Therefore, we used the CIBERSORT method to calculate the infiltration degree of immune cells in all samples; the results showed that resting T cells CD4 memory, resting dense cells, resting mast cell, and monocytes had higher infiltration degrees in the low-risk group than in the high-risk group. While activated T cells CD4 memory, macrophage M0, macrophage M1, and activated mast cells had higher infiltration degrees in the high-risk group compared with the low-risk group. From a scatterplot of the above eight different immune cells and platinum resistance genes, the average expression values of platinum resistance genes in prognosis were positively correlated with macrophage M0, macrophage M1 and activated T cells CD4 memory but negatively correlated with resting mast cell, monocytes, and resting T cells CD4 memory.
The TIDE score can be used to evaluate the potential clinical efficacy of immunotherapy in different immune-related gene prognostic model (IRGPI) subsets (Liu, 2018). The higher the TIDE prediction score, the higher the possibility of immune escape, suggesting that patients are less likely to benefit from immunotherapy. Compared with the high-risk group, the lowrisk group had higher TIDE scores, indicating that people in the low-risk group were more likely to experience immune escape and could not benefit from immunotherapy. N6-methyladenosine, also called m6A, is a base modification widely existing in mRNA. The internal modification of mRNA can affect the RNA splicing, translation, stability, and epigenetics of some non-coding RNAs (Meyer and Jaffrey, 2017). By analyzing the differences in m6A regulatory factors between high-and low-risk groups, it was found that FTO, GPm6A, METTL3, and YTHDC2 expression was higher in the low-risk group. Among these, studies have shown that FTO plays the role of an oncogene in acute myeloid leukemia by regulating the level of m6A and promoting the occurrence and development of leukemia. Later, other studies showed that FTO plays a role as m6A demethylase in various life processes (Gokhale et al., 2016;Xiang et al., 2017). However, in this study, FTO was highly expressed in the low-risk group. We suggest that FTO is not closely related to platinum drug resistance in LUAD. The expressions of HNRNPA2B1, HNRNPC, TGF2BP1, IGF2BP2, IGF2BP3, and RBM15B were higher in the high-risk group, indicating that the above m6A regulatory factors play an Frontiers in Genetics frontiersin.org important role in the mechanism of platinum resistance in LUAD. From drug sensitivity analysis, the high-risk group had lower IC50 to the drugs ABT7371910, Axitinib1021, Afatinib1032, Afuresertib1912, Axitinib1021, Ipatasertib1924, and Ibrutinib1799, indicating that the high-risk group is more sensitive to these drugs, but less sensitive to Afatinib1032, Bortezomib1191, Dasatinib1079, Docetaxel1007, Erlotinib1168, Gefitinib1010, Lapatinib1558, ZM4474391050, Paclitaxel1080, and Tozasertib1096. This has a certain clinical guiding significance for patients with LUAD after platinum resistance. Has-miR-374a-5p is related to the occurrence and development of pancreatitis (Wen et al., 2019), but the relationship between has-miR-374a-5p and platinum drug resistance of LUAD has not been explored. ceRNA analysis showed that has-miR-374a-5p is highly expressed in healthy individuals, and the higher the expression, the better the survival. RP6-24A23.7 is associated with lymphatic metastasis of LUAD (Yan et al., 2017). Morever, as we have demonstrated, may also be associated with platinum resistance in LUAD. The expression of RP6-24A23.7 was lower in the tumor group, and the survival was worse in the low expression group.These indicate that has-miR-374a-5p and RP6-24A23.7 were protective factors.
However, our research also had some limitations. First, in order to fully clarify the molecular mechanisms of resistance and development of platinum drugs in LUAD, microarray samples from platinum drugs in different degrees of LUAD are needed. Second, many biomarkers related to platinum resistance in LUAD still have no characteristics, and further bioinformatic analyses and experimental verifications are needed to clarify the biological function of these predictive genes in platinum resistance in LUAD. Unfortunately, because of the COVID-19 epidemic, our basic experimental process has been hindered. In future research, we will further use experiments to verify the mechanisms of drug resistance.
In summary, this study explored the characteristics of high-and low-risk groups by analyzing the biological process characteristics of platinum resistance genes in LUAD, established a prognosis model, and analyzed its m6A regulatory factors, immune infiltration, and drug sensitivity. Our results have significance for guiding clinical practice. We identified the potential targets and mechanisms of LUAD platinum resistance, laying the foundation for further research.
Ethics statement
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
Author contributions
ZW, LM, LX, and QL designed this research, and ZW, LM and HF analyzed the data and wrote their papers. JY, QW, and HZ conducted data analysis and explained the data. WY helped revise the manuscript of the article. All authors read and approve the final draft.
Funding
This work is supported by grants from Shanghai Hospital Development Center (Clinical Researsh Plan of SHDC, No. SHDC12020CR2052B), The funder played no role in the study design, data collection, analysis, management, and interpretation of data, or writing of the manuscript.
|
v2
|
2022-11-24T14:19:18.398Z
|
2022-11-24T00:00:00.000Z
|
253841510
|
s2orc/train
|
Recent developments in PD-1/PD-L1 blockade research for gastroesophageal malignancies
Gastroesophageal cancers (GECs) comprise malignancies in the stomach, esophagus, and gastroesophageal junction. Despite ongoing improvements in chemoradiotherapy, the clinical outcomes of GEC have not significantly improved over the years, and treatment remains challenging. Immune checkpoint inhibitors (ICIs) have been the subject of clinical trials worldwide for several years. Encouraging results have been reported in different countries, but further research is required to apply ICIs in the clinical care of patients with GEC. This review summarizes completed and ongoing clinical trials with programmed death 1 (PD-1)/programmed death-ligand 1 (PD-L1) pathway blockers in GEC and current biomarkers used for predicting PD-1/PD-L1 blockade efficacy. This review captures the main findings of PD-1/PD-L1 antibodies combined with chemotherapy as an effective first-line treatment and a monotherapy in second-line or more treatment and in maintenance therapy. This review aims to provide insight that will help guide future research and clinical trials, thereby improving the outcomes of patients with GEC.
Introduction
Among gastroesophageal cancers (GECs), gastric cancer (GC) ranks fifth in incidence and fourth in mortality worldwide, and the median survival for advanced GC is less than 12 months (1,2). Esophageal cancer (EC), another GEC, ranks seventh in incidence and sixth in overall mortality worldwide (1). In 2018, an estimated 570,000 individuals were diagnosed with EC worldwide, representing 3.2% of all cancer diagnoses and 5.3% of all cancer-related deaths (3). Over the past 30 years, the clinical benefits of conventional and emerging therapies have reduced GC mortality but have not improved EC survival (4, 5).
In certain western nations, adenocarcinoma has overtaken squamous cell carcinoma as the most prevalent type of EC, and its incidence continues to increase in other nations (6).
During the initiation of cellular immunity, antigens presented by the major histocompatibility complex on the surface of antigen-presenting cells (APCs) can selectively bind to cell receptors of the T-lymphocyte membrane, triggering further T-cell activation, proliferation, and differentiation. Activated T cells serve a vital function in the immune system (7). Under normal physiological conditions, programmed death 1 (PD-1), a negative costimulatory immune molecule also known as an immune checkpoint, is found on the surface of T, B, and myeloid cells. PD-1 specifically connects to programmed death-ligand 1 (PD-L1) on the surface of APCs to trigger immunosuppressive signal transduction, resulting in a decrease in T-cell activity. As cancer develops, tumor cells attach to vascular endothelial or perivascular cells, fibroblasts, and lymphocytes in the surrounding tissue, constituting the tumor microenvironment (TME) in combination with the extracellular matrix (8,9). The TME can disrupt the dynamic balance of the organism by blocking cell apoptosis and promoting angiogenesis and cell proliferation, leading to continued tumor cell development, immune escape, and distant metastasis. Tumor cells highly express PD-L1 to strengthen the PD-1/PD-L1 pathway, thereby exhausting T cells and permitting tumor cells to evade immune surveillance. Based on this principle, PD-1/PD-L1 antibodies were established to constrain the PD-1/ PD-L1 signaling pathway by binding to receptors on the surface of T lymphocytes or tumor cells in the late stages of peripheral tissue regulation of T-lymphocyte function, thereby disrupting the immune response, preventing tumor cell immune escape, and ensuring a normal immune response (10). The combined positive score (CPS), the most accepted PD-L1 scoring method, refers to the count of PD-L1-positive cells (including tumor cells, lymphocytes, and macrophages) divided by the total count of live tumor cells, multiplied by 100. The National Comprehensive Care Network (NCCN) recommends PD-L1 testing (i.e., CPS) for metastatic/advanced EC and GC.
Immunotherapy for GEC includes targeted blockade against immune checkpoints such as PD-1/PD-L1, cytotoxic T lymphocyte antigen-4 (CTLA-4), T cell immunoglobulin-3 (Tim-3), lymphocyte activation gene-3 (Lag-3) and chimeric antigen receptor T-cell (CAR-T) cell therapy; and therapeutic cancer vaccines. Among immune checkpoint inhibitors (ICIs), PD-1/PD-L1 antibodies have shown well applicability in EC and GC, thus dramatically changing the treatment outlook for these patients. An increasing number of PD-1/PD-L1 blockers have been authorized for use in EC and GC treatment. Exploring the administration conditions of known PD-1/PD-L1 inhibitors and developing new antibodies are key directions of current research, as well as evaluating and predicting PD-1/PD-L1 blockade efficacy. Although considerable research through clinical trials has been conducted in EC and GC, much less is known concerning the proper indication of the medicine and the patient selection criteria in these trials, which are often among the potential limitations of the study design. The assessment of the efficacy of PD-1/PD-L1 antibodies frequently employs biomarkers that could be used to select GC and EC patients; however, much work is yet to be discovered in this area. In this review, we present an update on and evaluate the results of current clinical trials with PD-1/PD-L1 antibodies in EC and GC and briefly describe the progress in developing common predictive biomarkers. By comparing previous clinical trials, we also highlight study design limitations that warrant consideration prior to establishing future clinical trials, with the hope of assisting patients in reaching a greater survival outcome.
Molecular and immunological basis of esophageal cancer and gastric cancer EC does not have clear molecular typing, but one study classified EC into low-and high-risk subtypes, which might be used as independent prognostic factors (11). The Cancer Genome Atlas (TCGA) classifies four molecular subtypes of GC: Epstein-Barr virus (EBV)-positive, high microsatellite instability (MSI-H), genomically stable (GS), and chromosomal instability (CIN) (12). PD-L1 and PD-L2 expression levels are amplified in EBV-positive GC. MSI in cancer genomes is caused by DNA mismatch repair (MMR) system deficiencies. High MSI in tumors leads to the accumulation of mutational load, which affects the tumor response to anti-PD-1 antibodies (13). The United States Food and Drug Administration (FDA) has authorized pembrolizumab for the treatment of previously treated MSI-H/mismatch repairdeficient (dMMR) solid tumors, including EC and GC (14).
EC and GC are highly immunogenic, and multiple tumor neoantigens have been identified (15,16). Owing to characteristics such as MSI and tumor mutational burden (TMB), tumor cells are highly susceptible to multiple genetic mutations, resulting in the production of specific neoantigens (17). These neoantigens can be taken up by APCs, which deliver the neoantigen to CD8+ T lymphocytes, initiating cytotoxic T lymphocytes (CTLs) and generating a key mechanism of antitumor immunity by killing tumor cells. In the TME, inflammatory factors, lymphocytes, monocytes, macrophages, and histiocytes comprise the tumor immune microenvironment. Tumor-infiltrating lymphocytes (TILs), consisting of T, B, and natural killer (NK) cells, infiltrate heavily in esophageal squamous cell carcinoma (ESCC) and gastric adenocarcinoma (18). TILs have been confirmed to be effective and independent prognostic factors during the antitumor immune response, and PD-1 expression on TILs correlates with adverse clinical outcomes in EC (19). Increased CD8+ TIL levels have been consistently detected in PD-L1-positive EC (20). Increased CD8+ TIL levels were closely associated with better survival, lower lymph node metastases, and higher PD-L1 expression levels; the combined evaluation of CD8+ TIL and PD-L1 expression has been used to predict patient responses to PD-1/ PD-L1 antibody treatment in a range of malignancies (21). Large numbers of CD20+ B cells are significantly correlated with both modest lymph node involvement and lower TNM stage as independent factors for GC prognosis (22). Moreover, tumorassociated macrophages (TAMs) can release cytokines that promote cancer cell motility and invasion (23-25). Overall, high TAM density is considered to be a negative prognostic factor in GC (26). TAMs often differentiate into M1-like TAMs with pro-inflammatory and tumor-suppressive functions and M2-like TAMs with anti-inflammatory and tumor-promoting functions (27). M1-like TAMs are an independent prognostic factor in GC, and CD68+CD163-macrophages, a group of representative M1-like TAMs, can be used as predictive biomarkers to guide PD-1/PD-L1 antibody treatment in GC (28). M2-like TAMs are involved in the inhibition of antitumor immune responses by increasing PD-L1 expression in tumors (29). Patients with EC who have high levels of M2-like TAMs had shorter overall survival (OS) (30,31). Thus, certain TAM subgroups could have prognostic value in gastric adenocarcinoma and esophageal adenocarcinoma (EAC) (32). Finally, through a variety of cytokines, cancer-associated fibroblasts (CAFs), valuable stromal cells in the TME, contribute to the growth, progression, and metastasis of EC (33,34). CAFs upregulate PD-L1 expression, thereby promoting cancer cell proliferation in GC (35). Furthermore, a study investigating CAFs in GC reported that extracellular matrix CAFs recruited M2-like macrophages and were associated with poor prognosis (36).
Clinical trials exploring PD-1/PD-L1 blockade in gastroesophageal cancers PD-1/PD-L1 inhibitors have been approved for clinical use in several countries. For example, the US FDA granted pembrolizumab, nivolumab, and dostarlimab-Gxly approval for the treatment of EC and GC under certain conditions in 2022. As a first-line therapy for ESCC, camrelizumab + chemotherapy has been approved by China in 2021. However, the findings of the few clinical trials that have tested PD-1/PD-L1 antibodies as first-line monotherapies so far are not encouraging. Chemotherapy combined with PD-1/PD-L1 antibodies is currently being investigated in clinical studies as the first-line therapeutic option. This section presents the outcomes of clinical trials with PD-1/PD-L1 antibodies in EC and GC, emphasizing progress and comparing application conditions.
PD-1/PD-L1 blockade as first-line treatment in esophageal cancer
Radical resection is the conventional first-line treatment for EC, with or without perioperative chemotherapy (37). Advanced EC is treatable with first-line chemotherapy, with an overall poor prognosis (38). Therefore, research has concentrated on the development of inhibitors for immune checkpoints. This section focuses on clinical trials exploring PD-1/PD-L1 antibodies combined with chemotherapy and introduces the application of new PD-1 antibodies as first-line treatments for EC (Table 1).
KEYNOTE-590 was the first clinical trial to evaluate the combination of PD-1 inhibition with chemotherapy as a firstline treatment for EC with significant survival benefits. In March 2021, pembrolizumab plus fluoropyrimidineand platinumbased chemotherapy was authorized by the FDA for the firstline treatment of patients with ESCC and EAC with CPS ≥10 (category 1, requires combination with cisplatin) and CPS <10 (category 2B) (39). The KEYNOTE-590 phase 3 trial enrolled 749 patients with advanced EC or Siewert type 1 gastroesophageal junction cancer (GEJC), among which 51% of the study population had CPS ≥10. The interventions included pembrolizumab or placebo plus chemotherapy (5fluorouracil plus cisplatin). Compared to the placebo arm, the pembrolizumab arm showed a considerably enhanced survival advantage and sustained antitumor response in the total population, advanced ESCC subgroup, and CPS ≥10 subgroup. In all three populations, the pembrolizumab arm maintained an advantage in Kaplan-Meier (KM) curves for OS, and pembrolizumab + chemotherapy treatment was roughly twice as effective as placebo + chemotherapy treatment at 24-month OS. Progression-free survival (PFS), 12-month PFS, and 18month PFS remained superior in all three populations treated with pembrolizumab plus chemotherapy. Additionally, the pembrolizumab + chemotherapy group had approximately 15% greater overall response rate (ORR), 2.3-month greater duration of response (DoR), and a nearly 3-fold increase in 24-month DoR than the placebo + chemotherapy group. No additional adverse events (AEs) were detected, indicating the safety of pembrolizumab combined with chemotherapy (40,41).
The CheckMate-648 study evaluated PD-1 antibody combination therapy, delivering three types of drugs to patients with ESCC (n = 970): nivolumab + chemotherapy (intravenous fluorouracil), nivolumab + ipilimumab (CTLA-4 antibody), and chemotherapy alone. In the randomized population and tumor-cell PD-L1 expression of ≥1% subgroup, the nivolumab + chemotherapy group maintained higher complete response (CR) rates and longer-lasting responses at the 13-month follow-up than the other treatment groups. The median overall survival (mOS) for >12 months of the nivolumab + ipilimumab group was 2.0-6.3 months longer than that of the chemotherapy group. In patients with tumor-cell PD-L1 expression of ≥1%, the nivolumab + chemotherapy group had a substantial PFS advantage over the chemotherapy group (6.9 vs. 4.4 months). In patients with CPS ≥1 (91%), both the nivolumab + chemotherapy [hazard ratio (HR), 0.69] and nivolumab + ipilimumab (HR, 0.76) groups achieved prolonged mOS compared with that in the chemotherapy group. The survival advantage of the nivolumab-based regimen was demonstrated in subgroups with tumor-cell PD-L1 expression of ≥1% thresholds of 1%, 5%, and 10%, all with HR <1. The AEs were mainly caused by chemotherapy (nausea, loss of appetite, and stomatitis) (42). Notably, the KEYNOTE-590 and CheckMate-648 clinical trials employed similar chemotherapy drug intensities (both included fluoropyrimidine) but did not use the same evaluation criteria for PD-L1 expression and subgroup analysis.
Camrelizumab, a monoclonal antibody against PD-1, has also been researched as a first-line combination treatment in EC. Patients enrolled in the ESCORT-1st trial received camrelizumab or placebo plus chemotherapy (paclitaxel-cisplatin). The camrelizumab arm showed a longer OS tendency than the placebo arm (mOS, 15.3 vs. 12.0 months). Fewer grade 3-4 treatment-related adverse events (TRAEs) in the camrelizumab + chemotherapy group compared with the placebo + chemotherapy group (63.4% vs. 67.7%) indicated lower toxicity, with the former group experiencing adverse immune reactions mainly due to reactive capillary endothelial proliferation often associated with camrelizumab (43). The findings of this clinical trial supported the approval of camrelizumab in China for first-line treatment of unresectable, locally advanced/recurrent, or metastatic ESCC. Toripalimab, an immunoglobulin G (IgG) PD-1 antibody, was evaluated in the JUPITER-06 trial, which enrolled 514 Chinese patients with advanced ESCC who received either toripalimab or placebo plus chemotherapy (paclitaxel plus cisplatin). PD-L1 expression was categorized as CPS ≥1 (PD-L1-positive) or CPS ≥10 (PD-L1 high expression). The toripalimab arm showed improved median progression-free survival (mPFS) (HR, 0.58) and mOS (HR, 0.58) compared to the placebo arm. The KM curves for PFS diverged early, with toripalimab retaining an advantage over the placebo. The 12month PFS was nearly four times greater in the toripalimab + chemotherapy arm than in the placebo + chemotherapy arm. In terms of the antitumor response, the ORR (69.3% vs. 52.1%, p = 0.001) and DoR (5.6 vs. 4.2 months) were considerably higher in the toripalimab arm than in the placebo arm. The safety profile of toripalimab was considered to be acceptable. The OS and PFS benefits of toripalimab with chemotherapy were statistically significant and independent of PD-L1 expression levels (44). Both the JUPITER-06 and ESCORT-1st trials enrolled Chinese ESCC patients only. However, the survival benefit in the ESCORT-1st trial corresponded with PD-L1 expression levels, in contrast to the JUPITER-06 trial. Different PD-L1 detection methods and scoring criteria may have affected the results.
Sintilimab is a human IgG4 anti-PD-1 monoclonal antibody. In the multicenter ORIENT-15 trial, patients with ESCC received either sintilimab or placebo plus chemotherapy (93% cisplatin and paclitaxel, 7% cisplatin and 5-fluorouracil). Chinese patients made up 97% (n = 640) of the patients. The sintilimab arm had markedly better OS (16.7 vs. 12.5 months), PFS (7.2 vs. 5.7 months), and ORR (66% vs. 45%) than those in the placebo arm. The KM curves of OS remained distinct for the two groups from the beginning. The sintilimab arm outperformed the placebo arm by 13% and 23% for 1-and 2year OS, respectively. Both tumor proportion score (TPS) and CPS for PD-L1 scoring were employed in the study. In the subgroup analysis, the survival advantage of sintilimab + chemotherapy was independent of PD-L1 expression levels (HR, 0.55 for TPS ≥10%; HR, 0.67 for TPS <10%; HR, 0.64 for CPS ≥10; HR, 0.62 for CPS <10) (45).
In the above clinical trials, PD-1 antibodies + chemotherapy were administered as a first-line combination therapy for EC. Although PD-1/PD-L1 antibody monotherapy has demonstrated good outcomes as a second-and third-line treatment, many challenges for its use as first-line treatment persist. The choice of the chemotherapeutic drug, patient distribution, inclusion criteria, and drug dose are factors that remain to be elucidated.
PD-1/PD-L1 blockade as first-line treatment in gastric cancer
The most common first-line treatment for metastatic and incurable GC is systemic therapy, with oxaliplatin frequently favored over cisplatin due to its reduced toxicity (46). Targeted therapies have also been used as first-line treatments for patients with specific types of GC. Patients with Human epidermal growth factor receptor 2 (HER2)-overexpressed gastric adenocarcinoma are recommended to receive pembrolizumab in combination with trastuzumab and chemotherapy (fluoropyrimidine and platinum) as first-line therapy. This recommendation is according to the results of the KEYNOTE-811 clinical trial. This ongoing international phase 3 trial is evaluating HER2-positive GC/GEJC in 692 patients treated with pembrolizumab or placebo plus trastuzumab and chemotherapy (capecitabine + oxaliplatin or fluorouracil + cisplatin). The trial employs MSI-H and PD-L1 as biomarkers. In the study population, 84.1% of patients had CPS ≥1, and large differences in ORR were reported. In the first interim analysis of 260 patients after an 8.5-month follow-up, the pembrolizumab arm had approximately 20% greater ORR than the placebo arm (74.4% vs. 51.9%) and maintained certain advantages in CR, disease control rate (DCR), and DoR, suggesting a more robust and durable response. Among the 433 patients examined for safety, the pembrolizumab group showed a lower incidence of grade 3-5 AEs and AEs leading to death than the placebo group. We look forward to updates from this trial (47, 48).
Based on the excellent clinical benefits and durable response achieved by nivolumab in combination with fluoropyrimidineand platinum-containing chemotherapy in patients suffering from unresectable HER2-negative GC, GEJC, and EAC, the FDA approved this therapy in April 2021 for first-line treatment of tumors with CPS ≥5 (category 1) and CPS <5 under certain circumstances (category 2B) (49). In the CheckMate-649 trial, the analysis of survival status and antitumor response was divided into CPS ≥1 and CPS ≥5 subgroups. The nivolumab arm achieved a more pronounced OS benefit than the chemotherapy arm in the CPS ≥5 cohort (mOS, 14.4 vs. 11.1 months), CPS ≥1 cohort (HR, 0.77), and in all random patients (HR, 0.80). In patients with CPS ≥5, the nivolumab arm had 1.7-month longer PFS than the chemotherapy arm (7.7 vs. 6.0 months) and 14% longer 1-year PFS. The follow-up study determined that the survival benefit of nivolumab + chemotherapy increased with higher CPS cutoff value. In patients with CPS ≥5, the nivolumab + chemotherapy group had 15% greater ORR and 2.5-month longer response duration than the chemotherapy group. The advantage of an intense and prolonged response was also reflected in the randomized population. Meanwhile, as per the number needed to treat (NNT) analysis, the nivolumab + chemotherapy group maintained a consistent advantage over the chemotherapy group on the basis of OS, PFS, and ORR in the whole population and the CPS ≥5 subgroup. The prevalence of TRAEs was considerably higher in the nivolumab + chemotherapy group than in the chemotherapy alone group (22% vs. 12%) with more grade 3-4 TRAEs (59% vs. 44%). However, the nivolumab arm showed a lower risk of deteriorating symptoms than the chemotherapy arm (CPS ≥5, HR, 0.64; overall patients, HR, 0.77). Additionally, the nivolumab + chemotherapy group was associated with improved quality-adjusted time without symptoms or toxicity (Q-TWiST) compared to the chemotherapy group. Improving quality of life (QOL) also helps clinicians better manage patients (50)(51)(52).
A similar trial, ATTRACTION-4, enrolled 724 Asian patients with GC/GEJC from Japan, Korea, and Taiwan. The trial evaluated either nivolumab or placebo plus chemotherapy (oxaliplatin + capecitabine or fluoropyrimidine S-1). Although the OS between the two arms did not differ significantly (p = 0.26), the mPFS of the nivolumab arm was nearly 2 months longer than that of the placebo arm (10.45 vs. 8.34 months; HR, 0.68). The KM curves for PFS separated early, and the nivolumab arm consistently had superior PFS rates than the placebo arm. Additionally, regardless of PD-L1 expression levels, the nivolumab arm had a better antitumor response. The ORR was nearly 10% greater in the nivolumab arm than that in the placebo arm (57% vs. 48%). The nivolumab arm was associated with improved survival and 4-month longer DoR than the placebo arm (12.91 vs. 8.67 months). Although the nivolumab + chemotherapy group had more frequent TRAEs than the placebo + chemotherapy group, including grade ≥3 TRAEs, serious TRAEs, and TRAEs leading to treatment discontinuation, the types of TRAEs were consistent with those previously associated with chemotherapy and nivolumab treatment. The researchers determined that the toxicity of chemotherapy plus nivolumab was manageable, and that nivolumab combined with chemotherapy helped maintain QOL (53,54). Compared to the CheckMate 649 trial, the ATTRACTION-4 trial enrolled Asian patients only and had more patients receiving subsequent anticancer drugs, which may be one of the reasons for the mOS difference between trials. Both trials added oxaliplatin as a chemotherapeutic agent and achieved good results, indicating that oxaliplatin works well in combination with nivolumab.
Pembrolizumab monotherapy was also explored as a first-line treatment for GC. The KEYNOTE-062 trial was established based on the positive outcomes of the KEYNOTE-059 and KEYNOTE-060 trials; however, KEYNOTE-062 did not achieve the desired results. The GC/GEJC population with CPS ≥1 was allocated to three arms: pembrolizumab or placebo plus chemotherapy (cisplatin combined with fluorouracil/capecitabine) and pembrolizumab alone. Analyses were performed based on CPS ≥10 (n = 281) and MSI-H (n = 50) subgroups. Among the overall study population with CPS ≥1, the pembrolizumab arm showed a lower OS compared with the chemotherapy arm (HR, 0.91) but approximately 1% and 6% higher 1-and 2-year OS, respectively. Pembrolizumab had a survival advantage over chemotherapy (HR, 0.91) and induced a longer DoR (13.7 vs. 6.8 months), suggesting that pembrolizumab had a long-term beneficial effect. In the CPS ≥10 cohort (n = 281), the pembrolizumab monotherapy arm seemed to have a clinical advantage over the chemotherapy arm, although the difference was not tested statistically (mOS, 17.4 vs. 10.8 months; HR, 0.62). The pembrolizumab arm had fewer TRAEs (54.3% vs. 91.8%) and grade ≥3 TRAEs (16.9% vs. 69.3%) than the chemotherapy arm. The overall population with CPS ≥1 was able to maintain health-related quality of life (HRQOL) when treated with pembrolizumab alone or pembrolizumab plus chemotherapy. A correlation between clinical efficacy and TMB in the pembrolizumab arm was proposed at a later stage of the study. The findings remained consistent at the 54.3-month follow-up, with the CPS ≥1 and CPS ≥10 subgroups treated with pembrolizumab having 8% and 18% greater 2-year OS than those treated with chemotherapy, respectively (55-58). Despite the lack of survival benefits compared to chemotherapy, pembrolizumab achieved better clinical benefit in the CPS ≥10 cohort than in the CPS ≥1 subgroup, suggesting that increased PD-L1 expression levels may improve OS for patients with GC. These findings seemed comparable to those in the CheckMate 649 trial. In contrast to the KEYNOTE-811 and ATTRACTION-4 trials, the KEYNOTE-062 trial used cisplatin rather than oxaliplatin, which may have led to differences in outcomes. In the ongoing KEYNOTE-859 trial, researchers are exploring the clinical effectiveness of pembrolizumab in combination with chemotherapy using 5-fluorouracil + cisplatin or capecitabine + oxaliplatin as the chemotherapeutic agents (59).
More trials investigating the combination of PD-1/PD-L1 antibodies and chemotherapy for GC/GEJC treatment are ongoing. The ORIENT-16 trial is exploring the clinical efficacy of sintilimab + oxaliplatin + capecitabine (60). The BGBA317305 trial (NCT03777657) is investigating the clinical efficacy of tislelizumab in combination with oxaliplatin + capecitabine or cisplatin + 5 fluorouracil (61). The above clinical trial results highlight that chemotherapy remains the mainstream first-line combination treatment for EC and GC for the time being. Studies exploring PD-1 antibody monotherapies have not yet demonstrated clinical advantages; however, the impact of different PD-L1 expression cutoffs on patient outcomes may influence future ICI studies.
PD-1/PD-L1 blockade as second-line or more treatment in esophageal cancer
Abundant PD-1/PD-L1 antibodies are involved in secondline treatment studies of EC and GC. Both monotherapies and combination therapies have demonstrated good applicability, and research is now focused on the possible applications of PD-1 antibody monotherapy as second-line or more treatments. Many of these agents have been approved by the FDA, including pembrolizumab, which has been approved for previously treated unresectable/metastatic MSI-H/dMMR or TMB-H solid tumors, including EC and GC (62, 63). Dostarlimab-Gxly is a second-line or more therapeutic option for MSI-H/dMMR GEC (64). Meanwhile, nivolumab is recommended for advanced ESCC (category 1), and pembrolizumab is also recommended for advanced ESCC with CPS ≥10 (category 1) ( Table 2).
Based on the positive outcomes of the KEYNOTE-180 and KEYNOTE-181 trials, the FDA approved pembrolizumab in 2019 as a second-line treatment for locally advanced/metastatic ESCC with CPS ≥10 (65). The phase II KEYNOTE-180 trial enrolled patients with advanced ESCC (n = 63) or EAC who had undergone second-line or more treatment, and patients were administered pembrolizumab for subsequent treatment. PD-L1-positive expression was defined as CPS ≥10. Antitumor responses were observed in the overall population (ORR, 9.9%), CPS ≥10 subgroup (ORR, 13.8%), and CPS <10 subgroup (ORR, 6.3%). Pembrolizumab conferred a significant survival advantage (OS, 5.8 months; 6-month OS, 49%; 12-month OS, 28%) and was deemed to be safe (TRAEs, 12.4%). The results suggested that PD-L1 expression levels may enhance the response to pembrolizumab in patients with ESCC or EAC (66,67). In the subsequent multicenter KEYNOTE-181 trial, 528 patients (63.9%) were treated with pembrolizumab or chemotherapy (irinotecan, paclitaxel, or docetaxel). The survival advantage of pembrolizumab was more pronounced than that of chemotherapy for Asian patients. Additionally, pembrolizumab did not prolong mOS in all patients but presented a notable survival benefit in the CPS ≥10 subgroup. Among the CPS ≥10 cohort, the pembrolizumab arm had an OS advantage of almost 2.6 months over the chemotherapy arm (9.3 vs. 6.7 months), 20% greater 1-year OS (43.0% vs. 20.4%), and reduced risk of death (PFS, HR, 0.73). Among patients with ESCC, the 12-month PFS increased by 7% (16.7% vs. 7.4%). The most significant improvement in survival was observed in patients with ESCC with CPS ≥10 (HR, 0.64). An antitumor response advantage was reported in the pembrolizumab arm over the chemotherapy arm in the patients with ESCC (ORR, 16.7% vs. 7.4%), CPS ≥10 subgroup (ORR, 21.5% vs. 6.1%), and the randomized population (ORR, 13.1% vs. 6.9%). The 9-month response rate to pembrolizumab was higher than that to chemotherapy (53.5% vs. 38.1%), indicating a longer duration of response. The pembrolizumab arm had almost 20% fewer TRAEs and grade ≥3 TRAEs than the chemotherapy arm, and both sets of patients had similar HRQOL values, suggesting that pembrolizumab had a superior safety profile. However, the cost of pembrolizumab treatment far exceeded that of chemotherapy by $37,201.68. Health practitioners may value the application of pembrolizumab as a second-line therapy for EC (68-70). Both trials supported pembrolizumab monotherapy as a second-line treatment for EC. Furthermore, pembrolizumab showed greater efficacy in ESCC.
A growing number of newly developed PD-1 antibody single agents are being investigated in ESCC, and most trials have been conducted in China, where ESCC is the major subtype of EC. In the multicenter RATIONALE-302 trial, tislelizumab or chemotherapy (irinotecan, docetaxel, or paclitaxel) were administered to patients with metastatic or advanced ESCC. Tislelizumab is a specific antibody designed to target PD-1. PD-L1 expression was estimated using tumor area positivity (TAP), with TAP ≥10% set as the criterion for positive PD-L1 expression. In the overall population, the tislelizumab arm displayed an OS advantage over the chemotherapy arm (8.6 vs. 6.3 months; HR, 0.70). The mPFS was shorter in the tislelizumab arm than in the chemotherapy arm, but the KM curves for PFS began to separate at 3 months and the PFS rates for the tislelizumab arm remained progressively higher than those of the chemotherapy arm (6-month PFS, 21.9% vs. 14.9%; 12-month PFS, 12.7% vs. 1.9%). The tislelizumab arm had an OS advantage over the chemotherapy arm in the TAP ≥10% subgroup (10.3 vs. 6.8 months; HR, 0.54), TAP <10% subgroup (HR, 0.82) and TAP unknown subgroup (HR, 0.67). The OS advantage was demonstrated regardless of PD-L1 expression levels, as determined by post-hoc interaction analysis. The ORR of the tislelizumab arm was 10% higher than that of the chemotherapy arm (20.3% vs. 9.8%), indicating a longer-lasting antitumor response. The tislelizumab arm experienced fewer TRAEs and grade ≥ 3 TRAEs than the chemotherapy arm. Patients with advanced ESCC treated with tislelizumab demonstrated clinical improvement in OS (HR, 0.70) and a lower decline in physical function, leading to extended HRQOL (71,72).
The phase 2 ORIENT-2 trial explored sintilimab as a secondline monotherapy for ESCC. The trial enrolled 190 patients with metastatic or advanced ESCC who were randomly assigned to the sintilimab or chemotherapy (paclitaxel or irinotecan) arms of the study. The mOS of the sintilimab arm was 1 month longer than that of the chemotherapy arm (7.2 vs. 6.2 months; HR, 0.70). The survival advantage of sintilimab over chemotherapy showed a longer tendency in the 12-month OS (37.4% vs. 21.4%) and 12month PFS (10.7% vs. 1.9%). The sintilimab arm also had a superior safety profile than the chemotherapy arm (grade ≥3 TRAEs, 20.2% vs. 39.1%). The restricted mean survival time (RMST) and Fleming-Harrington tests led to the conclusion that sintilimab treatment for ESCC was associated with prolonged response and possible long-term survival. Biomarker analysis revealed that patients with a low neutrophil-to-lymphocyte ratio (NLR) (NLR <3) 6 weeks after sintilimab treatment had a substantial survival benefit over those with NLR >3 (OS, 14.0 vs. 6.2 months; PFS, 2.9 vs. 1.5 months). Moreover, low molecular tumor burden index (mTBI) in peripheral blood was associated with PFS (HR, 0.55), demonstrating the clinical significance of mTBI in sintilimabtreated patients. Based on these findings, researchers recommended the combination of low mTBI with high T-cell receptor clonality and NLR <3 at 6 weeks after treatment as biomarkers for predicting survival outcomes (OS and PFS) of sintilimab-treated patients with ESCC (73).
In addition to these trials, the ESCORT trial investigated camrelizumab monotherapy as a second-line treatment for advanced/metastatic ESCC in China (74), while the ATTRACTION-3 trial explored nivolumab monotherapy as a second-line therapy for advanced/metastatic ESCC (75). The above trials supported the popularity of PD-1 antibodies as monotherapies in second-line or more therapy studies in EC because Asian patients accounted for the majority of participants in these studies. In addition, regional differences were reflected in the KEYNOTE-181 study with Asian patients benefiting more from PD-1 blockade treatment than non-Asian patients, although the RATIONALE-302 trial did not report the same results. Additionally, different trials used different PD-L1 expression criteria, and the ORIENT-2 trial did not predict the absolute benefit of sintilimab treatment despite the use of both TPS and CPS. The exploration of appropriate predictive markers remains a pending issue.
PD-1/PD-L1 blockade as first-line maintenance therapy and second-line or more treatment in gastric cancer Unlike EC, nivolumab and pembrolizumab monotherapies have not been authorized by the FDA as second-line treatments for GC. The conventional second-line treatment for GC is ramucirumab alone or in combination with paclitaxel (76); single-agent paclitaxel, docetaxel, and irinotecan are also suggested as category 1 therapies.
The phase 3 JAVELIN Gastric 100 trial explored the clinical effectiveness of avelumab applied to GC/GEJC as a maintenance therapy after primary induction chemotherapy. Avelumab did not markedly improve OS in either the PD-L1 expression on ≥1% of tumor cells (defined as PD-L1-positive) subgroup or randomized population. The KM curves for OS were lower in the avelumab arm than in the chemotherapy arm until 12 months. However, once the two curves crossed over, the avelumab arm preserved a trend toward higher OS, outperforming the chemotherapy arm by approximately 6% at 24-month OS (22.1% vs. 15.4%). The 1-year DoR and 2-year responses for the avelumab arm were approximately two and four times longer than those for the chemotherapy arm, respectively. In the CPS ≥1 subgroup, the mOS was comparatively higher in the avelumab arm than in the chemotherapy arm (HR, 0.72). Grade ≥3 AEs, TRAEs, and severe TRAEs occurred less frequently in the avelumab arm than in the chemotherapy arm. Although the JAVELIN Gastric 100 trial did not reach the primary endpoint of OS improvement, the potential survival benefits and excellent safety profile of avelumab in longterm treatment are informative (77). The JAVELIN Solid Tumor trial (78) also investigated the efficacy of avelumab as a first-line maintenance therapy for tumors. Although the trial data did not show a significant advantage over chemotherapy, the favorable 12month OS and PFS in the JAVELIN Solid Tumor trial suggest a lasting effect of avelumab in long-term first-line maintenance treatment for patients with GC.
As a second-line treatment, pembrolizumab monotherapy in the phase 2 KEYNOTE-059 trial demonstrated good efficacy in advanced GC/GEJC. The phase 3 KEYNOTE-061 trial enrolled 395 patients with GC/GEJC with CPS ≥1 for subsequent administration of pembrolizumab or chemotherapy (paclitaxel). In the overall population, pembrolizumab did not demonstrate superiority in terms of OS (HR, 0.82). In the longterm follow-up, the KM curves separated at 8 months, after which the pembrolizumab arm had greater 12-month (13%) and 18-month (11%) OS than that in the chemotherapy arm. The superior response time of the pembrolizumab arm compared to the chemotherapy arm (18.0 vs. 5.3 months) suggests a survival advantage in long-term therapy. In the CPS ≥10 cohort, the OS of the pembrolizumab arm was 2.4 months longer than that of the chemotherapy arm (HR, 0.64). Pembrolizumab was associated with fewer toxic events than paclitaxel, including TRAEs, grade ≥3 TRAEs, and AEs leading to treatment discontinuation. The pembrolizumab and paclitaxel arms had comparable HRQOL scores. In the CPS ≥1 subgroup, the pembrolizumab arm had prolonged mOS compared to the paclitaxel arm (HR, 0.81), and the pembrolizumab arm had approximately 15% greater ORR than the paclitaxel arm in the CPS ≥10 cohort. The difference in 2-year OS between the Additionally, the efficacy of pembrolizumab (PFS and ORR) progressively improved with increasing PD-L1 expression levels.
In the CPS ≥1 subgroup, patients with Eastern Cooperative Oncology Group performance status (ECOG PS) 0 fared better when treated with pembrolizumab than with paclitaxel (OS, 12.3 vs. 9.3 months), with different results observed for patients with ECOG PS 1 (OS, 5.4 vs. 7.5 months). These results suggest that patients with better ECOG PS may respond more favorably to pembrolizumab treatment. In the follow-up biomarker analysis, tissue TMB was suggested as a predictor of pembrolizumab treatment in GC, but there are also conflicting views (79)(80)(81)(82)(83)(84)(85).
Both the KEYNOTE-061 and KEYNOTE-062 trials achieved good and durable survival benefits in the CPS ≥10 subgroup, suggesting that patients with GC with high levels of PD-L1 expression may better respond to pembrolizumab, further supporting the use of PD-1 antibodies for patients with GC. The newly launched phase 3 KEYNOTE-063 trial was conducted after the KEYNOTE-061 trial. The KEYNOTE-063 trial enrolled 94 patients with advanced GC/GEJC with CPS ≥1 in Asia. This trial revealed superior results for the safety of pembrolizumab, although no definitive conclusions were reached regarding survival status and antitumor response (86). The use of PD-1 antibodies as second-line or more treatments in GC is worth further exploration. Both the ATTRACTION-2 and CheckMate-032 trials included nivolumab, and the results were of relative clinical value, while nivolumab in the CheckMate-032 had better clinical value than nivolumab plus ipilimumab, suggesting that nivolumab-related studies are deserving of future exploration. Nevertheless, further consideration needs to be given to appropriate control treatments, since conventional second-line chemotherapy drugs may be more comparable than placebo treatments.
PD-1/PD-L1 blockade as perioperative treatment
Combined treatment improves patient survival more than resection alone in patients with localized EC or esophagogastric junction cancer (EGJC) (87,88). Both perioperative and preoperative chemotherapy are routine regimens (89,90). Based on the findings of the CheckMate 577 trial, nivolumab monotherapy was licensed by the FDA in May 2021 for patients with residual disease following preoperative chemoradiation and R0 resection (category 1) (91). In the CheckMate 577 trial, patients with EC/GEJC who received neoadjuvant radiotherapy were recruited and given either nivolumab or switched to a placebo treatment schedule. PFS was roughly twice as long in the nivolumab arm as that in the placebo arm (22.4 vs. 11.0 months; HR for disease recurrence or death, 0.69). The two arms continued to diverge in the KM curves, with nivolumab being continuously superior to the placebo. More AEs were associated with nivolumab treatment than with placebo treatment, but the safety profile was consistent with that of earlier trials. In the subgroup analysis, similar HR values for disease recurrence or mortality were observed for tumor-cell PD-L1 expression ≥1% (HR 0.75) and <1% (HR, 0.73), indicating that the efficacy of adjuvant nivolumab treatment was independent of PD-L1 expression levels (92). According to the CheckMate 577 trial, the European Society of Molecular Oncology recommends nivolumab as standard therapy for patients with EC/GEJC undergoing neoadjuvant chemoradiotherapy, regardless of histologic subtype (93).
Localized GC can also be treated with combination therapy to improve survival. Clinical trials exploring PD-1 antibodies combined with chemotherapy as a neoadjuvant therapy in GC have been conducted. A phase 2 study explored neoadjuvant treatment with capecitabine, sintilimab, and oxaliplatin in locally advanced GC/GEJC before surgical resection. A pathological complete response (pCR) was considered to be a predictor of the long-term benefit of neoadjuvant treatment and was set as the primary endpoint of the study. pCR and major pathological response (MPR) was achieved in 19.4% and 47.2% of the study population, respectively. The researchers attributed the results to the multiple drug combination and a high proportion of the study population with CPS ≥1. The CPS ≥1 subgroup had higher pCR (28.6%) and MPR (57.1%) than the overall population, supporting the use of CPS as a predictive biomarker to screen those who might best benefit from neoadjuvant anti-PD-1 therapy (94). Although not as much attention has been given to PD-1 antibodies in neoadjuvant studies as in firstand second-line treatment studies, many trials are underway. For instance, the KEYNOTE-585 trial has confirmed the effectiveness of perioperative chemotherapy in combination with pembrolizumab in GC (95).
Predictive biomarkers of PD-1/PD-L1 blockade efficacy
As seen from the above clinical trials, many conditions limit the ability of PD-1/PD-L1 blockade to achieve good results, and a considerable number of patients do not respond to therapy. Predictive biomarkers are essential for screening patients before the start of treatment and avoiding adverse effects. This section presents a short summary of common biomarkers used in clinical trials and briefly introduces those that may predict the effectiveness of PD-1/PD-L1 antibodies.
PD-L1 and MSI-H are recommended by the NCCN as common biomarkers in GC and EC. As shown in multiple clinical trials, patients with different PD-L1 expression levels often exhibit differences in response to PD-1 antibodies. In the CheckMate 032 trial, the beneficial effects of nivolumab in combination with ipilimumab increased with higher CPS levels, suggesting the superiority of CPS as a biomarker (96). Although the effectiveness of PD-1 antibodies in some trials was independent of PD-L1 expression levels, this difference may stem from different PD-L1 detection methods, evaluation criteria, and location of the patient. As common molecular subtypes, EBV-positive GC and MSI-H GC were both associated with enhanced ORR and PD-L1/PD-1 antibody efficacy, with EBV-positive GC having close to 100% ORR (28). Patients with MSI-H GC may have shorter PFS and lower ORR when receiving first-line chemotherapy, but higher ORR and PFS was achieved after subsequent PD-1 antibody treatment, supporting the early use of ICIs in MSI-H GC (97). Genome sequencing demonstrated that both EBV-positive GC and MSI-H GC were associated with high PD-L1 expression levels and favorable response to pembrolizumab (98).
Other common biomarkers have also been explored in GC and EC. TMB is associated with better response to PD-1 antibody treatment in EC (99). NLR is one of the leading predictive indicators of nivolumab efficacy in GC, providing a straightforward, easily acquired, and cost-effective biomarker (100). Changes in the gut microbiome were found in the DELIVER trial, in which the mechanism for bacterial invasion of epithelial cells was related to nivolumab clinical outcomes and progressive disease, suggesting a potential novel biomarker for predicting treatment response to nivolumab in advanced GC (101). Numerous predictive biomarkers have been investigated in clinical trials of GC and EC, but practical biomarkers need to be validated by credible findings.
Conclusions and perspectives
The standard of care for EC and GC has long revolved around chemotherapy and surgery. Along with research progress in targeted therapies, PD-1/PD-L1 antibodies continue to be investigated in clinical trials as reliable ICIs. This review presents an overview of the molecular and immunological background of PD-1/PD-L1 antibody applications, summarizes recent clinical trials investigating PD-1/PD-L1 blockade in EC and GC/GEJC, and briefly introduces common predictive biomarkers that could be further investigated. However, the clinical trials described herein have various potential problems that complicate the evaluation of their results. For example, some trials specified PD-L1 expression levels as an inclusion criterion, whereas other trials only explored PD-L1 expression in subgroup analyses. Furthermore, subgroups with different CPS cutoff values yielded varied CPS scores for survival results, while different PD-L1 expression detection methods might further skew conclusions when comparing trial results. Moreover, small disparities between patient locations, cancer types, and control groups affected trial outcomes and the ability to draw meaningful conclusions across trials. Indeed, the proportion of Asian patients in the study population may affect study outcomes. In addition, some chemotherapeutic drugs may affect the TME and impact the effectiveness of PD-1/PD-L1 antibodies (102, 103). Although PD-1/PD-L1 antibody treatment can prolong the life of some patients with GEC, the increased incidence of adverse effects when combined with chemotherapy cannot be ignored, and patients may develop a reduced tolerance to the drug, thereby risking treatment discontinuation. Finally, PD-1/ PD-L1 antibodies are more expensive than conventional treatments, and both PD-L1 testing and dosing portals increase the cost of patient treatment. The above issues should be considered by investigators when designing future trials.
As immunotherapy research continues to advance, we believe that modalities of PD-1/PD-L1 blockade in EC and GC will further evolve. Here, we review and advise on common related issues ( Table 3). First-line treatment in EC and GC has been extensively studied in combination with chemotherapy, and the choice of chemotherapeutic agents has been compared for effectiveness, while treatment alone has not yielded good results. Along with radiotherapy (104), CTLA-4 (ipilimumab), HER2 [trastuzumab (105) and margetuximab], and vascular endothelial growth factor receptor-2 (VEGFR-2) (106) antibodies are also being explored in clinical trials; studies on PD-1/PD-L1 in combination with other therapeutic modalities are promising. In response to the poor results of classical PD-1 antibody in a first-line trial, it is possible to investigate the application of PD-1 monotherapy in a strictly screened range of patients, such as PD-L1 CPS cutoffs, molecular subtypes, pathological types, and immune cell levels. Moreover, studies of biomarker detection can be performed in parallel with trials on subgroup analysis. Many PD-1 antibodies have been used in clinical studies for second-line therapy, but only pembrolizumab is used as the first choice in CPS ≥10 ESCC, with the others suggested as second-line treatment options. Other PD-1 antibodies might be tested in trials to determine their suitability in a range of patients through subgroup analysis. The new PD-1 antibody tislelizumab/sintilimab monotherapy study focused mainly on Asian ESCC patients, and the new drug could be considered for validation in a large clinical trial, including EC patients worldwide. Non-Asian regions have different pathology type proportions. How to control the balance of patient proportions needs to be considered when enrolling patients in future studies. Considering that avelumab has not achieved a clear advantage in first-line maintenance therapy, conventional PD-1 antibodies could be taken into consideration. Perioperative therapy emphasizes the importance of PD-1/PD-L1 antibodies in neoadjuvant therapy, while PD-1 antibodies in neoadjuvant therapy are typically administered as a combination or monotherapy following chemotherapy. Future studies must focus on the effect of PD-1 antibodies alone and apply PD-1 antibodies to other stages of perioperative therapy. As PD-1/PD-L1 antibodies in the CPS ≥1 subgroup are analyzed effectively in neoadjuvant therapy, whether PD-L1 routine testing is applicable to patients who could receive neoadjuvant therapy should be further investigated. In terms of biomarkers, HER2, MSI-H, and PD-L1 are currently used in testing, but new potential biomarkers are needed for HER2-, MSI-H-, and PD-L1-negative patients. Bioinformatics analysis to screen tumor cell gene expression characteristics or molecular pathways, as well as cellular and cytokine changes in the TME, may provide suitable combinatorial biomarkers. Overcoming the abovementioned drawbacks and exploring the best therapeutic outcomes in patients with complex EC and GC will help future investigators design valuable clinical trials, yielding beneficial outcomes.
Publisher's note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
|
v2
|
2022-11-24T14:20:55.366Z
|
2022-11-24T00:00:00.000Z
|
253841584
|
s2orc/train
|
Identification and experimental validation of a tumor-infiltrating lymphocytes–related long noncoding RNA signature for prognosis of clear cell renal cell carcinoma
Clear cell renal cell carcinoma (ccRCC) is a common aggressive malignant tumor of the urinary system. Given the heterogeneity of the tumor microenvironment, immunotherapy may not fully exert its role in the treatment of advanced patients. Long noncoding RNA (lncRNA) has been reported to be critically associated with the differentiation and maturation of tumor-infiltrating lymphocytes (TILs), which work against tumor cells. In this study, we identified 10 TIL-related lncRNAs (AL590094.1, LINC02027, LINC00460, AC147651.1, AC026401.3, LINC00944, LINC01615, AP000439.2, AL162586.1, and AC084876.1) by Pearson correlation, univariate Cox regression, Lasso regression, and multivariate Cox regression based on The Cancer Genome Atlas (TCGA) database. A risk score model was established based on these lncRNAs. Next, a nomogram was constructed to predict the overall survival. By employing differentially expressed genes (DEGs) between groups with high and low risk scores, gene ontology (GO) enrichment analysis was performed to identify the major biological processes (BP) related to immune DEGs. We analyzed the mutation data of the groups and demonstrated that SETD2 and BAP1 had the highest mutation frequency in the high-risk group. The “CIBERSORT” R package was used to detect the abundance of TILs in the groups. The expression of lymphocyte markers was compared. We also determined the expression of two lncRNAs (AC084876.1 and AC026401.3) and their relationship with lymphocyte markers in the kidney tissue of ccRCC patients and showed that there was a positive correlation between AC084876.1 and FoxP3. Proliferation, migration, and invasion of AC084876.1-downregulated ccRCC cell lines were inhibited, and the expression of PD-L1 and TGF-β secretion decreased. To our knowledge, this is the first bioinformatics study to establish a prognostic model for ccRCC using TIL-related lncRNAs. These lncRNAs were associated with T-cell activities and may serve as biomarkers of disease prognosis.
Introduction
Renal cell carcinoma (RCC) is a common malignant tumor of the urinary system. Globally, 431,288 new cases (accounting for 2.2% of all cancer cases) and 179,368 new deaths (accounting for 1.8% of all cancer deaths) occurred in 2020, second only to prostate cancer and bladder cancer (1). In 2021, the number of new cases of kidney cancer in the United States was 76,080, ranking sixth in the male cancer incidence rate and ninth in the female cancer incidence rate (2). RCC has several subtypes, and about 70% of individuals receive a diagnosis of clear cell RCC (ccRCC). Although ccRCC is a disease that can be detected early and successfully treated by surgery, up to onethird of cases relapse or develop metastasis (3). ccRCC is not sensitive to radiotherapy and chemotherapy. The application of immunotherapy for different targets brings hope to these patients (4,5); however, the effective response rate of the treatment is limited. Therefore, it is necessary to find new effective immunotherapeutic targets to improve patients' prognoses.
Immune cells in the tumor microenvironment play an important role in regulating tumor progression and serve as attractive therapeutic targets (6). Moreover, ccRCC is prone to immune infiltration, and the characteristics of the tumor microenvironment strongly affect the response to immunotherapy (7). A successful antitumor immune response requires the activation and synergy of multiple tumorinfiltrating lymphocytes, including T cells, B cells, natural killer cells (NK cells), and their subtypes. These cells play a positive and negative regulatory role in the process of antitumor immunity and kill tumor cells (8). Available studies have constructed signatures related to immune infiltration based on different biological characteristics in ccRCC. Moreover, specific lymphocyte related signatures such as CD8 + T cells (9), CD39 + CD8 + T cells (10), and TNFRSF9 + CD8 + T cells (11) have been identified in ccRCC. However, previous studies did not focus on the impact of tumor-infiltrating lymphocyte profile on the prognosis of patients with ccRCC, so they could not accurately demonstrate the potential role of immune cells in the treatment of ccRCC.
Long noncoding RNAs (lncRNA) have a length of over 200 nucleotides and are dynamically expressed in the immune system and regulate the differentiation and function of immune cells (12). Previous studies have shown that lncRNAs are dysregulated in cancer and that they have important effects on tumor proliferation, angiogenesis, apoptosis, and metastasis by regulating the formation of the tumor immune microenvironment (13). For instance, lncRNA LIMIT has been identified to be associated with tumor-infiltrating T cells. Silencing LIMIT could impair antitumor immunity and blunt immunotherapy efficacy (14). Breast cancer-derived exosomal lncRNA SNHG16/miR-16-5p/SMAD5-regulatory axis can induce CD73 expression in gdT cells, thereby enhancing the immunosuppressive function (15). LINC00301 is highly expressed in non-small cell lung cancer and functions to increase regulatory T cells while decreasing the CD8+ T cell population (16). A clinical trial suggested that lncRNA was associated with immunotherapeutic overall survival benefits superior tumor alteration burden, programmed cell death ligand 1 (PD-L1) expression, and cytotoxic T-lymphocyte (CTL) infiltration (17). Based on the above findings, lncRNAs are considered a potential target for immunotherapy and have attracted extensive attention in the field of cancer treatment research.
With the maturation in the application of high-throughput sequencing technology, it has become possible to explore the expression levels of lncRNAs in ccRCC and their relationship with tumor-in filtrating lymphocytes; additionally, bioinformatics analysis can be applied to further explore the relevant mechanisms. In this study, we identified a signature consisting of 10 TIL-related lncRNAs to predict the prognosis of patients with ccRCC and explored the potential mechanism. Human renal tissues were used to detect the expression levels of related lncRNAs and immune markers. Finally, the biological functional effect of AC084876.1 on the ccRCC cell line was examined by in vitro experiments.
Ethical statement
This study was approved by the Ethics Committee of Guangzhou First People's Hospital, School of Medicine, South China University of Technology. Informed consent forms were signed by all patients. In accordance with the ethical and legal standards, seven pairs of matched frozen samples of ccRCC and benign renal tissue adjacent to cancer from ccRCC patients were handled and made anonymous.
Cell lines
A human ccRCC cell line 786-O was obtained from ATCC (American Type Culture Collection, Manassas, Virginia, USA). The cells were cultured at 37°C in a 5% CO 2 incubator with RPMI-1640 (MA0215, Meilunbio, Dalian, China) medium containing 10% fetal bovine serum, streptomycin, and penicillin.
Data and tissue processing
The RNA-sequencing data and corresponding clinical information of patients with ccRCC were downloaded from the C ancer Genome Atlas (TCGA-K IRC, https:// cancergenome.nih.gov/) database. A total of 507 patients with follow-up information were identified and used for further analysis. The total RNA expression data were standardized through log2 transformation.
Seven pairs of the matched frozen samples of ccRCC and benign renal tissue adjacent to cancer used in this study were all from the Guangzhou First People's Hospital. The included patients did not receive chemotherapy or radiotherapy before surgery. Each case was diagnosed and graded by two pathologists separately and re-examined by hematoxylin-eosin staining.
Construction of a prognostic TIL-related lncRNAs signature in ccRCC
The correlation between lncRNAs and TIL-related genes (18) was calculated using the Pearson correlation analysis. The correlation coefficient |R| > 0.3 and P< 0.001 were the criteria for TIL-related lncRNAs. Univariate Cox regression was applied to assess the prognostic value of TIL-related lncRNAs. TIL-related lncRNAs with P< 0.001 in univariate analysis were incorporated into least absolute shrinkage and selection operator (Lasso) regression. Then, TIL-related lncRNAs identified by Lasso were included in a multivariate Cox model to establish a risk score. Finally, we identified 10 TIL-lncRNAs associated with the prognostic risk to construct a prognostic risk score. The risk score of KIRC patients was calculated as follows: Risk score = o n i=1 b i  (expression of lncRNA i ). Patients in the cohort were divided into high-and low-risk groups according to the median risk score.
Validation of the TIL-related lncrnas signature in ccRCC
Based on the TCGA-KIRC cohort, the expression of TILrelated lncRNAs was measured by 'tinyarray' R package. The Kaplan-Meier (K-M) survival curves of overall survival (OS) were used to evaluate the clinical prognostic value of TIL-related lncRNAs. Cytoscape software 3.7.2 was applied to visualize coexpression networks between TIL-related lncRNAs and TILrelated genes. Sankey diagram was used to assess the association between prognostic TIL-related lncRNAs, TIL-related genes, and risk types. The K-M survival curves were created to compare the overall survival of high-and low-risk groups according to predictive signatures. The receiver operating characteristic (ROC) curve from the "survivalROC" R package was used to assess the sensitivity and specificity of the signature. Univariate Cox regression and multivariate Cox regression were applied to assess the prognostic value of the signature.
Construction and validation of a nomogram
Based on age, clinical stage, risk score, and pathological grade, a nomogram for OS was developed with the R package "rms" to predict the 1-year, 3-year, and 5-year relapse-free survival probability. The concordance index (C-index) was used to evaluate the consistency between the nomogrampredicted results and the actual observed results. A calibration curve was used to show the difference between the predictions of the model and the real outcomes. ROC analysis was performed to evaluate the predictive ability of the risk score. The web-based OS probability calculators were built using packages "DynNom" and "shiny" in R software.
Functional enrichment
Differentially expressed genes (DEGs) of mRNA between the high-risk and low-risk groups were identified using package limma in R, with thresholds of |log2 fold change (FC)| > 2 and adjusted P< 0.05. Then, gene ontology (GO) enrichment analysis was performed to find the major biological process (BP) related to immune DEGs. The visual GO enrichment maps of the annotation analysis results were obtained by R with the "ggplot2" and "GOplot" packages.
Somatic mutation analysis
Mutation data of the high-risk group and the low-risk group were analyzed and visualized using the "maftools" package. Mutation information for each gene in each sample was demonstrated by waterfall plots.
Landscape of tumor-infiltrating immune cells based on the signature
To evaluate the effect of the signature on immune cells, CIBERSORT (Cell-type Identification by Estimating Relative Subsets of RNA Transcripts; http://cibersort.stanford.edu) was used to measure the abundance of tumor-infiltrating immune cells in a gene expression matrix by linear support vector regression (19). The abundance of lymphocyte profiles in the TCGA-KIRC dataset was obtained via the "CIBERSORT" R package. The infiltration levels in the high-risk group and the low-risk group were visualized by the "ggplot" R package.
Real-time quantitative PCR assay
Total RNA was extracted from ccRCC and corresponding benign renal tissue samples using NucleoZOL reagent (740404.200, Meilunbio). Total RNA was then reversetranscribed into cDNA using HiScript ® III RT SuperMix for qPCR (R323-01, Vazyme, Nanjing, China). Real-time quantitative PCR (RT-qPCR) was carried out using the AceQ ® qPCR SYBR Green Master Mix (Q141-02, Vazyme). The relative expression of lncRNA was calculated based on the internal reference b-actin. All experiments were carried out in three replicates. The primers applied are listed in Supplementary Table S1.
Western blot assay
Quantitative analysis of protein expression in clinical tissues was performed using western blotting in accordance with the protocol of our previous study (20). The following antibodies were applied in the assay: b-Actin
Transfection of cell lines
Human AC084876.1-specific siRNA and negative control siRNA were purchased from Tsingke Biotechnology Co., Ltd. (Beijing, China). The siRNA sequences are shown in the Supplementary Table S1. RT-qPCR analysis was conducted 72 h after transfection to test the transfection efficacy.
Cell migration and invasion assay
Cell migration and invasion were measured by a wound healing assay and transwell assay in line with the protocol from our previous study (20).
Quantification of transforming growth factor-beta
The assay was conducted using a human TGF-b ELISA kit (JM-04929H1, Jingmei, Jiangsu, China). The cell culture medium was tested in accordance with the manufacturer's instructions. The results were normalized using cell counts.
Statistical analysis
All statistical analyses were performed in R software (version 4.03) and GraphPad Prism 8 (GraphPad Software, United States). Continuous variables are shown as the means ± standard deviations. Student's t test or analysis of variance (ANOVA) was used to determine the statistical significance of quantitative data. P< 0.05 was regarded as statistically significant.
Identifying 10 TIL-related lncRNAs in 507 patients with ccRCC
This study was performed following the workflow shown in Figure 1. A total of 589 candidate lncRNAs related to tumorinfiltrating lymphocytes (TIL) were identified by Pearson's correlation analysis (Supplementary Table S2). To further evaluate the prognostic value of these candidate lncRNAs, we performed univariate Cox regression analysis with the p-value of 0.05 as the cutoff threshold; 334 lncRNAs were detected as prognostic TIL-lncRNAs ( Figure 2A and Supplementary Table S3). Next, by leveraging the Lasso algorithm, these 334 TIL-lncRNAs were narrowed down to 20 with the optimal lambda of 0.06175523 (Figures 2B, C and Supplementary Table S4). Furthermore, among the 20 TIL-lncRNAs, 10 lncRNAs were found by multivariate Cox regression analysis to be independent prognostic factors in ccRCC ( Figure 2D and Supplementary Table S5). Among these 10 TIL-lncRNAs, seven lncRNAs, i.e., Flowchart of the study strategy. Next, correlation analysis was conducted. As shown in Figure 2E, AC084876.1, LINC00944, and AC026401.3 had an obvious positive correlation with other lncRNAs, whereas AP000439.2 had a negative correlation with other lncRNAs. Taken together, 10 TIL-lncRNAs, which significantly correlated with TIL and were significant in the univariate Cox regression model, LASSO algorithm, and multivariate Cox proportional hazards regression model, were selected as the most important TIL-lncRNAs for further analysis.
Evaluation of the 10 TIL-lncRNAs signature
Based on the median cutoff value, patients in the cohort were divided into the high-risk group (n = 253) and the low-risk group (n = 254) ( Figure 4A). As shown in Figure 4B, the death probability of high-risk patients was higher than that of low-risk patients. It was also found that the expression levels of AC026401.3, LINC00944, and AC084876.1 were visibly higher, and the expression levels of AP00439.2, AC14765.1, and LINC02027 were downregulated in the high-risk group ( Figure 4C).
The K-M survival curves showed that patients with a high risk score had significantly poorer overall survival than those with a low risk score ( Figure 4D). Then, time-dependent ROC curves were used to assess the predictive performance of the TILrelated lncRNA signature ( Figure 4E). The area under the ROC (AUC) was 0.750, indicating that the TIL-lncRNA signature was a reliable prognostic indicator for predicting OS in ccRCC. In addition, univariate Cox regression and multivariate Cox regression analysis were employed to assess the independent prognostic value of the signature with the following factors: risk score and relevant clinical factors (age, gender, grade, clinical stage, tumor stage [T], and metastasis stage [M]). Node stage was excluded as much data were missing. All of the factors except gender were significantly associated with OS in univariate analysis ( Figure 4F). Multivariate analysis indicated that risk score was still significantly related to OS, suggesting the TILrelated lncRNA signature could serve as an independent prognostic factor for patients with ccRCC ( Figure 4G).
Construction of a TIL-related lncRNA prognostic model to predict the survival
The risk score, age, clinical stage, and pathologic grade were included in the nomogram. As indicated in the nomogram, the risk score had the largest contribution to OS of patients with ccRCC ( Figure 5A). The C-index of the nomogram was 0.693. The calibration curve revealed good agreement between the predicted and observed probabilities. All calibration curves of 1-year, 3-year, and 5-year ( Figure 5B) OS were close to the 45degree line. Figures 5C-E, the area under curve (AUC) of the nomograms was 0.853, 0.802, and 0.755 for 1-, 3-, and 5-year OS, respectively. These AUC values of the nomograms were greater than those of every single clinical predictor (i.e., age, grade, and stage), indicating an advantage of combining these risk factors for ccRCC prognosis.
As shown in
For the clinical usability of the model, a dynamic nomogram was created for the prediction of OS probability in patients with ccRCC ( Figure 5F), which was convenient and intuitive for individual prognosis prediction based on the personal characteristics of ccRCC patients (https://zhonglab.shinyapps.io/dynnomapp/).
Functional analysis
As shown in the Volcano plots ( Figure 6A), DEGs were identified with fdr< 0.05 and |logFC| > 0.5. A total of 408 genes were upregulated and 455 genes were downregulated in the lowrisk group compared with the high-risk group (Supplementary Table S6). GO enrichment analysis ( Figure 6B) indicated that these DEGs were mostly related to the pathways involved in Tcell activation, differentiation, proliferation, co-stimulation, and migration.
CIBERSORT analysis ( Figure 6E) indicated that the abundance of activated CD4+ memory T cells, CD8+ T cells, CD4+ follicular helper T cells (TFH), and regulatory T cells (Tregs) was higher in the high-risk group. A higher abundance of activated NK cells was observed in the low-risk group. As shown in Figure 6F, immune markers CD4, CD8a, CD69, CD25, PD-1, LAG3, CD62L, CCR7, FoxP3, CD56, and CD16 were significantly upregulated in the highrisk group with a decrease in TIM-3.
Measuring the expression levels of AC084876.1 and AC026401.3 and immune markers in kidney tissues
Considering the above results, we performed an RT-qPCR assay to detect the expression of AC084876.1 and AC026401.3 in seven pairs of matched frozen samples of ccRCC and benign renal tissue adjacent to cancer from ccRCC patients. Moreover, a western blot assay was conducted to measure the expression of immune markers (including CD4, CD8a, PD-1, and FoxP3, Figure 6H). As shown in Figure 6G, the expression of AC084876.1 and AC026401.3 tended to be higher in cancer tissues although the threshold of statistical significance was not reached. The results of the western blot showed a higher trend for all of the immune markers detected in cancer tissues, but without reaching statistical significance ( Figure S1A). For the correlation analysis, we found that the expression of AC084876.1 positively correlated with FoxP3 ( Figure 6I and differentiation of tumor-infiltrating Tregs (22). To further analyze the potential role of AC084876.1 to regulate tumorinfiltrating Tregs, the expression level of PD-L1 and secreted TGF-b level were also examined. We found that the knockdown of AC084876.1 resulted in decreased expression of PD-L1 ( Figure 7E, P< 0.01) and decreased secreted TGF-b ( Figure 7F, P< 0.01).
Discussion
ccRCC is the most common pathological subtype of renal carcinoma. Although at an early stage patients can benefit from surgical treatment, advanced patients have finite treatment options (23). Immunotherapy has ignited their hope, but the treatment effectiveness is still limited. The important role of lncRNAs in tumor progression has been gradually explored. To date, several prognostic models based on the immune-related lncRNAs have suggested that lncRNAs are involved in the regulation of immune cell-mediated tumor killing in the tumor microenvironment (TME) (24)(25)(26). A large number of immune cells infiltrate the TME, and tumor-infiltrating lymphocytes (TIL) play a key role in the response to immunotherapy (27). Thus, it is worth noting that investigating the potential role of TIL-related lncRNA may identify novel biomarkers for the treatment and prognosis of patients with ccRCC.
In the present study, we established a novel risk coefficient model based on TIL-related lncRNAs through Pearson correlation analysis, univariate Cox regression, Lasso regression, and multivariate Cox regression in 507 ccRCC patients from the TCGA-KIRC dataset. Finally, 10 TIL-related predictor of overall survival in patients with ccRCC. In the nomogram model, risk score contributed the most, and the calibration curves of 1-, 3-, and 5-year survival prediction were close to the ideal value. Also, the AUC values of the 1-, 3-, and 5year nomograms were greater than those of every single clinical predictor, indicating that the nomogram could be clinically helpful. To make the model applicable for clinical use, we constructed a dynamic nomogram for convenient and intuitive individual prognosis prediction based on the personal characteristics of ccRCC patients. Consistent with previous studies based on the immunotherapy effectiveness in ccRCC (28,29), we found that the mutation frequencies of SETD2 and BAP1 were higher and their expression were both downregulated in the high-risk group. SETD2 and BAP1 mutations are associated with metastasis and poor prognosis in ccRCC (30,31). SETD2 mutations also play an important role in promoting ccRCC progression through cellular autophagy inhibition, DNA repair inhibition, and genomic stability perturbation (32, 33). Mutation of BAP1 may engender genomic instability and promote defects in DNA repair pathways (34). Thus, our novel TIL-related lncRNA signature may contribute to the understanding of the potential mechanism of the progression of ccRCC.
GO-BP analysis of DEGs showed that the immune-related biological processes were mostly involved in T-cell activation, differentiation, proliferation, co-stimulation, and migration. By applying CIBERSORT, we found that the abundance of activated CD4+ memory T cells, CD8+ T cells, TFH, and Tregs was higher in the high-risk group. CD8+ T cells are the main executors of killing tumor cells in the immune system. However, high abundance of CD8+ T cells was not associated with a favorable prognosis in ccRCC (35), and it has been reported that CD8+ T cells did not show functional status as exhausted CD8+ TILs (36). Consistently, our study showed that the exhausted T-cell markers, including PD-1, LAG3, and FoxP3, were upregulated in the high-risk group. Tumor-infiltrating CD8+T cells lose the ability to recognize antigens and activate proliferation under the long-term effect of inhibitory cells and factors, thereby leading to the failure of tumor-killing function (37). Moreover, as an important suppressive immune cell, Tregs can inhibit the proliferation of CD8+T cells by secreting TGF-b, IL-10, and IL-35 (38).
Our results demonstrated that the expression of AC084876.1 positively correlated with FoxP3 in ccRCC. FoxP3 is considered an important biomarker to characterize the Tregs, which are an immunosuppressive subset of CD4+ T cells (39,40). It is widely accepted that Tregs, the central mediators of immune suppression, are activated by PD-L1 and are induced to differentiate by TGF-b synthesized from tumor cells (22). PD-L1 expressed on tumor cells interacts with PD-1 on tumorinfiltrating lymphocytes, attenuating effector T-cell responses and allowing tumors to escape immune attack (41). In addition, Foxp3 in naive T cells could be induced by TGF-b, thereby promoting Tregs development (42). Interestingly, we found that knockdown of AC084876.1 resulted in decreased expression of PD-L1 and reduced level of secreted TGF-b. These results suggest that AC084876.1 may play an important role in the progression of ccRCC by regulating the activation and differentiation of Tregs. Tumor-infiltrating Tregs are highly activated (43) and exert suppressive activities on effector cells by inducing apoptosis and inhibiting activation/proliferation (44). It has been reported that Tregs promote tumorigenesis and immunosuppression via increasing consumption of IL-2 and upregulating inhibitory immune checkpoints (45). Importantly, the functional enrichment analysis suggested that our signature was related to T-cell activation, differentiation, and proliferation signaling pathways, and Tregs were significantly increased in the high-risk groups. Thus, we believe this may be related to the potential regulatory mechanism between AC084876.1 and tumor-infiltrating Tregs, which is worthy of further investigation.
Despite a number of immune-related lncRNA signatures, to our knowledge, the present study was the first to identify a TILrelated lncRNA signature to predict the prognosis of ccRCC patients. Compared with previous studies, which only established signatures (26, 46), our novel signature was used to establish a nomogram combined with clinical indicators to accurately predict survival. As for the study with nomogram construction (24), the potential mechanism of our signature was verified with experiments. We showed that the signature was mainly related to the function of T cells. Furthermore, we found that AC084876.1 may serve as a potential therapy target associated with the activation and differentiation of tumorinfiltrating Tregs. However, our study was retrospective, and a larger validation cohort is needed to confirm our conclusions. Moreover, the underlying mechanism of the identified lncRNAs in regulating TILs and tumor progression in ccRCC remains to be further explored.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
Ethics statement
The studies involving human participants were reviewed and approved by the Ethics Committee of Guangzhou First People's Hospital, School of Medicine, South China University of Technology. The patients/participants provided their written informed consent to participate in this study.
|
v2
|
2022-11-24T14:49:10.639Z
|
2022-11-24T00:00:00.000Z
|
253841210
|
s2orc/train
|
Zinc deficiency is associated with gynecologic cancer recurrence
Zinc deficiency can cause various symptoms, including hair loss, anemia, and taste disorders. Recently, the association between cancer and zinc deficiency has received much attention with respect to its antioxidant properties. However, only a few studies have investigated the association between gynecologic cancers and zinc; to date, no studies have evaluated serum zinc status at the onset of gynecologic cancer or the relationship between zinc and cancer recurrence. The objectives of the present study were to determine whether serum zinc concentrations are associated with the development of gynecologic cancer, to clarify serum zinc dynamics between the onset and recurrence of gynecologic cancer, and to identify the associated factors. Accordingly, we retrospectively determined serum zinc concentrations before treatment in gynecologic patients with benign disease or cancer at the Nippon Medical School Chiba Hokusoh Hospital. We investigated anemia and hypoalbuminemia—the most common causes of zinc deficiency—as indicators of hyponutrition to determine the causal relationship of this deficiency with chemotherapy, radiation therapy, and recurrence, which may affect zinc concentration during cancer recurrence. The results indicated that there was no difference in zinc concentration between preoperative cancer patients and noncancer patients and that serum zinc concentrations were not associated with developing gynecologic cancers. However, patients with gynecologic cancer exhibited significantly lower serum zinc concentrations following treatment, and patients with recurrent cancer were 4.8 times more likely to develop zinc deficiency than those with nonrecurrent cancer. A serum zinc concentration of <61 μg/dL was an independent predictor of recurrence. Once zinc deficiency occurred, the recurrence rate of zinc deficiency reached as high as 69%. Overall, our study indicates that zinc deficiency is associated with recurrence in gynecological cancers and physicians should monitor zinc levels during disease management.
Introduction
Zinc deficiency is common in developing countries, and it affects more than 2 billion individuals worldwide (1). Adequate but not excessive zinc intake benefits the general population, and dietary zinc intake may reduce the risk of gastrointestinal cancers, depression, and type 2 diabetes in adults. Zinc supplementation in adults reportedly improves depression, sperm quality, attentiveness, pregnancy rates, and diarrhea, reduces the risk of pneumonia in children, improves zinc deficiency, and promotes growth. Moreover, it ameliorates respiratory tract infections (including COVID-19) because of its antiviral, antioxidant, and anti-inflammatory effects (2).
Because the human body cannot store zinc, a deficiency can occur relatively rapidly, such as from an improper diet. Numerous epidemiological studies have demonstrated a relationship between dietary zinc content and cancer risk. The anticancer effects of zinc are mostly related to its antioxidant properties (3). Being a trace element, zinc is essential in various metabolic processes, including protein synthesis, immune response, and gene expression (4,5). Therefore, zinc deficiency can cause several disorders in the human body (Table 1) (4,(6)(7)(8)(9)(10)(11).
According to the Japan's Practical Guideline for Zinc Deficiency 2018 published by the Japanese Society of Clinical Nutrition, approximately 15% of the Japanese population has inadequate zinc intake (12,13). Several reports have linked digestive cancers to zinc, suggesting that its intake can reduce the risk of digestive, colorectal, and pancreatic cancers (14)(15)(16)(17); however, few studies have demonstrated a relationship between gynecologic cancers and zinc intake. Only three previous observational studies have shown that compared with healthy adults, patients with ovarian, endometrial, and cervical cancers exhibit lower zinc concentrations (18)(19)(20).
A comprehensive PubMed search using the search terms "zinc" (MeSH term) and "gynecologic cancer" (MeSH term) revealed only one report describing changes in serum zinc concentrations before and after treatment of gynecologic cancer. Yanazume et al. prospectively evaluated 28 patients with suspected zinc deficiency before chemotherapy and concluded that chemotherapy may be a risk factor for low zinc concentrations (21). No studies have evaluated serum zinc status at the onset of gynecologic cancer (i.e., zinc deficiency associated with the development of gynecologic cancer) or the relationship of zinc to cancer recurrence. Therefore, the objectives of the present study were to determine whether serum zinc concentrations are associated with gynecologic cancer development, to identify factors that affect serum zinc concentration at the time of recurrence, and to clarify serum zinc dynamics and associated factors between gynecologic cancer development and recurrence. We measured serum zinc concentrations in gynecologic patients with benign disease or cancer treated at the Department of Obstetrics and Gynecology, Nippon Medical School, Chiba Hokusoh Hospital; we d e t e r mi ne d t h e r e l a t i o n s h i p be t w e e n s e r u m z i n c concentrations and the development, treatment, and recurrence of gynecologic cancers.
Ethical approval
The purpose of the study was explained and written informed consent was obtained from all patients. In addition, approval was obtained from the Ethics Committee of the Nippon Medical School, Chiba Hokusoh Hospital (approval number: H-2022-003).
Patient history
From 1/1/2019 to 12/31/2021, serum zinc concentrations of 214 patients with gynecologic cancer were measured at the Nippon Medical School Chiba Hokusoh Hospital, Chiba, Japan. As controls, 120 patients with benign gynecologic diseases before surgery were enrolled. Patients diagnosed with zinc deficiency were treated with oral zinc supplements at 30 mg per day for 30 days (zinc acetate hydrate, Nobelpharma Co., Ltd., Tokyo, Japan) and they received nutritional guidance including a high zinc diet.
Study design
Serum zinc and blood hemoglobin (Hb) concentrations were measured before surgery, during chemotherapy, and at the follow-up examination after treatment in 214 patients with gynecologic cancer, which included 53 patients with cervical
Blood sample analysis
Blood samples were collected during preoperative testing, postoperative chemotherapy, and at follow-up visits after treatment completion. The samples were used to determine white blood cell, neutrophil, and lymphocyte counts using a multi-item automatic blood cell analyzer (XE-2100, Sysmex, Kobe, Japan). A chemistry autoanalyzer (LABOSPECT 008 a, Hitachi, Ibaraki, Japan) was used to measure serum albumin (Alb) levels based on the manufacturer's instructions. Anemia was defined as a blood Hb level of <12 g/dL and hypoalbuminemia as a serum Alb level of <4.1 g/dL. A flame atomic absorption spectrophotometry method (SpectrAA-240, Agilent Technologies, Inc., California, USA) was used to measure serum zinc concentrations within the scope of the Japanese health insurance program using zinc standard solutions (Zn 1000, Cat. No. 48096-1B, 2B, Kanto Chemical Co., Inc., Saitama, Japan); this test was accredited by IA Japan, a member of the International Association for Laboratory Accreditation Cooperation and the Asia-Pacific Accreditation Cooperation for Mutual Recognition. The limit of detection for this test is 0.0009 mg cm −3 with a relative expanded uncertainty of 0.5% (coverage factor k = 2; level of confidence, approximately 95%). A serum zinc concentration of <60 µg/dL was considered zinc deficiency. The timing of the serum zinc measurements was grouped into 6 periods: before treatment and 0-6, 7-12, 13-24, 25-36, and ≥37 months after treatment.
Diagnosis of zinc deficiency
Zinc deficiency can be reliably diagnosed using the following three criteria (12, 13). I. One or more symptoms and signs of zinc deficiency including dermatitis, aphthous stomatitis, hair loss, loss of appetite, taste disorder, hypogonadism in males, anemia, increased infection susceptibility, growth disturbances in terms of weight and height in children, and low serum alkaline phosphatase (ALP) levels. However, serum ALP levels are not always low in patients with liver disease, osteoporosis, chronic kidney disease, or diabetes mellitus. II. Ruling out other diseases associated with the above symptoms or low serum ALP levels. For example, conditions including contact dermatitis, atopic dermatitis, dermatitis resulting from vitamin A, biotin, or essential fatty acid deficiencies, alopecia areata, hair-pulling, short stature resulting from growth hormone deficiency, familial short stature, Turner syndrome, and congenital hypophosphatasia should be ruled out. III. Low serum zinc concentrations IV. 1. <60 µg/dL: zinc deficiency V. 2. 60-80 µg/dL: marginal zinc deficiency Zinc supplementation can be provided to patients who meet criteria I, II, and III. Symptoms for these patients can be improved with zinc supplementation.
Statistical analysis
Age and serum zinc concentration were analyzed using a ttest. Fisher's exact test and logistic regression analysis were used to compare zinc deficiency to anemia, hypoalbuminemia, chemotherapy, radiation therapy, and recurrence. The Kruskal-Wallis rank sum test was used to compare serum zinc levels in patients with cervical, uterine, and ovarian cancer after treatment. Prediction of the therapeutic effects of serum zinc was determined using a receiver operating characteristic (ROC) curve. The effectiveness of the ROC curve was evaluated by the area under the curve (AUC) and the threshold was set at the point in which the sum of the sensitivity and specificity was maximized. The log-rank test was used to evaluate recurrence by serum zinc concentration. All statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). It is a modified version of R commander, which is designed to add statistical functions frequently used in biostatistics (22).
Results
As controls, 120 patients with benign gynecologic diseases before surgery were enrolled. For the comparison of cancer and noncancer patients, BMI was substituted as a measure of nutritional status because the blood Alb concentration of noncancer patients was only partially measured. The mean age ± standard deviation of cancer and noncancer patients was 61.1 ± 15.8 and 46.9 ± 13.0, respectively (p<0.001), so patients were matched for age with a caliper of 0.2 standard deviation ( Table 2).
There were 214 patients with cancer in the analysis, including 53 patients with cervical cancer, 90 with endometrial cancer, 67 with ovarian cancer, and four with other cancers; their mean age was 58.8 years. Of these, 202 patients had undergone a total hysterectomy (types 1-3), 193 underwent bilateral adnexectomy, 117 were administered chemotherapy, 39 had undergone radiation therapy, and 35 experienced a recurrence. During the course of treatment or follow-up, serum zinc concentrations of <60 µg/dL were observed in 107 patients, and they were diagnosed with zinc deficiency (Table 3).
Preoperative serum zinc concentrations did not differ between noncancer (69.4 ± 11.1 mg/dL) and cancer (71.9 ± 10.3 mg/dL) patients (p = 0.259). Of the 214 patients, 35 patients with recurrent cancer (mean age, 58.4 years) and 179 with nonrecurrent cancer (mean age, 61.2 years) were observed. Following initial treatment, 107 patients (50%) were diagnosed with zinc deficiency. The mean time from cancer treatment to the onset of zinc deficiency was 30.8 months. The mean ± standard deviation of the lowest serum zinc concentration following treatment in the cervical, endometrial, and ovarian cancer groups was 59.8 ± 10.6, 59.4 ± 8.9, and 58.3 ± 11.4 mg/dL, respectively, with no significant difference among the three groups (Kruskal-Wallis rank sum test, p = 0.4485). Compared with pretreatment serum zinc concentrations, the concentrations at 6 months, 7-12 months, 13-24 months, 25-36 months, and ≥37 months post-treatment were significantly lower in patients with cancer ( Figure 1). There were no differences in zinc concentrations between the groups after treatment.
Because we observed significant differences in the association of zinc deficiency after cancer treatment with anemia (Hb < 12 g/dL), hypoalbuminemia (Alb < 4.1 g/dL), chemotherapy, and radiotherapy (Table 4), we performed a logistical regression analysis on binary variables ( Table 5). The results showed that only recurrence was significantly associated with zinc deficiency with an odds ratio of 4.8 (p = 0.0117). Of the 35 patients with recurrent cancer, 29 (83%) developed zinc deficiency. Of these, 27 developed zinc deficiency after recurrence, whereas 2 developed zinc deficiency before recurrence. The mean ± standard deviation time from recurrence to the onset of zinc deficiency was 16.9 ± 14.9 months, ranging from −5 to 63 months. The lowest serum zinc concentrations in patients with recurrent and nonrecurrent cancer were 54.0 ± 6.9 and 60.6 ± 10.1 mg/dL (p = 0.000378), respectively.
Next, we examined whether the decline in serum zinc concentrations in patients with recurrent cancer was influenced by the cancer treatment itself (chemotherapy or radiation therapy), weight changes because of cancer progression, or age. A comparison of 35 patients with recurrent and 179 with nonrecurrent cancer showed differences in BMI, chemotherapy, and radiotherapy (Table 6). We matched BMI with a caliper of 0.2 standard deviation, performed a matched-pair logistic regression analysis of chemotherapy and radiation therapy; we observed no difference in prior treatment between patients with recurrent and nonrecurrent cancer (Table 7). This indicates that zinc deficiency in patients with recurrent cancer is not caused by chemotherapy or radiotherapy but by the recurrence itself.
The relationship between the lowest serum zinc concentration and recurrence was evaluated using ROC curves. A cutoff serum zinc concentration of ≤61 mg/dL was associated with recurrence with a sensitivity of 0.417, a specificity of 0.941, and an AUC of 0.698 (95% confidence interval: 0.611-0.785) (Figure 2). The prognosis for recurrence-free survival was markedly worse in the group with a minimum zinc concentration of ≤61 mg/dL (p = 0.0000244) (Figure 3).
All patients who developed zinc deficiency were treated with an oral zinc supplement and received additional nutritional guidance for a zinc-rich diet. Nevertheless, among patients with recurrent cancer who had developed zinc deficiency once, 20 of 29 (69%) developed zinc deficiency again. Therefore, we examined the recurrence of zinc deficiency in patients with recurrent cancer using the lowest serum zinc concentration as a predictor. A ROC curve predicted recurrence of zinc deficiency using a cutoff of 55 mg/dL as the lowest serum zinc concentration (sensitivity: 0.667, specificity: 0.7, area under the ROC curve: 0.669, 95% confidence interval: 0.452-0.887) (Figure 4).
Discussion
In the present study, zinc concentrations decreased from pretreatment to post-treatment in patients with cancer. The development of zinc deficiency was significantly associated with recurrence, with 83% of patients with recurrent cancer developing zinc deficiency. We observed that serum zinc concentrations of ≤61 mg/dL were associated with cancer recurrence with an AUC of 0.698. Furthermore, a serum zinc concentration of ≤61 was associated with a worse prognosis compared with concentrations >61 mg/dL. The mean time from recurrence to the onset of zinc deficiency was approximately 17 months. Moreover, the risk of recurrent zinc deficiency was higher when the minimum serum zinc concentration was <55 mg/dL.
Concerning the relationship between zinc and gynecologic cancer, Lightman et al. (1986) first reported that zinc concentrations in ovarian tumor tissue were significantly lower compared with those in benign tissue specimens (23). Moreover, it was estimated that patients with cancer possess lower serum zinc concentrations than noncancer patients and that dietary zinc deficiency in terms of the antioxidant properties of zinc may result in DNA damage through oxidative modification, thereby increasing cancer risk (24). Previous studies have hypothesized that a decrease in serum zinc concentration resulting from diet or other factors triggers a reduced in its concentration in tumor tissue. This reduces the antioxidant effects of serum zinc and leads to the oxidative modification of DNA, thereby resulting in carcinogenesis.
Nonetheless, in the present study, the cancer patient group maintained similar preoperative zinc concentrations compared with the noncancer patient group. In addition, 69% of patients with recurrent cancer, who once developed zinc deficiency, developed zinc deficiency again, despite being treated with oral zinc supplements until their serum zinc concentrations improved and being provided nutritional guidance to adopt a zinc-rich diet. Therefore, in our cohort of patients with cancer, a decrease in zinc concentration because of an inappropriate diet did not contribute to carcinogenesis and recurrence was associated with zinc deficiency development.
Because our results showed a decrease in serum zinc concentration following cancer treatment, we predicted that factors associated with post-treatment nutritional status as well as other factors may have contributed to reducing serum zinc concentrations. Therefore, we considered chemotherapy and radiotherapy, which result in decreased dietary intake, as possible factors associated with zinc deficiency; moreover, factors such as anemia (25) and hypoalbuminemia (26), which are indicators of malnutrition and are correlated with zinc concentrations, were considered. However, we observed that only cancer recurrence was associated with zinc deficiency.
Because 87% of patients with recurrent cancer experienced zinc deficiency, it is possible that some of the symptoms associated with this condition, such as taste disorders and anemia, which also frequently occur in these patients, can be attributed to zinc deficiency (Table 1). However, our study did not retrospectively examine symptoms caused by zinc deficiency; hence, it is unclear the manner in which it affects the quality of life of patients with recurrent cancer. Therefore, further studies are required.
We observed that low serum zinc concentrations were independently associated with cancer recurrence after adjusting for age, anemia, hypoalbuminemia, and prior cancer therapy. Because zinc deficiency is associated with poor prognosis, measurement of serum zinc concentrations may be incorporated into treatment protocols for gynecologic cancers; however, further studies are required to determine whether zinc supplementation improves patient prognosis. Although the analysis of clinical symptoms associated with zinc deficiency in patients was not performed in the present study, patients with zinc deficiency who complained of alopecia or taste disorders not related to chemotherapy were rare. In contrast to some studies suggesting that low zinc concentrations trigger carcinogenesis, the occurrence of zinc deficiency occurred mostly after cancer recurrence, with a mean time between recurrence and occurrence of zinc deficiency of 17 months.
In patients with recurrent disease who exhibited zinc deficiency once, 20 of 29 (69%) developed zinc deficiency again. Moreover, zinc deficiency was likely to recur when the lowest serum zinc concentration fell below 55 mg/dL. Chemotherapy and radiation therapy as well as a low nutritional status of patients were not associated with zinc One limitation of our study is that we did not monitor the nutritional intake of patients with recurrent cancer; hence, we could not analyze whether reduced dietary intake resulting from cancer recurrence was associated with zinc deficiency. However, because 69% of the patients with zinc deficiency who received appropriate oral zinc supplements and nutritional guidance for a high zinc diet experienced recurrent zinc deficiency, it may be inferred that in patients with recurrent cancer, other causes of zinc deficiency besides diet can be involved. Because our study was a single-center, retrospective study, the impact of zinc deficiency on the quality of life of patients with cancer remains unknown and prospective studies at multiple centers are warranted.
Conclusions
No differences in serum zinc concentrations between patients with benign gynecologic disease and those with gynecologic cancer were observed before treatment. Patients with gynecologic cancer are at higher risk for developing zinc deficiency after initial treatment completion, and patients with recurrent cancer are at higher risk of zinc deficiency. Low serum zinc concentration is an independent factor associated with Receiver operating characteristic (ROC) curves predicting cancer recurrence using the lowest serum zinc concentrations. The ROC curve predicts cancer recurrence using a cutoff of 61 mg/dL as the lowest serum zinc concentration (sensitivity: 0.417, specificity: 0.941, area under the ROC curve: 0.698, 95% confidence interval: 0.611-0.785). Kaplan-Meier curves with a cutoff at the lowest blood zinc concentration of 61 mg/dL Kaplan-Meier curves show a worse prognosis for serum zinc concentrations of ≤61 mg/dL (p = 0.0000244).
FIGURE 4
Receiver operating characteristic (ROC) curve predicting recurrent zinc deficiency using the lowest serum zinc concentration. The ROC curve predicts the recurrence of zinc deficiency using a cutoff of 55 mg/dL as the lowest serum zinc concentration (sensitivity: 0.667, specificity: 0.7, area under the ROC curve: 0.669, 95% confidence interval: 0.452-0.887).
cancer recurrence and poor prognosis. Furthermore, once zinc deficiency develops, the risk of zinc deficiency recurrence is high. Therefore, gynecologic oncologists should actively measure serum zinc concentrations to improve prognosis and maintain quality of life in patients with recurrent cancer.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving human participants were reviewed and approved by Nippon Medical School Chiba Hokusoh Hospital Ethics Committee, Nippon Medical School Chiba Hokusoh Hospital. The patients/participants provided their written informed consent to participate in this study.
Author contributions
MT and GI contributed to the conception and design of the study. KN organized the database, performed experiments, and wrote the manuscript. SS approved the final draft of the article. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-24T14:53:23.708Z
|
2022-11-24T00:00:00.000Z
|
253840983
|
s2orc/train
|
Impact of sarcopenia on the future liver remnant growth after portal vein embolization and associating liver partition and portal vein ligation for staged hepatectomy in patients with liver cancer: A systematic review
Purpose The impact of sarcopenia on the future liver remnant (FLR) growth after portal vein occlusion, including portal vein embolization (PVE) and associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) has gained increasing interest. This systematic review aimed to explore whether sarcopenia was associated with insufficient FLR growth after PVE/ALPPS stage-1. Methods A systematic literature search was performed in PubMed, Embase, Web of Science, and Cochrane Library up to 05 July 2022. Studies evaluating the influence of sarcopenia on FLR growth after PVE/ALPPS stage-1 in patients with liver cancer were included. A predefined table was used to extract information including the study and patient characteristics, sarcopenia measurement, FLR growth, post-treatment complications and post-hepatectomy liver failure, resection rate. Research quality was evaluated by the Newcastle-Ottawa Scale. Results Five studies consisting of 609 patients were included in this study, with a sample size ranging from 42 to 306 (median: 90) patients. Only one study was multicenter research. The incidence of sarcopenia differed from 40% to 67% (median: 63%). Skeletal muscle index based on pretreatment computed tomography was the commonly used parameter for sarcopenia evaluation. All included studies showed that sarcopenia impaired the FLR growth after PVE/ALPPS stage-1. However, the association between sarcopenia and post-treatment complications, post-hepatectomy liver failure, and resection rate remains unclear. All studies showed moderate-to-high quality. Conclusions Sarcopenia seems to be prevalent in patients undergoing PVE/ALPPS and may be a risk factor for impaired liver growth after PVE/ALPPS stage-1 according to currently limited evidence. Systematic review registration https://inplasy.com/, identifier INPLASY202280038.
Introduction
Liver resection remains a mainstay treatment for patients with primary or secondary liver cancer (for instance, hepatocellular carcinoma or colorectal liver metastases) with curative intent (1). However, many patients have been at an advanced stage at first diagnosis, and only 15-25% of patients are indicative of liver resection (2). For those patients who are not eligible for surgery, a majority of them are due to the limited future liver remnant (FLR), which is the remaining part of the liver after liver resection, and it serves as a key determinant for extended liver resection (3). FLR has to be sufficient to maintain normal physiologic function after liver resection, otherwise a lethal complication, post-hepatectomy liver failure will occur (4). To prevent the occurrence of liver failure after liver resection, the FLR volume limit should be > 20% of the total liver in a normal liver, > 30% in the abnormal liver (such as steatosis or postchemotherapy), and at least 40% in the cirrhotic liver (5).
In clinical practice, many strategies have been proposed to increase the size of the FLR volume before extended liver resection. Portal vein embolization (PVE) is the commonly used technique and was first introduced by Masatoshi Makuuchi in the 1980s (6). At PVE, the branch of the portal vein leading the blood to the diseased lobes of the liver is occluded interventionally by using sponges or metal coils. By this interruption of the blood flow, the un-embolized lobes (i.e. the FLR) will be exposed to all the portal venous blood flow. This increase in flow, including exposure to nutrients, toxins, and oxygen triggers liver growth (7). Most often, after waiting for several weeks, a sufficient growth of the FLR volume has occurred and a radical liver resection can be performed safely. PVE is still the standard procedure before extended liver resection when the FLR volume is estimated to be insufficient (8). Typically, an FLR growth of 12-38% can be observed within 4-8 weeks after PVE (9). However, during the waiting period, approximately 20-40% of patients cannot proceed to hepatic resection due to insufficient liver growth or tumor progression (9,10). Furthermore, patients with poor liver growth after PVE also have an increased r is k of post-intervention complications (11).
Hepatobiliary surgeons have been committed to developing an improved method to overcome the above-mentioned limitations of PVE. In recent years, a novel strategy, called associating liver partition and portal vein ligation for staged hepatectomy (ALPPS) has been proposed (12, 13). It contains two steps: in the first step, after the branch of the portal vein to the diseased lobes has been ligated (PVL, rather similar to the PVE procedure), the liver parenchyma is transected between the ligated part and the unligated part (i.e. the FLR). Once the FLR volume has increased sufficiently, liver resection can be performed to remove the liver tumor in the second step (12, 14). Interestingly, ALPPS can trigger an accelerated FLR growth in a shorter time than PVE, with a 40-80% FLR increase in only 6-9 days (12-15). However, ALPPS has high perioperative morbidity and mortality due to major surgical trauma, and the FLR growth varies among patients after ALPPS stage-1 (14). Despite surgically successful ALPPS stage-1, not all patients can complete the liver resection (14).
It is therefore of clinical importance to identify pretreatment factors that indicate a risk for insufficient FLR growth, which might allow optimizing treatment management of patients with liver cancer. Many clinical variables have been identified to be predictive for insufficient liver growth after PVE/ALPPS stage-1, for example, age, body mass index, and the diseased liver parenchyma (16). Among those, body composition is drawing increasing attention and has been assumed to be a treatable, prognostic factor in several hepatopancreatobiliary cancers after surgery (17-19). Sarcopenia is characterized by a progressive, generalized loss of skeletal muscle mass and function with aging (20). Previous studies have demonstrated sarcopenia to be associated with poor overall survival, early tumor recurrence, prolonged intensive care unit, and hospital stay after liver resection (21,22). In recent years the influence of sarcopenia on the FLR growth after PVE/ALPPS stage-1 has been studied.
However, no research systematically summarizes the results of these studies to date. This study aimed to provide such a systematic review.
Methods and materials
The research protocol was prospectively registered at the public platform International Platform of Registered Systematic Review and Meta-analysis Protocols (https://inplasy.com/) with registration number INPLASY202280038. This study was carried out in accordance with the guidance of the Preferred Reporting Items for a Systematic Review and Meta-analysis (PRISMA) (23). The PRISMA checklist can be found in Supplementary Table S1.
Literature search and study selection
A systematic literature search was performed at four public databases: PubMed, Embase, Web of Science, and Cochrane Library and was last updated on 5 July 2022. A search strategy combining Medical Subject Headings (MeSH) terms and text words were adopted. The keywords for literature search included "sarcopenia", "body composition", "portal vein embolization", "portal vein ligation", "portal vein occlusion", "associating liver partition and portal vein ligation for staged hepatectomy". The detailed search queries are provided in Supplementary Table S2.
Records satisfying the following criteria were regarded as eligible: 1) prospective or retrospective observational studies; 2) patients with liver cancer who underwent PVE/portal vein ligation or ALPPS to induce FLR growth before liver resection; 3) FLR growth as the main outcome or one of the outcomes; 4) at least one index for sarcopenia or body composition assessment involved. Studies would be excluded if they were: 1) in the forms of narrative review, letter, reference abstract, editorial, and case report; 2) animal research.
The process of study selection was carried out by two researchers (Q.W & A.W) independently by reading the title and abstract first to screen potentially ineligible studies. After that, the full text of the screened studies was obtained to further check their eligibility in consensus. Previous reviews and the reference list of the eligible studies were also manually retrieved to detect potential eligible studies.
Data extraction and research quality evaluation
The same researchers (Q.W & A.W) independently extracted the data from the included studies and assessed the research quality. The extracted information included: study characteristics (first author, publication year, country, study design, single or multiple center studies, and sample size), patient characteristics (age, gender ratio, the procedure involved, indication, and whether also segment IV was embolized), sarcopenia related information (modality used, body composition measurement, body composition parameters, sarcopenia definition, and the incidence of sarcopenia), FLR growth (degree of hypertrophy and kinetic growth rate), independent risk factors for poor FLR growth, complications/post-hepatectomy liver failure, the liver resection rate, and the main finding of the study.
Research quality and risk of bias of the cohort or casecontrol studies were evaluated by using the Newcastle-Ottawa Scale tool, which is a validated and easy-to-use scale containing eight items within three domains (selection of study groups, comparability of groups, and ascertainment of exposure/ outcomes) (24). The maximum score of this tool is 9, with 7-9 indicating high quality, 4-6 moderate quality, and 0-3 low quality (24). Research quality and risk of bias of the crosssectional study were assessed by applying the Agency for Healthcare Research and Quality tool, which contains an 11item checklist (25). The quality grades were defined as follows: 8-11 (high quality), 4-7 (moderate quality), and 0-3 (low quality). Any disagreement in data extraction and research quality appraisal between the two researchers was solved by discussion or by consulting a senior researcher (T.B.B).
Study characteristics and research quality assessment
Systematic literature searching initially yielded 187 records from the four electronic databases. After the removal of ineligible studies (duplications (40), inappropriate form of research (71), animal research/case report (5), studies not related to portal vein occlusion or sarcopenia (63), and no liver growth indices available (3)), five studies remained for inclusion in this systematic review (26)(27)(28)(29)(30). The process of study selection is shown in Figure 1 and Supplementary Table S3.
The five studies were published between January 2020 and May 2022 and were all retrospectively designed. A total of 609 patients were evaluated, with a sample size ranging from 42 to 306 (median: 90) patients. Only one study was carried out at multiple medical centers, which were located in several European countries (29). Only one study was conducted in an Asian country (28). The Newcastle-Ottawa Scale score of the four cohort/case-control studies varied from 6 to 9 (median: 6.5) (moderate-to-high quality), while one cross-sectional study was assigned an Agency for Healthcare Research and Quality score of 6 (moderate quality) ( Table 1).
Patient characteristics
The average age of the included patients was between 56 and 68 years. A predominance of males was found in all studies (in total, 391/609 = 64%), which is typical of the diseases involved. Two studies exclusively focused on patients with colorectal liver metastases (26,27), while the patients in the other three studies had varying indications. Four studies evaluated the impact of sarcopenia in patients undergoing PVE while the remaining one evaluated patients with ALPPS (30) ( Table 1).
Skeletal muscle measurement and definition of sarcopenia
All body composition analyses were based on pretreatment computed tomography (CT) images: two studies stated the CT image phase used (one without contrast media and the other on images obtained in the portal venous phase) (26,28). A slice thickness of 5 mm was reported in three studies (26,29,30), while slice thickness was not reported in two (27, 28). All studies measured the skeletal muscle area (at the level of the third lumber vertebra), which was converted into skeletal muscle index by being divided by squared height (m 2 ). Three studies adopted the skeletal muscle index to define sarcopenia (27-29). All three studies used the same threshold levels, including the one from Japan; sarcopenia was defined by a threshold of skeletal muscle index < 41 cm 2 /m 2 in women, while in men two thresholds were used depending on the body mass index; at body mass index < 25 kg/m 2 a skeletal muscle index < 43 cm 2 /m 2 defined sarcopenia, while the threshold was skeletal muscle index < 53 cm 2 /m 2 when body mass index was > 25 kg/m 2 (27-29). The incidence of sarcopenia in the three studies ranged from 40% to 67% (median: 63%) (27-29). The study selection process of this study. A total of 187 records were initially identified in the four public databases. After the removal of 182 ineligible publications via reading the title, abstract, and full text, five studies were finally included in this systematic review. PVO, portal vein occlusion.
One study applied the parameter "muscularity" to comprehensively evaluate muscle quantity and quality (28). This parameter combines skeletal muscle index and intramuscular adipose tissue content to represent both skeletal muscle quantity and quality. Another study, which was an early study with a limited sample size, did not provide their definition of sarcopenia but explored the correlation between muscle indices and liver growth (26). With a case-control design, the ALPPS study dichotomized patients using a threshold of the kinetic growth rate of 7%/week (30). The difference in skeletal muscle index between the two groups was then compared. Detailed information about sarcopenia measurement can be found in Table 2.
Liver growth rate
The degree of hypertrophy and kinetic growth rate of the FLR are two parameters commonly used for the assessment of liver growth after PVE/ALPPS stage-1. Two studies reported liver growth in the whole cohort, with a degree of hypertrophy of 8.9% and 9.5% respectively (26,28). In three studies, the degree of hypertrophy in the sarcopenia and non-sarcopenia groups was evaluated, with a range of 8.0-8.3% and 10.8-15.2%, respectively (27-29). In the ALPPS research which dichotomized patients into low and high kinetic growth rate groups by a kinetic growth rate cutoff value of 7.0%/week, a degree of hypertrophy of 11% and 18% was observed in the two groups respectively (30) ( Table 3). Compared with a kinetic growth rate of 2.6-4.0%/week in the non-sarcopenia group, the sarcopenia group demonstrated a significantly lower kinetic growth rate of 2.0%/week in two studies (27, 29). One study reported an overall kinetic growth rate of 3.6%/week for the whole study cohort (26).
In the three studies that performed multivariable logistic regression analysis, all identified sarcopenia as an independent factor for poor FLR growth (28)(29)(30). The other independent variables detected were initial FLR volume, total bilirubin level, and body mass index. All studies concluded that sarcopenia was associated with poor FLR growth after PVE/ALPPS stage-1 (Table 3).
Post-treatment complications, post-hepatectomy liver failure, and resection rate Two studies reported complications after PVE intervention, with a major complication (≥ III Clavien-Dindo classification) of 52% and 31% in the sarcopenia group versus 41% and 33% in the non-sarcopenia group respectively (both statistically nonsignificant) (28,29). The incidence of post-hepatectomy liver failure was reported in two PVE studies and one ALPPS study. An opposite result was observed in the two PVE studies where the incidence of postoperative liver failure was 38% and 16% in the sarcopenia group versus 17% and 22% in the non-sarcopenia group respectively (one significant while the other not) (28,29). The ALPPS study reported an incidence of post-hepatectomy liver failure of 20% and 7% after ALPPS stage-1 in the low and high kinetic growth rate groups respectively, but also here the difference was not statistically significant (30). There was a significant difference of the post-hepatectomy liver failure incidence between the low and high kinetic growth rate groups after ALPPS stage-2 (i.e. liver resection) with an incidence of 31% and 7%, respectively (p < 0.05) (30).
Overall resection rate was reported to be 83% and 73% respectively in two of the PVE studies (28,29) and 87% in the ALPPS study (30). One study reported a significantly lower resection rate after PVE in the sarcopenia group, compared with the non-sarcopenia group (66% vs 87%) (29). Interestingly, as a study evaluated factors that might affect liver growth after PVE, one study excluded patients with insufficient FLR growth after PVE and only included patients who proceed to liver resection (28). In that study, 26 patients did not undergo liver resection. In the ALPPS study, the resection rate was 84% in the low kinetic growth rate group, but that was not statistically significantly less than the 93% resection rate in the high kinetic growth rate group (30). Detailed information can be found in Table 3.
Discussion
The present study systematically reviews the association between skeletal muscle loss and FLR growth after PVE/ ALPPS stage-1. A high incidence of sarcopenia among patients undergoing PVE was observed, and sarcopenia was associated with impaired FLR growth after PVE/ALPPS stage-1. However, its relationship with post-treatment complication rate, posthepatectomy liver failure as well as surgical resection rate remains unclear.
The median incidence of sarcopenia among patients undergoing PVE in the included studies was 63%, which was higher than the reported incidence in patients with colorectal liver metastases (17-26%) (18,31) or hepatocellular carcinoma (30-54%) (18), the two most common indications for PVE/ ALPPS. Generally, the incidence of sarcopenia in patients with cancer has a wide variation due to different tumor types, tumor stages, measuring methods, and indices and criteria used (32). In the case of PVE/ALPPS, the indications usually vary among centers, which may also contribute to a varying and higher incidence of sarcopenia. Another explanation for the high observed incidence of sarcopenia is that the patients requiring PVE/ALPPS often have a chronically diseased liver such as liver cirrhosis (33) or have experienced several cycles of chemotherapy (e.g. neoadjuvant chemotherapy in patients with colorectal liver metastases) (31,34). Considering that patients undergoing PVE/ALPPS experience two major interventions in a relatively short time, the patients may be at a high risk of malnutrition if additional calories and protein cannot be supplemented in time. These additional clinical conditions may further impair patients' nutritional status. The higher incidence of sarcopenia also implies that the evaluation of body composition in these patients should be paid more attention to the perioperative assessment. All included studies applied CT-based measurement for muscle mass assessment. This is reasonable given that CT is a commonly used imaging modality for the diagnosis and staging of patients with liver cancer. Furthermore, in the setting of PVE/ ALPPS, CT is also widely applied for liver volumetry in pretreatment evaluation and to evaluate liver volume change after intervention. That is to say, the evaluation of body composition does not pose an extra burden for these patients. Dual-energy X-ray absorptiometry and bioelectrical impedance analysis are the other two commonly used methods for body composition measurement (35), but, to the best of our knowledge, they have not been employed in predicting liver remnant growth after surgery.
However, when analyzing the body composition, only limited details on CT imaging were provided in the included studies. Only two studies reported the imaging phase and just three studies described the slice thickness. It has been shown that these factors exert a considerable impact on the results of body composition analysis (36). In a study by Morsbach et al, the influence of contrast media and slice thickness on CT body composition segmentation was evaluated (37). They found that the skeletal muscle mass area, adipose tissue area, and muscle and fat attenuation (expressed in Hounsfield Units) showed a significant change after contrast media administration. There also was a significant effect on the area measurements (skeletal muscle mass area and adipose tissue area) when the slice thicknesses were adjusted. A systematic review summarized a group of CT-related factors which may affect the sarcopenia assessment (38). The CT parameters that according to the review can potentially affect the assessment included the use of contrast media, kilovoltage, CT manufacture and model, patient position, and slice thickness (38). Considering such many potential confounders, researchers need to bear them in mind when measuring body composition before transferring their results into clinical implementation. Besides, to make the findings reproducible and to increase the comparability among different studies, it also seems necessary to provide such information when reporting the body composition results.
Recent research has identified muscle quality, which can be determined by the infiltration of fat into muscle, as an independent prognostic factor in several types of cancer (39)(40)(41). In the present review, only one study adopted a composite index that combined skeletal muscle quantity and quality (named "muscularity") (28), while the others only assessed skeletal muscle quantity. Theoretically, muscularity should have a better performance in the prediction of the clinical outcomes, including the FLR growth after PVE/ALPPS stage-1, but this needs to be confirmed by further research. Besides, as highlighted by the European Working Group on Sarcopenia in Older People 2 evaluation of muscle strength and physical performance is equivalent to the evaluation of muscle quantity and quality in the diagnosis of sarcopenia (20). Future research assessing the impact of sarcopenia on liver growth and the clinical outcomes in patients with cancer can also consider taking these components into account. Even though an obvious heterogeneity was displayed in the included studies, all of them drew a similar conclusion that sarcopenia had a negative influence on liver growth after PVE/ ALPPS stage-1. Furthermore, sarcopenia seemed to have an association with a higher risk of post-hepatectomy liver failure and lower surgical resection rate in the patient who underwent PVE/ALPPS, although the results were inconsistent in this review. Sarcopenia is also a risk factor for poor overall survival in patients with liver cancer (21). But the impact of sarcopenia on the overall survival of patients who undergo PVE/ALPPS remains unknown.
Tumor progression is another common reason for patients not being able to reach curative surgery after PVE. Its incidence is even greater than that of insufficient liver growth contributing to a "failed" PVE, 19% vs 11%, as reported in the international DRAGON trial (42). This may be partly due to the slow growth after PVE, approximately 4-8 weeks to induce an FLR growth of 12-38% (9,43). During this long waiting interval, the tumor is likely to progress, leaving the patient not eligible for surgery anymore. Until now, only one study explored the influence of sarcopenia on the resectability in patients undergoing PVE. It showed that sarcopenia (defined as psoas muscle index < 500 mm 2 /m 2 ) was a risk factor for unresectability (44). However, the sample size of that research was limited (only 88 patients). On the other hand, a meta-analysis that included 13 studies revealed that sarcopenia was also significantly associated with tumor recurrence (adjusted hazard ratio: 1.76) (21). Whether sarcopenia results in both impaired liver growth and increased tumor progression after PVE, and whether improvement of patient sarcopenia status can increase resectability and longterm prognosis are still unclear.
Even though all studies claimed that sarcopenia impaired liver growth, it is of note to point out that sarcopenia should only be considered as a cofactor that undermines FLR growth after PVE/ALPPS stage-1. To put it another way, other vital clinical variables also determine liver growth after PVE/ALPPS stage-1. Three of the included studies also detected initial FLR volume, total bilirubin level, and body mass index as independent risk factors for insufficient liver growth (28)(29)(30). As a prognostic factor, the initial FLR volume was also reported in previous studies (45)(46)(47). Other reported indicators include age (45,48), embolic agent (49-51), segment IV embolization (52, 53), chemotherapy (45), and portal collaterals (54,55). It is assumed that a combination of these risk factors may improve the predictive accuracy for the FLR growth after PVE/ALPPS stage-1.
There are some limitations in this study. This review was first limited by the small number of included studies, all with a limited sample size (median: 90). Also, there were no prospective studies and only one multicenter study. The lack of large prospective multicenter studies may undermine a convincing conclusion drawn from this systematic review. Second, due to the limited study number and methodological heterogeneity, it was not possible to perform a meta-analysis and synthesize the results to provide a pooled relative risk value of sarcopenia for poor FLR growth. Third, the limited number of studies and the research heterogeneity also made it difficult to identify the most accurate and reliable parameter for sarcopenia assessment, given that a variety of indices were used for muscle mass evaluation. As summarized in a review, as many as 14 methods are currently available for sarcopenia assessment (32). Nevertheless, an index combining skeletal muscle quantity and quality evaluation (for example, muscularity) seems more rational and effective. Future studies can be designed to compare these indices. Lastly, there seems to be a need to improve the research and reporting quality of studies on sarcopenia. For example, detailed information on CT imaging during body composition measurement is required to ensure a reproducible and reliable study.
Conclusions
Research on the impact of sarcopenia on liver growth after PVE/ALPPS stage-1 is still in its initial stage. Based on currently available evidence, sarcopenia seems to have a high incidence in patients undergoing PVE/ALPPS and it may impair FLR growth. Its relationship with post-treatment complications, posthepatectomy liver failure, and resection rate requires further comprehensive research.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author/s.
|
v2
|
2022-11-24T14:55:26.138Z
|
2022-11-24T00:00:00.000Z
|
253841655
|
s2orc/train
|
Machine learning-based analysis of a semi-automated PI-RADS v2.1 scoring for prostate cancer
Background Prostate Imaging-Reporting and Data System version 2.1 (PI-RADS v2.1) was developed to standardize the interpretation of multiparametric MRI (mpMRI) for prostate cancer (PCa) detection. However, a significant inter-reader variability among radiologists has been found in the PI-RADS assessment. The purpose of this study was to evaluate the diagnostic performance of an in-house developed semi-automated model for PI-RADS v2.1 scoring using machine learning methods. Methods The study cohort included an MRI dataset of 59 patients (PI-RADS v2.1 score 2 = 18, score 3 = 10, score 4 = 16, and score 5 = 15). The proposed semi-automated model involved prostate gland and zonal segmentation, 3D co-registration, lesion region of interest marking, and lesion measurement. PI-RADS v2.1 scores were assessed based on lesion measurements and compared with the radiologist PI-RADS assessment. Machine learning methods were used to evaluate the diagnostic accuracy of the proposed model by classification of PI-RADS v2.1 scores. Results The semi-automated PI-RADS assessment based on the proposed model correctly classified 50 out of 59 patients and showed a significant correlation (r = 0.94, p < 0.05) with the radiologist assessment. The proposed model achieved an accuracy of 88.00% ± 0.98% and an area under the receiver-operating characteristic curve (AUC) of 0.94 for score 2 vs. score 3 vs. score 4 vs. score 5 classification and accuracy of 93.20 ± 2.10% and AUC of 0.99 for low score vs. high score classification using fivefold cross-validation. Conclusion The proposed semi-automated PI-RADS v2.1 assessment system could minimize the inter-reader variability among radiologists and improve the objectivity of scoring.
Introduction
Prostate cancer (PCa) is one of the most common cancers in men and the fifth leading cause of cancer-related death globally (1). Over the past several years, multiparametric MRI (mpMRI) has shown the ability to improve the early detection of clinically significant PCa and patient selection for biopsy (2). Prostate Imaging Reporting and Data System version 2.1 (PI-RADS v2.1) was published to simplify the reporting rules, modify imaging sequences, and define clinically significant cancer to reduce the variability in imaging, interpretation, and reporting (3). The PI-RADS assessment system is a qualitative scale with higher values indicating higher suspicion of PCa (3). PI-RADS includes a lesion size-based decision criterion (cutoff = 1.5 cm) to differentiate between score 4 and score 5 and also provides a minimal requirement for the measurement of lesion volume (>0.5 cc for clinically significant PCa) (3). Martorana et al. found that as PI-RADS scores increase, the probability of detecting a clinically significant PCa proportionally increases with increase in lesion volume (4).
Primarily, diffusion-weighted imaging (DWI) in the peripheral zone (PZ) and T2-weighted imaging (T2WI) in the transition zone (TZ) are used to assign the PI-RADS score (3). However, PI-RADS scoring is challenging due to its inherent technical difficulties to visualize a small lesion on MRI, and its subjectivity. Currently, the PI-RADS score is assessed qualitatively by a radiologist, which makes the PI-RADS process time-consuming as reporting time has become an important performance indicator in healthcare (5). Previous studies have shown poor inter-reader variability in the assessment of PI-RADS scores but with the potential to detect clinically significant PCa (6,7). Artificial intelligence-based workflow systems have shown similar or improved performances in detecting clinically significant PCa compared to radiologists (8) and have the potential to assist radiologists in the screening process by reducing inter-reader variability and evaluation time (9). Recently, one study by Dhinagar et al. (2020) presented a deep learning-based semi-automated model for PI-RADS scoring with limited area under the receiver-operator characteristics curve (AUC) of 0.70 (10). This model was trained and validated to classify lesions with only PI-RADS score 4 and score 5. An automated or semi-automated PI-RADS scoring system for PCa should be able to accurately classify all scores (score 2, score 3, score 4, and score 5) with good accuracy, as each class has a different prognosis for different PI-RADS scores (11). The objectives of this study were (i) to develop a semiautomated model for PI-RADS v2.1 scoring in order to speed up and simplify the reporting process and (ii) to analyze the diagnostic performance of the proposed model by classifying PI-RADS scores using machine learning methods.
MRI data-acquisition
An MRI dataset of 59 men (mean age: 65 ± 8.5 years) with clinically proven PCa (PI-RADS v2.1 score 2 = 16, score 3 = 10, score 4 = 18 and score 5 = 15) was used in this retrospective study with the prior approval from the institutional review board (IRB) of All India Institute of Medical Sciences (AIIMS), New Delhi. Informed consent was waived off by IRB for this study because of the retrospective nature of the study. This study was performed in accordance with institute guidelines and regulations. All prostate MRI examinations were acquired using a 1.5T scanner (Achieva, Philips Health Systems, the Netherlands). T2W images were acquired using a turbo spinecho sequence with TR/TE = 3330/90 ms, field of view (FOV) = 250×250 mm 2 , reconstructed matrix = 320×320, voxel size = 0.49×0.49×3 mm 3 , slice thickness = 3 mm, slice gap = 3 mm, and number of slices = 36. Diffusion-weighted images were acquired using echo-planar imaging with TR/TE = 6831/81 ms, FOV = 292×292 mm 2 , reconstructed matrix = 112×112, voxel size = 2.6×2.6×3 mm 3 , slice thickness = 3 mm, slice gap = 3 mm, and number of slices = 36, with five b-values of 0, 500, 1000, 1500, and 2000 s/mm 2 . Apparent-diffusion coefficient (ADC) maps were calculated using all five b-values with the least squareoptimization to the mono-exponential model using the vendorprovided algorithm at the clinical workstation (12).
Data processing
MRI data in DICOM format were transferred to a workstation (DELL Precision Tower 3620, using Intel ® Xeon ® CPU E3-1245 v5 @3.50GHz processor and 32GB RAM) and processed using MATLAB ® (MathWorks Inc., v2018, Natick, MA). The midgland region of the prostate was used for processing and this region consisted of approximately five to eight slices for each subject. Since most of the prostate cancers (70%-75%) originate in the PZ, this study focused only on this region. In this study, T2WI, DWI, and ADC images were considered for each patient, which was acquired in the MRI examinations.
PI-RADS v2.1 scoring model
The pre-processing steps involved automatic prostate gland and zonal segmentation, 3D image registration, lesion region of interest (ROI) marking, and lesion measurement. The Chan-Vese active contour model along with morphological opening operation was used for prostate gland segmentation and a probabilistic atlas with a partial volume correction algorithm was used to segment the prostate zones into the peripheral zone and transition zone (13). The prostate and its zonal segmentation were performed on the DWI dataset and then the T2W to DWI were registered so that the same segmentation results can be used for both the sequences. An Affine transformation method with a mutual information similarity index was applied for 3D registration of the prostate gland ROIs of T2WI and DWI.
All lesion ROIs were manually marked on each slice for all subjects, as per PI-RADS v2.1 guidelines (3) with the help of an expert radiologist (>20 years of experience in prostate imaging).
The ROI marking was first demonstrated in a few subjects by the radiologist. The PhD student (or expert) then marked the ROIs, which were verified by the radiologist, and changes to the marked ROIs were made as needed. ROIs were marked on the peripheral zone of DWI (b = 2000 s/mm 2 ) data and the same ROIs were used for ADC and registered T2W images of the respective subject. Representative examples of the lesion ROIs for different PI-RADS scores 2 to 5 are shown in Figure 1.
PI-RADS v2.1 introduced the lesion maximum diameter and lesion volume-based parameters for evaluating the aggressiveness of lesions. The ellipse fitting-based automatic algorithm was used in this research for the measurement of the lesion maximum diameter and volume, the same as proposed in (14). Maximum diameter was defined as the major axis of the best fitted ellipse. Lesion volume was determined by the multiplication of the slice profile (slice thickness + slice gap) with the summation of all lesion areas in the 2D plane. For comparison, the radiologist manually measured the maximum diameter and volume of the lesion using the image processing software ImageJ (v.1.48; National Institute of Health, Bethesda, USA). In the current study, PI-RADS v2.1 scores were assessed based on the lesion maximum diameter and lesion volume from the fitted ellipse. The workflow of the proposed model for PI- Lesion region of interests delineation from high b-value (b = 2000 s/mm 2 ) DWI and ADC of the representative patients with different PI-RADS scores 2 to 5. Figure 2. The detailed description of all the steps of the proposed model and its parameters are provided in annexure
Performance of the proposed model
Three machine learning methods (linear discriminant analysis (LDA), linear support-vector machine (SVM), and Gaussian SVM) were used to evaluate the diagnostic performance of the proposed model. In this study, the model proposed two different classification approaches: the first approach was to classify PI-RADS scores into four classes: score 2 (n = 16) vs. score 3 (n = 10) vs. score 4 (n = 18) vs. score 5 (n = 15), and the second approach was to classify into two classes: low score (score 2 and score 3) vs. high score (score 4 and score 5).
Statistical analysis
The PI-RADS scores obtained from the proposed model were compared with the radiologist's assessment using the Pearson correlation coefficient (r). Sensitivity, specificity, accuracy, and area under the receiver-operating characteristic curve (AUC) were measured to evaluate the performance of the proposed model. The accuracy of the proposed model was validated using stratified fivefold cross validation.
Lesion measurement
Lesion maximum diameter and lesion volume measured by the ellipse fitting approach were 0.47 ± 0.06 cm and 0.13 ± 0.03 cc for score 2, 0.67 ± 0.11 cm and 0.29 ± 0.12 cc for score 3, 0.96 ± 0.18 cm and 0.66 ± 0.28 cc for score 4, and 1.45 ± 0.15 cm and 0.99 ± 0.25 cc for score 5. Table 1 shows the maximum diameter and volume of the lesion measured with manual assessment and automated ellipse fit method for different scores.
Semi-automated model-based PI-RADS scoring
The proposed model-based PI-RADS v2.1 assessment showed 50 out of 59 subjects correctly matched (~85%) with the radiologist assessment. The proportion of the correct classification rate was 93.75% for score 2, 90% for score 3, Proposed workflow of the semi-automated model for PI-RARS v2.1 assessment.
83.34% for score 4, and 73.35% for score 5, which are also shown in Figure 3. The correct classification rate was also evaluated for the low score (score 2 and score 3) and high score (score 4 and score 5). The correct classification rate was 92.30% for the low score and~79% for the high score. Semi-automated PI-RADS v2.1 assessment showed a strong positive correlation (r = 0.94, p < 0.05) with the radiologist's assessment.
3.3 Diagnostic performance using machine learning methods Table 2 presents the performance of both classifications approaches, i) four-class classification (score 2 vs. score 3 vs. score 4 vs. score 5) and ii) two-class classification (low score vs. high score) using three different classifiers. The LDA classifier achieved the highest performance with sensitivity of 85.50 ± 1.95%, specificity of 75 ± 1.10%, accuracy of 88.00 ± 0.98%, and AUC of 0.94 in the four-class classification, and linear SVM classifier achieved the highest performance with sensitivity of 91.45 ± 3.65%, specificity of 95.85 ± 1.25%, accuracy of 93.20 ± 2.10%, and AUC of 0.99 in two-class classification using fivefold cross-validation. Figure 4 shows the ROC graphs for the fourclass and two-class classifications using three different classifiers.
Discussion
The PI-RADS v2.1 has emerged as a technical and reporting standard for uniform interpretation of prostate MRI. However, PI-RADS was challenged by the limited reproducibility among radiologists and medical centers due to its inherent subjectivity in scoring prostate lesions and lack of quantitative metrics (2,15). An automatic or semi-automatic PI-RADS scoring could assist the radiologist to perform initial screening, speed up reporting, and reduce errors in misclassifying lesions. In this study, a semi-automated PI-RADS v2.1 scoring model was developed for the diagnosis of PCa and validated the Proportion of the detection rate across all PI-RADS v2.1 scores.
Singh et al. 10.3389/fonc.2022.961985 Frontiers in Oncology frontiersin.org diagnostic performance of the proposed model using machine learning methods. The automatic non-invasive measurement of the prostate lesion could significantly improve to determine the prognosis and assist in PI-RADS assessment. Few approaches for measuring prostate lesions have been reported in earlier studies (16, 17). The maximum diameter and volume of the lesion provide an important information in determining the clinically significant PCa (18, 19). In this study, an automatic ellipse-fitting-based method was used to measure the dimensions of the lesions, and it was found that the results were similar to those obtained from the expert manual measurements. This is because PI-RADS uses only a twodimensional approach for lesion marking and measurement.
Sanford et al. utilized a convolution neural network (CNN) to evaluate the performance of the automated PI-RADS v2 scoring with an accuracy of 60% for low score vs. high score classification (20). A recent study has presented a semiautomated PI-RADS scoring system using CNN with an AUC of 0.70 (10). However, both studies were limited to a two-class classification only. The proposed semi-automated PI-RADS v2.1 scoring model here outperformed the existing methods with the correct classification rate of 85% and AUC of 0.94 for four-class classification (score 2 vs. score 3 vs. score 4 vs. score 5) and AUC of 0.99 for two-class classification [low score (score 2 and score 3) vs. high score (score 4 and score 5)]. Table 3 shows a comparison of the performance of the current study with the recent related papers.
The three supervised classifiers (LDA, linear SVM, and Gaussian SVM) employed in this study are based on the literature, as these three methods are the most commonly used and have been demonstrated to provide better classification performance compared to other classifiers for the prostate MRI (21,22). These classifiers are fast and have been shown to work well with small datasets (23).
The proposed model for semi-automated PI-RADS v2.1 scoring is relatively objective and could be helpful for nonexpert radiologists in terms of reporting accuracy. However, there are some limitations of the study. First, the proposed model is evaluated only for the lesions in the peripheral zone. The second limitation is that the sample size was small. The small sample size in this study provides the preliminary evidence to establish this new automated analysis approach. However a much larger cohort from multiple clinical centers will be Multiple receiver-operating characteristic graphs for (A) score 2 vs. score 3 vs. score 4 vs. score 5 classification and (B) low score (2 and 3) vs. high score (4 and 5) classification. essential for wider comparison and clinical acceptability which is beyond the scope of the current manuscript. Third, the reference measurement was done by only one radiologist; inter-observer and intra-observer variability were not evaluated. In this study, it was found that the correct classification rate for scores 4 and 5 is lower than for scores 2 and 3, possibly because the extra prostatic invasion of the lesion was not evaluated and because of inherent curvature bias in ellipse-fitting (24, 25), which needs to be further investigated. An automated lesion ROI marking may improve the model significantly and minimize the workload of radiologist. However, its impact on optimizing clinical workflow in hospital settings was not evaluated, which would require a much larger clinical study.
Conclusion
The proposed semi-automated model for PI-RADS v2.1 scoring achieved a high classification accuracy of 88% in fourclass classification and 93% in two-class classification. This model could reduce the inter-reader variability among radiologists and improve the objectivity of scoring in a screening setting of prostate cancer.
Annexure
Step 1: Prostate gland segmentation A semi-automated method based on a level set formulation of the Mumford-Shah function developed by Chan and Vese was used for segmentation of prostate gland (26). Chan and Vese proposed a pure region-based model to segment image where, u is the segmentation image and c 1 and c 2 are the average intensities of the two regions partitioned by the curve C. During the minimization of equation (1), the image is divided into two regions: inside and outside of the curve. The level set framework is combined to minimize the energy function shown in equation (1). The steps of the proposed method for segmentation of the prostate gland using a DWI image (b = 2000s/mm 2 ) were as follows: 1) a rectangular mask was constructed manually depending on the prior shape information of the prostate. This mask was further used in all slices of DWI in a subject; 2) a segmentation step to estimate the prostate by Chan-Vese active contour model combining the shape prior, and the range of total iterations was set to 80-100; 3) a refining step to smooth the prostate surface; few surrounding pixels were also found to be labeled as prostate gland. A morphological opening operation (structuring element= disk and size= 2) was applied to remove these speckle pixels. Please refer to (27) for more details of prostate gland segmentation.
Step 2: Image registration The prostate gland segmentation was performed on the DWI images (b = 2000s/mm 2 ). The segmentation of the prostate gland was also performed individually on the T2W images. The ROIs of T2WI and DWI were used for 3D registration, not the original images. The affine transformation method with a mutual information similarity index was applied for 3D registration.
The affine transformation with 12 degrees of freedom (DoF) allows shearing, scaling, translation, and rotation in three directions. In 3D, affine transformation can be expressed as T: (x, y, z)= ð q 11 q 12 q 13 where the coefficients parameterize the 12 DoF of the transformation.
Step 3: Prostate zonal segmentation This algorithm proposes a novel method for the subsegmentation of the prostate into peripheral zone and transition zone. The prostate zonal segmentation was performed on the DWI images (b = 2000s/mm 2 ). The method is based on probabilistic atlas approach with partial volume (PV) correction algorithm [ii]. Preprocessing steps were carried out for the registration and template creation. Followed by atlas construction, finally, zonal segmentation was performed. The statistical atlas was created by averaging the intensity of images of all training subjects aligned to the image of the target subject based on the corresponding spatial information of PZ and TZ. After that, probabilistic atlases were obtained by counting the number of occurrences of individual pixels in zonal statistical atlases (separate probability map for PZ and TZ) and normalizing the result.
The zonal segmentation of the registered data of the test subject was performed using a probabilistic map at 50% threshold probability (probability= 0.5) for both zones. When binary masks of PZ and TZ were applied to the registered dataset of the test subject, it was observed that some pixels in between the two zones could not be assigned to either of the zones. These pixels were having either equal or less than 0.5 probability values for belongingness to two zones; these pixels constitute the PV zone.
The PV correction algorithm was thus developed for correct assignment of these pixels in the PV zone to either PZ or TZ. For every pixel in the PV zone a belongingness value was calculated separately for both the zones, PZ and TZ, by optimizing the cost function for normalized intensity difference, probability, and Euclidian distance from the corresponding zone. The mask of PZ and TZ was further used in all slices of T2WI and ADC for each subject. Please refer to (27) for more details of prostate zonal segmentation and PV correction steps.
Step 4: Lesion region of interest (ROI) marking All lesion ROIs (size range: 50 to 200 voxels) were manually delineated from each slice of all subjects for lesion measurement, based on PI-RADS v2.1 guidelines (3). ROIs were outlined on the PZ of the DWI (b = 2000 s/mm 2 ) data with the help of a radiologist (with >20 years of experience in prostate MRI) and the same ROIs were used for the ADC and T2W images of the respective subject.
Step 5: Lesion measurement In this study, an automatic ellipse-fitting approach was used for the measurement of lesion maximum diameter and volume. This function employs the least-squares (LS) criterion to estimate the best fit the ellipse to a given set of points (x, y) of the lesion.
The LS estimation is done for the conic representation of an ellipse.
Conic ellipse representation= a à x 2 + b à x à y + c à y 2 + d à x + e à y +f ¼ 0 Maximum diameter was defined as the major axis of the best fit ellipse. Lesion volume was determined by the multiplication of slice profile (slice thickness + slice gap) with the summation of all lesion areas in the two-dimensional plane.
For comparison, the radiologist manually measured the maximal diameter and volume of the lesion using the image processing software ImageJ. (v.1.48; National Institute of Health, Bethesda, MD, USA).
Step 6: PI-RADS v2.1 assessment In the current study, PI-RADS v2.1 scores were assessed based on the lesion maximum diameter and lesion volume from the fitted ellipse.
Step 7: Machine learning based validation Three machine learning methods (Linear discriminant analysis (LDA), linear support-vector machine (SVM), and Gaussian SVM) were used to evaluate the diagnostic performance of the proposed model. These machine learning methods were implemented using default parameters so as not to introduce a bias or overfit the model on the given data.
LDA is supervised and computes the directions ("linear discriminants") that will represent the axes that maximize the separation between multiple classes. The LDA approach in detail is shown in (28).
SVM creates a hyper plane in the feature space to divide the data into two classes with the maximum margin. Using a positive semidefinite function, the feature space can map the original features (x, y) into a higher-dimensional space.
x, y ð Þ ! k x, y ð Þ The function k (·, ·) is called the kernel function. Here, we implemented two standard kernel SVM classifiers. k x, y ð Þ ¼ x: y ð Þ Linear ¼ exp (-( ∥ x -y ∥ )=s 2 Þ Gaussian (5) where s is the width of Gaussian. The importance of this parameter relates to cost of constraint violation during the SVM training. The accuracy of the proposed model was validated using stratified fivefold cross validation.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Ethics statement
The studies involving human participants were reviewed and approved by Institutional Review Board, All India Institute of Medical Sciences (AIIMS), New Delhi. The ethics committee waived the requirement of written informed consent for participation.
Author contributions
DS contributed significantly to the data collection, examination, and analysis and wrote the manuscript. VK contributed to the conception of the study and the revision of the manuscript. CD contributed to the data collection and conception of the study research design. AS contributed to conception of the study and the revision of the manuscript. AM contributed to the conception of the study and the revision of the manuscript and provided constructive discussions. All authors read and approved the final version of the manuscript.
|
v2
|
2022-11-25T06:17:28.964Z
|
2022-11-24T00:00:00.000Z
|
253838826
|
s2ag/train
|
Molecular Characterization of Adult Tumors Diagnosed as Cerebellar Glioblastomas Identifies Subgroups Associated With Prognosis
Adult tumors diagnosed as cerebellar glioblastoma (cGBM) are rare and their optimal classification remains to be determined. The aim of this study was to identify subgroups of cGBM based on targeted molecular analysis. cGBM diagnosed between 2003 and 2017 were identified from the French Brain Tumor Database and reviewed according to the WHO 2021 classification. The following molecular alterations were studied: IDH1/2, H3F3A, FGFR1, BRAF, TERT promoter mutations, EGFR amplification, MGMT promoter methylation, and alternative lengthening of telomere status. DNA methylation profile was assessed in a subset of cases. Eighty-three cGBM were included and could be classified into 6 mutually exclusive subgroups associated with median age at diagnosis (MA) and prognosis: TERT-mutant and/or EGFR-amplified tumors (n=22, 26.5%, MA=62 y, median overall survival [OS]=4 mo), H3K27M-mutant tumors (n=15, 18.1%, MA=48 y, median OS=8 mo), mitogen-activated protein kinases (MAPK) pathway–activated tumors (FGFR1, BRAF mutation, or occurring in neurofibromatosis type I patients, n=15, 18.1%, MA=48 y, median OS=57 mo), radiation-associated tumors (n=5, 6%, MA=47 y, median OS=5 mo), IDH-mutant tumors (n=1), and unclassified tumors (n=25, 30.1%, MA=63 y, median OS=17 mo). Most MAPK pathway–activated tumors corresponded to high-grade astrocytomas with piloid features based on DNA methylation profiling. In multivariate analysis, MAPK pathway–activating alterations, ATRX loss of expression, and alternative lengthening of telomere positivity were independently associated with a better outcome and TERT/EGFR alterations with a worse outcome. cGBM display an important intertumoral heterogeneity. Targeted molecular analysis enables to classify the majority of tumors diagnosed as cGBM into mutually exclusive and clinically relevant subgroups. The presence of MAPK pathway alterations is associated with a much better prognosis.
|
v2
|
2022-11-25T06:17:30.517Z
|
2022-11-24T00:00:00.000Z
|
253838260
|
s2ag/train
|
NR4A1 mediates NK-cell dysfunction in hepatocellular carcinoma via the IFN-γ/p-STAT1/IRF1 pathway.
BACKGROUND
Hepatocellular carcinoma (HCC) is one of the most fatal tumors worldwide and has a high recurrence rate. Nevertheless, the mechanism of HCC genesis remains partly unexplored, while the efficiency of HCC treatments remains limited.
METHODS
The present study analyzed the expression of nuclear receptor subfamily 4 group A member 1 (NR4A1) in tumor-infiltrating natural killer (NK) cells derived from both human patients with HCC and tumor-bearing mouse models, as well as the features of NR4A1high and NR4A1low NK cells. In addition, knockout of NR4A1 by CRISPR/Cas9 and adoptive transfer experiments were applied to verify the function of NR4A1 in both tumor-infiltrating NK cells and anti-PD-1 therapy.
RESULTS
The present study found that NR4A1 was significantly highly expressed in tumor-infiltrating NK cells, which mediated the dysfunction of tumor-infiltrating NK cells by regulating the IFN-γ/p-STAT1/IRF1 signaling pathway. Knockout of NR4A1 in NK cells not only restored the antitumor function of NK cells but also enhanced the efficacy of anti-PD-1 therapy.
CONCLUSION
The present findings suggest a regulatory role of NR4A1 in the immune progress of NK cells against HCC, which may provide a new direction for immunotherapies of HCC.
|
v2
|
2022-11-25T06:17:30.544Z
|
2022-11-24T00:00:00.000Z
|
253837807
|
s2ag/train
|
Overcoming AZD9291 Resistance and Metastasis of NSCLC via Ferroptosis and Multitarget Interference by Nanocatalytic Sensitizer Plus AHP-DRI-12.
The acquired resistance to Osimertinib (AZD9291) greatly limits the clinical benefit of patients with non-small cell lung cancer (NSCLC), whereas AZD9291-resistant NSCLCs are prone to metastasis. It's challenging to overcome AZD9291 resistance and suppress metastasis of NSCLC simultaneously. Here, a nanocatalytic sensitizer (VF/S/A@CaP) is proposed to deliver Vitamin c (Vc)-Fe(II), si-OTUB2, ASO-MALAT1, resulting in efficient inhibition of tumor growth and metastasis of NSCLC by synergizing with AHP-DRI-12, an anti-hematogenous metastasis inhibitor by blocking the amyloid precursor protein (APP)/death receptor 6 (DR6) interaction designed by our lab. Fe2+ released from Vc-Fe(II) generates cytotoxic hydroxyl radicals (•OH) through Fenton reaction. Subsequently, glutathione peroxidase 4 (GPX4) is consumed to sensitize AZD9291-resistant NSCLCs with high mesenchymal state to ferroptosis due to the glutathione (GSH) depletion caused by Vc/dehydroascorbic acid (DHA) conversion. By screening NSCLC patients' samples, metastasis-related targets (OTUB2, LncRNA MALAT1) are confirmed. Accordingly, the dual-target knockdown plus AHP-DRI-12 significantly suppresses the metastasis of AZD9291-resistant NSCLC. Such modality leads to 91.39% tumor inhibition rate in patient-derived xenograft (PDX) models. Collectively, this study highlights the vulnerability to ferroptosis of AZD9291-resistant tumors and confirms the potential of this nanocatalytic-medicine-based modality to overcome critical AZD9291 resistance and inhibit metastasis of NSCLC simultaneously.
|
v2
|
2022-11-25T06:17:31.658Z
|
2022-11-24T00:00:00.000Z
|
253837989
|
s2ag/train
|
Superiority of ceftazidime off-label high-dose regimen in PK/PD target attainment during treatment of extensively drug resistant Pseudomonas aeruginosa infections in cancer patients.
AIM
The objective of this study was to evaluate off-label high-dose ceftazidime population pharmacokinetics in cancer patients with suspected or proven extensively drug resistant (XDR) Pseudomonas aeruginosa infections, and then to compare the achievement of the PK/PD target after standard and off-label high-dose regimens using population model-based simulations. Further aim was to clinically observe the occurrence of adverse effects during the off-label high-dose ceftazidime treatment.
METHODS
In patients treated with off-label high-dose ceftazidime (3 g every 6 hours), blood samples were collected and ceftazidime serum levels measured using LC-MS/MS method. Pharmacokinetic population model was developed using nonlinear mixed-effects modeling approach and Monte Carlo simulations were then used to compare standard and high-dose regimens for PK/PD target attainment.
RESULTS
A total of 14 cancer patients with serious infection suspicious of XDR P. aeruginosa etiology were eligible for PK analysis. XDR P. aeruginosa was confirmed in 10 patients as causative pathogen. Population ceftazidime volume of distribution was 13.23 L, while clearance started at the baseline of 1.48 L/h and increased by 0.0076 L/h with each 1 mL/min/1.73 m2 of eGFR. High-dose regimen showed significantly higher probability of target attainment (i.e. 86% vs. 56% at MIC of 32 mg/L). This was translated into very low mortality rate of 20%. Only one case of reversible neurological impairment was observed.
CONCLUSION
We proved superiority of ceftazidime off-label high-dose regimen in PK/PD target attainment with very low adverse effects occurrence. The off-label high-dose regimen should be used to optimize treatment of this XDR P. aeruginosa infections.
|
v2
|
2022-11-25T14:35:27.029Z
|
2022-11-24T00:00:00.000Z
|
253841320
|
s2orc/train
|
Case Report: Differential diagnosis for tuberous sclerosis and neurofibromatosis type 1 diagnostic pitfall of aggressively enlarged right upper limb
Tuberous sclerosis complex (TSC) is an inherited disorder that typically presents with seizures, developmental delay, cutaneous lesions, and facial angiomas. Clinical diagnosis of TSC based on symptoms is sometimes challenging due to its clinical similarities with neurofibromatosis type 1 (NF1), another type of neurogenetic tumor syndrome. Differential diagnosis should be carefully performed on the basis of clinical presentations, imaging, laboratory, and genetic testing. Here, we presented a case of a patient with an aggressively enlarged right upper limb in the NF1 clinic, who was initially suspected of a giant plexiform neurofibroma. However, differential diagnosis revealed TSC as the final diagnosis. The treatments for NF1 and TSC vary significantly, and misdiagnoses can lead to serious threat to the patients’ health. We also systematically reviewed all previous cases regarding differential diagnoses between NF1 and TSC. This case report can help clinicians make more accurate diagnoses and benefit the potential patient community.
Introduction
Tuberous sclerosis complex (TSC) is a rare genetic disease characterized by seizures, developmental delay, and facial angiomas (Vogt's triad) (1). Neurofibromatosis type 1 (NF1) is another neurogenetic tumor syndrome caused by the mutation of the NF1 gene (2). There are similarities in clinical symptoms between TSC with NF1, which might cause misdiagnosis. Here, we described a case of a young patient with TSC with an enlarged right arm who was first diagnosed as NF1 during first visit to the clinic. Genetic testing revealed the TSC1 mutation, and TSC was confirmed as the final diagnosis. Furthermore, the radiological imaging showed lung lymphangioleiomyoma and poor blood supply in the right arm. Last, we discussed the clinical treatment and follow-up recommendation for this patient (Figure 1).
Case presentation
A 20-year-old man presented to our institution, complaining that the right upper limb was much thicker than the left from birth, along with the enlargement of his left index finger and middle finger (Figures 2A-C). The volume of the right upper limb increased significantly during puberty. Five months before this visit, the patient suffered from an ulcer on the right hand that had failed to heal. He did not pay much attention to the enlarged right arm as he has got used to the condition for over 20 years. The main complaint for this visit was the non-healing right-hand ulcer, which he thought could be solved after simple debridement. He showed little awareness and had limited knowledge regarding the disease. Upon questioning, the patient declared no history of seizures and showed no signs of amentia. His parents also reported that the patient had no history of autism.
On clinical examination, the right arm was significantly more prominent than the left, with the soft tissue being much thicker than the opposite side. Meanwhile, there were sebaceous adenomas on the back, depigmented macules on the face, and fibromas under the nail bed (periungual fibromas) and gingiva ( Figures 2D-G). Other skin lesions of tuberous sclerosis such as ash-leaf spots and shagreen patches were not evident. The physical examinations of cardiovascular and central nervous systems were also normal.
We initially suspected the case as plexiform neurofibroma with NF1 due to the enlargement of the right arm. However, the facial spots described by the patient as "café ' au lait spots" are actually depigmented macules. Whole-body imaging was then conducted for the patient. The head CT revealed subependymal, cortical, and subcortical nodules ( Figure 3A). Meanwhile, we found lymphangioleiomyomatosis on the chest CT, focal sclerosis, and bone cysts in multiple bone regions such as the spine ( Figures 3B, C). Although MRI might help clarify the nature of the enlarged right arm, it could not be conducted as the volume of the patient's right arm exceeded the maximum size limit of our MRI machine. No hemorrhage and retinal hamartomas were found on ophthalmic examination, and mild mitral regurgitation was found on echocardiography.
Genetic testing was conducted for further differential diagnosis. No NF1 mutation was found. Furthermore, there was a heterozygous mutation on the TSC1 gene.
On the basis of these features, a diagnosis of TSC was established, and the non-healing ulcer was considered to be associated with the vascular blockage caused by the thickened soft tissue. Digital subtraction angiography (DSA) was used to analyze the blood flow condition of the right arm (Supplementary File 1). The result showed insufficient blood supply in the distal FIGURE 1 Timeline of this case. right upper limb, and amputation appeared to be the only treatment plan for the patient. During consultation, the patient showed difficulties accepting this treatment plan and stated that he needed time for consideration. Approximately 4 months later, he finally accepted the amputation surgery of the right arm as the ulcer condition barely improved. After the surgery, we recommended life-long follow-up and monitoring of potentially fatal complications. Annual ophthalmology examination was also essential for the risk of hemorrhage. As the patient also presented with lung lymphangioleiomyoma and cyst, high-resolution CT should be performed every year and the annual pulmonary function examination was necessary.
Discussion and conclusions
TSC is caused by the pathogenic variants in either the TSC1 or TSC2 tumor suppressor genes, located on chromosomes 9q34 and 16p13, respectively (3). TSC1 and TSC2 encode the protein hamartin and tuberin, which together suppress the activity of the mammalian target of rapamycin (mTOR) pathway (4). Loss of their function leads to continuous activation of the mTOR pathway, which results in cell proliferation (4). Typical clinical symptoms of TSC include periungual fibromas, renal angiomyolipoma, benign interstitial expansion of lung pulmonary smooth muscle cells, and neurological symptoms like seizures and neurodevelopmental delay (1).
Neurofibromatosis type 1 (NF1) is also an autosomal dominant genetic disorder caused by another tumor suppressor gene NF1 (2). The NF1 gene encodes neurofibromin, a negative regulator of the RAS-mitogen-activated protein kinase pathway. The loss of NF1 contributes to the activation of this pathway and finally leads to tumor growth and development (5). Classic clinical features include café -au-lait macules, skinfold freckling, benign neurofibromas like cutaneous neurofibroma and plexiform neurofibroma, brain tumors, iris hamartomas, and typical bony lesions. Meanwhile, there are also reports of neurological manifestations such as epilepsy (6).
Comparing these two diseases, both may have vague and overlapping clinical presentations that can lead to missed diagnosis and finally cause delayed treatment. Coincidentally, both NF1 and TSC were first described by Von Recklinghausen, which proved from the side that these two diseases have some similar symptoms (7,8). The major clinical diagnostic criteria of TSC such as hypomelanotic macule, angiofibroma, and shagreen patch are sometimes difficult to distinguish from typical NF1 phenomenon like café -au-lait macules, skinfold freckling, and benign neurofibroma (Table 1). Although NF1 and TSC are the first and second most common neurogenetic tumor syndromes, NF1 only has a prevalence of approximately 1:2,500 to 1:3,500, whereas TSC is even rarer and affects 1 in 5,500 to 1 in 10,000 live births (9). The limited amount of clinical cases further caused difficulties in clinical differential diagnosis, especially for doctors in remote areas. Moreover, it was shown that, in some cases, both NF1 and TSC could occur in a single individual, which might further confuse the diagnosis. We that suffered from both NF1 and TSC but reported a family history of NF1 only (12). After a definite diagnosis of TSC was made for this girl, her parents were carefully re-examined and also confirmed the diagnosis of TSC (12). This case further indicates the similarities of some clinical symptoms, and one definite diagnosis might further inhibit the recognition of another. Genetic testing and pathological biopsy may be necessary for a definite diagnosis of TSC. Moreover, as TSC and NF1 are both hereditary diseases, first-degree relatives should undergo a clinical assessment and a three-generation family history is required.
Current clinical management of TSC is insufficient as most of them are symptomatic treatments, especially for seizures. Early therapeutic intervention for children who present with infantile spasms correlated with TSC is essential, and the first-line treatment is vigabatrin (15). Furthermore, the mTOR inhibitors such as everolimus have provided great therapeutic promise. A clinical study showed that treatment with everolimus in patients with TSC resulted in a sustained decrease in seizure frequency in children and adolescents (16). Nonetheless, this patient did not have any history of seizures. As he reached the age of 20 and presented with vascular blockage indicated by the non-healing ulcer, surgical approaches were considered according to the 2021 updated TSC international recommendations (17). DSA examination further showed signs of ischemia. As a result, amputation was the only surgical choice for him. Compared to the limited drug choices for patients with TSC, the development of drug therapies for patients with NF1 progressed rapidly. Selumetinib, a Mitogen-activated extracellularsignal-regulated kinase (MEK) inhibitor, was reported effective for plexiform neurofibroma and was approved for clinical usage by FDA (18). Early distinguishment of these two diseases is essential for further proper clinical treatment.
Another essential issue for the treatment of patients with TSC is life-long follow-up recommendations. TSC influences multiple organs in the body, and some manifestations are life-threatening. For example, our patient has lymphangioleiomyoma, which might cause respiratory dysfunction. According to the 2021 updated TSC international recommendations, we recommended routine serial pulmonary function test annually on chest CT (17). Meanwhile, we had several recommendations regarding other possible TSC features, including continuous sun protection for depigmented macules, detailed dental examination every 6 months, electrocardiography every 3 years for mild mitral regurgitation, annual ophthalmic evaluation, and renal function assessment. For patients with NF1, life-long follow-up is also essential, but they are less concerned with respiratory dysfunction. A definite diagnosis is essential for further proper follow-up.
Definite and accurate diagnosis is the basis for proper and effective treatment, and misdiagnosis might cause further damage to the patients and waste medical resources. It is far
Diagnosis Criteria
An individual must have two or more of these features. Definite TSC: Two major features or a major feature with two minor features; probable TSC: one major feature and one minor feature; possible TSC: one major feature or two or more minor features.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
Ethics statement
The studies involving human participants were reviewed and approved by the institutional review board of Shanghai Ninth People's Hospital, Shanghai Jiaotong University School of Medicine. The patients/participants provided their written informed consent to participate in this study. Written informed consent was obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.
Author contributions
C-JW wrote the manuscript and was involved in the diagnostic and therapeutic clinical processes. L-LP analyzed radiology medical images and contributed to the diagnostic and therapeutic processes. M-HC made substantial contribution during revision. Z-CW and BW helped in the diagnostic process and critically revised the manuscript and were responsible for the diagnosis and treatment of the patient.
|
v2
|
2022-11-26T06:16:11.954Z
|
2022-11-24T00:00:00.000Z
|
253879253
|
s2ag/train
|
A Close Follow-Up Strategy in the Short Period of Time after Helicobacter pylori Eradication Contributes to Earlier Detection of Gastric Cancer.
INTRODUCTION
The purpose of this study was to optimize the surveillance frequency and period for efficient detection of early gastric cancer (EGC) after Helicobacter pylori (HP) eradication.
METHODS
Data from patients with eradicated HP infection were extracted from the endoscopy databases of two institutions from January 2016 to March 2021. The patients were divided into a close follow-up group with frequent surveillance after eradication and an open follow-up group with an intermittent surveillance method, and the cases of post-eradication EGC found in the two groups were analyzed.
RESULTS
Thirty-six out of 9,322 patients (0.39%) in the close follow-up group and 20 out of 11,436 patients (0.17%) in the open follow-up group were found to have EGC. The cumulative incidence of EGC after eradication was significantly higher in the close follow-up group (p = 0.004). The duration between eradication and EGC detection was significantly shorter in the close follow-up group (51.7 vs. 90.5 months, p = 0.002). A logistic regression model revealed that duration after eradication was an independent predictor for detecting EGC in the close follow-up group (p = 0.045). A Cox proportional hazards model revealed that the close follow-up strategy was effective in patients with an eradication duration of less than 65 months to identify EGC (p = 0.015), but there was no difference between the two strategies in patients with an eradication duration of more than 65 months (p = 0.624).
DISCUSSION/CONCLUSIONS
Frequent surveillance after HP eradication is efficient for the early detection of EGC during the first 65 months.
|
v2
|
2022-11-26T14:44:47.211Z
|
2022-11-24T00:00:00.000Z
|
253882352
|
s2orc/train
|
Multicentre, randomised, open-label, parallel-group, clinical phase II study to evaluate immunonutrition in improving efficacy of immunotherapy in patients with metastatic non-small cell lung cancer, undergoing systematic nutritional counseling
Background Nutritional support, including nutritional counseling and oral nutritional supplements (ONS), has been recommended as a first-line strategy in patients with non-small cell lung cancer (NSCLC). Evidence on the efficacy of immunonutrition during immunotherapy in these patients is positive, but still limited some secondary endpoints, such as treatment toxicity and tolerance. We hypothesize that early systematic provision of ONS with a high-protein-high calorie mixture containing immunonutrients (Impact®) in addition to nutritional counseling, compared to nutritional counseling alone, is beneficial to patients with NSCLC receiving immunotherapy with or without chemotherapy. We designed the present study to evaluate the efficacy of early systematic provision of ONS enriched with immunonutrients compared to nutritional counseling alone, in patients with NSCLC undergoing immunotherapy. Study endpoints were: treatment response (primary endpoint: progression-free survival), treatment tolerance and toxicity, body weight, body composition, protein-calorie intake, quality of life, fatigue, muscle strength and immunological profile. Methods This is a pragmatic, multicentre, randomized (1:1), parallel-group, open label, controlled, pilot clinical trial (N = 180). Discussion The improvement of efficacy of nutritional support in oncology still deserves many efforts. Immunonutrition represents a promising approach also in patients with NSCLC, but evidence on its efficacy on clinical outcomes during immunotherapy is still inconclusive. The present pilot study, which guarantees early high-quality nutritional care (assessment and treatment) to all patients in agreement with current guidelines and recommendations, could represent one of the first proofs of efficacy of early oral immunonutrition in patients with cancer undergoing immunotherapy. Further large randomized trials addressing the improvement of supportive care could be hypothesized, accordingly. Trial registration This study is registered on ClinicalTrials.gov Identifier: NCT05384873.
Background
Since diagnosis, patients with non-small cell lung cancer (NSCLC) frequently present a variable impairment of nutritional status. Nutritional derangements are attributable to multiple factors, including those related to tumor location as well as to systemic features, i.e. inflammatory mediators responsible for tissue wasting, anorexia and weight loss. Anticancer treatments themselves (e.g. radiotherapy, chemotherapy and surgery) can enhance the deterioration of nutritional status, by increasing energy requirements, reducing food intake and/or impairing nutrients absorption [1][2][3]. An altered nutritional status is associated with a worse prognosis and the more frequent need to suspend/delay anticancer therapies [4].
The international guidelines, addressing nutritional care in oncology [1][2][3], agree on the usefulness of nutritional support -whenever necessary -in improving clinical outcomes. Previous studies have shown that nutritional counseling is able to increase protein-calorie intake, prevent the deterioration of nutritional status and quality of life (QoL) in patients with head-and-neck (H&N) cancer [5]. Nonetheless, two recent studies suggested that, while some H&N cancer patients may present with pretreatment normal nutritional status, early nutritional counseling is essential for the improvement of treatment tolerance and survival [6,7]. Recently, our group has demonstrated that the systematic use of oral nutritional supplements (ONS) in combination with dietary counseling in these patients enables to maintain the nutritional status, to recovery QoL and, more notably, to favor the feasibility of chemoradiotherapy (CT-RT) [8]. This effect was substantially attributed to the increase in protein-calorie intake, but the possible anti-inflammatory role of omega-3 fatty acids could not be excluded, as it is known that the modulation of inflammation by omega-3 fatty acids and other nutrients could be helpful during anticancer treatments [9].
The use of immunonutrition in patients with cancer has progressively gained attention, as a ONS enriched with immunonutrients (arginine, nucleotides and omega-3 fatty acids; Impact ® -Nestlè Health Science -Creully Sur Seulles -France), was proven to be effective in reducing the risk of post-operative complications (e.g. infections, fistulas, etc.) and length of hospital stay in patients undergoing major surgery for cancer (abdominal and H&N) [10,11]. An interest in the modulation of inflammation and immunosuppression in the tumor microenvironment is also growing [12].
Nutritional support interventions in patients with cancer should be urgently addressed through high-quality clinical trials. To date, only two trials have investigated the effectiveness of immunonutrients-enriched ONS in H&N cancer patients during adjuvant CT-RT in comparison with an isonitrogenous and isocaloric control supplement [13][14][15]. The immunomodulating formula failed to reduce severe mucositis during CT-RT, but a trend to improved overall survival (OS) and progression-free survival (PFS) was observed, particularly in patients more compliant to the intervention.
In recent years, along with traditional CT, immunotherapy has become the cornerstone of non-oncogene addicted advanced NSCLC treatment. However, the therapeutic efficacy of immuno-nutrient-enriched blends in patients with NSCLC undergoing immunotherapy has been not assessed yet.
The aim of the present randomized, controlled, openlabel study is to assess the effect of high-protein, high-calorie ONS containing immunonutrients, in association with nutritional counseling, on immunotherapy activity and immunological profile in patients with metastatic NSCLC treated in first-line setting with immunotherapy, alone or in combination with CT.
We hypothesize that this combination will improve the efficacy of immunotherapy.
Methods/design
The study methods and design described partially correspond to those of our ongoing trial in H&N cancer patients receiving immunonutrition [16].
Standard protocol approval, registration, and patient consent
The study will be conducted in accordance with good clinical practice and ethical standards laid down in the 1964 Declaration of Helsinki and its later amendments. during immunotherapy is still inconclusive. The present pilot study, which guarantees early high-quality nutritional care (assessment and treatment) to all patients in agreement with current guidelines and recommendations, could represent one of the first proofs of efficacy of early oral immunonutrition in patients with cancer undergoing immunotherapy. Further large randomized trials addressing the improvement of supportive care could be hypothesized, accordingly.
Trial registration: This study is registered on Clini calTr ials. gov Identifier: NCT05384873.
A informed consent will be obtained from each patient enrolled in the study. At any time, patients will have the right to withdraw their consent without modifying their current or future care. The progresses of the study will be shared with general practitioners.
Design
A multicentre, randomised (1:1), open-label, parallelgroup, open label, controlled clinical trial will be conducted. At inclusion, patients will be allocated to study treatments using a computer-generated and centralized randomization list. The concealment will be attained through a web-based randomization.
Inclusion criteria
• Confirmed histological diagnosis of metastatic NSCLC (both squamous and non-squamous histology); • First-line treatment with immunotherapy (alone or in combination with chemotherapy) for metastatic disease by investigators' choice within the framework of good clinical practice and in agreement with current guidelines; • Willingness to participate by signing written informed consent; • Availability to administer oral supplements and immunotherapy with or without chemotherapy; • ECOG Performance Status ≤2; • Life expectancy ≥6 months.
Exclusion criteria
• Age < 18 years; • Inability to sign an informed consent; • Indication to or ongoing artificial nutrition support (totally compromised spontaneous food-intake) and incapacity or unavailability to consume ONS.
Assessments
Demographic and clinical data, including tumor site, histology, stage, as well as scheduled anticancer treatment, will be collected. In addition, in agreement with our ongoing trial in H&N cancer patients receiving immunonutrition [16], the following assessments will be performed: Anthropometry -according to standard procedures, body weight [to the nearest 0.1 kg] and height [to the nearest 0.5 cm] will be measured, and body mass index (BMI) will be calculated [17]. Unintentional weight loss (WL) during the last 6 month will be also recorded.
Calorie and protein intakes -Intakes of energies and proteins will be estimated at all treatment visits using a 3-day quantitative food diary and the 24-hour dietary recall method (including weekdays and weekends) and by consulting validated atlas of food portions and collecting information on brand names of commercial and ready-to-eat-foods, method of preparation, and the use of dressings or added fat [18,19]. Total protein-calorie intakes throughout the study will be estimated taking also into consideration the use of ONS. They will be considered achieved if total energy and protein consumption will attain ≥90% of estimated requirements and ≥ 1.5 g/kg/day, respectively.
Since the energy content differs between the two ONS, the difference will be taken into account in the calculation of total intakes.
Nutritional screening risk and malnutrition -Nutritional risk will be evaluated at the screening visit using the Nutrition Risk Index 2002 (NRS-2002) screening tool, which is based on data collected on a routine basis (BMI, 6-month unintentional WL and oral food intake, diagnosis and age) [20]. Malnutrition will be diagnosed according to phenotypic and etiologic criteria proposed by the Global Leadership Initiative on Malnutrition (GLIM) [20]. Body composition -Body composition will be assessed using bioelectrical impedance vector analysis (BIVA) and the NUTRILAB Software (Akern srl; Florence, Italy). Specifically, resistance and reactance will be measured and used to calculate phase angle (PhA), standardized PhA (SPA) and hydration index (HI) [21][22][23][24]. Operative procedures will be standardized and the same devices will be used to ensure a homogenous collection of data. The measurement of skeletal muscle mass (SMM) will be performed using computed tomography. To this purpose, muscle area will be quantified on scans at L3 and T12 [25,26] collected at baseline disease staging and subsequent reassessments, as scheduled by the oncologists for the evaluation of response to CT. The SliceOmatic software v5.0 (TomoVision, Montreal, QC, Canada) will segment radiological images. Assessment of SMM at the level of L3 and T12 is easy, robust and a validated imaging procedure in patients with NSCLC [26]. Muscle strength -A digital hand dynamometer (DynEx ™ , Akern / MD Systems) will be used for the measurement of muscle strength (handgrip [HG]). Quality of life -At baseline and at the end of treatment, quality of life will be evaluated using the European Organization for Research and Treatment of Cancer (EORTC) Core QoL Questionnaire (QLQ-C30), the EORTC QoL Lung Cancer 13-item module (QLQ-LC13) and the Functional Assessment of Cancer Therapy-Lung (FACT-L) Questionnaire [27,28]. Fatigue -At baseline and at the end of treatment, selfreported fatigue and its impact on activities of daily living and functional status will be assessed through the 40-item Functional Assessment of Chronic Illness Therapy -Fatigue (FACIT-F) scale [29]. Symptoms -Patients will be administered the Edmonton Symptom Assessment Scale (ESAS) [30]. Presence or onset of symptoms potentially influencing food intake will be addressed accordingly. Physical activity -Self-reported physical activity level, will be assessed using the adapted version of the Godin's Shepard Leisure Time Exercise Questionnaire before diagnosis, at baseline and post-intervention, using the adapted version of the Godin's Shepard Leisure Time Exercise Questionnaire [31]. Efficacy -Tumor evaluation will be performed during immunotherapy as per local institutional standard of care with imaging techniques; Response Evaluation Criteria in Solid Tumors (RECIST v1.1) assessments will be performed based on local institutional imaging results, using CT/MRI assessments of the brain, chest and abdomen. Adverse complications and events -All adverse complications and events attributable to nutritional interventions (namely gastrointestinal side effects), including unplanned hospitalizations and their duration, will be recorded. Immunologic profile -Measurements obtained using multiple tools will be integrated with the aim of analyzing different cell subsets, their functionality, and soluble molecules in the peripheral blood. The profile will be based on the following parameters: amount of T cells (CD3+, CD4+, CD8+), of lymphocytes, neutrophils, neutrophil to lymphocytes ratio (NLR), LDH, C Reactive Protein, T helper cells (CD4+, CD25+, FOXP3+ or CD4+, CD25hi+, CD39+), myeloid derived suppressor cells (CD11b+/CD33+/ CD15+ or CD11b+/CD14+/HLADR−/CD15-), amount of plasmocytoid dendritic cells (CD303+, CD123+, CD45RA+), amount of slan+ dendritic cells (CD16 and Slan-M-DC8), IL-1β, IL-6, TNF-alfa.
A summary of study assessments and related endpoints is provided in Table 1.
Anti-cancer treatments
Immunotherapy with or without chemotherapy will be prescribed by investigator's choice within the framework of good clinical practice and in agreement with current Italian Association of Medical Oncology guidelines [32]. Particularly, first-line treatment options will consist of:
Nutritional counseling is the current standard of care and will begin 2 weeks before starting immunotherapy. A registered dietitian will prepare a specific dietary program according to anthropometry, energy requirements and diet history which will be thoroughly investigated at the first visit. Dietetic plan will include both qualitative and quantitative data regarding suggested foods intake and distribution of meals. The texture of the diet will be adapted according to the presence of dysphagia for solids or liquids. Dietary prescription may include ONS, which are usually recommended when patients are unable to maintain adequate spontaneous food intake (less than 50% of the requirement for more than 1 week or only 50-75% of the requirement for more than 2 weeks) [1]. Therefore, while in the experimental arm the administration of ONS enriched with immunonutrients will begin 2 weeks before starting immunotherapy, in the control arm the use of isonitrogenous standard blend ONS will be considered only on the basis of the regular assessment of food intake. Nutritional counseling will be provided for all the length of study and will continue after the evaluation of primary endpoint according to patient's needs. However, ONS provision will continue up to first disease re-assessment (12-14 weeks) and prolonged according to patient's needs. Adherence to nutritional interventions will be assessed and monitored by the caregiver and the dietitian through daily recording of the bottles consumed. The safety of ONS will be also monitored by the occurrence of any potential gastrointestinal side effect.
Total daily energy requirements will be calculated by adjusting the estimated resting energy expenditure (from Harris-Benedict equation) for a correction factor of 1.5. Similarly daily protein requirements will be set at 1.5 g/kg of actual body weight. Every 7 days, a registered dietitian will perform regular consultations by face-to-face interviews and food intake will be quantified through a 3-day food diary and a 24-hour recall. The patient will have the opportunity to contact the local Clinical Nutrition Unit by telephone for any specific clarifications and advice.
During the first visit, each subject will be evaluated and consecutively allocated to one of the nutritional interventions. Data will be collected and the follow-up planned according to the checks of treatment protocol.
Stratification factors: • Recruiting center • Histology (squamous vs non-squamous) • Type of treatment (immunotherapy alone vs chemoimmunotherapy) • PD-L1 TPS (< 1% vs ≥ 1%) It is not possible to adopt a blind design, both for the investigators and for patients. However, statistical analysis will be carried out blinded to treatment group.
Efficacy endpoints
Primary objective: to compare immunonutrition to nutritional counseling in terms of efficacy of immunotherapy with respect to Progression-free Survival (PFS). Primary end-point: PFS is defined as the time from randomization to the first documented progression of the disease (PD) or death due to any cause, whichever occurs first.
Secondary objectives: • To assess the safety and tolerability of immunotherapy in association with immunonutrition compared to that observed with nutritional counseling alone; • To assess the duration of response (DOR) with immunotherapy in association with immunonutrition compared to that observed with nutritional counseling alone; • To compare immunonutrition to nutritional counseling in terms of changes in body composition (assessed by TC scan at L3 and T12 level and bioelectrical impedance vector analysis [BIVA]); • To compare immunonutrition to nutritional counseling alone in terms of efficacy of immunotherapy with respect to Overall Survival (OS); • To compare immunonutrition to nutritional counseling alone in terms of efficacy of immunotherapy with respect to quality of life (QoL); • To compare immunonutrition to nutritional counseling alone in terms of efficacy of immunotherapy with respect to self-reported fatigue; • To evaluate the self-reported physical activity level, before diagnosis, at baseline and post-intervention, using the adapted version of the Godin's Shepard Leisure Time Exercise Questionnaire.
Secondary end-points: • Adverse events (AEs) and discontinuation due to AEs; • DOR, defined as the time from the first documented evidence of response until progression or death due to any cause, whichever occurs first; • Changes in body composition; • OS, defined as the time from randomization to the date of death due to any cause; • QoL, assessed by the EORTC QLQ-C30, the QLQ-LC13 a and the FACT-L Questionnaire; • Fatigue and its impact upon daily activities and function will be assessed at baseline and at the end of treatment using the 40-item Functional Assessment of Chronic Illness Therapy -Fatigue (FACIT-F) scale.
Benefit for participants
All participants will be receive early and tight nutritional assessment and support. Their nutritional status will be regularly monitored and nutritional support will be continuously optimized according to treatment tolerance and possible side-effects.
This study may result in significant improvements in nutritional care, which will prevent or ameliorate the impact of anticancer treatments in patients with NSCLC.
Potential risks and burdens for research participants
No risks and burdens for participants are expected in the context of the present research.
Dissemination
The results of the study will be presented at local, national and international medical meetings. The findings will be published in peer reviewed medical/scientific journals and made open-access on acceptance. Information may also be disseminated to the general population via public engagement and community outreach programs.
Statistics
For a future large study, we hypothesize a difference between a proportion of success 63.5% in the control arm and of 75% in the treatment arm to be clinically relevant. This hypothesis is based on the literature (KEY-NOTE-024, 042, 189 and 407 studies) [33][34][35][36].
Sample size
Sample size calculations are based on the primary endpoint. In a confirmatory study we would enroll 504 patients (252 in each arm), when the power is 80% and the type I error is 5%, and the proportion of success is expected 63.5% in control, and 75% in treated at 12 months.
An external pilot study of an overall trial designed with a power 80% and a type I error 5%, would aim at showing whether the treatment estimate is larger than zero. Using the one sided-90% confidence interval approach, with 154 patients (77 per arm), the lower 90% confidence limit for a zero difference would be 9.9%, excluding the 11.5% treatment effect estimate. In this case, the pilot study would point towards the presence of a treatment effect.
Accounting for a 15% dropout rate, we may enroll up to 180 patients (90 per group).
Calculations are performed following the approach by Cocks et al. [37], based on the confidence interval. Since this is a pilot study that will give elements to help in the decision to proceed with a confirmatory study, we will use a 90% one-tailed confidence interval (type I error of 10%). With this approach, the confidence interval is calculated under the H0 assumption of no difference between arms, using the expected sample size for the pilot study. If the upper limit of the interval excludes the hypothesized treatment effect in a confirmatory study, then consideration can be given to designing a confirmatory study.
Analysis set
Data analysis The Stata software (release 17, Stata-Corp, College Station, TX, USA) is used for sample size calculation, generation of the randomization list and data analysis.
Being this a pilot study, all analyses are exploratory and meant to guide the decisional process to proceed to a confirmatory study. We will use the mean and standard deviation or the median and quartiles to describe continuous data and the count and percent to describe categorical data; we will report rates per 100 person year to describe time to event data. The time horizon is 12 months.
If needed, normalizing transformations will be applied to the data prior to model fitting.
Analysis of the primary endpoint The risk difference between groups will be computed with a generalized liner model extended to the binomial family, with identity link (command binreg), together with its 90% confidence interval. Patients' dropout is expected to be very low. For them multiple imputation of PFS will be performed using all baseline characteristics. In a second model, treatment will be adjust for Enrolling center, Histology (squamous vs non-squamous), Type of treatment (immunotherapy alone vs chemo-immunotherapy) and PD-L1 TPS (< 1% vs ≥ 1%).
Randomization
Patients will be randomized 1:1 by the treating physician to one of the two study arms according to a computergenerated random blocks randomization list. Randomization will be stratified by center, in order to maintain the 1:1 ratio at center level. It will be performed via web, using the REDCap at Fondazione IRCCS Policlinico San Matteo. The system will assign the patient to the treatment arm after an initial check on the eligibility criteria to be answered by the treating physician. The randomization list, with random blocks, will be generated by and is kept at the Clinical Epidemiology & Biometry Unit of the coordinating center.
The Stata software (release 16, StataCorp, College Station, TX, USA) is used for sample size calculation, generation of the randomization list and data analysis.
Handling of missing data and drop-outs
Patients starting artificial nutrition and patients undergoing oncologic surgery during oncological treatments qualify as dropouts.
Study organization
The Fondazione IRCCS Policlinico San Matteo, Pavia, Italy, is responsible for the project management of the trial. The study was planned by the Clinical Nutrition and Dietetics Unit, the Medical Oncology Unit and the Clinical Epidemiology and Biometry Unit of the Fondazione IRCCS Policlinico San Matteo and the board of oncologists from other institutions listed as co-authors. Periodic board meetings will be scheduled (approximately every 3 months), in order to harmonize study procedures and to monitor and share the study progression.
Discussion
Malnutrition in oncology still represents an overlooked problem [38][39][40], which negatively affects clinical outcomes, and this is particularly relevant in patients with NSCLC [41]. The evidence supporting the efficacy of nutritional support in patients affected by NSCLC is promising, but still scanty and mainly focused on nutritional endpoints, while the impact on survival and treatment feasibility still requires confirmation.
Immunonutrition represents a promising approach in cancer care. It has been gaining attention in the last decades, particularly in the surgical gastrointestinal setting, where it has been demonstrated to reduce the rate of infectious complications and the length of hospitalization, without affecting mortality [42,43].
The present study ensures the early provision of nutritional assessment and support to all the enrolled patients, in agreement with recent evidence-based guidelines and recommendations. Besides, it would help clarifying the hypothesized advantages of immunonutrition during immunotherapy for patients with NSCLC.
Toxicity frequently requires the prolongation and/or the reduction of planned systemic treatments, resulting in reduced response rates and worse prognosis [44]. Therefore, tight nutritional support with immunonutrients since treatment initiation, aimed at fully and continuously satisfying estimated energy and protein requirements, may enable not only to achieve the maintenance/improvement of nutritional status and QoL, but may also have a positive and decisive impact on the adherence to anticancer treatment and the related curative intent.
Positive results from this pilot trial would stimulate further larger randomized -hopefully international -trials, potentially resulting in the improvement of the quality of supportive care for patients with NSCLC, and in the expansion of the number of patients who may benefit from immunonutrition also in the non-surgical oncologic setting.
Finally, the immune response is emerging as a key factor affecting the efficacy of treatments also in NSCLC. Therefore, we will also evaluate how the immunological profile could change during immunotherapy according to nutritional intervention.
This approach may help to initiate the exploration of the interactions between the immune system and the supplementation with immunonutrients. This new area of research could lead to the discovery of new molecular mechanisms regulating the immune system during anticancer treatments and, potentially, the development of new therapeutic strategies aimed at enhancing the efficacy of anticancer treatments themselves.
A possible practical critical aspect of the study could be the standardization of nutritional counseling. However, to achieve this across participating center, dietitians will share their local protocols and will clarify potential discrepancies.
|
v2
|
2022-11-26T14:44:47.319Z
|
2022-11-24T00:00:00.000Z
|
253882269
|
s2orc/train
|
Prognostic model for overall survival that includes the combination of platelet count and neutrophil–lymphocyte ratio within the first six weeks of sunitinib treatment for metastatic renal cell carcinoma
Background The association between the combination of platelet count and neutrophil–lymphocyte ratio (COP-NLR) at the time of adverse events during sunitinib treatment and prognosis is unclear, and prognostic models combining the prognostic factors of sunitinib have not been well studied. Thus, we developed a prognostic model that includes the COP-NLR to predict the prognosis of patients with metastatic renal cell carcinoma (mRCC) treated with sunitinib. Methods We performed a retrospective cohort study of 102 patients treated with sunitinib for mRCC between 2008 and 2020 in three hospitals associated with Showa University, Japan. The primary outcome was overall survival (OS). The collected data included baseline patient characteristics, adverse events, laboratory values, and COP-NLR scores within the first 6 weeks of sunitinib treatment. Prognostic factors of OS were analyzed using the Cox proportional hazards model. The integer score was derived from the beta-coefficient (β) of these factors and was divided into three groups. The survival curves were visualized using the Kaplan–Meier method and estimated using a log-rank test. Results The median OS was 32.3 months. Multivariable analysis showed that the number of metastatic sites, Memorial Sloan Kettering Cancer Center risk group, number of metastases, non-hypertension, modified Glasgow Prognostic Score, and 6-week COP-NLR were significantly associated with OS. A higher 6-week COP-NLR was significantly associated with a shorter OS (p < 0.001). The β values of the five factors for OS were scored (non-hypertension, mGPS, and 6-week COP-NLR = 1 point; number of metastatic sites = 2 points; MSKCC risk group = 3 points) and patients divided into three groups (≤ 1, 2–3, and ≥ 4). The low-risk (≤ 1) group had significantly longer OS than the high-risk (≥ 4) group (median OS: 99.0 vs. 6.2 months, p < 0.001). Conclusions This study showed that the COP-NLR within the first 6 weeks of sunitinib treatment had a greater impact on OS than the COP-NLR at the start of sunitinib treatment. The developed prognostic model for OS, including the 6-week COP-NLR, will be useful in decision-making to continue sunitinib in the early treatment stage of patients with mRCC. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-022-10316-w.
Background
Renal cell carcinoma (RCC) accounts for 5% of all cancers in men and 3% of cancers in women worldwide, representing the 6th and 10th most frequently diagnosed cancers, respectively [1,2]. The 5-year survival rate is 74% overall, decreasing to 8% among patients with metastatic disease (stage IV) [3,4]. Early-stage RCC is often asymptomatic, although the presence of systemic symptoms is frequently associated with advanced or metastatic RCC (mRCC).
The treatment selection for patients with mRCC widely uses the Memorial Sloan Kettering Cancer Center (MSKCC) and International Metastatic Renal Cell Carcinoma Database Consortium model. The advent of immune checkpoint inhibitors has broadened the treatment options for mRCC. Although sunitinib is one of the first-line treatment regimens for patients with low-risk mRCC, it must be discontinued if it causes severe bone marrow suppression. However, it has been reported that long-term treatment with sunitinib at a dose that reduces tumor size in the early stage is critical to maximize the potential efficacy of sunitinib treatment [5]. Therefore, determining the clinical benefit of continuing sunitinib prior to the occurrence of serious adverse events (AEs) leads to an appropriate treatment option for mRCC.
On the other hand, the combination of platelet count and neutrophil-lymphocyte ratio (COP-NLR), which is calculated using inflammatory markers, such as the NLR and platelet count (PLT), has been shown to be useful as a prognostic factor in gastrointestinal cancer and nonsmall cell lung cancer [16][17][18][19][20][21]. The COP-NLR before surgery or targeted therapy has also been associated with prognosis in patients with RCC [22,23]. Additionally, the COP-NLR values are affected by neutropenia and thrombocytopenia as AEs related to sunitinib treatment. In particular, these AEs are more likely to occur within the first 6 weeks of sunitinib treatment.
However, the association between the COP-NLR at the time of AEs during sunitinib treatment and prognosis is unclear, and models combining the prognostic factors of sunitinib have not been well studied. If a prognostic model could be developed, the clinical benefit of continuing sunitinib in the early stage could be determined, leading to the avoidance of serious AEs and longer survival based on long-term treatment with sunitinib. Therefore, we investigated the prognostic factors, including the COP-NLR at the time of AEs within the first 6 weeks of sunitinib treatment, and developed a prognostic model to predict the prognosis of patients with mRCC treated with sunitinib.
Study patients
We performed a retrospective cohort study of 102 patients treated with sunitinib for mRCC at Showa University Hospital, Showa University Northern Yokohama Hospital, and Showa University Fujigaoka Hospital, between June 2008 and August 2020. The data collection limit date was September 30, 2020. All patients were diagnosed with mRCC based on computed tomography (CT) / magnetic resonance imaging (MRI), and, when appropriate, brain imaging, and bone scintigraphy. This study was approved by the Ethics Committee of the Showa University School of Pharmacy.
Collection of patient data
Patient data and baseline laboratory values were collected from medical records. AEs within the first 6 weeks of sunitinib treatment were collected. To assess early response to treatment, laboratory values within the first 6 weeks of sunitinib treatment were collected. dose intensity (RDI) during the first 6 weeks of sunitinib treatment (6-week RDI), and duration of therapy. Blood test data included levels of hemoglobin (Hb), calcium (Ca), aspartate aminotransferase (AST), albumin (Alb), CRP, lactate dehydrogenase (LDH), alkaline phosphatase (ALP), mGPS, neutrophil-lymphocyte ratio (NLR), PLT and COP-NLR.
To investigate prognostic factors, including COP-NLR within the first 6 weeks of sunitinib treatment, we used the MSKCC classification that did not include neutrophil and PLT levels, which are components of COPNLR. COP-NLR within the first 6 weeks of sunitinib treatment is an item reflecting AEs and the early response to sunitinib treatment.
Definitions
The MSKCC model was based on five pretreatment variables (Karnofsky PS, LDH concentration, Hb concentration, serum Ca concentration, and time from initial diagnosis to start of systemic treatment) and divided into three risk groups: favorable-risk (0 risk factor), intermediate-risk (1, 2 risk factors), and poor-risk (≥ 3 risk factors) groups. Hypertension was defined as ≥ 140/90 mm Hg. Hypothyroidism was defined as elevated thyroidstimulating hormone levels with normal triiodothyronine and thyroxine levels. mGPS was defined as follows: patients with elevated CRP levels (> 0.5 mg/dL) and hypoalbuminemia (< 3.5 g/dL) were allocated mGPS 2, patients with only one factor were allocated mGPS 1, and patients with neither factor were allocated mGPS 0. The COP-NLR was defined as follows: patients with elevated platelet levels (> 310 × 10 9 /L) and NLR > 3.5 were allocated COP-NLR 2, patients with only one factor were allocated COP-NLR 1, and patients with neither factor were allocated COP-NLR 0.
Division
CRP and Alb levels were divided into two groups according to the lower limit of normal values. AST and ALP levels were divided into two groups according to the upper limit of the normal values. LDH was divided into two groups based on the LDH levels (333 U/L) of the MSKCC model. mGPS and COP-NLR were divided into two groups: moderate (score, 1) or higher.
Assessment of response
The response was assessed by CT/MRI performed at 2to 3-month intervals. Response data presented according to the Response Evaluation Criteria in Solid Tumors (RECIST) v.1.1. were collected from medical records. Progressive disease (PD), which is treatment response data, was collected to calculate PFS.
Adverse events
The following AEs related to sunitinib treatment were collected: hypertension, hand-foot syndrome, stomatitis, dysgeusia, oedema, nausea/vomiting, hemorrhage, constipation, diarrhea, fatigue, hypothyroidism, leukopenia, thrombocytopenia, anemia, elevation of AST, elevation of serum creatinine, and elevation of ALP. AEs related to sunitinib treatment were evaluated using the National Cancer Institute Common Terminology Criteria for Adverse Events version 5.0.
Outcome
The primary outcomes were time to treatment failure (TTF), PFS, and OS. Tumor progression was evaluated based on PD using RECIST.
Time-to-event variables were estimated using the Kaplan-Meier method. TTF was defined as the duration from the first day of sunitinib treatment until the date of discontinuation of sunitinib treatment or death from any cause, whichever came first. PFS was defined as the duration from the first day of sunitinib treatment to the date of tumor progression or death from any cause or the last follow-up visit, whichever came first. OS was defined as the duration from the first day of sunitinib treatment to the date of death from any cause or the last follow-up visit.
Statistical analysis Baseline and 6-week laboratory value changes
NLR and PLT values at the baseline and within the first 6 weeks of sunitinib treatment were compared by Wilcoxon rank sum tests.
The Kaplan-Meier method
Survival curves were estimated using the Kaplan-Meier method. The log-rank test was used to compare survival times between the two groups.
Univariate and multivariable analyses
Univariate and multivariable analyses were performed using the Cox proportional hazards model. Significant variables (p < 0.05) extracted by univariate analysis were entered into the multivariable analysis. Significant independent variables contributing to the prognosis of patients with mRCC treated with sunitinib were extracted using a stepwise selection method. These data were analyzed by using the SPSS software, version 27 (IBM, Tokyo, Japan). Statistical significance was set at P < 0.05.
Prognostic model and assessment
Prognostic model Each prognostic model was developed using prognostic factors extracted by multivariable analysis. The β values for these factors were derived from the smallest β value among the prognostic factors, approximated to the nearest integer. For each factor, the approximate β values were scored as integers. For each patient, the scores were calculated as the sum of the scores for each factor. Patients were divided into three groups (low-, intermediate-, and high-risk) based on the distribution of their scores. Survival curves of the three groups were estimated using the Kaplan-Meier method. The log-rank test was used to compare survival times among the three groups in prognostic models for TTF, PFS, and OS.
Prognostic nomogram A nomogram for possible prognostic factors was formulated to provide visualized risk prediction using R software with the rms package. A nomogram was established through Cox regression model analysis according to prognostic factors of OS (i.e., the number of metastatic sites, MSKCC risk group, nonhypertension, mGPS, and 6-week COP-NLR).
Assessment Calibration of the prognostic model and nomogram for OS was performed by comparing the predicted outcomes with the observed outcomes. The performance of the prognostic model and nomogram for predicting survival was evaluated with Harrell's concordance index (c-index) which is a measure of discrimination. The maximum value of the c-index is 1.0, which indicates perfect discrimination. The c-index of 0.5 indicates a random chance to correctly discriminate the outcome.
Patient characteristics
The characteristics of the 102 patients are shown in
Outcome
The cumulative survival curve for all patients is shown in Fig. 1. The median TTF, PFS, and OS were 4.9, 5.8, and 32.3 months, respectively. During the follow-up period, 17 patients (16.7%) discontinued sunitinib due to AEs, 87 patients (85.3%) experienced disease progression, and 55 patients (53.9%) died of any cause.
Univariate and multivariable analyses
The results of univariate and multivariable analyses are summarized in Table 2. Among the factors that were significant in the univariate analysis, multivariable analysis was performed, except for those that were correlated. In the multivariable analysis, the number of metastatic sites, AST, ALP, 6-week RDI, and 6-week COP-NLR were significantly associated with TTF. Additionally, the number of metastatic sites, MSKCC risk group, non-hand-foot syndrome, and 6-week COP-NLR were significantly associated with PFS. Moreover, the number of metastatic sites, MSKCC risk group, non-hypertension, mGPS, and 6-week COP-NLR were significantly associated with OS.
Survival curves according to the 6-week COP-NLR
The Kaplan-Meier curves of TTF, PFS, and OS according to the 6-week COP-NLR are shown in Fig. 2. A higher 6-week COP-NLR was significantly associated with shorter TTF, PFS, and OS (p < 0.001).
Prognostic model
The integer scores assigned from the β value of prognostic factors for TTF were as follows: 1 point for the number of metastatic sites, AST, ALP, and 6-week COP-NLR; and 2 points for 6-week RDI. The sum of the scores of the five factors, ranging from 0 to 6, was calculated for all patients. The patients were divided into three groups: low-risk group (≤ 1 point; n = 37), intermediate-risk group (2-3 points; n = 36), and highrisk group (≥ 4 points; n = 20). Additionally, the integer scores assigned from the β value of prognostic factors for PFS were as follows: 1 point for non-hand-foot syndrome; and 2 points for the number of metastatic sites, MSKCC risk group, and 6-week COP-NLR. The sum of the scores of the five factors, ranging from 0 to 7, was calculated for all patients. The patients were divided into three groups: low-risk group (≤ 1 point; n = 34), intermediate-risk group (2-3 points; n = 36), and high-risk group (≥ 4 points; n = 26). Moreover, the integer scores assigned from the β value of prognostic factors for OS were as follows: 1 point for non-hypertension, mGPS, and 6-week COP-NLR; 2 points for the number of metastatic sites; and 3 points for the MSKCC risk group. The sum of the scores of the five factors, ranging from 0 to 8, was calculated for all patients. The patients were divided into three groups: low-risk group (≤ 1 point; n = 30), intermediate-risk group (2-3 points; n = 32), and highrisk group (≥ 4 points; n = 34). The Kaplan-Meier curves of TTF, PFS, and OS according to the prognostic models are shown in Fig. 3. There were significant differences among the three groups in the prognostic models for TTF, PFS, and OS (p < 0.001). For internal validation, the bootstrapped calibration plot of the model predicting 1-year OS performed well with the ideal model (Supplemental Fig. 1). The C-index of model was 0.757 (95% confidence interval [CI]: 0.699-0.816).
Prognostic nomogram
Each factor in the nomogram was assigned a weighted number of points, and the sum of points for each patient was in accordance with a specific predicted 1-and 3-year OS (Supplemental Fig. 2). For internal validation, the bootstrapped calibration plot of the nomogram predicting 1-and 3-year OS performed well with the ideal model (Supplemental Fig. 3). The C-index of nomogram was 0.821 (95% CI: 0.762-0.880).
Discussion
In this study, we first demonstrated that the COP-NLR within the first 6 weeks of sunitinib treatment had a greater impact on OS than the COP-NLR at the start of sunitinib treatment. The 6-week COP-NLR, which reflects the early response to sunitinib treatment and bone marrow suppression, may be a useful prognostic indicator of the benefit from sunitinib. To the best of our knowledge, this is the first study to reveal the relationship between prognosis and the 6-week COP-NLR. Prognostic factors reflecting the early response to sunitinib treatment, such as the 6-week COP-NLR, have not been previously reported. Therefore, this study had a high clinical application value. Moreover, the developed prognostic model for OS with the addition of 6-week COP-NLR to the existing prognostic factors accurately predicted the prognosis of patients with mRCC treated with sunitinib. Thus, this model may provide clinical criteria for the continuation of sunitinib treatment in the early stages of mRCC. The higher 6-week COP-NLR indicated that sunitinib did not reduce the number of platelets and neutrophils in the blood. Sunitinib exhibits a dose-and time-dependent antitumor effect [24]. In the absence of the occurrence of thrombocytopenia, the antitumor effect of vascular endothelial growth factor receptor (VEGFR) inhibition is not achieved and may lead to a shorter OS. In tumor progression, neutrophils and lymphocytes, which are components of the COP-NLR, are associated with the tumor microenvironment. Neutrophils are involved in tumor progression, and lymphocytes play a role in antitumor immunity [25]. Platelets induce epithelial-to-mesenchymal transition in cancer and promote metastasis from the primary site [26]. Angiogenic factors and growth factors released from platelets promote tumor angiogenesis, tumor growth, and metastasis [27]. Therefore, in the absence of neutropenia or thrombocytopenia, cytokines released from neutrophils may cause tumor growth and progression. NLR and PLT levels within the first 6 weeks of sunitinib treatment were significantly reduced compared to those at baseline. This suggests that the early response of NLR and PLT levels was associated with improved prognosis. Therefore, the 6-week COP-NLR is a useful prognostic factor combined with bone marrow suppression and early response to sunitinib treatment.
In addition to the 6-week COP-NLR, the MSKCC risk group, number of metastases, non-hypertension, and mGPS were significantly associated with OS. These prognostic factors were similar to those reported previously [6,8] Additionally, sunitinib-induced hypertension is correlated with the effects of VEGFR inhibition [12]. In the absence of the occurrence of hypertension, the effect of VEGFR inhibition is not achieved and may lead to a shorter OS. Non-hypertension is an important prognostic indicator because it has been previously reported as a prognostic factor [13].
We showed that the developed OS prognostic model accurately predicted the prognosis of patients with mRCC treated with sunitinib. In this developed prognostic model, integrating the 6-week COP-NLR into the existing prognostic factors may improve discrimination between groups and thus improve individual risk prediction. Although COP-NLR has previously been shown to be useful at baseline prior to TKI [22], this study is the first to show that alterations of these values may provide additional prognostic information.
In this study, a prognostic nomogram and model for OS were developed; the C-index of the nomogram was higher than that of the model. However, the advantage of the developed prognostic model is that it is easy to stratify prognostic risk into three groups based on simple scores. The model is simple and easy to use in clinical practice, making it a useful tool to assist providers in determining appropriate treatment according to their prognostic risk for patients with mRCC. Therefore, the model is useful tool for decision-making to continue sunitinib in the early treatment stage for patients with mRCC. The low-risk group achieved an antitumor effect from the VEGFR inhibitor sunitinib, which is expected to lead to a longer OS. On the other hand, in the highrisk group, a longer OS cannot be expected even if sunitinib is selected, so it is necessary to consider changing to other molecular-targeted agents or immune checkpoint inhibitors.
In this study, we also investigated the impact of TTF and PFS on prognosis. In mRCC, it is important to use a highly effective drug at an early stage for as long as possible as the effect of tumor burden reduction in the early stage of sunitinib treatment affects subsequent prognosis [5]. Additionally, it has been reported that long-term treatment at a dose to achieve tumor burden reduction is associated with a favorable prognosis. Therefore, PFS associated with tumor growth and TTF associated with treatment continuation are considered to have a strong impact on the prognosis of mRCC. The prognostic factors of TTF and PFS may be important indicators for selecting a targeted agent for mRCC.
Non-hand-foot syndrome, high AST (> 30 U/L), ALP (> 322 U/L) levels, and 6-week RDI (< 60%) were extracted as prognostic factors for PFS and TTF, respectively. Hand-foot syndrome is a favorable prognostic factor for sunitinib [28]. High AST and ALP levels are associated with liver and bone metastases and indicate poor PS [7,29]. The high AST group had liver metastases in 26.3% of cases, and the high ALP group had bone metastases in 31.6% of cases (data not shown). Additionally, because sunitinib is metabolized in the liver, early liver toxicity is likely to lead to discontinuation of sunitinib at the early stage.
Limitations
The present study has two limitations. First, there were few patients treated with sunitinib as first-line therapy; therefore, a prognostic model could not be developed for patients with mRCC treated with sunitinib as first-line therapy. Second, a prognostic model that included the severity of AEs could not be developed.
Conclusions
This study showed that the COP-NLR within the first 6 weeks of sunitinib treatment had a greater impact on OS than the COP-NLR at the start of sunitinib treatment. We showed that the developed prognostic model for OS with the addition of 6-week COP-NLR to the prognostic factors at baseline accurately predicted the prognosis of patients with mRCC treated with sunitinib. The developed prognostic model for OS, including the 6-week COP-NLR, will be useful in decision-making to continue sunitinib in the early treatment stage of patients with mRCC. The lowrisk group can achieve the antitumor effect of the VEGFR inhibitor sunitinib, which is expected to lead to longer OS.
|
v2
|
2022-11-26T16:06:57.631Z
|
2022-11-24T00:00:00.000Z
|
253911856
|
s2orc/train
|
Anti-Inflammatory Effects Exerted by 14-Methoxyalternate C from Antarctic Fungal Strain Pleosporales sp. SF-7343 via the Regulation of NF-κB and JAK2/STAT3 in HaCaT Human Keratinocytes
Atopic dermatitis (AD) is a chronic inflammatory skin disease with a profound negative impact on patients’ quality of life. Four known secondary fungal metabolites were found in the chemical study of the Antarctic fungus Pleosporales sp. SF-7343, including 14-methoxyalternate C (1), 5′-methoxy-6-methyl-biphenyl-3,4,3′-triol (2), 3,8,10-trihydroxy-4-methoxy-6-methylbenzocoumarin (3), and alternariol monomethyl ether (4). Additionally, we identified the skin anti-inflammatory composition from the SF-7343 strain. Interleukin-8 and -6 Screening results showed that compound 1 inhibited IL-8 and IL-6 in tumor necrosis factor-α/interferon-γ stimulated HaCaT cells. Compound 1 showed inhibitory effects on MDC and RANTES. It also downregulated the expression of intercellular adhesion molecule-1 (ICAM-1) and upregulated the expression of involucrin. The results of the mechanistic study showed that compound 1 inhibited the nuclear translocation of nuclear factor-kappa B p65 and STAT3. In conclusion, this study demonstrates the potential of the Antarctic fungal strain SF-7343 as a bioactive resource to inhibit skin inflammation, such as AD.
Introduction
Atopic dermatitis (AD) is a usual chronic skin inflammatory disease. Approximately 10% of adults and 20% of children suffer from this disease, which is characterized by a compromised immune system, excessive inflammation, and skin barrier disruption [1,2]. Genetic factors, immune disorders, and epidermal barrier dysfunction are all causes of AD, the pathogenesis is very complex. Severe AD is now generally treated with monoclonal antibodies. Long-term use of steroids and antihistamines has significant side effects. Skin fragility and thinning, suppression of melanocytes and gastrointestinal side effects. Therefore, there is a need to find alternative strategies for the treatment of mild or moderate AD [3][4][5]. In many instances, skin diseases caused by infections or other tissue damage are indistinguishable from inflammation [6]. The two major causes of AD are skin immune system disorder and epidermal barrier destruction [7]. A damaged epidermal barrier allows the intrusion of various allergens, triggering immune imbalance and aggravating the development and deterioration of AD. At the onset of AD, resident and infiltrating cells, such as keratinocytes, langerhans cells, and neutrophils, overexpress chemokines [8,9], which activate type 2 helper T (Th2) cells and induce the generation of Th2-type cytokines.
This results in the aggravation of the epidermal barrier, impaired keratinocyte differentiation, persistent skin itching, and impairment of stratum corneum permeability [10]. Keratinocytes are the dominant kind of type in the epidermis cells. Compounds with antiinflammatory effects play a key role in the progress and pathogenesis of AD by enabling communication with other cells. Hence, in vitro keratinocyte models have been widely used to study the possibility of various natural and synthetic substances as anti-inflammatory candidates to regulate inflammation in the keratinocytes [11].
The Antarctic microorganisms are of particular interest because of the enormous potential for isolating new biologically active and valuable components that are yet unexplored [12]. This can be attributed to the unique and harsh environment of the Antarctic, which induces unusual metabolic properties and the production of unusual metabolites [13]. In our previous studies, we isolated four compounds from the metabolites of Pleosporales sp. SF-7343, and elucidated the anti-inflammatory effects of alternate C in skin inflammation in human keratinocytes [14]. In this study, we isolated additional components from the fungal strain Pleosporales sp. SF-7343 and investigated the regulation of skin inflammatory response in human keratinocytes exerted by these compounds.
causes of AD are skin immune system disorder and epidermal barrier destruction [7]. A damaged epidermal barrier allows the intrusion of various allergens, triggering immune imbalance and aggravating the development and deterioration of AD. At the onset of AD, resident and infiltrating cells, such as keratinocytes, langerhans cells, and neutrophils, overexpress chemokines [8,9], which activate type 2 helper T (Th2) cells and induce the generation of Th2-type cytokines. This results in the aggravation of the epidermal barrier, impaired keratinocyte differentiation, persistent skin itching, and impairment of stratum corneum permeability [10]. Keratinocytes are the dominant kind of type in the epidermis cells. Compounds with anti-inflammatory effects play a key role in the progress and pathogenesis of AD by enabling communication with other cells. Hence, in vitro keratinocyte models have been widely used to study the possibility of various natural and synthetic substances as anti-inflammatory candidates to regulate inflammation in the keratinocytes [11].
The Antarctic microorganisms are of particular interest because of the enormous potential for isolating new biologically active and valuable components that are yet unexplored [12]. This can be attributed to the unique and harsh environment of the Antarctic, which induces unusual metabolic properties and the production of unusual metabolites [13]. In our previous studies, we isolated four compounds from the metabolites of Pleosporales sp. SF-7343, and elucidated the anti-inflammatory effects of alternate C in skin inflammation in human keratinocytes [14]. In this study, we isolated additional components from the fungal strain Pleosporales sp. SF-7343 and investigated the regulation of skin inflammatory response in human keratinocytes exerted by these compounds.
Determination of the Molecular Structure of the Isolated Compounds from Metabolites
To obtain 14-methoxyalternate C (1), 5′-methoxy-6-methyl-biphenyl-3,4,3′-triol (2), 3,8,10-trihydroxy-4-methoxy-6-methylbenzocoumarin (3), and altenuene (4). Multi step chromatographic analysis were carried on using the fungal strain dry fermentation extract. Based on comparing the results of 1D and 2D NMR and MS analyses with data reported in the literature, their structures are shown in Figure 1. Compound 1, based on the HRESIMS peak at m/z 357.0956 [M + Na] + (calcd. for C 17 H 18 O 7 Na, 357.0950) the molecular formula was determined to be C 17 H 18 O 7 . The 13 C-, DEPT-, and HMQC-NMR spectra of compound 1 revealed signals of a carbonyl carbon (δ C 171.5), 12 olefinic carbons [containing four oxygenated carbons at δ C 163.6, 162.9, 145.1, and 144.6], three methoxy carbons (δ C 57.5, 55.8, and 52.0), and an oxygenated methylene carbon (δ C 72.5). The requied 9 degrees unsaturation, these molecular formulas occupied for 7, thereby indicating that compound 1 is a biphenyl derivative. Additionally, the NMR data of compound 1 were quite similar to those of alternate C [15], suggesting that compound 1 had the same biphenyl skeleton made of two tetrasubstituted benzenes. A detailed comparison of the 1D NMR data of compound 1 and alternate C revealed that the difference in the signal is due to the presence of methoxymethylene protons instead of a methyl group at C-13. This structural difference was supported by the Heteronuclear Multiple Bond Correlation (HMBC) of H-12/C-14, H-14/C-13, C-14-OCH 3 , and H-14-OCH 3 /C-14 (Table S1). Therefore, the structure of Compound 1 is thought to be 14-methoxyalternate C. (Table S2). A comparison of the 1D NMR data of compound 2 in the literature helped determine the structure of compound 2, as shown in Figure 1B [16].
Cell Viability of Isolated Compounds 1, 2, 3, and 4 in HaCaT Cells
We using a MTT assay to examined the cytotoxicity of our isolated compounds 1, 2, 3, and 4. Results are shown in Figure 2; for the subsequent experiments, the cells were co-treated with these four compounds in a safe concentration (10-40 µM).
Inhibitory Effects of the Four Compounds on IL-6 and IL-8 Production in TNF-α/IFN-γ-Treated HaCaT Cells
The secretion of IL-6 and IL-8 were to select the compound which we isolated is with anti-inflammatory effect. The results are shown in Table 1, wherein compound 1 inhibited the secretion of IL-8 and IL-6. However, compounds 2, 3, and 4 had no obvious inhibitory effect on IL-8 or IL-6 secretion. Therefore, we performed further experiments using only compound 1.
Inhibitory Effects of the Four Compounds on IL-6 and IL-8 Production in TNF-α/IFN-γ-Treated HaCaT Cells
The secretion of IL-6 and IL-8 were to select the compound which we isolated is with anti-inflammatory effect. The results are shown in Table 1, wherein compound 1 inhibited the secretion of IL-8 and IL-6. However, compounds 2, 3, and 4 had no obvious inhibitory effect on IL-8 or IL-6 secretion. Therefore, we performed further experiments using only compound 1.
Effects of Compound 1 on the Level of MDC and RANTES in TNF-α/IFN-γ-Stimulated HaCaT Cells
The main fuction of chemokines is to regulate the recruitment of inflammatory cells at the site of infection or inflammation. The effects of compound 1 on RANTES and MDC are shown in Figure 3A,B, wherein the secretion of MDC and RANTES was significantly increased by co-stimulated TNF-α and IFN-γ compared to that in the control groups, which showed a dose-dependent reduction on pretreatment with compound 1. Int
Effects of Compound 1 on the JAK2/STAT3 Signaling Pathways in HaCaT Cells
JAK/STAT pathway regulates the major signaling cascades for many cytokines, chemokines and growth factors, and it also participates in intracellular signal transduction and expression [21]. As shown in Figure 5A,B, when cells were co-treated with TNF-α and IFN-γ, compound 1 significantly downregulated the phosphorylation of JAK2 and STAT3. As shown in Figure 5C, the nuclear translocation of STAT3 was inhibited by compound 1, consistent with the results of Western blotting. These results suggest that compound 1 regulates the JAK2/STAT3 signal pathways in HaCaT cells. JAK/STAT pathway regulates the major signaling cascades for many cytokines, chemokines and growth factors, and it also participates in intracellular signal transduction and expression [21]. As shown in Figure 5A and B, when cells were co-treated with TNF-α and IFN-γ, compound 1 significantly downregulated the phosphorylation of JAK2 and STAT3. As shown in Figure 5C, the nuclear translocation of STAT3 was inhibited by compound 1, consistent with the results of Western blotting. These results suggest that compound 1 regulates the JAK2/STAT3 signal pathways in HaCaT cells.
Effects of Compound 1 on NF-κB Signaling Pathways in HaCaT Cells
In resting cells, inhibitor protein IκB forms a complex with inactive NF-κB, retaining it in the cell cytoplasm. Upon stimulation, IκB is phosphorylated and subsequently degraded by the proteasome, causes the translocation of NF-κB into the nucleus, where it regulates the transcription of specfic genes encoding proinflammatory cytokines [22,23]. In Figure 6 A and B, co-treatment with TNF-α and IFN-γ significantly increased the p65 and p-IκBα levels compared with that in the control group. Pre-treatment with compound 1 decreased IκBα phosphorylation and p65 activation. The p-IκBα/IκBα ratio also decreased. Additionally, in Figure 6 C, the results of immunofluorescence showed that the p65 nuclear translocation was inhibited by compound 1, demonstrating that compound 1 can alleviate inflammation in HaCaT cells through NF-κB signaling pathways.
Effects of Compound 1 on NF-κB Signaling Pathways in HaCaT Cells
In resting cells, inhibitor protein IκB forms a complex with inactive NF-κB, retaining it in the cell cytoplasm. Upon stimulation, IκB is phosphorylated and subsequently degraded by the proteasome, causes the translocation of NF-κB into the nucleus, where it regulates the transcription of specfic genes encoding proinflammatory cytokines [22,23]. In Figure 6A,B, co-treatment with TNF-α and IFN-γ significantly increased the p65 and p-IκBα levels compared with that in the control group. Pre-treatment with compound 1 decreased IκBα phosphorylation and p65 activation. The p-IκBα/IκBα ratio also decreased. Additionally, in Figure 6C, the results of immunofluorescence showed that the p65 nuclear translocation was inhibited by compound 1, demonstrating that compound 1 can alleviate inflammation in HaCaT cells through NF-κB signaling pathways. Immunofluorescence assay was performed as described in the methods section. Data are represented as the mean ± SD (n = 3). * p < 0.05, ** p < 0.01, vs. TNF-α/IFN-γ-treated group.
JAK inhibitors have been investigated as possible solutions for treating AD [24]. The inhibition of pro-inflammatory cytokine or chemokine secretion from the JAK-STAT pathway ultimately ameliorates the symptoms of AD, leading to improved quality of the patient's life [25]. STATs are phosphorylated upon cytokine stimulation by JAK, causing the dimerization of STAT, followed by translocation of STAT to the nucleus through the nuclear membrane to combine and regulate the expression of their target genes. The imbalance in T helper (Th)-2 cells is an important immunopathological characteristics of AD. JAK-STAT pathway will activated be when the cytokine binds to their specific receptor, and then start the phosphorylation cascade on the cell cytoplasmic side, leading to the transcription of the target genes [26][27][28]. Compound 1 showed a strong regulatory Figure 6. Effects of compound 1 on NF-κB signaling pathways in HaCaT cells (A-C). The expression of p65, p-IκBα, and IκBα in the fractions were determined by using Western blotting. Immunofluorescence assay was performed as described in the methods section. Data are represented as the mean ± SD (n = 3). * p < 0.05, ** p < 0.01, vs. TNF-α/IFN-γ-treated group.
JAK inhibitors have been investigated as possible solutions for treating AD [24]. The inhibition of pro-inflammatory cytokine or chemokine secretion from the JAK-STAT pathway ultimately ameliorates the symptoms of AD, leading to improved quality of the patient's life [25]. STATs are phosphorylated upon cytokine stimulation by JAK, causing the dimerization of STAT, followed by translocation of STAT to the nucleus through the nuclear membrane to combine and regulate the expression of their target genes. The imbalance in T helper (Th)-2 cells is an important immunopathological characteristics of AD. JAK-STAT pathway will activated be when the cytokine binds to their specific receptor, and then start the phosphorylation cascade on the cell cytoplasmic side, leading to the transcription of the target genes [26][27][28]. Compound 1 showed a strong regulatory effect on the activity of the JAK2/STAT3 signaling pathway and may be a good promising candidate for the treatment of AD.
Skin provides protection for our body; it is the barrier of our body and the external environments, prevent the invasion of external factors and maintain moisture [29]. Impairment of skin barrier function and the subsequently increased penetration of allergens into the skin, cause the allergic inflammatory response, this is the main feature of the Th2 inflammation in AD area [30]. FLG is a basic structural protein for the maintenance and development of the skin barrier function. It induces the outermost skin cell's structural proteins to form tight bundles that flatten and strengthen the cells to create a strong and solid skin barrier [31]. Mutations in the FLG gene encoding filaggrin cause other atopic diseases, increased risk of ADs, and exacerbation of some conditions [32]. IVL is a protein component of human skin. It is a structural protein precursor of the keratinocyte-cornified envelope, a protective sheath of covalently crosslinked proteins that is formed during the final stages of keratinocyte differentiation [33]. In this study, compound 1 upregulated IVL expression but failed to upregulate the FLG expression, which was decreased by TNF-α and IFN-γ treatment. These results demonstrate that the protective activity of compound 1 was not exerted through modulation of filaggrin expression.
Skin inflammation is one of the key features in the pathogenesis of AD. Agents exhibiting anti-inflammatory activity could be potential therapeutic candidates for AD treatment. Our study on fungal metabolites from the Antarctic fungal strain Pleosporales sp. SF-7343 led to the identification of 14-methoxyalternate C (1), which exerted inhibitory effects on inflammation in the TNF-α and IFN-γ-induced HaCaT cells. This metabolite could be further developed as an agent to prevent inflammation in human keratinocytes.
Cell Culture and Reagents
HaCaT cells were cultured in DMEM media added with 10%FBS. For the source of antibody and other reagent, refer to our published articles [34].
MTT Assay
HaCaT cells were seeded at a density of 2 × 10 4 cells in a 48-well plate for 24 h and then treated with compounds 1-4 (20-80 µM). The determination method of MTT assay is performed according to the previous paper [34].
Measurement of Cytokines and Chemokines
The culture supernatants were used to check the secretion of IL-6, IL-8, RANTES, and MDC by specific ELISA kits, according to the manufacturer's instructions.
Extraction of Total, Nuclear, and Cytosolic Protein
HaCaT cells were pretreated with compound 1 for 3 h. For the total protein analysis, cells were lysed by RIPA buffer. For nuclear and cytoplasmic proteins, the cells were extracted by using the Nuclear Extraction Kit according to the manufacturer's instructions.
Western Blot Analysis
For the separation of protein samples using SDS-PAGE gels and transfer to NC membranes, the experimental steps are according to our published articles [34].
Immunofluorescence
For the translocation of NF-κB and STAT3, HaCaT cells were cultured on glass chamber slides. After treatment, the cells were fixed in paraformaldehyde, permeabilized with 0.01% TX-100, and probed first with NF-κB and STAT3 antibodies and second with FITC-labelled secondary antibodies. Then, the cells were treated with DAPI solution for 10 min and washed, and then, coverslips were covered on glass slides with an anti-fade reagent. Pictures were taken under a Nikon fluorescence microscope (ECLIPSE Ts2; Nikon Optical Co, Tokyo, Japan).
Statistical Analysis
Results for each group are represented as the mean ± standard deviation (SD) (n = 3). one-way analysis of variance was carried out by the GraphPad Software. The significance results were followed by Duncan's multiple comparison tests. Statistical significance was set to * p < 0.05, ** p < 0.01, *** p < 0.001 vs. the TNF-α/IFN-γ-treated groups.
Conclusions
This time, the four compounds isolated from the fungal strain Pleosporales sp. SF-7343, their skin anti-inflammatory effects were examined in TNF-α-and IFN-γ-treated keratinocytes. Among them, 14-methoxyalternate C showed inhibitory effects on inflammatory cytokines and chemokines, decreased ICAM-1 expression, and increased IVL expression. The anti-inflammatory effect may have been exerted through the regulation of the two signaling pathways, JAK2/STAT3 and NF-κB. These results indicated that the compounds extracted from the metabolites of the fungal strain SF-7343 has the potential to become a preventive or therapeutic agent for AD patients.
|
v2
|
2022-11-26T16:24:54.539Z
|
2022-11-24T00:00:00.000Z
|
253943094
|
s2ag/train
|
Dyskeratosis congenita: a case report
Dyskeratosis congenita is a rare hereditary disorder characterized by skin pigmentation, nail dystrophy and leukoplakia along with bone marrow failure and increased predisposition to malignant tumours. Here we describe a 7-year-old child who presented with classic triad of pigmentation, nail changes, leukoplakia along with manifestations of bone marrow failure. He was initially put on androgen therapy with plan for a possible matched related allogenic HSCT. A haematologist should be aware of the complications of therapy, bronchopulmonary complications, futility of IST and dangers of using myeloablative transplant in such a patient.
|
v2
|
2022-11-26T16:45:16.487Z
|
2022-11-24T00:00:00.000Z
|
253929733
|
s2ag/train
|
Public health risk assessments associated with heavy metal levels in panga fish fillets imported from Vietnam
Pangasius hypophthalmus (panga fish) is farmed in the Mekong River (Vietnam), which is known as a polluted river, and exported to many countries. The present study aimed to determine heavy metal levels in frozen panga fillets imported from Vietnam as well as the risks of heavy metals to human health. Panga fillets belonging to four brands were bought from three supermarkets in Adana city, Turkey, and heavy metals (As, Cd, Hg, Pb, Ni, Cu, Mn and Co) were analyzed. To analyze the potential risks to human health, EWI (estimated weekly intake), THQ (target hazard quotient), and CR (lifetime cancer risk) values were calculated to assess the potential risks to consumer health of the metal content in panga fillets. The health risk assessment values were calculated for children and adults according to the frequency of consumption once, three and seven times a week. The results revealed that the presence of heavy metals in the studied panga fillets was below permissible limits indicated by WHO (World Health Organization), EPA (United States Environmental Protection Agency) and TKB (Turkish Fisheries Laws and Regulations). The EWI, THQ or ∑ THQ and CR values were below PTWI (provisional tolerable weekly intake), 1 and 10-5, respectively. Remarkably, the highest values of the EWI/PTWI ratio and THQs were found for children.
|
v2
|
2022-11-26T16:56:56.603Z
|
2022-11-24T00:00:00.000Z
|
253923721
|
s2ag/train
|
Synchronous Bilateral Flexure Tumor Causing Ileus and Requiring Surgical Treatment
This case report presented a simultaneous right colon tumor detected perioperatively in a patient who developed ileus due to a metastatic left colon tumor in the preoperative period. A seventy-six-year-old man was admitted with epigastric pain, nausea, and vomiting. There was tenderness and defense on deep palpation on the epigastrium. On computed tomography, there were multiple hypodense lesions on the liver, a mass at the level of the splenic flexure that obliterates the lumen. In addition, there were numerous air-fluid levels due to tumoral mass on splenic flexure. Emergency surgery was performed, and during surgery, there were tumoral masses at the hepatic flexure and splenic flexure. Palliative total abdominal colectomy with end ileostomy was performed due to megacolon. The patient died due to sudden cardiac arrest on the 1st postoperative day.
|
v2
|
2022-11-26T16:59:15.797Z
|
2022-11-24T00:00:00.000Z
|
253932490
|
s2orc/train
|
Numb Chin Syndrome in Sickle Cell Disease: A Systematic Review and Recommendations for Investigation and Management
Numb chin syndrome (NCS) is a rare sensory neuropathy resulting from inferior alveolar or mental nerve injury. It manifests as hypoesthesia, paraesthesia, or, rarely, as pain in the chin and lower lip. Several case reports suggest that sickle cell disease (SCD) could be a cause of NCS. However, information about NCS is scarce in this population. Our objectives were to synthesize all the available literature relevant to NCS in SCD and to propose recommendations for diagnosis and management based on the best available evidence. A systematic review was performed on several databases to identify all relevant publications on NCS in adults and children with SCD. We identified 73 publications; fourteen reports met the inclusion/exclusion criteria. These described 33 unique patients. Most episodes of NCS occurred in the context of typical veno-occlusive crises that involved the mandibular area. Radiological signs of bone infarction were found on some imaging, but not all. Neuropathy management was mostly directed toward the underlying cause. Overall, these observations suggest that vaso-occlusion and bone infarction could be important pathophysiological mechanisms of NCS. However, depending on the individual context, we recommend a careful evaluation to rule out differential causes, including infections, local tumors, metastatic disease, and stroke.
Introduction
Sickle cell disease (SCD), the most common hemoglobinopathy worldwide, is a group of inherited blood cell disorders associated with episodes of acute illness and progressive organ damage [1]. Early historical data suggests that the prevalence of the sickle cell allele responsible for sickle cell disease is highest in Sub-Saharan Africa as well as some parts of the Mediterranean, the Middle East, and India reaching frequencies of up to 34%. Genetic studies and the strong geographic relationship between the distribution of the sickle cell allele and malaria suggest that the sickle cell allele confers a selective advantage for improving the survival of mutation carriers against severe and lethal malaria. Due to the modern globalization process, migration, and the forced relocation of millions of Africans to the Americas driven by the slave trade, means that the distribution of the sickle cell allele has expanded across the globe. Thus SCD has become a worldwide concern [2].
SCD has been associated with several central nervous system disorders, mainly through vascular phenomena [3,4]. Patients are at risk of silent ischemic infarcts, Moya Moya syndrome, aneurysms, and ischemic and/or hemorrhagic strokes [5][6][7][8]. An increased Diagnostics 2022, 12, 2933 2 of 11 prevalence in the early onset of the neurocognitive disorder is observed, even in those without ischemic lesions, possibly because of reduced cerebrovascular reserve [9,10]. Patients with the SC genotype also commonly report auditory disturbances [11]. Despite peripheral neuropathies being relatively rare in SCD, there have been several cases associated with numb chin syndrome (NCS).
NCS is a rare sensory neuropathy characterized by inferior alveolar or mental nerve damage, which manifests as hypoesthesia, paraesthesia, or pain in the chin and lower lip ( Figure 1) [12,13]. These nerves provide sensory information from the front of the chin, the lower lip, and the labial gingivae of the mandibular anterior teeth and the premolar teeth [14,15]. Although more commonly associated with malignancies, traumatic injuries, and dental interventions, NCS has also been described in patients suffering from SCD [16]. The few available case reports suggest that NCS was caused by SCD a complication of acute painful crises in the mandibular bone and infarction in the nerve passage within the mandibular canal. During an episode of vaso-occlusive crisis (VOC), it has been hypothesized that NCS can be triggered by local ischemia due to thrombosis of the vasa vasorum.
of the sickle cell allele has expanded across the globe. Thus SCD has become a worldwide concern [2]. SCD has been associated with several central nervous system disorders, mainly through vascular phenomena [3,4]. Patients are at risk of silent ischemic infarcts, Moya Moya syndrome, aneurysms, and ischemic and/or hemorrhagic strokes [5][6][7][8]. An increased prevalence in the early onset of the neurocognitive disorder is observed, even in those without ischemic lesions, possibly because of reduced cerebrovascular reserve [9,10]. Patients with the SC genotype also commonly report auditory disturbances [11]. Despite peripheral neuropathies being relatively rare in SCD, there have been several cases associated with numb chin syndrome (NCS).
NCS is a rare sensory neuropathy characterized by inferior alveolar or mental nerve damage, which manifests as hypoesthesia, paraesthesia, or pain in the chin and lower lip ( Figure 1) [12,13]. These nerves provide sensory information from the front of the chin, the lower lip, and the labial gingivae of the mandibular anterior teeth and the premolar teeth [14,15]. Although more commonly associated with malignancies, traumatic injuries, and dental interventions, NCS has also been described in patients suffering from SCD [16]. The few available case reports suggest that NCS was caused by SCD a complication of acute painful crises in the mandibular bone and infarction in the nerve passage within the mandibular canal. During an episode of vaso-occlusive crisis (VOC), it has been hypothesized that NCS can be triggered by local ischemia due to thrombosis of the vasa vasorum. Given the limited published evidence on NCS in SCD, it remains a poorly characterized disease. Thus, information about the causes, disease course, and treatment is limited. Our objective was to review the available literature to synthesize all relevant studies that discuss NCS in SCD patients. A secondary objective was to propose recommendations for the management of NCS based on the best available evidence.
Materials and Methods
A systematic review of the currently available literature was performed following our pre-specified protocol that was registered on PROSPERO (CRD42021239583).
A number of databases were searched by a professional librarian for relevant studies on January 5th, 2021. The databases include MEDLINE (via Ovid, 1946 to 5 January 2021); Embase (via Ovid, 1974 to 3 January 2021), EBM Reviews (via Ovid up to 5 January 2021) and CI-NAHL Complete (via EBSCO, 1937 to 5 January 2021). The search strategies, designed by a librarian (CS), used text words and relevant indexing terms to identify studies concerning NCS in patients with SCD. The reports describing adults and/or children with any SCD genotype were included. Reports pertaining to patients with SCD traits were excluded. The MEDLINE strategy supplementary Table S1was applied to all databases, with modifications to the search terms as necessary. No language limits were applied to the search. Conference abstracts of the relevant scientific meetings were manually searched with no language limits applied. Reviews, notes, editorials, or comments were excluded from the search criteria. The results were uploaded to Covidence, and duplicates were removed automatically. A two-stage screening process was performed to extract all Given the limited published evidence on NCS in SCD, it remains a poorly characterized disease. Thus, information about the causes, disease course, and treatment is limited. Our objective was to review the available literature to synthesize all relevant studies that discuss NCS in SCD patients. A secondary objective was to propose recommendations for the management of NCS based on the best available evidence.
Materials and Methods
A systematic review of the currently available literature was performed following our pre-specified protocol that was registered on PROSPERO (CRD42021239583).
A number of databases were searched by a professional librarian for relevant studies on 5 January, 2021. The databases include MEDLINE (via Ovid, 1946 to 5 January 2021); Embase (via Ovid, 1974 to 3 January 2021), EBM Reviews (via Ovid up to 5 January 2021) and CI-NAHL Complete (via EBSCO, 1937 to 5 January 2021). The search strategies, designed by a librarian (CS), used text words and relevant indexing terms to identify studies concerning NCS in patients with SCD. The reports describing adults and/or children with any SCD genotype were included. Reports pertaining to patients with SCD traits were excluded. The MEDLINE strategy Supplementary Table S1 was applied to all databases, with modifications to the search terms as necessary. No language limits were applied to the search. Conference abstracts of the relevant scientific meetings were manually searched with no language limits applied. Reviews, notes, editorials, or comments were excluded from the search criteria. The results were uploaded to Covidence, and duplicates were removed automatically. A two-stage screening process was performed to extract all the existing relevant reports. In the first step, two independent reviewers, (LT and MB) performed a title and abstract screening to identify, select, and filter the pertinent reports. In the second screening step, the two independent reviewers, (LT and MB) performed a full-text examination to further identify and select the pertinent reports. Excluded reports were sorted and designated as being "Non-relevant", having the "Wrong study design", or discussing an "Animal" study. A cross-comparison of the autonomously screened texts was carried out by the reviewers LT and MB. Disagreements were resolved by a consensus or with the input of a third reviewer (SF).
The extraction of the data from the final approved list of relevant texts was performed by the two independent reviewers (LT and MB). Discrepancies occurring within the data extraction step were resolved by a consensus or with the input of a third reviewer (SF). The study data that was extracted included the year of publication, country, study design, sample size, and follow-up duration. Patient demographics (including age, sex, country, and sickle cell type), clinical presentation, investigations, and outcomes were also extracted.
Results
After the removal of duplicates, 73 studies were identified using the pre-specified systematic search strategy. Following two levels of screening, 11 studies were included for extraction, and three publications were found and extracted following a manual search of references ( Figure 2) [17][18][19][20][21][22][23][24][25][26][27][28][29][30]. The included studies were published between 1972 and 2021. Overall, 33 distinct patients were included in this report (Supplementary Table S1). The geographical distribution of patients was as follows: 39.4% of patients were from Jamaica, 15.2% from Ghana, 21.2% from the United States, 9.1% from France, 6.1% from India, 6.1% from England, and 3% from Turkey. NCS due to SCD was reported across a wide range of ages (from 11 to 60 years), including five patients under the age of 18. The mean age at presentation was 28.3 years (SD = 11.7) ( Patients presented with either unilateral or bilateral chin numbness (Supplementary Table S1). The symptom distribution was as follows: 11 patients reported left-sided numbness or pain of the jaw or lip, 7 patients reported numbness on the right side, 10 patients presented with bilateral NCS, and symptoms were unclear for 5 patients. One patient reported the evolution of symptoms from the right side to both sides of the lower face and was included in the bilateral count. The duration of the neurological symptoms of 21 patients was reported in the literature, with a slight majority reporting the complete resolution of symptoms (52%, n = 11). The duration of symptoms ranged from 1.5 h to more than 168 months.
A number of acute medical conditions were occurring concomitantly or in the days prior to the NCS. Indeed, all but one case of NCS was associated with a vaso-occlusive crisis (VOC) and/or acute chest syndrome (ACS) (97%). In the case of VOC, the pain was described in the mandibular region in most cases (83%, n = 24). However, in five reports (27%), the pain was described in other regions, such as the hips, lumbar region, or knee, without involving the mandibular region. One case occurred post-dental surgery, which led to a VOC. One case was associated with osteomyelitis complicated by bone necrosis, and four others were associated with systemic infections. One patient presented in the context of multi-organ failure. In addition, some patients presenting with NCS had significant comorbidities. These included rheumatoid arthritis treated with prednisone (n = 1), pregnancy (n = 1), type II diabetes (n = 1), metastatic breast cancer treated by chemotherapy (n = 1), membranoproliferative glomerulonephritis (n = 1), asthma (n = 1), retinal detachment (n = 1), and splenectomy (n = 1).
Patients with NCS were investigated using a number of different imaging modalities (Supplementary Table S1). A specific clinical workup was reported for 16 patients. A mandibular X-ray was used in 10 patients, making it the most used imaging modality. Five of these patients presented with specific findings, including (1) focal radiolucency, (2) diffuse lytic changes, and (3) stepladder trabeculations, consistent with (1) bone infarction, (2) osteomyelitis and (3) reactive bone changes typical of sickle cell disease. Images for the remaining five patients displayed no abnormalities. Seven patients were examined using a head CT scan. In cases where a patient was tested with both a mandibular X-ray and a head CT scan, the head CT scan did not reveal any notable difference or offer any alternative explanations for the mental nerve neuropathy. In addition, seven patients were examined with an MRI (one MRI exam was self-reported by a patient who claimed to have been previously assessed; the data was unavailable). Abnormalities were reported in 2 of these patients, including one incidental brain lesion unrelated to the NCS. Furthermore, a radionuclide bone scan was performed on 3 patients, which positively detected abnormalities in 2 patients. One patient had slightly increased tracer uptake in the right mandibular molar region, indicative of bone infarction. The other patient had diffusely increased tracer uptake in the skull and the periarticular regions of the long bones, consistent with reactive bone marrow hyperplasia. A lumbar puncture was performed on 1 patient showing no abnormalities suggestive of infection or malignancy. Finally, a dental pulp test was performed on 1 patient, which marked the non-vitality of several teeth.
The treatments were mainly symptomatic, with the management of the underlying cause of the neuropathy mainly being VOC/ACS. Treatments were reported for 18 patients. These included blood transfusions in two patients, tooth extractions in one patient, antibiotics in the case of infections in two patients, and standard treatment of VOC, including hydration, oxygen administration, and analgesics in 15 patients. Of note, one patient required no treatment.
Discussion
Our systematic review presents the largest collection of patients with NCS secondary to SCD. NCS seems to affect patients of all ages, sex, and SCD genotype. The majority of events occurred in the context of VOC/ACS, highlighting the important relationship between SCD and NCS pathophysiology.
Pathophysiology
Some pathophysiological mechanisms may be common to both VOC and NCS. Indeed, in an early report on the topic of this mental-nerve neuropathy, it was noted that the mandible was the site of merely 4% of the 100 consecutive adult admissions for a painful sickle-cell crisis [17]. However, in a more recent case series, 13 patients were presented with a number of concurrent diagnoses, such as active infection, rheumatoid arthritis, and diabetes, that could have either (1) contributed to the development of the VOC which in turn caused the NCS, or (2) these could be independent causes of NCS regardless of VOC [27]. This highlights the importance of careful clinical evaluation to identify both causes of VOC and NCS.
Vaso-occlusion and the resulting tissue ischemia are unique features of SCD resulting from the presence of the pathological hemoglobin S. A point mutation in the beta-globin gene on chromosome 11 leads to the replacement of a glutamic acid residue with a valine residue on the surface of the protein. The resulting hemoglobin is characterized by abnormal hydrophobic interactions with the adjacent chains, leading to polymer bundling and the distorted red blood cell shape distinctive of the sickle cells. The passage of these crescentshaped red blood cells through the narrow blood vessels is impaired, which may lead to ischemia and hemolysis [31]. Common triggers of vaso-occlusion are inflammatory or infectious conditions, such as infection, hypoxia, dehydration, and acidosis [1].
The anatomical location of the vaso-occlusion in the case of NCS could be the mandibular bone. The inferior alveolar nerve that branches into the mental nerve and the inferior alveolar artery, and the main blood supply to the mandible bone, are characterized by their passage into the mandibular foramen, through a narrow mandibular bony canal, before exiting at the mental foramen. The nature of this confined anatomical structure makes it susceptible to compression, ischemic infarction, and infection. Sickle red blood cells could block the small inferior alveolar artery and prevent perfusion into the inferior alveolar and mental nerves [14,23,24,27]. Additionally, edema due to bone infarction and osteomyelitis could compress these structures. Numbness would thus typically occur in the area of the chin and lower lip, where the mental nerve, the terminal branch of the alveolar nerve, provides sensory innervation [14]. Some studies described in this review have documented imaging consistent with bone infarction in this anatomical location [19,26].
The vasa nervorum is another potential locus of vascular injury in the case of NCS associated with SCD. SCD leads to a generalized inflammatory state and widespread vascular changes that can, in turn, result in vascular complications typical to both large (e.g., stroke) and small (e.g., retinopathy) blood vessels. In other medical conditions, such as diabetes or connective tissue disease, vasculitis in these small blood vessels compromise the vascular supply and can lead to neuropathy involving single nerves [32,33].
Theoretically, it is conceivable that hyperviscosity could also contribute to NCS pathophysiology [34]. In SC disease, a subtype of sickle cell disease was characterized by more elevated hemoglobin and a more silent and insidious disease course; it has been proposed that phlebotomies and the resulting anemia could prevent the occurrence or progression of neurosensory hearing loss and other complications attributed to hyperviscosity [35,36].
In addition, SCD patients are significantly more susceptible to infections, severe infectious complications, and osteomyelitis [37]. Serious bacterial infections remain a major cause of death in low-and medium-income countries [38]. Indeed, it is postulated that early and progressive injury to the spleen in SCD patients as a consequence of repeated vaso-occlusion followed by ischemia leads to progressive fibrosis and atrophy of the spleen [39]. For instance, one patient affected by local osteomyelitis led to NCS [21]. The resultant swelling and necrosis may have compressed and damaged the distribution of the left inferior alveolar and mental nerves resulting in mandibular neuropathy. Thus, a combination of VOC and a vulnerability to infection and osteomyelitis may predispose SCD patients to NCS.
In the general population, NCS has been associated with metastatic disorders. In our review, there was only one reported case of a patient developing NCS in the context of metastatic disease to the liver secondary to breast cancer. The associated pathophysiological mechanism is different than in the case of VOC, as neuropathy is caused by the compression or infiltration of the mental or inferior alveolar nerves by neoplastic cells [18]. In this reported case, NCS may be symptomatic of perineural invasion relating to metastatic liver cancer.
Differential Diagnosis
Before concluding that a case of NCS is caused by SCD, a number of alternative diagnoses that can cause NCS in the general population must be rigorously considered. These diagnoses have been comprehensively reviewed elsewhere [16]. In particular, dental procedures and local trauma are the most common cause of NCS and are usually obvious in medical history. Moreover, based on anatomy, the differential diagnosis must include odontogenic infection, osteomyelitis, osteosarcoma, primary lymphoma of the bone, and central mucoepidermoid carcinoma. In the absence of clear anatomic etiology, infectious, autoimmune, inflammatory, toxic, and malignant etiologies should be considered. Epidemiological risk factors and clinical suspicion will guide the relevant work-up. Of interest, sarcoidosis has been reported as one of the common causes of NCD. Patients of African descent are at increased risk of developing sarcoidosis [40].
In addition, systemic infections can trigger VOC/ACS and must be considered in patients presenting with NCS. Given the association of NCS with metastasis, systemic neoplasms should be considered in all patients, although a neoplasm was noted in only one case in this review. SCD patients are at risk of ischemic and hemorrhagic strokes, but vascular events would unlikely present with only sensory symptoms in the mental nerve distribution. Patients are also at risk of electrolyte disorders, namely hypocalcaemia, secondary to blood transfusions which can present with circumoral numbness or paresthesia. These are typically not lateralized or as delineated as NCS and resolve with the correction of the hypocalcaemia.
Imaging
The most frequently used imaging modalities in these studies were mandibular X-rays and head CT scans, followed by cerebral MRI and bone scans. One patient who was examined using a CT scan and whose results appeared normal was also assessed using a cerebral MRI that displayed subperiosteal fluid collection in the mandibular rami. The imaging by CT scan offers a higher level of detail than X-rays while still maintaining high imaging speeds and is generally accessible. However, in this review, data from patients tested using both imaging modalities showed no significant difference in the findings. CT scans would still be recommended over X-ray imaging, if available, for a more sensitive and accurate diagnosis of bone destruction and masses [16]. In turn, cerebral MRIs are tools that offer more detailed imaging of brain parenchyma and the pathway of the trigeminal nerve compared to CT scans [41,42]. Importantly, the mandibular region must be included in the scanned region, which is not always the case with a standard brain MRI; otherwise, important anatomical findings may be missed [43]. Although central neurological causes of the numb chin are probably extremely rare, a thalamic lacunar infarct can present such circumscribed symptoms [44]. Since patients with sickle cell disease have a heightened risk of stroke, this possibility should be considered. In our study, MRIs were performed only in the United States, France, England, and India. This suggests that the geographical location may have influenced access to investigations. MRIs may be less accessible in regions where SCD is most prevalent [45]. In these contexts, other imaging modalities should be used, including mandibular X-ray and/or CT scans.
Treatment
Currently, there is no specific treatment for NCS, and the literature is even scarcer with respect to cases with SCD. Indeed, the management of sensory neuropathy was directed toward resolving the underlying cause. Patients typically received the standard treatment for VOC to resolve their crisis, including hydration and analgesia. Some patients receive blood transfusions. However, this is not a standard treatment for uncomplicated VOC. Furthermore, the role of transfusions in the treatment of NCS is unknown. The patient who received a tooth extraction suffered from a VOC and infection in the area between the first and third molar, which were promptly treated. In this respect, comorbidities, such as infections or malignancies, must be treated parallel to the treatment of the VOC. Indeed, several series highlights the importance of assessing concurrent illnesses within SCD patients who present with NCS, as it can be the first manifestation of systemic disorders [27].
Finally, in the case of neuropathic pain, Pregabalin or Gabapentin could be tried in analogy to herpes zoster reactivation for pain relief [46]. Other therapeutic options include tricyclic antidepressants (ex: Amitriptyline, Nortriptyline), other antiepileptics (ex: Lamotrigine), and neural blockade. However, no data is available on the effectiveness of these medications in the case of NCS related to SCD.
Recommendations
Based on the limited available evidence and our collective clinical expertise, we recommend a thorough evaluation of patients presenting with NCS in SCD by a neurologist and a clinician with expertise in SCD (hematology, internal medicine, pediatrician) ( Table 2). We consider a detailed oral cavity examination by a dentist or ENT essential, especially in the case of a history of recent dental procedures, trauma, or pain. A detailed medical history should also be assessed for symptoms of systemic infection, potential malignancy, or auto-immune disease, along with common VOC/ACS precipitants. In addition to a general physical examination, a complete neurological exam is necessary to delineate the sensory deficit and rule-out other focal deficits. For all patients, laboratory studies should minimally include a complete blood count to assess the severity of the anemia, biochemistry to screen for kidney failure, and a pregnancy test since pregnancy can be a trigger of VOC and other acute complications of sickle cell disease. Furthermore, certain antibiotics that are commonly used to treat SCD complications are to be avoided in pregnancy (e.g., tetracyclines and fluoroquinolones). Therefore, according to the physicians' judgment, a pregnancy test may be ordered. In the case of fever, chest pain, or any respiratory sign/symptom, an urgent X-Ray should be performed to identify acute chest syndrome. This is a potentially deadly medical emergency mandating specific care not covered in this article but described extensively elsewhere [47]. Further testing that can be consider in cases of atypical presentations are highlighted in Table 2.
Conclusions
This study represents the largest review to date of the clinical manifestations, pathophysiology, investigations, and management of NCS in patients with SCD. NCS occurred most often in the context of VOC/ACS. The data extracted from this systematic review reinforces the hypothesis that a vaso-occlusion affecting the inferior alveolar or mental nerves in the microvasculature traversing through the mandibular bone could be a pathophysiological mechanism of NCS. In most cases, the management consisted of treating the acute trigger of the NCS and providing adequate analgesia. Careful evaluation is warranted to rule out alternative differential causes of NCS, including local infection, primary neoplasm, or metastatic disease. Imaging serves to narrow down the differential diagnoses and identify treatable causes. This study brings attention to a rare and underrecognized complication of SCD that deserves further investigation to optimize diagnosis and management.
|
v2
|
2022-11-26T17:01:06.635Z
|
2022-11-24T00:00:00.000Z
|
253913014
|
s2orc/train
|
DNA Damage Response in Cancer Therapy and Resistance: Challenges and Opportunities
Resistance to chemo- and radiotherapy is a common event among cancer patients and a reason why new cancer therapies and therapeutic strategies need to be in continuous investigation and development. DNA damage response (DDR) comprises several pathways that eliminate DNA damage to maintain genomic stability and integrity, but different types of cancers are associated with DDR machinery defects. Many improvements have been made in recent years, providing several drugs and therapeutic strategies for cancer patients, including those targeting the DDR pathways. Currently, poly (ADP-ribose) polymerase inhibitors (PARP inhibitors) are the DDR inhibitors (DDRi) approved for several cancers, including breast, ovarian, pancreatic, and prostate cancer. However, PARPi resistance is a growing issue in clinical settings that increases disease relapse and aggravate patients’ prognosis. Additionally, resistance to other DDRi is also being found and investigated. The resistance mechanisms to DDRi include reversion mutations, epigenetic modification, stabilization of the replication fork, and increased drug efflux. This review highlights the DDR pathways in cancer therapy, its role in the resistance to conventional treatments, and its exploitation for anticancer treatment. Biomarkers of treatment response, combination strategies with other anticancer agents, resistance mechanisms, and liabilities of treatment with DDR inhibitors are also discussed.
Introduction
DNA and other biological molecules are susceptible to damage, but DNA damage can have much more complex consequences due to its function. Unlike other molecules, which are synthesized and degraded depending on their necessity, DNA is constantly present and replicating when the cell is in division. Therefore, there is a great need for a reparation system to maintain DNA integrity. Estimations suggest that every day, about 10 5 lesions occur in the cell [1]. In addition to exogenous threats, such as irradiation, chemical pollutants, and chemical agents, endogenous processes increase reactive oxygen species (ROS), damaging DNA directly or indirectly. DNA damage includes single-and double-strand brakes, inter-and intra-strand links, abasic sites, bulky adducts, and base changes, such as 8-deoxyguanosine [1]. Due to the diversity of DNA damage, there are different repair mechanisms implicating many proteins. The activation of different repair mechanisms with the primary goal to restore DNA integrity is collectively known as the DNA damage response (DDR). These proteins/genes were historically identified in Fanconi's anemia (FA), a rare genetic disorder characterized by bone marrow failure, skeletal malformation, and increased cancer incidence. Mutations in this rare disease include genes in the FA pathway that are fundamental genes involved in DNA damage repair [2]. If DNA damage is not repaired or misrepaired, genomic instability and mutations will be established, which are among the hallmarks of cancer [3]. The DDR plays a relevant role not only in cancer development but also in cancer treatment. Defects in DDR genes are known as cancer drivers, and cells with deficient DDR show a higher sensitivity to DNA-damaging agents [4]. In this review, we summarize recent evidence of DDR pathways in cancer therapy, its role in the resistance to conventional treatments, and its exploitation for anticancer treatment. Since DDR is involved in cancer development and is a molecular target of cancer treatment, biomarkers of treatment response, combination strategies with other anticancer agents, resistance mechanisms, and liabilities of treatment with DDR inhibitors are also discussed.
The DNA Damage Response
The DNA damage response pathways are composed of an intricate system of sensors, transductors, and effectors involved in DNA repair and cell cycle checkpoint control that manage the execution of DNA replication and cell proliferation. The wide variety of DNA lesion types requires multiple and different DNA repair mechanisms. To repair single-strand breaks (SSBs), the mismatch repair (MMR), base excision repair (BER), and nucleotide excision repair (NER) are activated, while homologous recombination (HR) and non-homologous end joining (NHEJ) pathways repair the double-strand breaks (DSBs) [5]. DDR also includes damage tolerance processes and the consequent signaling control of cell decisions on senescence or death. In parallel, DDR affects epigenetics and gene expression regulation preferentially related to the induction of apoptosis [6,7].
Single-Strand Break Repair
The need for MMR occurs during DNA replication, where polymerases are prone to mistakes. Therefore, a mismatch repair system repairs wrongly matched bases to obtain replicative fidelity. The mismatch repair system in humans includes several proteins: DNA mismatch repair protein Mlh1 (MLH1), DNA mismatch repair protein Msh2 (MSH2), DNA mismatch repair protein Mlh3 (MLH3), DNA mismatch repair protein Msh6 (MSH6), DNA mismatch repair endonuclease PMS2 (PMS2), and DNA mismatch repair protein Msh3 (MSH3) [8,9]. These proteins act as heterodimers MSH2 with MSH6 or MSH3 (MutSα or MutSβ complexes, respectively) and MLH1 with PMS2 or MLH3 (MutLα, MutLβ, or MutLγ complexes, respectively) [9]. In addition to these protein complexes, proliferating cell nuclear antigen (PCNA) with the help of MSH2/MSH3 and MSH2/MSH6 complexes recognizes and binds to the mispaired region [10]. Furthermore, single-strand DNA-binding protein RPA (replication protein A) and EXOI (exonuclease I) both contribute to MMR by protecting the gap from excision, while EXOI is needed for the repair of the break located either 5 or 3 to the mispair [11]. The evidence strongly suggests that polymerase δ (pol δ) is required for MMR [12], and DNA ligase I is needed for the final step in MMR [13].
NER is the mechanism of bulky adduct reparation. The offset of this mechanism can be initiated by global genome NER or transcription-coupled NER [14]. Like other repair systems, there are two steps involved: recognition of the damage and the reparation step. Bulky adducts cause DNA distortions recognized by XPC-RAD23B, but only if the nucleotide opposing the lesion is not missing [15]. Once XPC-RAD23B recognizes the bulky adduct, a small bubble is formed, and TFIIH, a complex of 10 proteins, is recruited [16]. Then, XPB translocase and XPD helicase open the bubble even more (22 to 25 base pairs) [17], which allows binding of XPA and RPA. RPA protects the undamaged strand while XPA verifies the damage and further recruits XPF-ERCC1 endonuclease that cleaves the 5 end, while XPF cleaves the 3 end [18]. The resulting gap is filled by DNA polymerase δ and ε, replication factor C (RFC), PCNA, and RPA [19].
Both c-NHEJ and HR repair start with binding Ku70-Ku80 heterodimer to a doublestrand break [40]. The binding of Ku70-Ku80 recruits other factors: DNA-dependent protein kinase catalytic subunit (DNA-PKcs), DNA ligase IV (LIG4), and the associated scaffolding factors of DNA repair protein XRCC4 (XRCC4), XRCC4-like factor (XLF), and paralogue of XRCC4 and XLF (PAXX), which bring the two ends closer, enabling end processing by Artemis and DNA polymerases λ and µ [37]. The HR mechanism is usually associated with cancer and BRCA1 and BRCA2 mutations, which are linked to hereditary breast and ovarian cancer [41]. However, new evidence suggests that the basis of hypersensitivity of BRCA-deficient tumors is not double-strand breaks induced by chemotherapy but rather single-strand breaks [42]. The HR process includes numerous steps, the first one being recognition by two kinases, ATM and ATR, which phosphorylate targets: CHEK2, P53, BRCA1, and H2AX. BRCA1 serves as a scaffold that recruits other proteins [
Epigenetic Control in DNA Damage Response
To properly understand how cells control complex DDR, epigenetics and miRNA regulations cannot be omitted. Epigenomic alterations are known to significantly affect gene expression and overall tumor heterogeneity. Therefore, it is not surprising that DNA repair processes are also affected by epigenetic chromatin regulation. Histone deacetylases (HDACs) are important players in chromatin preparation to promote DSBs repair through HR and NHEJ. For example, PARP1 recruits the nucleosome remodeling deacetylation (NuRD) complex by attaching a PAR chain signal essential for DSB repair [ [53], gastric cancer [54], ovarian cancer [55], and BRCA1 gene in breast cancer [56], bladder cancer [57], NSCLC [58], and gastric cancer [59]. Additionally, the methylation status of some DDR genes has been used as diagnostic, prognostic, and therapy response biomarkers in various cancer types. MLH1 methylation has been indicated not only as a diagnostic biomarker and an indicator of good prognosis in several cancers, including colorectal, ovarian, and breast cancers, but also as a therapy response biomarker associated with platinum compounds, temozolomide, and epirubicin resistance and with methotrexate sensitivity [60].
DNA Damage Response Inhibitors in Cancer Therapy
The DDR mechanisms are involved in the control of core processes of cell fate, survival, and genome maintenance. In cancer, accumulated genetic defects compromise the cell response to physiological growth control and promote uncontrolled division and evasion of apoptosis. Vogelstein and collaborators identified approximately 130,000 different mutations in more than 3000 individual drivers of tumorigenesis, including both oncogenes and tumor suppressors. From these, about 330 genes were identified as drivers involved in the regulation of cell survival, genome maintenance, and overall DDR. These genes are valuable targets for new approaches to cancer treatment [87].
The effect of primary anticancer therapies, including ionizing radiation and different chemotherapeutic agents that damage both nuclear (nDNA) and mitochondrial (mtDNA) DNA, are expected to drive the cancer cell, directly or indirectly, towards death. Cisplatin represents such a highly efficient DNA-damaging agent with substantial anticancer effects. Despite the discovery of its cytotoxic effects and its first Food and Drug Administration (FDA) approval for treating testicular cancer in 1978, cisplatin is still used as first-line chemotherapy in numerous solid tumors. The cell response to cisplatin is complex and includes mechanisms regulating its entry, exit, accumulation, and detoxification and mechanisms modulating DNA repair, cell survival, and the tumor microenvironment [88].
DNA Damage Response Inhibitors in Single-Agent Approaches
Cancer cells show a higher level of endogenous DNA damage and increased replication stress than normal cells and usually have one or more DDR pathways disabled. Such deficiencies are attractive points for novel cancer treatments development, mostly those exploiting synthetic lethality concepts [89,90]. In principle, synthetic lethality in cancer treatment is based on targeting and inhibiting the DDR pathway that is left functional. If one DDR pathway is compromised and not functional, e.g., due to mutations in one or more DNA repair genes, the cancer cell attempts to restore the DNA damage utilizing the backup repair mechanism. However, if this backup mechanism is pharmacologically targeted, the cancer cell has no functional DNA repair pathway and is doomed. In 2014, in both Europe and the USA, Olaparib-a PARP inhibitor (PARPi)-was the first DDR inhibitor (DDRi) approved for cancer treatment [89]. Shortly after, in 2017, two other PARP inhibitors, Rucaparib and Niraparib, were FDA-approved for use in cancer patients with BRCA mutation as well as for non-carriers of BRCA mutations to treat primary peritoneal cancer, fallopian tube, or recurrent epithelial ovarian cancer resistant to cisplatin chemother-apy [91,92]. Later, Talazoparib was approved by the FDA to treat patients with germline BRCA (gBRCA) mutations or metastatic breast cancer [93,94]. In 2020, the FDA approved Veliparib for use in combination with gamma-ray radiotherapy or chemotherapy for advanced lung squamous cell carcinoma [95] and in combination with paclitaxel/carboplatin for the treatment of recurrent ovarian, breast, and lung cancer patients [96][97][98]. Veliparib and another PARPi Iniparib in combination with gemcitabine/carboplatin for breast and lung cancer treatment reached phase III trials [99].
In clinical practice, the conventional treatment of women with OC is based on debulking surgery and selecting first-line platinum-based chemotherapy, followed by second-line platinum chemotherapy in case of relapse. Further management of OC patients diverges with the decision for maintenance treatment or active surveillance watching until the third relapse [100]. If OC patients harbor BRCA mutation, targeted therapy with PARPi may replace chemotherapy to maintain the response, delay disease progression, and prolong the period between treatment cycles [91,101]. Recently, Olaparib has been approved for maintenance treatment in the first-line setting for women with a BRCA mutation [102]. Similarly, rucaparib was approved in the treatment setting for patients with relapsed BRCA-mutated platinum-sensitive OC [103] and has been shown to be beneficial in both the maintenance as well as treatment settings. The substantial benefit of PARPis in the first-line setting has been demonstrated in randomized phase III trials (SOLO-1, PAOLA-1, PRIMA, VELIA) [104][105][106][107]. However, besides the significant contribution of PARPis to better treatment management of OC patients, the side effects of PARPis on patients' quality of life must be thoroughly monitored.
PARP is the best-known element of the DDR. Functional PARP identifies single-strand breaks and utilizes nicotinamide adenine dinucleotide (NAD + ) to form poly ADP-ribose chains that open chromatin to allow DNA repair proteins to access the DNA [108]. PARPi prevents the formation of PAR chains and keeps PARP on the DNA at SSBs. Consequently, formed PARP-DNA complexes stall or collapse the replication fork and generate DSBs. DSBs can be repaired by HR, but if HR is missing or defective in cancer cells, as in BRCA1 mutation-carrying cancers, the cell must use error-prone NHEJ, leading to genomic instability and cancer cell death [108][109][110]. PARPs are involved in the repair of SSBs through BER and DSBs, through HR, NHEJ, and alt-NHEJ (or microhomology-mediated end joining, MMEJ). Along with successful validation on patients carrying BRCA1 and BRCA2 mutations [111], positive effects of PARPi were also observed in patients without BRCA mutations with high-grade serous or poorly differentiated ovarian carcinoma or TNBC [112]. Accordingly, PARP inhibition as a therapeutic approach successfully expanded to other cancers, including pancreatic, endometrial, prostate, urothelial, colorectal, lung, and glioblastoma [113]. Overall, DDR involves more than 450 proteins [89,114,115], and several are being investigated as potential novel therapeutic targets. The most promising include DNA damage sensors (MLH1), damage signaling molecules (ataxia telangiectasia mutated (ATM), ATM-and RAD3-related (ATR), CHK1, CHK2, DNA-dependent protein kinase, catalytic subunit (DNA-PKcs), and WEE1), or effector proteins for DNA repair (POLQ, RAD51, or PARG) [113]. Several DDRi are currently in preclinical and clinical trials ( Table 1). Clinical trials have been initiated to test their targeting with single agents or in combination therapy. In parallel, different technologies are being explored to screen for synthetic lethal combinations, including small interfering RNA (siRNA) or exploiting CRISPR-Cas9-based strategies to target them for anticancer treatment purposes.
DNA Damage Response Inhibitors in Combinatory Therapies
To gain maximum efficiency of anticancer treatments, combinations of individual agents and strategies are tested and validated. Concerning the interplay between epigenetics and DNA repair, the most explored therapeutic approach is the combination of epigenetic inhibitors with chemotherapeutic agents. Additionally, several clinical trials have been initiated to evaluate the efficacy of combined administration of HDAC and PARP inhibitors. Such combinations have been examined in pre-clinical models and effectively kill prostate, ovarian, and breast cancer cells [116][117][118][119]. Similarly, combined administration of DNMT1 and PARP in AML and breast cancer showed a synergistic activity [120]. Chromatin remodeling inhibitors also prevent HR repair and sensitize different cancers cells towards DNA-damaging agents [121], while HDAC, DNMT, and LSD1 inhibitors restore chemosensitivity in different solid tumors [122]. A synergism of combinations has been identified between radiotherapy and epigenetic inhibitors [123][124][125][126][127]. However, due to high toxicity and limited patient benefits, these approaches still require more investigation and clinical validation. In recent years, much attention has been paid to investigating the therapeutical effect of combinations of PARP inhibitors with antiangiogenic therapy. Multi-kinase inhibitors targeting VEGFR, PDGFR, and FGFR were shown to sensitize tumor cells to PARP inhibitors via induction of hypoxia and triggering HR defects [128].
Combination regimens between two DDRi have also been investigated. The synergic effect of PARPi in combination with ATR inhibitor (ATRi) was reported in HR-deficient HGSOC in vivo [129]. The therapeutic combinatory approach of non-toxic concentrations of a CHK1 inhibitor (CHK1i; PF-00477736) with a WEE1 inhibitor (WEE1i; MK-1775) showed a synergistic effect in breast, ovarian, colon, and prostate cancer cell lines in a P53 status-independent manner [130]. The combinations of the CHK1is (PF-00477736 or AZD-7762) with the WEE1i (AZD-1775) showed a synergistic effect in all the diffuse large B-cell lymphoma cell lines independently of the molecular subtype and MYC status [131]. The combination of CHK1i with WEE1i also showed a strong synergy in mantle cell lymphoma [132], lung, prostate, and erythroleukemia [133]. The synergistic effect of DDRi in B-cell lymphomas was also observed in combinations of ATRi with CHK1i and ATRi with WEE1i [134]. Moreover, the combination of ATR (VE-821) and CHK1 inhibitors (AZD7762) induced replication fork arrest, ssDNA accumulation, replication collapse, and synergistic cell death in osteosarcoma, breast, and lung cancer cells in vitro and in vivo [135]. The co-administration of Olaparib and AZD1775 (WEE1 inhibitor) demonstrated a synergistic antiproliferative effect in TNBC cell lines and significantly inhibited tumor growth in a xenograft model of BC [136]. Preclinical studies showed that ATR inhibitor synergizes with WEE1i in TNBC [18,137]. This therapeutic association reduced cell proliferation and induced cell death in several BC cell lines [137,138] and tumor remission, increased survival, and inhibited metastasis in orthotopic BC xenografts mouse models [138]. Bukhari et al. also demonstrated the therapeutic potential of this association in mammospheres, reporting similar sensitivities to the combined treatment in cancer stem cells [138]. Kim and colleagues demonstrated, using acquired and de novo PARPi and platinum-resistant models, that PARPi (AZD2281) in combination with ATRi (AZD6738) synergistically decreased cell viability and colony formation using doses with minimal off-target effects [139]. This combination also induced tumor regression and a significant increase in overall survival in HGSOCs patient-derived xenograft (PDX) models.
The metabolic vulnerability of cancer cells is a highly relevant dimension that can be exploited for therapeutical targeting and potential overcoming therapy resistance. Mammalian target of rapamycin (mTOR) kinase, mTORC1, and mTORC2 complexes are considered critical drivers of cancer drug resistance that integrate signaling pathways driving cell metabolism and growth [140,141]. Both mTOR complexes belong to effectors of the most oncogenic drivers, including RAS-driven MAPK and PI3K-AKT pathways. Sustained mTOR signaling contributes to resistance to therapeutics targeted against the driving oncogenes [142] or chemotherapy resistance, for example, by inducing the FA DNA repair pathway [143] and modulating other proteins that are essential in chromosomal integrity and DNA damage response [144,145]. Deregulation of mTOR has been found in various human cancers [146], including resistant ones such as TNBCs [147,148]. Inhibitors of mTOR are therefore considered a valuable addition to chemotherapy or targeted cancer therapy, either as an option for relapsed patients or as a frontline combination therapy to prevent or delay the development of resistance due to sustained mTOR signaling [142,149].
Similarly, using DDR inhibitors and/or radiation as sensitizers provide new potential to increase immunotherapy efficacy. Immunotherapy attracts much attention and is considered a breakthrough in the field of cancer treatment. Individual DNA repair pathways' defects were associated with immune checkpoint blockade response. DNA damage induced in cancer cells upon radiation or chemotherapy leads to the release of chromosome fragments or small pieces of DNA that activate an immune response. When DDRi (e.g., PAPRi) are used, more DNA fragments are released, making tumor cells more immunogenic and more sensitive to immunotherapy. For example, defects in MMR result in neoantigen generation [150] that is associated with better anti-PD-1/PD-L1 immunotherapy outcomes [151]. The benefits of multiple combination therapies involving immune checkpoint inhibitors with DDRi are undergoing clinical trials.
A novel, although limited, field of anticancer approaches is opened via targeting mtDNA repair pathways. The capacity of mtDNA repair significantly contributes to therapeutical cancer cell response. BER is the main repair pathway used by mitochondria to repair mainly ROS-induced lesions. mtDNA carries many mutations that usually correlate with cancer progression [152]. Defective mtDNA repair pathways or downregulated mtDNA repair-associated proteins, such as mitochondrial transcription factor A (mtTFA) and POLγ, together with administered DDR inhibitors, can result in higher sensitivity of cancer cells to radio-or chemotherapy [153,154].
Biomarkers of DNA Damage Response Inhibitors
As mentioned previously, a significant number of new DDR inhibitors and DDRbased therapeutic strategies have arisen recently. The remarkable clinical success of PARP inhibitors in patients with BRCA1 and BRCA2 gene mutations showed that the clinical utility of DDRi relies on establishing response biomarkers that select patients who will benefit from these therapies. PARPi has shown efficacy in HR-deficient cancers, including those with RAD51C, RAD51D, and PALB2 mutations [155] and with a "BRCAness" phenotype [156], but not all HR alterations have the same impact on the efficacy of these inhibitors [157,158]. The "BRCAness" phenotype is defined by the lack of BRCA1/2 mutations in tumors with similar molecular phenotypes. This phenotype can result from mutations and epigenetic modifications of HR-related genes that cause homologous recombination deficiency (HRD), such as RAD51C, RAD51D, ATM, BARD1, PALB2, BRIP1, and MRE11 mutations and BRCA1 hypermethylation [155,156,[159][160][161][162]. The sensitivity to platinum-based chemotherapy is considered a surrogate biomarker of "BRCAness" phenotype to PARPi, and FDA approved this biomarker as a response biomarker for Olaparib therapy in maintenance settings [163]. However, not all patients who respond to platinum-based therapy will respond to PARPi, and some patients resistant to these conventional therapies will respond to PARPi [164]. Moreover, gene alterations, mutations, or functional loss of proteins involved in DDR mechanisms result in defective ATR, CHEK1, CHEK2, DSS1, MRE11A/NBS1, Fanconi anemia complementation group (FANC family of genes), EMSY, XRCC2, XRCC3, or PTEN, predispose patients to the success of PARP inhibitors for cancer treatment [165][166][167][168].
Throughout the years, several HRD assays have been developed to try to identify patients who will benefit from DDRi. These tests include the mutational status of DDR genes that identify specific causes of HRD, "genomic scars" or mutational signatures that identify HRD cancers, and functional assays that provide a readout of HRD or homologous recombination proficiency [169]. HRD cancers are expected to have genomic instability, and patients with these features are identified as "BRCAness". For example, tumors with BRCA1/2 mutations were associated with loss of heterozygosity (LOH), large genomic deletions, large-scale transitions (LST), and telomeric allelic imbalance (TAI) [170][171][172][173][174], while microsatellite instability (MSI) is characteristic of MMR deficiency [175]. A combination of these genomic scars, LOH, LST, and TAI, robustly predicted the "BRCAness" phenotype and the sensibility of PARPi and was the basis of myChoise HDR (Myriad Genetics) and FoundationOne CDx (Foundation Medicine) commercial assays [158,176]. Additionally, a genomic mutational signature, "signature 3", was significantly associated with BRCA1/2 mutations [177]. The reduced nuclear RAD51 foci have been associated with BRCA1/2 mutations and with PARPi responses. However, currently, the functional assays of HRD, as an estimation of nuclear RAD51 amount being the most used system, have insufficient evidence to establish their clinical value to predict PARPi response [169]. Several other genomic alterations have been proposed as potential biomarkers of DDRi response. For example, decreased CHK1 phosphorylation, increased expression of γH2AX, and increased replication fork instability are associated with ATR inhibitors response, while P53 deficiency and replication stress promoting genomic charges, including CCNE1 and MYC amplification, are associated with WEE1 inhibitors response [158]. KRAS mutations and the overexpression of CCNE1, CCND2, and MYC genes induce hypersensitivity to ATR inhibitors in cancer cell lines [178][179][180]. Moreover, tumor cells with loss of H3K36me3 due to mutation in SETD2, the gene that encodes a histone lysine methyltransferase, or mutation in histone H3 itself showed HR, NHEJ, and MMR impairment and sensitivity to WEE1, CHK, or ATR inhibitors [181]. These tumors were also sensitive to PARP and ATR inhibitors [158].
According to the European Society for Medical Oncology (ESMO) Translational Research and Precision Medicine Working Group, the most useful predictive biomarkers for HRD and indicate the PARPi benefit in the clinic are single-gene aberrations and/or genomic scars [169,182]. These tests reflect HRD phenotype and aim to identify patients who may benefit from PARPi. The germline and tumor (incorporating germline and somatic) BRCA mutation testing exhibit adequate clinical validity by consistently identifying the subgroup of OC patients who benefit more from PARPi therapy and remain the goldstandard predictive biomarker for PARPi. Additionally, HRD tests using genomic scars incorporating scores of allelic imbalances (GIS or LOH) are also reasonable since this test identifies a subgroup of BRCA wild-type, platinum-sensitive cancers that will benefit from PARPi therapy in some settings [169]. However, HRD biomarkers able to evaluate cancer evolution and provide a real-time read-out of homologous recombination proficiency still need to be developed and optimized [169,183], and further studies are required to clinically validate the existing ones.
Liquid biopsy approaches may also facilitate the selection of therapy and predict chemoresistance in cancer patients by identifying mutations in genes implicated in DNA repair mechanisms. In HER+ breast cancer patients, circulating tumor DNA (ctDNA) profiling identified ERBB2, TP53, EGFR, NF1, and SETD2 mutations contributing to trastuzumab resistance; in the same retrospective study, genetic aberrations in TP53, PIK3CA, and DNA damage repair genes were found in HER2-negative BC patients resistant to chemotherapy [184]. The components of the DNA repair machinery may also be utilized as biomarkers for evaluating tumor mutational burden (TMB). In a retrospective study of NSCLC, nextgeneration sequencing (NGS) was applied to different specimens, including small biopsy and cytology specimens; genomic alterations were found in genes implicated in DNA repair, including TP53 and BRCA2 [185]. Liquid biopsies have also been used to identify BRCA1/2 reversions in several pathologies [186]. One example is the identification of BRCA1/2 reversions in circulating free DNA (cfDNA) in HGOC patients treated with rucaparib [186]. In this study, 18% of platinum-resistant and 13% of platinum-refractory patients had BRCA1/2 reversions pretreatment in comparison with 2% of platinum-sensitive patients. This study supports the potential clinical use of liquid biopsies prior to initiating PARPi therapy since this test allowed the identification of patients who benefit from rucaparib therapy (patients without pretreatment cfDNA BRCA1/2 reversions had 2-fold higher PFS on rucaparib; 9.0 versus 1.8 months; p < 0.001) [187]. Another advantage of liquid biopsies is the detection of clonal heterogeneity of reversion events [186]. Lin et al. described the detection of eight BRCA1 mutation reversions in cfDNA, but only one of them was detected in the tumor biopsy [187]. The development of liquid biopsy approaches to detect specific aberrations in genes involved in DDR would provide a non-invasive and efficient means to improve treatment selection and disease outcome.
DNA Damage Response as a Mechanism of Cancer Therapy Resistance
Chemotherapy and radiotherapy rely on the cytotoxic DNA-damaging effects that for proliferating cancer cells, already burdened by genomic instability and defective DNA repair pathways, represent an induction of unrepairable genome-wide DNA damage leading to apoptosis. Therefore, chemotherapy and radiotherapy are still used as a first-line approach for many unresectable or metastatic malignancies. However, due to the large capacity of cancer cells to resist anticancer agents and adapt, the DDR is dysregulated and can lead cancer cells to genotoxic hypersensitivity or resistance development. Defective DDR allows for tumor heterogeneity development by preselecting subclones with intrinsic or acquired resistance, driving cancer progression and tumor relapse [188][189][190]. For example, in the case of cisplatin, despite its consistent rate of initial responses in multiple solid tumors [191], the treatment often results in the development of chemoresistance and therapeutic failure. Platinum salts such as cisplatin cause DNA inter-and intrastrand crosslinks, with DNA lesions repaired by a combination of NER and HR pathways. Higher expression and activity of DNA damage repair enzymes are observed in cisplatin-resistant tumor cells, and NER inhibition enhanced their sensitivity to cisplatin [192,193]. Highgrade serous ovarian cancers with germline or somatic mutations in BRCA1 or BRCA2 genes and hypermethylation of the BRCA1 gene and NSCLC without ERCC1 are sensitive to platinum compounds [194,195].
Temozolomide (TMZ) is the standard treatment in glioblastoma that acts mainly through O 6 -methylguanine (O 6 -meG) lesions. Other lesions caused by TMZ, such as N 3 -meA and N 7 -meG DNA adducts, are easily repaired by BER enzymes such as O-6methylguanine-DNA methyltransferase (MGMT) [196]. The MGMT gene promoter hypermethylation has been associated with longer survival in glioblastoma patients treated with TMZ [197,198]. The MGMT enzyme removes alkyl groups from guanine at the O 6 position, reducing the effect of TMZ and suggesting that MGMT activity is likely a biomarker of alkylating agents' sensitivity [199]. On the other hand, Oldrini and colleagues (2021) reported that MGMT genomic rearrangements carried by a subset of recurrent gliomas led to MGMT overexpression and TMZ resistance in vitro and in vivo, independently from changes in its promoter methylation [200].
Radiotherapy response is also modulated by DDR machinery. In TNBC patients, low expression of 53BP1, an NHEJ pathway protein, is associated with radioresistance [201]. Glioblastoma patients with nuclear PTEN phosphorylation show reduced sensitivity to radiation by enhancing DNA repair [202]. Overexpression or activation of the BER pathway is observed in radioresistant cells. Glioma cell lines with higher endogenous APE1 endonuclease are more radioresistant, and the APE1 ectopic expression increases radioresistance [203]. Moreover, radioresistant cancer cells and biopsies from radioresistant cancer patients show low expression of GADD45α in cervical cancer [204]. Cervical cancer patients with low expression of Ku80 respond better to radiotherapy, and hypopharyngeal squamous cell carcinoma patients with low Ku70 or XRCC4 proteins have better sensitivity to chemoradiotherapy [205,206]. The residual carcinoma from patients with cervical cancer after radiotherapy showed increased expression of DNA-PKcs, Ku70, and Ku86 genes, components of NHEJ pathways, compared to the counterpart primary tumors [207]. HR is also involved in radioresistance in cancer cells. For example, overexpression of BRCA1, BRCA2, RAD51, and RPA1 was observed in hypopharyngeal and nasopharyngeal carcinoma cells resistant to radiotherapy [208,209].
DNA Damage Response Inhibitors to Overcome Therapy Resistance
Tumor cells exhibit several therapy resistance mechanisms, but probably the most relevant ones are the inter-and intrapatient heterogeneity and intratumor heterogeneity [184]. Several strategies could be implemented to circumvent drug resistance, including adjustment of drug dose, optimization of therapies sequence, and targeting bypass mechanisms or alternative molecular targets by combinatorial approaches. The DDRi are potential alternative strategies to overcome cancer resistance (Table 2), and their combination with radiation, cytotoxic, or targeted agents can maximize the benefits of DDR targeted therapies.
Inhibition of PARP
In the absence of functional BRCA, targeting PARP is effective as monotherapy [210] and also sensitizes cancer cells to other drugs. In Palbociclib-resistant breast cancer cells, PARP inhibition combined with a STAT3 specific inhibitor re-sensitized cells to the agent, suggesting that concurrent targeting of DDR mechanisms and the IL6/STAT3 pathway could effectively treat acquired resistance to Palbociclib [211]. Epigenetic modifying agents have also been combined with PARP to improve therapeutic effectiveness. Specifically, the combination of DNA methyltransferase Gadecitabine and PARPi Tlazoparib were found to synergize in PARPi-resistant breast and ovarian cancer cells irrespective of BRCA status [212]. PARP inhibitors have been approved to treat high-grade serous ovarian cancer (HGSOC); however, women often develop resistance to treatment. The combination of antiangiogenic agent Cediranib with PARP inhibitor Olaparib was evaluated with varied results in a clinical trial of women with HGSOC who had developed resistance to therapy [213]. In HGSOC cell lines and PDX animal models, the checkpoint kinase 1 (CHK1) inhibitor Prexasertib showed efficacy as monotherapy but also sensitized cells to PARP inhibition [214]. PARP inhibition via Olaparib was able to reverse Sorafenib resistance in hepatocellular carcinoma by suppressing DDR mechanisms. In addition, Olaparib caused CHD1L-mediated chromatin condensation in the promoter region of transcription factors that promote cancer pluripotency [215]. Increased DNA repair in multiple myeloma is one of the main reasons for treatment resistance. Although not the standard of care for multiple myeloma, melphalan (MEL) is still used in combinatory strategies to treat this disease. To study whether PARP inhibition would reverse resistance to MEL, the agent was combined with several PARP inhibitors (Veliparib, Olaparib, and Iraparib) in multiple myeloma cell lines. The combination of MEL and PARPi in MEL-resistant cells showed an enhanced effect. However, MEL resistance, possibly caused by HR and NHEJ pathways, was not completely reversed by PARPi [216]. To discover the therapeutic potential of PARPi in combination treatment with other agents, a systems approach was developed by performing reverse-phase protein arrays to characterize adaptive responses to different therapies. The results indicate that the combination of PARPi with MEK/ERK, WEE1/ATR, and PI3K/AKT/mTOR inhibitors would show efficacy; this was also evident in various preclinical models. Based on this approach, several combinational therapies using PARPi are being assessed in clinical trials [217]. As mentioned before, TMZ, such as N 3 -meA and N 7 -meG DNA adducts, are easily repaired by BER [196]. To block BER and sensitize cells to TMZ, the agent may potentially be combined with PARP inhibitors that sustain N 3 -meA and N 7 -meG TMZ-induced lesions and improve the drug's efficacy. A BER-independent function of PARP inhibitor Veliparib has also been shown to re-sensitize MMR deficient cells to TMZ [218]. In addition, knock-down of HR-involved proteins such as BRCA1 or RAD51 has been shown to improve the efficacy of TMZ [219].
Inhibition of ATM-ATR Complexes and Downstream Effectors
Resistant cancer cells may also be sensitized to treatment following exposure to ATM inhibitors. Kinases ATM, ATR, and their downstream effector kinases CHK1 and CHK2 are activated in response to DNA damage, leading to cell cycle arrest. The ATM-CHK2 axis is involved in G1 checkpoint control, whereas ATR-CHK1 controls the S and G2 checkpoints. Both ATM and ATR can convey their effects through P53, either directly or via activation of checkpoint kinase 2 (CHK2). P53 induces the CDK2 inhibitor P21 preventing damaged cells from entering the S phase [220].
Agent 2-morpholin-4-yl-6-thianthren-1-yl-pyran-4-one (KU-55933) acts as an inhibitor of ATM; specifically, it blocks phosphorylation of ATM and inhibits its downstream targets. KU55933 sensitized radioresistant breast cancer cells to ionizing radiation (IR). Specifically, breast cancer cells with a defective disabled homolog 2-interacting protein (DAB2IP) are often aggressive and resistant to radiation. KU-55933 improved the efficacy of IR against siDAB2IP breast cancer cells by targeting ATM and impairing DNA repair mechanisms [221]. Improved ATMi analogs with enhanced bioavailability, such as KU-60019 and AZ32, were also found to radio-sensitize human cancer glioma cells [222,223]. Similarly, VE-822, an ATR inhibitor, decreased the viability of pancreatic cancer cells following exposure to irradiation or to gemcitabine both in vitro and in vivo. The effect of VE-822 was achieved through dysregulation of cell cycle checkpoints and maintenance of DNA damage. Notably, the agent did not display cytotoxicity against normal cells. VE-821, another ATR inhibitor, successfully sensitized bone and ovarian cancer cells to radiation in vitro, forcing irradiated cells to divide into daughter cells and decreased survival selectively in cancer cells [224]. A detailed review describing the different approaches to sensitize cancer cells to radiation therapy by targeting DNA damage response components was recently published [225].
VX-970, a small-molecule ATR inhibitor, is currently being tested with promising results in many clinical trials in combination with chemotherapeutic drugs against resistant and aggressive cancers [226]. VX-970 also displayed radio-sensitizing effects in TNBC cells and PDX models. Specifically, the agent inhibited the ATR-CHK1-CDC25a axis signaling, sustained DNA double-strand breaks, and reduced colony formation following radiotherapy in TNBC cells. These effects were selective to cancer cells compared to normal epithelial breast cells [227]. A combination of CHK1 inhibitor PF-00477736 with Ibrutinib showed synergistic effects in vitro in several mantle cell lymphoma (MCL) cell lines. Ibrutinib is a Bruton's tyrosine kinase (BTK) inhibitor that has been approved for refractory MCL. The study showed that in MCL cells resistant to Ibrutinib, the combination with CHK1 inhibitor led to enhanced effects [228]. The ATR inhibitor NU60 induces G2/M arrest and impairs homologous recombination, leading to increased sensitivity of breast cancer cells to DNA-damaging agents, such as cisplatin, and PARP inhibitors [229].
The APC tumor suppressor gene is inactive in 70% of sporadic breast cancers; APCdeficient tumors resemble the aggressive TNBC subtype. APC deficiency decreases sensitivity to doxorubicin (DOX), which is attributed to the inactivation of ATM, CHK1, and CHK2 and increased DNA repair in the presence of DOX. Concurrent inhibition of ATM and DNA-PK enhanced DOX-induced apoptosis in resistant cells [230]. These findings support that inhibition of the ATM-ATR-CHK axis is a promising approach to enhance radiation or chemotherapy therapeutic efficacy [231]. Importantly, synthetic lethality with ATM, ATR, and DNA-PK inhibitors is being evaluated to target HR-proficient cells [232,233].
Inhibition of WEE1
A recent review highlights the potential of WEE1 inhibition in radio-and chemosensitization [234]. WEE1 is a protein kinase mainly localized in the nucleus. It negatively regulates the G2/M transition following the detection of DSB [235,236]. It affects the CDK1cyclin B complex by phosphorylating and inactivating Cyclin B on Tyr15, causing cell cycle arrest at G2. When errors happen during replication, this mechanism blocks the cell cycle to allow for repair; downregulation of WEE1, either by decreased synthesis or through proteolytic degradation, promotes entry into mitosis [237,238]. The role of WEE1 as a gate-keeper for the G2/M transition suggests that it acts as a tumor suppressor gene; however, WEE1 was overexpressed in patients with hepatocellular carcinoma, medulloblastoma, and glioma, and its levels were further elevated after exposure to chemotherapy in patients with ovarian cancer [239][240][241][242]. Overexpression of WEE1 in melanoma cells has been correlated with proliferation markers, including Ki-67 [243]. Therefore, it is postulated that the expression of WEE1 allows cancer cells to repair DNA damage following chemo-or radiotherapy, develop resistance, and continue to proliferate. In addition, cancer stem cells adopt high WEE1 expression as a protection mechanism against therapeutic agents [244]. It is well established that cancer stem cells convey resistance to DNA-damaging treatments; their percentage increases in the tumor cell population following the progressive deterioration of non-stem cells.
WEE1 inhibition represents an attractive approach for radio-and chemotherapy potentiation. Several pharmacological inhibitors belonging to different chemical classes have been developed against WEE1 and are described in a recent review [245]. AZD1775 is a WEE1 inhibitor currently in clinical trials, combined with DNA damage agents or radiotherapy. AZD1775 has been found to have a radio-sensitizing effect in pancreatic cancer, pontine gliomas, and glioblastoma [246][247][248]. Some studies suggest that AZD1775 s ability to sensitize cells to therapy is effective only in TP53-deficient tumors [249][250][251]. The combination of AZD1775 with cisplatin sensitized squamous cell carcinoma of the head and neck (HNSCC) cells to the latter in in vitro and in vivo models; importantly, HNSCC cells carrying high-risk TP53 mutations became sensitive to cisplatin treatment by the selective WEE1 kinase inhibitor [252]. In conclusion, inhibition of WEE1 may sensitize cells to DNA damage therapy; although P53 has been reported to affect the effectiveness of this approach, other studies support that WEE1 inhibition sensitizes cancer cells to chemotherapeutics independently of P53 function [253].
Inhibition of DNA-Dependent Protein Kinase to Re-Sensitize Cells
DNA-PK belongs to the PI3K-related protein kinase (PIKK) superfamily. It participates in NHEJ to repair DSBs in DNA [254]. DNA-PK may play a role in resistance to chemotherapy and radiotherapy [255,256]. Several inhibitors, including small molecules, have been developed to target DNA-PK, block the DSB repair pathway, and sensitize cells to therapy [199,257]. The small molecule DNA-PK inhibitor, PI-103 or NU7441, combined with the third-generation epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) Osimertinib, led to synergistic effects in TKI-resistant lung cancer cells. The enhanced effect was attributed to the prolongation of DNA damage and cell cycle arrest [258]. The DNA-PKcs inhibitor NU7441 blocked glioma stem cell tumorsphere formation in vitro. In addition, in human-derived glioblastoma xenograft mice, the inhibitor blocked tumor growth and sensitized cancer cells to radiotherapy [259].
Mechanisms of Resistance to DNA Damage Response Inhibitors
As with most target therapies, some patients are primarily resistant to DDRi and others eventually develop acquired resistance (Figure 1), with the latter being more frequent in patients with advanced disease [260]. Since PARPi are the only DDRi approved for clinical use, most known resistance mechanisms are associated with these inhibitors.
Resistance to PARP Inhibitors
The mechanisms of resistance to PARPi can be credited to several factors, including restoration of the mechanisms controlled by BRCA, such as HR repair and/or stabilization of replication forks [261,262]. Like other systemic chemotherapies, cancer cell develops PARPi resistance via several different mechanisms: (i) increased expression of multidrug resistance pumps (MDRs), enhancing the efflux of the PARPi out of the cell [263], (ii) reduced PARP1 binding affinity to DNA due to mutations and functional alterations of the PARP1 protein and/or disrupted PARylation [264,265], or (iii) restored HR and/or replication fork stabilization [266][267][268][269]. Mechanisms of resistance to DDR inhibitors. Cancer cells develop resistance to DDR through several mechanisms. The molecular mechanisms of resistance to PARPi include HR capacity restoration, decreased trapping of PARP1, stabilization of replication forks, and P-gp-mediated drug efflux. The resistance to WEE1 inhibitor is induced by AXL overexpression, mTOR signaling, CHK1 activation, and through the overexpression of MYT1 that decrease CDK1 activity. The resistance to CHK1 inhibitor is associated with increased E2F/G2M/SAC expression and reduced replication stress. The resistance to ATR inhibitor is induced by the loss of PGBD5 and CDC25A deficiency. Finally, the DNA-PK inhibitor resistance is caused by the loss of MLH1/MSH3 and the overexpression of ABCG2.
In BRCA-deficient tumors, the most frequent acquired resistance mechanism to PARPi is the re-establishment of BRCA1 or BRCA2 functionality by secondary intragenic mutations; specifically, genetic alterations may reinstate the open reading frame (ORF), leading to the expression of functional BRCA [270]. In addition, restoration of the wild-type BRCA protein may occur via a secondary mutation that reverses the inherited mutation or by the demethylation of the BRCA1 promoter; both these events may lead to restoration of the wild-type BRCA protein [160,271]. Specific mutations in the BRCA gene, including the BRCA1-C61G mutation, may also confer PARPi and cisplatin resistance [272]. Another possible way of HR restoration is due to loss of the shieldin complex, consisting of REV7, c20orf196 (SHLD1), FAM35a (SHLD2), and FLJ26957 (SHLD3), which normally prevents DSB resection and facilitates NHEJ. However, if lost, shieldin can promote PARPi resistance even in the absence of BRCA [273,274]. ATPase TRIP13 inactivates the shieldin complex, triggering the 5 to 3 resection of double-strand breaks and promoting HR [275]. In many BRCA-deficient tumors, TRIP13 is upregulated, contributing to the intrinsic PARPi resistance. Inhibiting the ATPase domain of TRIP13 can stabilize the shieldin complex to promote NHEJ, block HR and overcome intrinsic PARPi resistance. Inhibiting TRIP13 might be useful to treat BRCA-deficient tumors with intrinsic but also acquired PARPi resistance [275]. Secondary mutations restoring BRCA function were found in patients with germline BRCA mutation-associated ovarian and breast cancer upon acquired resistance to PARPi and/or cisplatin [276]. Reversion mutations of BRCA1 can also exhibit the MMEJ signature, pointing to the potential involvement of POLQ in driving resistance [277]. Consequently, inhibitors of POLQ can suppress PARPi resistance in HR and NHEJ-deficient cancers [277]. A detailed review of the mechanisms of BRCA re-activation in PARPi resistant cells was recently published [261].
In addition, BRCA1/2-deficient cancer cells may develop PARPi resistance by protecting their replication forks; they achieve this by blocking the recruitment of nucleases, MRE11 or MUS81, to the stalled fork, thereby resulting in fork protection [267,268]. These studies indicate that PARPi resistance is achieved without restoring HR repair. Furthermore, other mechanisms of resistance to PARPi have been reported, such as downregulation of PARP levels and increased levels of the P-glycoprotein efflux pump [260,278]. Overexpression of ABCB1 has been reported in PARPi-resistant human ovarian cancer cells; administration of MDR1 inhibitors such as Verapamil and Elacridar reversed resistance to PARPi [279]. It has been reported that most clinical PARP inhibitors induce cytotoxicity by trapping PARP1 at sites of DNA damage [280]. Resistance to PARP inhibitors can emerge through point mutations in PARP1 that alter PARP1 trapping, highlighting the importance of PARP1 intramolecular interactions in PARPi-mediated cytotoxicity [264].
Cell Cycle Regulators in DNA Damage Response Inhibitors in Resistant Cells
Several reports suggest that the silencing of cyclins may confer resistance to DDR inhibitors. There are two major classes of G1 cyclins that regulate cell cycle progression during the G1 phase: cyclin D, which cooperates with either CDK4 or CDK6, and cyclin E, which binds CDK2. During cell cycle progression in the early G1 phase, cyclin D-CDK4/6 complexes phosphorylate the Rb protein. Complete phosphorylation of the Rb protein is achieved at the end of the G1 phase by the CDK2/cyclin E complex. Fully phosphorylated Rb protein is inactive and releases the E2F factor, allowing the expression of S phase genes, leading the cells through the G1/S checkpoint. In response to DNA damage, p21 levels are increased; p21 binds both to the cyclin and the CDK subunits of the CDK/cyclin complex and disrupts the interaction between CDK and its substrates, blocking cell cycle progression [281].
Downregulation of cyclin D has been shown to confer resistance to CHK1 inhibition. CHK1, a serine/threonine kinase that acts as an ATM-ATR effector, is activated following exogenous DNA damage, including nicks caused by chemotherapeutic drugs. CHK1 activates the S and G2 checkpoints by controlling different mechanisms of DNA repair, including activation of homologous recombination repair or apoptosis if DNA damage is too severe [282,283]. In a MCL cell line that was made resistant to the CHK1 inhibitor PF-00477736, the re-expression of cyclin D1 partially re-sensitized cells to the agent. This suggests that low levels of cyclin D1 confer resistance to CHK1 inhibitors and that reestablishment of this protein may re-sensitize cells [284].
Cell division cycle 25A (CDC25A) is a dual-specificity phosphatase implicated in cell cycle control by inhibiting CDK phosphorylation and causing the formation of cyclin-CDK complexes. Following DNA damage, CDC25A is degraded, leading to cell cycle arrest. CDC25A is overexpressed in cancer and promotes tumorigenesis [285]; interestingly, a genome-wide CRISPR screen showed that the absence of CDC25A leads to ATR inhibitor resistance. Loss of CDC25A led to cell cycle arrest in cells treated with ATR inhibitor, diminishing the DNA damage caused by ATR inhibitors might otherwise generate; resistance was reversed using a WEE1 inhibitor that forced mitotic entry [286].
Dysfunctional apoptosis is one of the hallmarks of cancer. The increased levels and/or activity of anti-apoptotic proteins and, concurrently, the inactivation of pro-apoptotic molecules convey resistance to many anticancer drugs [287]. P53, a key molecule controlling cell cycle fate following DNA damage, is silenced in most human cancers. However, restoration of P53 following inhibition of MDM2 by Nutlin conveyed resistance to the cytotoxic effects of WEE1 inhibitor AZD1775 [288].
Activation of Alternative DNA Repair Pathways
Several studies report that the activation of alternative pathways to repair DNA damage is responsible for the observed resistance to DDR inhibitors. In P53-deficient cells, the induction of DSBs using a radiomimetic agent and DNA-PK inhibition led to an increased DSB burden in the S-phase; however, a subset of the cell population exhibited resistance to this combination therapy, which was caused by the recruitment of DNA polymerase theta (Pol θ or POLQ). Pol θ mediated end joining repair to improve cell viability following therapy-induced DNA damage. Concurrent inhibition of Pol θ and DNA-PK sensitized p53-deficient breast cancer cells to therapy [289].
Liabilities upon Treatment with DDR Inhibitors
Target therapies, including PARPis, contribute to important therapeutic breakthroughs in oncology, improving the quality of life and increasing the life expectancy of cancer patients. As mentioned, PARPis were demonstrated to be clinically effective, with acceptable tolerability and safety, in a specific range of solid tumors, which led to FDA and European Medicines Agency (EMA) approval of Olaparib, Rucaparib, Niraparib, and Talazoparib [31,290]. However, a consolidated body of evidence from studies of PARPi in patients has identified several adverse events and specific indications for their prevention, monitoring, and management [291][292][293][294]. PARPi display several on-and off-target toxicities, with hematological and gastrointestinal toxicities among the most common adverse events. Pneumonitis and therapy-related myeloid neoplasias (t-MN), such as AML and myelodysplastic syndromes (MDS), have been reported with PARPi, but despite their rare frequency, they are potentially life-threatening, often fatal, and deserve particular attention due to their severity [291]. The t-MN is typically a late complication of some chemo-and radiotherapy, and the subtype and latency period are usually treatment-dependent [295].
The link between PARPi and the development of t-MN is not fully understood. The pretreatment presence of clonal hematopoiesis of indeterminate potential (CHIP) with TP53 mutations [296], a hematopoietic cell population with one or more somatic mutations/copy number alterations that can expand with time and under positive clonal selection pressures [297], have been proposed as a possible explanation. Kwan et al. also analyzed the risk of t-MN development in patients with HR gene alterations and found a higher prevalence in patients with high-grade ovarian cancer that harbored a deleterious mutation in BRCA1, BRCA2, RAD51C, or RAD51D (4.1%) compared to those with mutation-containing cancers (1.0%) and without mutations (1.0%) [296]. Mutations of DDR genes (e.g., TP53, PPM1D, and CHEK2) involved in CHIP occur with increased frequency in cancer patients exposed to platinum compounds/topoisomerase II inhibitors or radiation therapy [296,298]. Additionally, previous treatments with platinum and alkylating agents may increase the risk of t-MN development in BRCA-associated high-grade ovarian cancer patients treated with PARPi as maintenance therapy [299]. PARPi may potentiate t-MN in patients with preexisting CHIP by selecting clones with DDR gene mutations that improve the competitive fitness of the cells under these conditions [31,299]. Oliveira et al. (2022) performed a comprehensive analysis of the pathologic and genetic characteristics of PARPi-related t-MN patients, showing that these patients have complex karyotypes and frequently have pathogenic TP53 mutations [300].
Most data available about t-MN arise from gynecologic cancer patients treated with Olaparib, with an estimated frequency of t-MN development between 1% in the PAOLA-1 study [105] and 8% in the SOLO-2 trial [301]. A recent study by Morice et al. (2021) evaluated the safety profile of 31 randomized controlled trials with PARPis as one arm in different tumor types and settings [302]. In this systematic review, PARP inhibitors significantly increased the risk of AML and MDS in comparison with placebo treatment (Peto OR 2.63 [95% CI 1.13-6.14], p = 0.026); the incidence of these t-MN across PARPi groups of 0.73% and placebo groups was 0.47%, with a median latency between first PARPi and the t-MN onset of 17.8 months [302]. The risk of t-MN development was small but more than doubled, even after controlling for prior platinum-based chemotherapy.
In a meta-analysis, Nitecki et al. (2021) did not find an increased incidence of t-MN in PARPi-treated patients compared to control treatments [303]. However, patients who received a PARPi as frontline treatment and those who received fewer than two prior lines of chemotherapy showed a higher risk of t-MN [303]. In a pharmacovigilance analysis of the FDA adverse event reporting system, Ma [304]. Current studies showed several limitations, including the cross-over between control and PARPi arms. This scenery may overestimate the incidence of t-MN in control/placebo arms since the subsequent therapies are not regularly reported [303].
Clinicians need to be aware of these late but potentially fatal adverse events, especially in the front-line maintenance settings, and pharmacovigilance and mechanistic studies should be implemented to improve the understanding of the risk factors that predispose to t-MN. Identifying biomarkers that discriminate patients at high risk of t-MN development upon PARPi treatment from those who benefit from frontline PARPi will improve treatment outcomes and prevent undesired adverse events.
Conclusions and Perspectives
Cancer cells display several defects in DDR pathways, offering a chance to explore these deficiencies clinically. DDR-based cancer treatments and combinatory regimens provide potential therapeutic approaches that exploit deficiency DDR pathways via synthetic lethality strategies. Despite the success of PARPi in HR-deficient cancers, such as breast, ovarian, and pancreatic cancers, several patients present serious toxicities or developed resistance to DDRi. A variety of DDRi resistance mechanisms have already been identified in preclinical models and patients, but clinical data are still scarce, and this remains an open field of research. Another challenge in DDR-based cancer treatments is the identification of genetic and functional biomarkers that define the patients who will be most suitable, suffer fewer side effects and toxicity, and benefit more from these therapeutic options. Moreover, although the higher benefits of DDRi are observed in patients with impaired DDR machinery, patients with proficient cancers can also benefit from these therapeutic approaches. Thus, further investigation is warranted to identify differential strategies for these patients, including combinatory approaches with targeted therapies such as immunotherapies (e.g., immune checkpoint inhibitors and non-specific immunotherapies), anti-angiogenic agents (e.g., VEGF inhibitors), and metabolic drugs (e.g., IDH inhibitors), among others. Another possible strategy is to combine different DDRi (e.g., PARPi with ATM, ATR, WEE1, or CHK1/2 inhibitors). Currently, PARPi is the maintenance therapy of choice for some cancers, such as ovarian, fallopian tube, primary perineal, and pancreatic cancer, showing manageable toxicity profiles. However, it should be highlighted that PARPi treatment increases the risk of AML and MDS development. This is a rare but frequently fatal event, and prescribing clinicians should remain vigilant about this complication. Additional research, including long-term pharmacovigilance studies, is needed to identify toxicity-predisposing factors and susceptibility biomarkers to further refine and personalize DDRi treatment and prevent t-MN development in front-line and maintenance settings. Furthermore, a better understanding of the molecular mechanisms of resistance to DDRi and the development of strategies to prevent or delay the acquisition of resistance are needed.
|
v2
|
2022-11-26T17:02:36.747Z
|
2022-11-24T00:00:00.000Z
|
253935590
|
s2orc/train
|
A Novel, Simple, and Low-Cost Approach for Machine Learning Screening of Kidney Cancer: An Eight-Indicator Blood Test Panel with Predictive Value for Early Diagnosis
Clear cell renal cell carcinoma (ccRCC) accounts for more than 90% of all renal cancers. The five-year survival rate of early-stage (TNM 1) ccRCC reaches 96%, while the advanced-stage (TNM 4) is only 23%. Therefore, early screening of patients with renal cancer is essential for the treatment of renal cancer and the long-term survival of patients. In this study, blood samples of patients were collected and a pre-defined set of blood indicators were measured. A random forest (RF) model was established to predict based on each indicator in the blood, and was trained with all relevant indicators for comprehensive predictions. In our study, we found that there was a high statistical significance (p < 0.001) for all indicators of healthy individuals and early cancer patients, except for uric acid (UA). At the same time, ccRCC also presented great differences in most blood indicators between males and females. In addition, patients with ccRCC had a higher probability of developing a low ratio of albumin (ALB) to globulin (GLB) (AGR < 1.2). Eight key indicators were used to classify and predict renal cell carcinoma. The area under the receiver operating characteristic (ROC) curve (AUC) of the eight-indicator model was as high as 0.932, the sensitivity was 88.2%, and the specificity was 86.3%, which are acceptable in many applications, thus realising early screening for renal cancer by blood indicators in a simple blood-draw physical examination. Furthermore, the composite indicator prediction method described in our study can be applied to other clinical conditions or diseases, where multiple blood indicators may be key to enhancing the diagnostic potential of screening strategies.
Introduction
In recent years, newly diagnosed cases of kidney cancer have increased year by year [1], with 90% of the cases being clear cell renal cell carcinoma (ccRCC) [2]. Previous studies have pointed out that the incidence rate of renal cancer in the elderly is much higher than in the younger population [3]. With the rapid explosion and growth of the global newborn population after the Second World War, all countries have faced the ageing population problem in the past decade [4,5]. As a senile disease, the incidence of renal cell carcinoma (RCC) is highest in people aged 60 to 70 [6]. The number of confirmed cases of renal cancer has recently increased yearly [7,8]. For ccRCC patients, the five-year survival rate of early-stage (TNM 1) ccRCC is 96%, while that of advanced-stage (TNM 4) patients is only 23% [9,10]. Consequently, as a potentially fatal disease, early screening is extremely important for the successful treatment of kidney cancer, and no low-cost screening strategies are currently available. This problem is especially important in lowincome countries or countries with limited opportunities for physical examination and imaging of a large number of people. In these cases, early screening of renal cancer through simple blood tests may significantly impact life-saving strategies.
In these cases, early screening of renal cancer through simple blood tests may have a significant impact on a life-saving strategies [11]. From a clinical point of view, Urology is facing future challenges, such as implementing the wide use of biomarkers. Along with medical imaging, like ultrasound and computed tomography, biomarkers, if reliable, could be successfully used for early diagnosis of kidney cancer [12,13]. In most developing countries, people rarely have whole-body physical examinations [14]. Generally, patients will not go to the hospital for examination until they have symptoms, such as physical weakness and gross hematuria [15]. However, when kidney cancer has such symptoms, patients are basically in the advanced stage of cancer, which significantly increases the difficulty of treatment and results in a lower survival rate [16]. Therefore, simple and efficient means of early cancer screening can effectively assist in cancer diagnosis. Previous studies have pointed out that cancer patients' physical function and blood biochemical indicators will change significantly compared with healthy people [17]. The possible clinical utility of biomarkers in urology has been investigated and well-reviewed in Mancini et al. [11]. For renal cancer, He et al. [18] investigated the ability of preoperative serum albumin (ALB) to globulin (GLB) ratio (AGR), which predicts the long-term mortality of RCC patients, and proved AGR is an inexpensive survival predictor to be considered for routine clinical use. Shah et al. [19] proved that the level of haemoglobin (HB) and the survival rate of renal cell carcinoma are significantly related. Checheriţȃ et al. [20], Lazich and Bakris [21] pointed out that blood potassium (K + ) imbalance is widespread in patients with kidney disease, especially in patients with the renal tubular disease or reduced glomerular filtration rate. Many studies have proved that blood urea nitrogen (BUN), serum creatinine concentration (CREA), and uric acid (UA) are important indicators of kidney function [22][23][24]. In addition, a high level of high-sensitivity C-reactive protein (hs-CRP) in the blood is one of the reasons for the shortened survival time of patients with chronic kidney diseases and cancers [25,26]. All the above-mentioned studies propose some blood indicators that show different patterns of expression in patients with kidney cancer, detectable with a simple blood test. The aim of our study was to combine these indicators and utilize them as an eight-indicator panel, with potential predictive value for diagnosis of kidney cancer in the general population.
We collected peripheral blood samples from people diagnosed with early-stage ccRCC (TNM 1) and advanced-stage ccRCC (TNM 4). After that, we measured the indicators of peripheral blood and compared them with the indicators of blood samples collected from healthy people. Meanwhile, we analysed the correlation and differences between patient's indicators. Then, the diagnosis and prediction of ccRCC were performed using a naive Bayesian model (single blood indicator) and a random forest model (mixed blood indicators). As shown in Figure 1, the random forest prediction model was trained to test the prediction performance of each indicator in the blood. All ccRCC-related peripheral blood indicators were combined for comprehensive prediction. The prediction differences of early and late indicators are compared simultaneously to improve the accuracy and prediction performance of the model. Early screening for renal cancer by simple test blood indicators is finally realised. Figure 1. The scheme of early screening for kidney cancer and research protocol. We collected data from healthy people and patients diagnosed with ccRCC. The blood samples were measured by biochemical instruments for eight main indicators: ALB, TP, HB, K + , BUN, CREA, UA and hs-CRP. After the data were measured, the indicators were introduced into the RF model for training. Finally, the prediction model's performance was tested to achieve accurate classification and prediction performance.
Patients and Samples
In this study, we collected 743 blood samples from patients with, renal cancer from September 2017 to September 2022 in Dazhou Central Hospital, including 505 blood samples from patients with early-stage ccRCC (TMN 1) and 237 blood samples from patients with advanced ccRCC (TMN 4). The age of cancer patients is mainly around 57 years old, and the interquartile range is 47-65 years old. Therefore, we randomly collected samples from 500 patients aged 50-65 as the control set, with an average of 56 years old. We collected venous blood samples from patients with medically and radio-logically confirmed ccRCC, using heparin as an anticoagulant. We centrifuged the collected sample at 4 • C at 3000 rpm for 30 min and separated the blood cells to prepare for the determination of blood indicators. Patient ALB, total protein (TP), HB, renal function indicators (BUN, CREA and UA), and inflammation indicators hs-CRP in blood samples were obtained and collected by the automatic biochemical analyser LABOSPECT 006 (Hitachi, Ltd., Tokyo, Japan) and Coulter Ac•T 5diff AL (Autoloader) Hematology Analyser (Beckman Coulter, Ltd., Indianapolis, IN, USA) using a MedicalSystem test kit (MedicalSystem Biotechnology Co., Ltd., Ningbo, China). We confirm that this study was conducted in accordance with the Declaration of Helsinki, and was approved by the ethics committee of Dazhou Central Hospital (protocol code 2022(052)).
Statistical Analyses
The clinical report was completed with the "gtsummary" package [27] in R language, in which the number and percentage of male and female populations are expressed. In contrast, age and other biochemical indicators are displayed by mean and interquartile range (IQR). The different analyses between biochemical indicators were carried out with the "limma" tool package (version 3.52; https://bioconductor.org/packages/limma (accessed on 10 October 2022)) developed by Ritchie et al. [28]. Continuous variables are presented as mean ± SD, and categorical variables are expressed as a number (percentage) or visualisation through R studio and Python. For correlation coefficient, 1 means positive correlation and −1 means negative correlation, which can measure the strength of the variable relationship. The closer to 0 correlation, the weaker the correlation. All the indicators were analysed and calculated. The following is the formula for calculating the correlation coefficient: where Corr is the correlation coefficient, x i is the values of the x-variable in a sample,x is the mean of the values of the x-variable, y i is the values of the y-variable in a sample, andȳ is the mean of the values of the y-variable.
For the significance analysis of statistical differences, we consider an observed teststatistic t from unknown distribution T. Then, the p-value p is what the prior probability would be of observing a test-statistic value at least as "extreme" as t if null hypothesis H 0 were true [29]. Therefore, the calculation of p is as follows: In statistics, naive Bayes classifiers are a class of simple probabilistic classifiers based on the application of Bayes' theorem and the assumption of strong (naive) independence between features [30,31]. The conditional probability model probability can be decomposed as: where C k is the class variable, corresponding with vector x.
In model research, the ROC curve is often used to evaluate a model's effectiveness and test whether the model has practical value [32,33]. The "pROC" package (version 1.18; https://cran.r-project.org/web/packages/pROC/ (accessed on 10 October 2022)) visualizes the ROC curve and AUC of our model, where AUC is a critical index in ROC curve that tests whether positives are ranked higher than negatives. The AUC is equivalent to the Wilcoxon or Mann-Whitney U test [34] statistic, with the relation as follows:
Modelling of Early Screening Models
We constructed a naive Bayesian classification model [30] to predict a single index and introduced a single biochemical indicator of healthy people and ccRCC patients into the model for training to obtain the ROC prediction performance curve and the corresponding AUC of each of the eight indicators. The random forest (RF) classifier [35] is an integrated machine learning method that is a collection of decision trees. The final decision of RF is to make a majority vote in all trees to produce a more accurate classification, and it has been widely used to solve classification difficulties. Compared with other popular classifiers [36], RF is recognised as a good classification method. Meanwhile, a naive Bayesian model is suitable for building and further analysing enormous data sets. This model is a straightforward compound classification method that can classify well even in complex situations. In this study, eight blood indicators were normalised, and the data were grouped by random sampling. We divided healthy individuals and ccRCC patients into training sets (80%) and verification sets (20%). The RF model is based on Python (version 3.9; https://www.Python.org (accessed on 10 October 2022)) and the "sklearn" library (version 1.1.2; https://scikit-learn.org/stable (accessed on 10 October 2022)). The GridSearchCV module was used to automatically adjust the parameters of the RF model with about 100 trees, each with eight randomly selected variables and a maximum tree depth of 50 to achieve the best results for the model. We collected the results and selected the model with the best performance while measuring the prediction accuracy on the test set. At the same time, the prediction accuracy was measured on the test set. Then, the model was optimised for the number of variables selected for each tree. In this process, the over-fitting of the RF model during parameter adjustment was prevented by crossvalidation, so as to keep the stability and practicability of the model. Model performance evaluation is based on the ROC curve and corresponding AUC value.
The Demographic Characteristics of All Samples
In this study, the ratio of males to females in the healthy population is approximately 1:1. In early-stage ccRCC, the ratio of males is 273:505, accounting for 54.1%, which is slightly higher than that of females, while in advanced ccRCC, the ratio of males is much higher than that of female, accounting for 60.8%. The detailed distribution is shown in Figure 2A-C. The average age of healthy people is 56 years old and that of people with cancer is 57 years old, but the age fluctuation range is slightly different. The age range of early-stage ccRCC is 47-65 years old, while that of advanced cancer is 54-65 years old. This also shows that advanced cancer is more likely to occur in older people.
Furthermore, we analysed the correlation between eight key biochemical indicators in healthy people and early-stage ccRCC patients. In the results, only ALB and TP have a strong correlation in healthy people, because TP is the sum of ALB and GLB and the value of TP is partly determined by ALB. Notably, BUN and CREA also have a strong positive correlation ( Figure 2D). In fact, BUN can reflect the kidney condition like CREA can, because BUN is the same as blood creatinine. It is one of the ultimate protein metabolic products, mainly through the filtering function of glomerular balls to discharge in vitro. Therefore, BUN and CREA show a strong correlation.
In the cancer population, the renal function indicators (BUN, CREA, and UA) are all positively correlated with the patient's age, indicating that the renal function indicators also increase with the increase of age. As shown in Figure 2E, BUN and CREA have the same trend in the healthy population, showing a strong correlation. The difference is that the correlation between BUN and CREA in the healthy population is lower than that between ALB and TP. In comparison, the correlation between BUN and CREA in cancer patients is far more significant than that between ALB and TP. In addition, there is a positive correlation between BUN, CREA, K + , and HB indicators, among which the correlation between CREA and BUN, and CREA and HB are the most obvious. Clinically, urea nitrogen and creatinine in the blood are products of protein metabolism; therefore, both have a positive correlation with protein (HB) rising.
Analysis of Blood Biochemical Indicators
After comparing the correlation of blood indicators of patients in each group, we also analysed eight indicators previously showed to be involved in kidney diseases. Due to the differences between men and women in constitution, endocrinology, and normal range of blood indicators [37], we compared the statistical significance between men and women in the analysis process, as shown in Figure 3A. *** means p < 0.001 statistically significant; ** means p < 0.01 is statistically significant; * means p < 0.05 is statistically significant; insignificance is denoted by ns. In terms of protein indicator results, ALB and TP indicators between healthy men and women were not significantly different. ALB and TP of early-stage ccRCC began to show statistical significance (p < 0.05), while the male and female patients with advanced ccRCC (TNM 4) showed high statistical significance (p < 0.01). On the other hand, there is a significant difference in haemoglobin between healthy men and women. With the development of cancer, This may be because more advanced cancer consumes a large amount of HB. Therefore, more advanced tumours better balance the differences between men and women. This result is also reflected in hs-CRP, and there is no significant difference between men and women in advanced ccRCC patients.
As a result, for early diagnosis, women could benefit more from the panel proposed in this study, while in case of advanced tumours, women and men are equally well performing regarding the diagnostic ability of the panel. However, there is a statistical significance (p < 0.001) between male and female patients in renal function indicators, and there is a statistical significance (p < 0.05) in both healthy people and cancer patients. Another key point is that in the analysis of male and female indicators, the values of ALB, TP, HB and K + of healthy people are above the average (dashed line in Figure 3A). The indicators of cancer patients decreased significantly compared with those of healthy people. On the contrary, renal function indicators (BUN, CREA, and UA) in the healthy population are significantly lower than the mean, while those in cancer patients are significantly higher. Moreover, as the cancer cycle increases, the inflammatory indicator hs-CRP also shows the same situation, indicating that renal cancer causes the rise of renal function indicators and increases inflammatory factors in patients. By combining the above parameters with our model, the published data can be machine-learned to improve the accuracy of prediction. The statistical significance analysis of ccRCC and healthy people. The statistical significance analysis of indicators with protein, renal function, inflammation, etc. (*** means "p < 0.001 statistically significant"; ** means "p < 0.01 is statistically significant"; * means "p < 0.05 is statistically significant"; insignificance is denoted by ns).
As well as the analysis of male and female blood indicators, we compared the significance of differences between indicators in different groups of samples from healthy and cancer populations. This allows us to identify specific differences between healthy people and people with cancer. The protein groups of early-and late-stage ccRCC showed statistical significance. The protein indicators of cancer patients were significantly lower than those of the healthy population. In addition, advanced cancer consumed relatively more endogenous proteins. Therefore, ALB, TP, and HB indicators showed a significant decline. For indicators K + and CREA, there is no significant difference, indicating that more advanced kidney cancer does not affect the indicators of K + and CREA. TP and BUN begin to show statistical significance (p < 0.05), while other indicators show extreme statistical significance (p < 0.001). Previous studies have pointed out that early-stage renal cancer does not affect renal function, and our results show that UA also shows consistent characteristics. In addition, ha-CRP mainly shows inflammation in patients. It can be clearly found from the indicator comparison that the inflammation indicator in cancer patients is much higher than that in the healthy population, and the hs-CRP indicator of advanced TNM 4 patients is much higher than that of early-stage cancer patients.
The Clinical Significance of the Blood Indicator Ratio
We systematically analysed eight related indicators of peripheral blood measurement, where TP level mainly reflects the loss of protein caused by renal lesion and serum ALB is also be greatly reduced in cancer patients. GLB is closely related to human immunity. When the human body is invaded by viruses or cancer cells, GLB will rapidly increase [38]. As we know, the total protein is the sum of albumin and globulin, so ALB/GLB ratio (AGR) is one of the key indicators in the combination of indicators for screening cancer. Detection of abnormalities can be made earlier by examining the AGR so that patients can prevent kidney complications and cut off the development process of renal cancer earlier. In particular, the AGR is often used as a critical nutritional reference value before clinical operation. In general, an albumin/globulin ratio between 1.2 and 2.5 is considered normal, although this may vary depending on the laboratory tests [39]. Human blood usually contains a little more albumin than globulin, which is why the normal ratio is slightly higher than 1. An AGR lower than 1.2 indicates that patients have severe nutritional problems with protein. As shown in Figure 4A, the red line indicates the demarcation line of AGR of 1.2, and the grey line indicates the demarcation line of the AGR ratio of 1:1, which also indicates that protein deficiency is severe. The results showed that the value for healthy people was above 1, while the value for patients with early T1 ccRCC was below 1, suggesting a serious problem. Moreover, the warning line of 1.2 passes through the high-density area of the cancer population. Half of the patients with advanced T4 are below the red line, which also indicates that malignant renal cancer is negatively correlated with AGR. Suh et al. [40] have also demonstrated that a lower AGR is associated with a higher risk of death in cancer patients.
In renal function indicators, CREA and BUN can reflect the degree damage to glomerular filtration function to some extent. BUN is also affected by extra-renal factors, such as a high-protein diets, gastrointestinal bleeding, dehydration, and high catabolism, which can cause BUN to increase. CREA is more accurate than BUN because it mainly depends on glomerular filtration capacity when exogenous creatinine intake is stable, and creatinine production in vivo is constant [41]. Therefore, the observation of the BUN/CREA ratio (BCR) in serum has certain clinical significance. Furthermore, the level of BCR reflects the quality of renal function. Figure 4B shows the distribution of BUN and CREA in healthy people and cancer patients. The green area in the figure shows the range of normal values (20-100) of BCR. Healthy people are basically in the normal range. More patients with cancer are in the normal range, and the proportion of ccRCC in TNM 4 is higher outside the range than that in the patient population with TNM 1. Its significance is mainly that when BUN or creatinine CREA values are increased, it can be used as the judgment of the difference between the causes of renal or pre-renal (extra-renal) cancer. A BCR less than 20 indicates a high risk of renal disease [42]. The influence of the tumour increases the urea nitrogen value, so cancer may lead to a BCR greater than 100.
A B
Healthy T1 ccRCC T4 ccRCC Healthy T1 ccRCC T4 ccRCC Many correlations and interactions between indicators were challenging to mine individually. Therefore, we performed the pairwise comparative analysis on patient age and the distribution of the eight key indicators, and conducted an in-depth analysis of the indicators through the scatter diagram and density diagram. The upper part of Figure 5 shows the density distribution of the 8 indicators. The scatter distribution of samples between groups is shown in the lower part of Figure 5, where the distribution of each patient and the indicator relationship between groups can be seen. In the figure, green indicates the healthy population, blue indicates early-stage ccRCC, and orange indicates advanced ccRCC. Among cancer patients, the age distribution curve ranges widely, but the peak is concentrated around the age of 60. In addition to the UA indicator of the healthy population basically covering cancer patients, the healthy population in other indicators was included in the scope of cancer patients. + Figure 5. The paired relationship between biochemical indicators with ccRCC patients and healthy population. Correlation analysis of pairwise parameter combinations of the eight indicators was conducted to visualise the distribution differences between cancer patients and healthy people.
Performance Test of Cancer Prediction Model
An ROC curve is a graphical technique that is often applied to visualise classifier performance. For the single indicator prediction model, ROC was used to determine the prediction effect of the model. In this study, we tested the classification prediction performance of single indicator data of healthy population and early-stage ccRCC patients using the prediction model, as shown in Figure 6. The results showed the prediction performance of all eight key indicators. Except for the fact that UA in the renal function indicators had little predictive performance, all other indicators had high AUC values, among which HB, K + , and BUN had generally good performance, with AUCs around 0.6. The indicators of ALB, TP, and CREA had good performance, with AUCs above 0.7. Among them, ALB and TP have reasonable specificity above 90%, while CREA has poor specificity, but a high sensitivity of 91.1%. In addition, among the single-indicator prediction models, the inflammatory indicator hs-CRP performed best. The AUC of the model was up to 87.3%, the specificity was up to 93.4%, and the sensitivity was also 76.6%. Finally, the results show that the single-indicator prediction models of the ALB, TP, CREA, and hs-CRP indicators have good prediction effects. We trained the RF model with data sets of healthy people and cancer patients. The AUC of the RF prediction model was verified by combining eight key renal cancer prediction indicators. Our model has good prediction performance, with an AUC of 0.932 (Figure 7), a sensitivity of 88.2%, and a specificity of 86.3%. These findings indicate that random forest-based prediction can provide a satisfactory alternative biopsy method for ccRCC patients. In particular, high specificity and sensitivity may make our method useful for the early screening of renal cancer.
Discussion
In this study, the peripheral blood of early-stage ccRCC (TNM 1) was collected, and healthy people and advanced ccRCC (TNM 4) blood samples were used as control groups. After measuring the indicators of peripheral blood samples, the correlation between the indicators of early-stage ccRCC patients and the differences of healthy people were analysed. The results showed that cancer patients had lower protein levels, while renal function and inflammation indicators were higher than healthy ones. In addition, there was a high positive correlation between renal function indicators and a large difference between male and female indicators. However, more advanced tumours narrowed the difference between male and female indicators. Analysis of the blood indicators of healthy people and cancer patients can preliminarily classify the patients, with samples whose indicators exceed the standard have a higher risk of cancer. Correspondingly, we applied a naive Bayesian model (single blood index) and a random forest model (mixed blood index) to diagnose and predict ccRCC. The results are shown in Figure 6. A random forest prediction model was trained to test the prediction performance of each indicator in the blood. All the peripheral blood indicators related to ccRCC were combined for comprehensive prediction, and a good prediction performance was obtained. The AUC of the RF prediction model was verified at 0.932 (Figure 7). The sensitivity and specificity of the RF model in the validation queue of eight indicators were 88.2% and 86.3%, respectively. In summary, the early screening of renal cancer was realised by simple blood tests.
Gender differences exist in the incidence of kidney cancers, hormones, blood indicators, and the prognosis of the disease [43,44]. Nevertheless, clinical trials and studies of renal cancer are always unbalanced in terms of gender. Most data on gender differences in kidney cancer comes from studies published in developed countries [44]. In this study, we analysed the differences between male and female indicators in detail through the combination of blood indicators and models to improve the diagnostic potential of screening in women. Therefore, our model, when applied in countries where women have limited access to healthcare, can enhance the screening possibilities for this specific portion of the population, so that they could be diagnosed at an early stage. At more advanced stages, the differences between men and women tend to be less pronounced. Cancer patients have lower protein levels than healthy people [45], Furthermore, the inflammatory indicator hs-CRP has been proven to increase [46]. The changes in indicators in ccRCC patients in our results ( Figure 3) are consistent with the trends in previous studies. Yim et al. [47] proved that the renal function of patients with early-stage renal cancer is basically unaffected, and uric acid tends to be low with tumour development. Norberg et al. [48] also pointed out that renal cancer increases levels of creatinine, urea nitrogen, and other indicators with the development of tumours. This kind of situation also existed in patients with renal cancer in our research results ( Figure 3B). High CREA and BUN levels in blood represent the decrease of glomerular filtration rate and renal detoxification function. At this time, the kidney function of patients with nephropathy begins to be damaged and enters the stage of renal insufficiency. With the continuous increase of CREA and BUN in blood, the degree of renal function damage will become much more serious. However, the decrease of UA may be due to the gradual weakening of the metabolism and intracellular enzyme activity, resulting in the weakening of biochemical reactions of UA metabolism. Previous studies have used serum albumin/globulin ratio (AGR) to predict long-term mortality [18,49] and predict the prognostic effect of screening by the same method [50] in ccRCC. A low albumin/globulin ratio may put patients at risk for cancer. In an observational study of nearly 27,000 people, participants with an AGR below 1.2 had an increased risk of cancer [40], even if they were otherwise healthy. In addition to being associated with risk of cancer, AGR may also indicate the extent to which cancer patients respond to treatment [51]. Moreover, He et al. [52] used 13,890 patients from 24 articles for an analysis of overall survival (OS); compared to lower AGR patients, higher AGR patients had better OS. Our results also show that with the development of cancer, more populations deviate from normal values of AGR and BCR (Figure 4). These results also prove the levels of AGR and BCR are greatly affected by cancer.
Prediction of the clinical behaviour of cancer by artificial intelligence is a hot topic in contemporary research [53]. The RF model has been previously used in disease and cancer prediction, and most cancer prediction models test model performance by the AUC value of the ROC curve. Wang et al. [54] used four genes to predict ccRCC to guide immunotherapy and their RF model's AUC was 0.78. Erdim et al. [55] constructed a random forest model with an AUC as high as 0.916; this model was recognised as a suitable method to distinguish benign and solid renal tumours. Our eight-indicator RF prediction model shows good predictive performance, with an AUC reaching 0.932. Therefore, the RF model in this study can be used as an effective tool for the early screening of renal cancer.
Our study has some limitations: the data comes from a single center, the patients' number is limited, and all the patients belong to the same race. Moreover, the statistical data is relatively simple. We can improve the statistical power of the study in the next future, by adding more data and more variability, and by introducing more research on the patients' follow-up.
Conclusions
This study focused on developing a low-cost, patient-friendly, and effective strategy for kidney cancer screening. A potential clinically useful combination of eight biomarkers, all "renal" indicators detectable with a simple blood analysis, were studied through an RF prediction model, which achieved a sensitivity of 88%, a specificity of 86.3%, and an AUC of 0.932.
These findings indicate that random forest-based prediction can be used as a reliable method for screening with high specificity and sensitivity, potentially acting as a low-cost liquid biopsy for ccRCC patients. This strategy, useful for early screening worldwide, could become particularly crucial in the large part of the world where more expensive screening programs are not possible and portions of the population are excluded from physical medical examinations or imaging.
Additionally, the comprehensive indicator prediction method applied in our research can also be used to predict the risk of harbouring other diseases whose presence has been correlated to alterations in specific blood indicators, as is the case in kidney cancer.
Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the ethics committee of Dazhou Central Hospital (protocol code 2022(052)).
Informed Consent Statement: Not applicable.
Data Availability Statement: The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.
Acknowledgments: This work was funded by Sun Yat-sen University and Dazhou Central Hospital. Thanks to Dazhou Central Hospital for supporting our research through instrumentation and determination of participants' data.
Conflicts of Interest:
The corresponding author declares no competing interests. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest, including any methods, additional references, Nature Research reporting summaries, source data, extended data, supplementary information, acknowledgements, peer review information; details of author contributions and competing interests.
|
v2
|
2022-11-26T17:08:34.001Z
|
2022-11-24T00:00:00.000Z
|
253909774
|
s2ag/train
|
Assessment of denosumab treatment efficiency in patients with giant cell tumor of bone using CT and MRI (own results and literature review)
Relevance. Giant cell tumor of the bone is most common in people of working age, which determines the high social significance of successful treatment of this category of patients. The main method of treatment is surgical. Currently, the targeted drug denosumab has appeared, the criteria for evaluating the effectiveness of therapy for which, according to the data of radiation methods, are not clearly defined.Target. To analyze and compare the possibilities of CT and MRI in evaluating the effectiveness of denosumab therapy for giant cell tumors.Materials and methods. The data of CT and MRI of 19 patients with giant cell tumor of tubular bones on the background of denosumab therapy were analyzed.Results. Before treatment, the extraosseous component was determined in 57.9 % (n = 11), after – 31.6 % (n = 6). The decrease occurred in 100 %, the disappearance – in 45 % (n = 11) of cases. The thickness of the extraosseous component before treatment ranged from 4 to 43 mm (Me = 15 mm), after treatment it ranged from 0 to 30 mm (Me = 8 mm). The decrease occurred in the range from 4 to 14 mm (M ± SD = 7 ± 4 mm). In 100 % of cases, a sclerotic rim appeared, the thickness of which after treatment ranged from 1 to 5 mm (Me = 3 mm). In the structure of the tumor, fibrosis occurred in 95 % (n = 18), a decrease in the cystic component occurred in 82 % (n = 9) of cases. Perifocal changes decreased in 100 % of cases. In 100 %, the average tumor density increased. The mean tumor density before treatment ranged from 27 to 65 HU (M ± SD = 42 ± 11 HU), after treatment it ranged from 69 to 500 HU (Me = 150 HU). The increase in density occurred in the range from 41 to 454 HU (Me = 101 HU). All differences are statistically significant (p < 0.05).Conclusions. Evaluation of effectiveness with the definition of quantitative and qualitative indicators is possible according to the data of both CT and MRI; with CT, changes are recorded longer, and more indicators available for quantitative measurement are determined.
|
v2
|
2022-11-26T17:18:35.530Z
|
2022-11-24T00:00:00.000Z
|
253921235
|
s2orc/train
|
STIL Promotes Tumorigenesis of Bladder Cancer by Activating PI3K/AKT/mTOR Signaling Pathway and Targeting C-Myc
Simple Summary Currently; there are no reports on the role of STIL in bladder cancer. Using public databases; we observed that STIL is highly expressed in bladder cancer and is closely related to the cell cycle. The results of tumor formation experiments on UMUC3 and T24 bladder cancer cell lines indicate that STIL promotes the growth of bladder cancer cells in vivo and in vitro. Mechanistically, the cell cycle after STIL knockout was arrested in the G0/G1 phase; and cell cycle-related proteins (cyclin D1 and CDK2/4/6) were also reduced. RNA-seq results and immunoblotting experiments confirmed that STIL enhanced the PI3K/AKT/mTOR pathway, resulting in increased c-myc expression which ultimately promoted the occurrence and progression of bladder cancer. Our results suggest that STIL may be a promising potential therapeutic target for bladder cancer. Abstract SCL/TAL1 interrupting locus (STIL) regulates centriole replication and causes chromosome instability, which is closely related to malignant tumors. The purpose of our study was to investigate the role of STIL in bladder cancer (BC) tumorigenesis for the first time. The public database indicated that STIL is highly expressed and correlated with the cell cycle in BC. Immunohistochemistry staining showed that STIL expression is significantly elevated in BC tissues compared with paracancer tissues. CRISPR-Cas9 gene editing technology was used to induce BC cells to express STIL-specific sgRNA, revealing a significantly delayed growth rate in STIL knockout BC cells. Moreover, cell cycle arrest in the G0/G1 phase was triggered by decreasing STIL, which led to delayed BC cell growth in vitro and in vivo. Mechanically, STIL knockout inhibited the PI3K/AKT/mTOR pathway and down-regulated the expression of c-myc. Furthermore, SC79 (AKT activating agent) partially reversed the inhibitory effects of STIL knockout on the proliferation and migration of BC cells. In conclusion, STIL enhanced the PI3K/AKT/mTOR pathway, resulting in increased expression of c-myc, ultimately promoting BC occurrence and progression. These results indicate that STIL might be a potential target for BC patients.
Introduction
The incidence of bladder cancer (BC) ranks eleventh among all malignant tumors in the world [1]. BC is divided into two types, the majority being non-muscle-invasive diseases (90%) and the minority being muscle-invasive diseases [2,3]. Non-muscle-invasive BC frequently recurs (47%) [4], which greatly affects patients' quality of life. Furthermore, muscle-invasive BC has a poor prognosis and a five-year survival rate of <50% [5]. Despite advances in chemotherapy and surgery, the prognosis has not changed significantly, and new entry points for treatment options are urgently needed [6]. Therefore, it is important for BC diagnosis and prognosis to discover novel biomarkers.
SCL/TAL1 interrupting locus (STIL), a key regulator involved in the regulation of centriole duplication, is an important checkpoint protein in mitosis [7]. Chromosomal instability caused by ectopic centriolar amplification is the main feature of human cancer [8]. Elevated expression of STIL has been reported in various cancers, such as colorectal cancer, pancreatic cancer, gastric cancer, prostate cancer, lung cancer, and nasopharyngeal carcinoma [9][10][11][12][13][14]. STIL promotes the development of many cancers; however, there is no detailed investigation of STIL in BC.
In our study, we found that STIL was abnormally expressed in BC patients and predicted a poor prognosis. Moreover, STIL knockout markedly blocked proliferation and migration of BC cells in vitro and inhibited proliferation of tumor in vivo, and triggered cell cycle arrest in the G0/G1 phase. Mechanistically, STIL promoted the PI3K/AKT/mTOR pathway and increased c-myc expression, thereby facilitating BC progression.
Tissue Microarray and Immunohistochemistry
TMA (tissue microarrays) (product code: HBlaU079Su01, Outdo Biotech, Shanghai, China) included 63 BC tissues and 16 paracancer tissues. The samples were collected from May 2007 to January 2011. TMA slices were dehydrated with different concentrations of gradient alcohol. After being made transparent with a transparent agent (xylene), the slices were soaked in 3 pots of paraffin at 60 • C one by one for 1 h and finally embedded and sectioned. Antigen repair was performed in an electrochromic oven with a repair solution (0.01 mM EDTA buffer, pH 9.0). Three-percent hydrogen peroxide was added to block endogenous peroxidase, and goat serum was blocked at room temperature for 30 min. Then, the slices were incubated overnight (12 h) with a primary antibody. Next, the slices were incubated with the second antibody at 37 • C for 30 min, and then with a freshly prepared DAB chromogenic solution. After redyeing with hematoxylin, the slices were dehydrated, sealed, scanned, and photographed under a microscope. Immunohistochemical scores were performed independently by two experienced pathologists. In general, the expression of STIL was evaluated on the criteria of the intensity and positive staining rate of immunostaining of the tumor tissue. The positive signal was brownish yellow or brown. The staining intensity was scored as 0, 1, 2, and 3. The positive staining rate score was defined as 0 (0%), 1 (1-25%), 2 (26-50%), 3 (50-75%), and 4 (75-100%). Histoscore was calculated by multiplying the staining intensity score and the staining positive rate score.
Cell Culture
The human BC cell lines include the UMUC3, EJ, J82, T24, and SCABAR cell lines, HEK293T cell line, and normal human bladder cell line SV-HUC-1. These cells were donated by the Stem Cell Bank, CASS, China. One hundred U/mL penicillin, 10% fetal bovine serum (Invitrogen, Shanghai, China), and 0.1 g/mL streptomycin sulfate were added to RPMI medium 1640 (Invitrogen, Shanghai, China) or DMEM high glucose medium (Gibco, Shanghai, China). SV-HUC-1, EJ, and T24 cells were cultured in mixed RPMI medium 1640; UMUC3, J82, SCABAR, and HEK293T cells were cultured in mixed DMEM high-glucose medium. All cells were incubated at 37 • C with 5% CO 2 and 95% air.
STIL Knockout in UMUC3 and T24 Cells
We performed STIL knockout in UMUC3 and T24 cells with the CRISPR-Cas9 system. In this system, small guide RNAs (sgRNAs) of STIL were cloned into lenti-v2 (Addgene, 92062, Shanghai, China). HEK293T cells were co-transfected with recombinant lenti-CRISPR-v2 and package plasmid to generate lentivirus for 48 h. The supernatant was collected and added to UMUC3 and T24 cells, and incubated for 36 h; subsequently, 1000 ng/mL puromycin was added for 4 days. We used a limited dilution method to obtain the cloned cells (UMUC3 and T24 cells were immobilized on five 96-well plates and grown for 3 weeks). We screened monoclonal cells by Western blotting to obtain accurate STIL knockout cells. The sgRNA sequences were sgRNA-F, 5 -caccGGGGTTATTTCTAGGCATTC-3 ; sgRNA-R: 5 -aaacGAATGCCTAGAAATAACCCC-3 .
CCK-8 Assay
The cells with STIL knockout were inoculated into 96-well plates with 2 × 10 3 cells per well. The absorbance of 96-well plates was measured on days 1, 2, 3, 4, 5, and 6, respectively. Ten µL of CCK-8 solution (BS350A, biosharp, Hefei, China) was added to each well and cultured in a cell incubator for 3 h. The optical density was measured at 450 nm using a multifunctional enzyme marker (PE Enspire, PerkinElmer, Singapore).
Cell Colony Formation
The cells were inoculated into 6-well plates at a density of 1.2 × 10 3 cells per well and cultured for about 3 weeks until clearly visible to the naked eye. Finally, they were dyed with 0.3% crystal violet (dissolved by anhydrous ethanol), washed gently with water, and photographed.
Transwell Migration Assays
Medium containing 20% fetal bovine serum was added to the lower chamber of the 24-well plates, and 5 × 10 4 cells with serum-free medium were added to the upper chamber of a transwell chamber (Thermo Fisher Scientific, Shanghai, China) to culture for 24 h. The migrated cells were fixed with 4% paraformaldehyde (biosharp, Shanghai, China), stained with 0.3% crystal violet, photographed, and counted with an inverted microscope.
Soft Agar Assay
One-point-four-percent low-melting-point agarose (Promega, Madison, WI, USA) was added to 6-well plates and solidified at room temperature. Ten thousand cells were mixed with 0.7% low-melting-point agarose and overlaid on 1.4% low-melting-point agarose. After solidifying at room temperature again, they were put into an incubator and incubated for about 4 weeks before being photographed under a microscope.
Cell Cycle Assay
We used a cell cycle staining kit (Multisciences, 70-CCS012, Shanghai, China). After collecting 5 × 10 5 cells, the cells were washed with PBS. Then, 1 mL of DNA staining solution and 10 µL of osmotic solution were added to the cells. After incubating at room temperature for 30 min, the cells were detected on a flow cytometer (CytoFlex S, Beckman Coulter, Wuhan, China) with the lowest flow rate.
The EDU (5-Ethynyl-2-deoxyuridine) Assay
EDU cell proliferation detection kit (C0071S, Shanghai, China) was purchased from Beyotime. Cells were incubated with EDU working solution for 2 h, then fixed with 4% paraformaldehyde and treated with permeabilization solution (P0097, Immunostaining Strong Permeabilization Solution, Beyotime, Shanghai, China) for 30 min at room temperature. According to the protocol of the manufacturer, cells were incubated in click reaction solution (CuSO 4 , Click Reaction Buffer, Click Additive Solution, Azide 488) at room temperature in the dark for 30 min. Nuclei staining was conducted with Hoechst (33342, Beyotime, Shanghai, China) reagent. After staining, the cells were observed under a confocal fluorescence microscope or prepared as cell suspension samples for flow cytometry.
Immunofluorescence Assay
The Ki67 cell proliferation kit (E607238-0100, sangon, Shanghai, China) was used to measure cell proliferation. For specific steps, please refer to the instruction manual. Briefly, cells were incubated with the diluted primary antibody Ki67 overnight at 4 • C and incubated with the fluorescent-labeled secondary antibody at 37 • C for 30 min. Nuclei were then counterstained with 4 ,6-diamidino-2-phenylindole (DAPI). Finally, immunofluorescence staining was observed under an inverted fluorescence microscope.
RT-qPCR
RNA was extracted from cells using trizol reagent (Invitrogen, Shanghai, China); then, reverse transcription was performed using the cDNA reverse transcription kit (Toyobo Life Science, Shanghai, China) to synthesize cDNA. RT-qPCR was performed using the SYBR Green PCR kit (TSE201, Tsingke Biotechnology, Wuhan, China) and a Bio-Rad CFX96 PCR system to detect CT values. Primer sequences are in Supplementary Table S1.
Western Blotting
Protease inhibitors were added to the RIPA buffer (Beyotime, Wuhan, China) to lyse cells on ice and eventually extract proteins. The proteins were electrophoretic in 8-15% SDS-PAGE gel and transferred to PVDF membrane (IPVH00010, Sigma-Aldrich, Shanghai, China), after which they were sealed with blocking buffer (P30500, Ncm Biotech, Suzhou China) for 15 min. The membrane was incubated overnight in the designated primary antibody at 4 • C. After washing with TBST, the membrane was incubated with the corresponding secondary antibody at room temperature for 1 h. Subsequently, the membrane was imaged on a chemiluminescence image analysis system (Tanon 5200, Wuhan, China). STIL antibodies were purchased from Santa (SC-271910, Dallas, TX, USA). Original blots see File S1.
Tumor Xenografts
The experimental plan was approved by the Animal Biosafety Level III Laboratory of Wuhan University. We injected cells via syringe into the backs of nude mice living in a pathogen-free environment (6 × 10 6 cells/nude). The nude mice were purchased from China Shanghai Laboratory Animal Research Center. Tumor volume was measured every 3 days after surgery and calculated by the formula: length × width × height × π/6. On the 26th postoperative day, the tumors of mice were collected, photographed, and weighed.
2.14. Statistical Analysis SPSS 25.0 was used for statistical analysis. All experiments were performed at least 3 times. One-way ANOVA was used for three or more groups. A two-sided Student's t-test was used for the significance of the difference between the two groups. p < 0.05 was considered statistically significant (* p < 0.05, ** p < 0.001, and *** p < 0.001).
RNA-Seq Analysis of N-butyl-N-(4-hydroxybutyl) Nitrosamine (BBN)-Treated Bladder Cancer Mouse Models
We obtained the raw sequencing data from the NCBI SRA Database with Bioproject: PRJNA587619 [15]. STAR was used to align RNA sequencing reads [16], and featureCounts was used for quantification of gene and transcript levels [17]. And differentially expressed genes were screened between the BBN-treated group and control group by R package Deseq2 [18].
STIL Was Highly Expressed in BC and Was Inextricably Intertwined with the Cell Cycle
To analyze the expression levels of STIL in various tumors, we analyzed TCGA RNA sequencing (RNA-seq) data and found that STIL expression was significantly elevated in various cancer types, including bladder urothelial carcinoma (BLCA), uterine corpus Figure 1A). To further verify the expression levels of STIL in bladder cancer (BC), we analyzed the RNA-seq datasets of several BC tissues from TCGA. The results showed that STIL mRNA levels were significantly higher in BC tissues than in normal bladder tissues, and significantly higher in high-grade BC than in low-grade BC ( Figure 1B,C). Similarly, in the GSE13507 database, STIL mRNA leveler was higher in BC tissues than normal bladder tissues, and higher in muscle-invasive bladder cancer (MIBC) than non-muscle invasive bladder cancer (NMIBC) ( Figure 1D,E). In addition, BC patients with high expression of STIL had poor prognoses ( Figure 1F). Interestingly, the results showed that cell cycle-related genes (CCNB1, MKI67, CDK1, CCNA2, CCNB2, CCNE2) were significantly positively correlated with STIL ( Figure 1G-L). The results from the KEGG pathway enrichment analysis showed that the genes correlated with STIL were significantly enriched in the cell cycle ( Figure 1M). Marina Degoricija et al. checked the expression of the STIL gene in the most common mouse model of bladder cancer, the so-called BBN model in already published papers in which RNA-seq analysis was carried out on that model [15]. On these RNA-seq data, we found that STIL mRNA expression was significantly up-regulated in BBN-induced bladder cancer tissues, compared with control samples. Moreover, some other genes including AKT, c-myc, MCM4, and EIF2S2 were also significantly elevated in cancer tissues. These results were mostly consistent with our findings from human bladder cancer. We presented our data in the Supplementary Materials, Figure S4.
Expression of STIL Was Significantly Up-Regulated in BC
Immunohistochemical staining was performed on 63 BC tissues and 16 paracancer tissues from the tissue microarray; the findings revealed that the expression of STIL was significantly up-regulated in BC tissues compared to paracancer tissues (Figure 2A,B). Subsequently, we examined mRNA and protein levels in the normal bladder cell line (SV-HUC-1) and five BC cell lines; results showed that the expression levels of STIL in BC cells were significantly higher than in normal cells ( Figure 2C,D). The UMUC3 and T24 cell lines showed the highest STIL expression ( Figure 2E). Therefore, UMUC3 and T24 cells were used in the subsequent cell experiments in our study.
Knockout of STIL Inhibited BC Tumorigenesis In Vitro
We performed a CRISPR-Cas9-based STIL knockout assay in the UMUC3 cell line. The STIL knockout cells, especially STIL −/− 2# and STIL −/− 21#, were screened for the best knockout efficiency by Western blot (STIL +/+ : wild-type cell. STIL −/− : STIL knockout cell) ( Figure 3A). In the Cell Counting Kit-8 (CCK-8) assay, STIL knockout revealed remarkable inhibition of cell proliferation ( Figure 3B). The colony formation assay showed that STIL knockout significantly reduced cell proliferation ( Figure 3C,D). Similarly, cell migration was significantly reduced by STIL knockout in the transwell migration assay ( Figure 3E,F). Moreover, the colony size and the number on soft agar, which always indicate cell growth, were reduced in STIL knockout cells ( Figure 3G,H). We also performed knockout of STIL in the T24 cell line using CRISPR-Cas9 technology, and finally obtained STIL −/− 8# and STIL −/− 16# ( Figure S1A). Similarly, we performed a CCK-8 assay ( Figure S1B), colony formation assay ( Figure S1C,D), transwell migration assay ( Figure S1E,F), and soft agar assay ( Figure S1G,H). The results show that the ability of T24 cell to proliferate and migrate was decreased after STIL knockout. significantly up-regulated in BC tissues compared to paracancer tissues (Figure 2A,B). Subsequently, we examined mRNA and protein levels in the normal bladder cell line (SV-HUC-1) and five BC cell lines; results showed that the expression levels of STIL in BC cells were significantly higher than in normal cells ( Figure 2C,D). The UMUC3 and T24 cell lines showed the highest STIL expression ( Figure 2E). Therefore, UMUC3 and T24 cells were used in the subsequent cell experiments in our study.
Knockout of STIL Inhibited BC Tumorigenesis In Vitro
We performed a CRISPR-Cas9-based STIL knockout assay in the UMUC3 cell line. The STIL knockout cells, especially STIL −/− 2# and STIL −/− 21#, were screened for the best knockout efficiency by Western blot (STIL +/+ : wild-type cell. STIL −/− : STIL knockout cell) ( Figure 3A). In the Cell Counting Kit-8 (CCK-8) assay, STIL knockout revealed remarkable inhibition of cell proliferation ( Figure 3B). The colony formation assay showed that STIL knockout significantly reduced cell proliferation ( Figure 3C,D). Similarly, cell migration was significantly reduced by STIL knockout in the transwell migration assay ( Figure 3E,F). Moreover, the colony size and the number on soft agar, which always indicate cell
Cell Cycle of STIL Knockout Cells Was Arrested in the G0/G1 Phase
To understand the mechanisms of how STIL knockout inhibits BC tumorigenesis, we examined the cell cycle progression of STIL knockout UMUC3 cells and T24 cells. The results showed that STIL knockout markedly reduced the percentage of Ki67 positive cells (Figures 4A,C and S2A,C) and 5-ethynyl-2-deoxyuridine (EDU) positive cells ( Figures 4B,D and S2B,D). Flow cytometry analysis showed that proliferation cells (EDU positive cells) were significantly reduced after STIL knockout ( Figure 4E,F). To explore the relationship of STIL to the cell cycle in proliferation, we assessed changes in the cyclin-dependent kinases (CDKs). At the protein level, CDK2/4/6 and cyclin D1 were decreased after STIL knockout (Figures 4G and S2E). Flow cytometry analysis showed that STIL knockout cells were stagnant in the G0/G1 phase to prevent cell proliferation ( Figures 4H,I and S2F,G). growth, were reduced in STIL knockout cells ( Figure 3G,H). We also performed knockout of STIL in the T24 cell line using CRISPR-Cas9 technology, and finally obtained STIL −/− 8# and STIL −/− 16# ( Figure S1A). Similarly, we performed a CCK-8 assay ( Figure S1B), colony formation assay ( Figure S1C,D), transwell migration assay ( Figure S1E,F), and soft agar assay ( Figure S1G,H). The results show that the ability of T24 cell to proliferate and migrate was decreased after STIL knockout.
Cell Cycle of STIL Knockout Cells Was Arrested in the G0/G1 Phase
To understand the mechanisms of how STIL knockout inhibits BC tumorigenesis, we examined the cell cycle progression of STIL knockout UMUC3 cells and T24 cells. The results showed that STIL knockout markedly reduced the percentage of Ki67 positive cells
The Proliferative Function of STIL Was Confirmed by the Xenotransplantation Model
To demonstrate the function of STIL in vivo, we constructed a xenograft model. STIL +/+ , STIL −/− , 2#, and STIL −/− 21# cells were implanted into nude mice. On the 26th day, tumors were removed from the nude mice. The solid tumor volume of STIL +/+ was significantly larger than that of STIL −/− 2# and STIL −/− 21# ( Figure 5A-C). In addition, through dynamic observation of tumor growth, it was found that the tumor growth of the STIL knockout cells was slowed ( Figure 5D). These results suggested that STIL knockout has a dramatic effect on the regulation of BC tumorigenesis in vivo.
(E) quantitative analysis of EDU positive cell count. Ten thousand cells were counted in each sample. The ordinate of the second peak is the number of EDU-positive cells, and the fluorescence intensity (abscissa of the graph) was 10 6 at this point. (G) Representative images of Western blotting of cyclin D1 and cell cycle-related proteins CDK2/4/6. (H) Representative images of flow cytometry for propidium iodide (PI) staining and (I) quantitative analysis of cell cycle phase. EDU, 5-ethynyl-2-deoxyuridine (** p < 0.01 and *** p < 0.001; NS, no significance).
The Proliferative Function of STIL Was Confirmed by the Xenotransplantation Model
To demonstrate the function of STIL in vivo, we constructed a xenograft model. STIL +/+ , STIL −/− , 2#, and STIL −/− 21# cells were implanted into nude mice. On the 26th day, tumors were removed from the nude mice. The solid tumor volume of STIL +/+ was significantly larger than that of STIL −/− 2# and STIL −/− 21# ( Figure 5A-C). In addition, through dynamic observation of tumor growth, it was found that the tumor growth of the STIL knockout cells was slowed (Figure5D). These results suggested that STIL knockout has a dramatic effect on the regulation of BC tumorigenesis in vivo.
STIL Knockout Down-regulated the PI3K/AKT/mTOR/c-myc Signaling Pathway
To further disclose the potential mechanism behind STIL-driven malignant behaviors in bladder cancer, we analyzed the differentially expressed genes in STIL-deficient UMUC3 cells and wild-type UMUC3 cells. The results of RNA-seq showed that the PI3K/AKT/mTOR/c-myc signaling pathways were significantly decreased in STIL-deficient cells compared to the wild-type, as found through GSEA analysis based on KEGG pathways ( Figure 6A-C). We also checked the RNA sequencing of 412 BC patients from the TCGA database and divided the patients into a high-expression group and a low-expression group according to median STIL expression. We found that PI3K/AKT/mTOR/cmyc signaling pathways were significantly enriched in the high-expression group ( Figure 6D,E). Our RNA-seq analysis results showed that the downstream molecules of the transcription factor c-myc (including MCM4, HSPD1, EIF2S2, DDX21, DDX18, CBX3, ABCE1, etc.) were significantly down-regulated in BC cells after STIL knockout ( Figure 6F). To
STIL Knockout Down-Regulated the PI3K/AKT/mTOR/c-myc Signaling Pathway
To further disclose the potential mechanism behind STIL-driven malignant behaviors in bladder cancer, we analyzed the differentially expressed genes in STIL-deficient UMUC3 cells and wild-type UMUC3 cells. The results of RNA-seq showed that the PI3K/AKT/mTOR/c-myc signaling pathways were significantly decreased in STIL-deficient cells compared to the wild-type, as found through GSEA analysis based on KEGG pathways ( Figure 6A-C). We also checked the RNA sequencing of 412 BC patients from the TCGA database and divided the patients into a high-expression group and a low-expression group according to median STIL expression. We found that PI3K/AKT/mTOR/c-myc signaling pathways were significantly enriched in the high-expression group ( Figure 6D,E). Our RNA-seq analysis results showed that the downstream molecules of the transcription factor c-myc (including MCM4, HSPD1, EIF2S2, DDX21, DDX18, CBX3, ABCE1, etc.) were significantly down-regulated in BC cells after STIL knockout ( Figure 6F). To confirm this result, we verified by RT-qPCR, showing that the expression of these molecules was significantly decreased after STIL knockout ( Figure 6G). To further demonstrate the relationship between STIL, the PI3K/AKT/mTOR pathway, and c-myc in BC, we performed an immunoblotting analysis. PI3K/AKT/mTOR protein levels did not fluctuate significantly after STIL knockout, but p-PI3K/p-Akt/p-MTOR/c-myc protein levels significantly decreased in STIL knockout BC cells ( Figure 6H,I).
confirm this result, we verified by RT-qPCR, showing that the expression of these molecules was significantly decreased after STIL knockout ( Figure 6G). To further demonstrate the relationship between STIL, the PI3K/AKT/mTOR pathway, and c-myc in BC, we performed an immunoblotting analysis. PI3K/AKT/mTOR protein levels did not fluctuate significantly after STIL knockout, but p-PI3K/p-Akt/p-MTOR/c-myc protein levels significantly decreased in STIL knockout BC cells ( Figure 6H,I).
SC79 Treatment Partially Reversed the Effects of STIL Knockout on Cell Tumorigenesis
To further verify the relation between PI3K/AKT/mTOR/c-myc signaling pathway and STIL-driven malignant behaviors in BC, we activated the AKT pathway in STIL +/+ and STIL −/− cell lines with SC79 (AKT activating agent). After activation of the PI3K/AKT/mTOR pathway by SC79, the STIL-driven inhibition of migration and proliferation was partially reversed, as shown by colony formation assays (Figures 7A-C and S3A-C). Next, we performed a transwell migration assay; the results of this assay were similar to the colony formation experiment ( Figures 7D,F and S3D,F). In addition, the soft agar assay also showed that SC79 increased cell proliferation in STIL −/− cells more than in STIL +/+ cells ( Figures 7E,G and S3E,G). Effectively, after activation of the PI3K/AKT/mTOR pathway, the elevated tumorigenesis levels in STIL knockout BC cells were partially restored.
SC79 Treatment Partially Reversed the Effects of STIL Knockout on Cell Tumorigenesis
To further verify the relation between PI3K/AKT/mTOR/c-myc signaling pathway and STIL-driven malignant behaviors in BC, we activated the AKT pathway in STIL +/+ and STIL −/− cell lines with SC79 (AKT activating agent). After activation of the PI3K/AKT/mTOR pathway by SC79, the STIL-driven inhibition of migration and proliferation was partially reversed, as shown by colony formation assays (Figures 7A-C and S3A-C). Next, we performed a transwell migration assay; the results of this assay were similar to the colony formation experiment ( Figures 7D,F and S3D,F). In addition, the soft agar assay also showed that SC79 increased cell proliferation in STIL −/− cells more than in STIL +/+ cells ( Figures 7E,G and S3E,G). Effectively, after activation of the PI3K/AKT/mTOR pathway, the elevated tumorigenesis levels in STIL knockout BC cells were partially restored.
Discussion
Previous studies have shown that STIL causes abnormal centriole expansion, which, in turn, leads to chromosomal instability [7]. Chromosomal instability is the hallmark of many cancers [8]. STIL is considered to be an oncogene whose expression is elevated in many types of cancers [9][10][11][12][13]. However, there is little in the literature on the question of STIL in BC. Our study found that the mRNA expression levels of STIL were significantly elevated and were positively correlated with the cycle-related gene (CCNB1, CDK1, CCNA2, CCNB2, and CCNE2) in BC. In addition, BC patients with high expression of STIL had poor prognoses. Demonstrated by microarray detection and immunohistochemistry, we confirmed an increase in STIL expression in clinical specimens. Previous studies have suggested that STIL is critical for cancer cell migration [19]. Our experimental results show that STIL has similar results in BC: STIL knockout inhibited BC cells proliferation and migration in vitro and blocked proliferation of tumor in vivo. These results strongly showed that STIL is related to the development and occurrence of BC. STIL is a cell-proliferation-related gene involved in cell cycle regulation [20]. We analyzed cell proliferation by detecting cell proliferation markers (EDU, Ki67), as well as observing changes in cell cycle and cycle-related proteins.
According to previous studies, the ability of the G0/G1 phase to enter the S phase is mainly determined by CDKs, including CDK2/4/6 [21,22]. Cyclin D and CDK4/6 are highly related and readily form conjugates. By activating CDK2, the cyclin D-CDK4/6 complex promotes DNA replication and cell proliferation [22]. In many tumors, reductions in CDK2/4/6 and cyclin D1 often coexist with G0/G1 cell cycle arrest [23][24][25]. And some studies have shown that STIL is closely related to G1 phase arrest [26]. A similar phenomenon emerged in our study: STIL knockout resulted in cell cycle arrest in the G0/G1 phase in BC cells. At the protein level, CDK2/4/6 and cyclin D1 were decreased in STIL knockout BC cells. STIL knockout showed different blockade phases in different tumor types, and reduced the proliferation of cervical and colon cancer cells by inhibiting Cyclin B1/CDK1, which, in turn, induced cell cycle arrest in the G2/M phase [11,27,28].
To explore the effects of STIL on proliferation in vivo, we used a xenograft model and confirmed that STIL knockout inhibited the growth of xenograft tumor cells. Our RNA sequencing results and gene set enrichment analysis (GSEA) showed a significant decrease in PI3K/AKT/MTOR pathway enrichment and c-myc target enrichment after STIL knockout. Heat map analysis showed that the downstream molecules of c-myc were significantly down-regulated in STIL knockout BC cells. RT-qPCR and Western blotting were used to confirm that the downstream p-PI3K/p-AKT/p-mTOR/c-myc pathway was down-regulated after STIL knockout. These results suggest that STIL plays a cycle-related gene vital role in BC through the PI3K/AKT/mTOR pathway and c-myc. Significant research has shown that the PI3K/AKT signaling pathway plays a significant role in promoting invasion, migration, proliferation, and other malignant characteristics of human cancers [29,30]. C-myc, as a proto-oncogene, is fundamental to regulating cell proliferation [31,32]. Several studies have documented that the activation of c-myc is mediated by the AKT/mTOR pathway [33][34][35]. Furthermore, c-myc is up-regulated in many tumors and is important for proliferation [36,37]. Additionally, c-myc is crucial to the cell cycle progression of tumor cells [38]. C-myc acts as a transcription factor to stimulate cell cycle progression and cell proliferation [39,40]. It has been proposed that c-myc regulates cell cycle progression through CDK2/4 and cyclin D1 [41,42]. These previous studies are consistent with our findings. Thus, we conclude that STIL reduces c-myc through the PI3K/AKT/ mTOR signaling pathway, leading to a decrease in CDK2/4/6 and cyclin D1, thereby causing cell cycle arrest in the G0/G1 phase and ultimately inhibiting cell proliferation. Therefore, in the future, it may be possible to inhibit tumor proliferation by inhibiting STIL, reducing c-myc, and arresting the G0/G1 phase. Additionally, PI3K/AKT/mTOR signal transduction is modulated by SC79 in BC cells, which significantly reduces the down-regulation of STIL knockout. We proved that changes in STIL drive fluctuations in c-myc, which affect the cell cycle. However, we still do not know the direct target of STIL; we will address this question in future research.
In summary, we are the first to confirm that STIL may promote development in BC. Our results suggest that STIL is strongly associated with prognosis in BC patients and mediates c-myc through the PI3K/AKT/ mTOR signaling pathway, ultimately promoting the BC cell cycle, proliferation, invasion, and metastasis. These findings may shed new light on the role of STIL in human malignancies and provide new targets for treating BC.
Conclusions
STIL enhanced the PI3K/AKT/mTOR pathway, resulting in increased expression of c-myc, ultimately promoting BC occurrence and progression. These results indicate that STIL may be a potential target for treating BC.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/cancers14235777/s1, Table S1: real-time quantitative PCR primer sequences. Figure S1. In the T24 cell line, knockout of STIL inhibited BC tumorigenesis in vitro. Figure S2. In the T24 cell line, STIL played a vital role in the proliferation and cell cycle of BC. Figure S3. In the T24 cell line, SC79 treatment partially reversed the effect of STIL knockout on cell tumorigenesis. Figure S4. STIL and some other genes mRNA expression was significantly upregulated in BBN induced bladder cancer tissues, compared with control samples. The original WB can be found in File S1.
Data Availability Statement:
The data presented in this study are openly available in GEO, reference number GSE211756. The data can be found here: https://www.ncbi.nlm.nih.gov/geo/query/acc. cgi?acc=GSE211756.
Conflicts of Interest:
The authors declare no conflict of interest.
|
v2
|
2022-11-26T17:25:20.446Z
|
2022-11-24T00:00:00.000Z
|
253913613
|
s2orc/train
|
Endothelial cell‑derived connective tissue growth factor stimulates fibroblast differentiation into myofibroblasts through integrin αVβ3
Connective tissue growth factor (CTGF) is expressed at high levels in blood vessels, where it functions as a regulator of a number of physiological processes, such as cell proliferation, angiogenesis and wound healing. In addition, CTGF has been reported to be involved in various pathological processes, such as tumor development and tissue fibrosis. However, one of the main roles of CTGF is to promote the differentiation of fibroblasts into myofibroblasts, a process that is involved in disease progression. Therefore, the present study aimed to investigate the possible mechanism by which pathological changes in the microvasculature can direct the activation of fibroblasts into myofibroblasts in the context of hypoxia/reoxygenation (H/R). Human umbilical vein endothelial cells (HUVECs) and normal human dermal fibroblasts were used in the present study. The expression levels of CTGF were determined by western blot analysis and reverse transcription-semi-quantitative PCR. To analyze the paracrine effect of HUVECs on fibroblasts, HUVECs were infected with CTGF-expressing adenovirus and then the culture supernatant of HUVECs was collected to treat fibroblasts. The formation of α-smooth muscle actin (α-SMA) stress fibers in fibroblasts were observed by immunofluorescence staining. It was found that H/R significantly increased CTGF expression in HUVECs. CTGF was also able to directly induce the differentiation of fibroblasts into myofibroblasts. In addition, the culture supernatant from CTGF-overexpressing HUVECs stimulated the formation of α-SMA stress fibers in fibroblasts, which was inhibited by treatment with a functional blocking antibody against integrin αVβ3 and to a lesser degree by a blocking antibody against α6 integrin. The mechanism of CTGF upregulation by H/R in HUVECs was then evaluated, where it was found that the CTGF protein was more stable in the H/R group compared with that in the normoxic control group. These findings suggest that CTGF expressed and secreted by vascular endothelial cells under ischemia/reperfusion conditions can exert a paracrine influence on neighboring fibroblasts, which may in turn promote myofibroblast-associated diseases. This association may hold potential as a therapeutic target.
Introduction
Fibroblasts are tissue-resident stromal cells that are important for maintaining the structural integrity of tissues (1). They function to synthesize and integrate structural proteins, such as collagen and elastin, into the extracellular matrix (ECM) of connective tissues (1). In environments where this homeostasis is disturbed, such as during wound healing, chronic inflammation and cancer, fibroblasts are activated to proliferate and upregulate ECM production (2,3). Activated fibroblasts acquire various smooth muscle features, including enhanced formation of contractile stress fibers and expression of α-smooth muscle actin (α-SMA) (2,3). Cells with these characteristics are known as myofibroblasts (2,3). Although transient acquisition of this myofibroblast phenotype confers beneficial effects on normal tissue repair processes, persistence of myofibroblasts is associated with the development of diseases mediated by tissue stiffening and deformation (4). Stiff scar tissues adversely alters normal organ function (4). In addition, fibrosis is characterized by the abnormally excessive accumulation of ECM proteins, which contributes to organ failure in various chronic diseases affecting the liver, kidney, skin, lungs and the heart (5). By contrast, activated fibroblasts, especially cancer-associated fibroblasts (CAFs) in the tumor stroma, serve an important role in tumorigenesis by stimulating angiogenesis, cancer cell proliferation and invasion (6). Activated CAFs can also produce a variety of growth factors and proinflammatory cytokines, such as TGF-β, vascular endothelial growth factor, IL-6 and CXC-chemokine ligand 12, to promote tumor progression (6)(7)(8). CAFs have been reported to contribute to ECM remodeling and cancer cell invasion by secreting connective tissue growth factor (CTGF), collagen, fibronectin and elastin (9), implicating CAFs to be targets for anti-cancer therapy (6)(7)(8).
CTGF, also known as cellular communication network factor 2, is a regulatory protein that has been demonstrated to be involved in several biological processes, such as cell proliferation, angiogenesis and wound healing (10). In addition, CTGF has been associated with a number of pathological processes, such as tumor development, cardiovascular diseases, inflammatory diseases and tissue fibrosis in major organs (10). CTGF was first discovered as a protein secreted by endothelial cells during angiogenesis under normal conditions (11). CTGF expression is generally higher in blood vessels compared with that in other organs or tissues (11). CTGF mRNA is expressed at particularly high levels in developing blood vessels and in the large blood vessels of the adult heart, suggesting that CTGF is involved in the development and maintenance of blood vessels (12). However, one of the main roles of CTGF is considered to be the promotion of myofibroblast differentiation and angiogenesis (13)(14)(15). CTGF is typically secreted into the extracellular environment, where it interacts with cell surface receptors, growth factors and the ECM (13)(14)(15). Subsequently, CTGF mediates downstream effects by binding to heterodimeric cell-surface integrin complexes, such as α 6 , β 1 , α V and β 3 integrins (13)(14)(15).
The present study aimed to investigate whether CTGF from endothelial cells after hypoxia/reoxygenation (H/R) could stimulate the differentiation of neighboring fibroblasts into myofibroblasts in a paracrine manner.
Materials and methods
Cell culture conditions. HUVECs (passages 4-10; Lonza Group, Ltd.) were cultured in EGM™-2 Endothelial Cell Growth Medium-2 BulletKit™ (Lonza Group, Ltd.) containing all the included supplements at 37˚C in a humidified atmosphere containing 5% CO 2 . Normal human dermal fibroblasts (PromoCell GmbH) were cultured in DMEM (Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (Thermo Fisher Scientific, Inc.) at 37˚C in a humidified atmosphere containing 5% CO 2 . For the H/R conditions, the cells were first incubated at 37˚C for 16 h in a hypoxic incubator (Thermo Scientific 3131 Forma Incubator; Thermo Fisher Scientific, Inc.) filled with 1% O 2 and 5% CO 2 , balanced with N 2 , before being placed under normoxic conditions at 37˚C for 2 h for reoxygenation treatment. For the preparation of conditioned media (CM), the medium of HUVECs was changed with EBM™-2 Basal Medium (Lonza Group, Ltd.) containing 1% FBS at 37˚C for 18 h before it was collected, and filtered using a Minisart ® Syringe Filter (0.25 µm; Sartorius AG).
Vector construction. The coding region of human CTGF (NM_001901) was PCR-amplified from HUVEC cDNA (1 µl) using TaKaRa Ex Taq ® DNA Polymerase (cat. no. RR01AM; Takara Bio, Inc.) according to the manufacturer's instructions, cloned into the pGEM ® -T Easy Vector Systems (cat. no. A1360; Promega Corporation), before being subcloned into the pShuttle-CMV vector in the AdEasy adenoviral vector system (cat. no. 240009; Agilent Technologies, Inc.) to produce CTGF-expressing adenovirus. The primers are as follows: cdsCTGF forward, 5'-GAGTCGACAGTGCCAACCATGACCGC-3' (nucleotides plus SalI adapter) and cdsCTGF reverse, 5'-GACTCGAGCTGGCTTCATGCCATGTC-3' (nucleotides plus XhoI adapter). PCR was performed for 27 cycles at 94˚C for 1 min, 60˚C for 1 min and 72˚C for 1 min. All PCR-amplified fragments and cloning junctions were verified by DNA sequencing performed by SolGent Co., Ltd. Adenoviral CTGF cloning was performed according to the manufacturer's protocols. Production and harvesting of adenoviruses were performed as described (16,17). The pShuttle-CMV vector containing CTGF (1 µg) was cotransformed with pAdEasy-1 vector (100 ng) into BJ5183 competent cells (20 µl; supplied in the in the AdEasy adenoviral vector system), where homologous recombination occurred. The recombinant adenoviral vector expressing human CTGF was transfected into 293A cells to obtain viral particles. 293A cells (cat. no. R70507; Thermo Fisher Scientific, Inc.) were cultured in DMEM (Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Thermo Fisher Scientific, Inc.) and 1% penicillin/streptomycin (Thermo Fisher Scientific, Inc.) at 37˚C. 293A cells (5x10 6 cells/100 mm dish) were plated 24 h before transfection. Subsequently, 10 µg recombinant adenoviral vector DNA was used for transfection with MetaFectene PRO (cat. no. T040-1.0; Biontex Laboratories GmbH) according to the manufacturer's instruction. Transfected cells were incubated at 37˚C for 7-10 days, the adenovirus-producing 293A cells were collected and the virus particles were purified. The infection into HUVECs was performed as previously described (18). The harvested adenoviruses (25 MOI) were added to cells in endothelial basal medium (Lonza Group, Ltd.) containing 1% FBS at 37˚C for 4 h, then virus-containing medium was removed and growth medium was added. After 24 h, medium was removed and the cells were washed one time with endothelial basal medium, and then incubated in endothelial basal media containing 1% FBS at 37˚C for 18 h before collecting CM.
Control HUVECs were infected with a control adenovirus generated with control shuttle vector (pShuttle-CMV-lacZ) and pAdEasy-1 vector.
Statistical analysis. All data are expressed as the mean ± standard error of the mean from three or four independent experiments. All of the significance analysis was performed using the SigmaPlot version 14.0 software (SPSS, Inc.). The statistical differences were compared using one-way ANOVA followed by the Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.
CTGF treatment causes the differentiation of fibroblasts into myofibroblasts.
It was reported in our previous study that endothelial cells undergo endothelial-to-mesenchymal transition (EndMT) when subjected to ischemia/reperfusion, influencing neighboring fibroblasts to actively participate in cardiac fibrosis (23). Although it was found in this previous study that CTGF from EndMT cells has a paracrine influence on fibroblast activation (23), the direct effects of CTGF on the activation of fibroblasts into myofibroblasts or its mechanism were not examined. Therefore, in the present study, the effects of hypoxia/reoxygenation (H/R) on CTGF expression in HUVECs was first assessed (Fig. 1A). It was found that H/R significantly increased the expression of the CTGF protein (Fig. 1A). Culture supernatants from both normoxic and H/R HUVECs were then obtained and were used to treat fibroblasts (Fig. 1B). The immunofluorescence of α-SMA stress fibers was examined, which indicates the generation of myofibroblasts (2). Fibroblasts treated with CM from normoxic HUVECs (N CM) showed punctate or patchy α-SMA immunofluorescence. By contrast, the treatment of fibroblasts with CM from H/R HUVEC (H/R CM) led to the formation of more intense and fibrous α-SMA immunofluorescence, typical of stress fibers (Fig. 1B). The effect of soluble CTGF on fibroblast differentiation was then examined using rhCTGF (Fig. 2). Fibroblasts were treated with 0.5 or 1 µg/ml rhCTGF, before α-SMA immunofluorescence was observed. Treatment with rhCTGF dose-dependently stimulated the formation of α-SMA stress fibers in fibroblasts, indicating that rhCTGF directly induced the differentiation of fibroblasts into myofibroblasts.
Function-blocking antibody against integrin α V β 3 abolishes the effect of CM from CTGF-overexpressing HUVECs on α-SMA fiber formation.
To investigate how CTGF from endothelial cells affects neighboring fibroblasts, function-blocking antibodies against integrin α 6 and α V β 3 were used alongnside CM from CTGF-overexpressing HUVECs (Fig. 3). CTGF mediates downstream effects by binding to integrins, such as α 6 , β 1 , α V and β 3 (15). HUVECs were first infected with a control adenovirus (Ad-mock) or an adenovirus expressing human CTGF (Ad-CTGF), before CTGF overexpression was confirmed using western blot analysis (Fig. 3A). HUVEC CM was then collected and used to treat fibroblasts and immunofluorescence staining for α-SMA was performed (Fig. 3B). CM from CTGF-overexpressing HUVECs potently stimulated α-SMA stress fiber formation, which was significantly inhibited by a function-blocking antibody against integrin α V β 3 (Fig. 3B). The function-blocking antibody against integrin α 6 could not block stress fiber formation ( Fig. 3B and C), suggesting that CTGF from endothelial cells stimulates the differentiation of fibroblasts to myofibroblasts through integrin α V β 3 .
CTGF protein stability is increased under H/R in HUVECs.
Subsequently, the mechanism underlying CTGF upregulation by H/R in HUVECs was evaluated (Fig. 4). The CTGF mRNA level was first examined to test whether H/R could affect the transcription of CTGF. However, CTGF mRNA was not changed under H/R conditions (Fig. 4A), although CTGF protein expression was markedly increased by H/R (Fig. 4B). This suggests that increased CTGF protein expression was not due to any changes in CTGF mRNA levels under H/R conditions. The protein synthesis of CTGF by H/R was therefore Figure 3. CM from CTGF-overexpressing HUVECs stimulates α-SMA stress fiber formation through α V β 3 -integrin in fibroblasts. (A) CTGF protein expression was increased in HUVECs infected with Ad-CTGF. CTGF overexpression was confirmed using western blot analysis. Ad-mock HUVECs were infected with a control Ad containing a control shuttle vector (pShuttle-CMV-lacZ). (B) Function-blocking Int α V β 3 diminished the effects of CTGF CM on α-SMA fiber formation (green) in fibroblasts. Intα 6 , Int α V β 3 or purified mouse IgG (each 10 µg/ml) was added to the CM directly. Nuclei were stained with DAPI (blue). Magnification, x400. (C) Quantification of the α-SMA + area (n=5-6 per group). *** P<0.001 vs. mock CM. ### P<0.001 vs. CTGF CM only. CM, conditioned medium; Ad, adenovirus; CTGF, connective tissue growth factor; IgG, immunoglobulin G; neu, neutralizing; Ab, antibody; Intα 6 , anti-integrin α 6 antibody; Int α V β 3 , anti-integrin α V β 3 antibody; SMA, smooth muscle actin. tested after blocking protein degradation with the proteasome inhibitor MG132 (Fig. 4B). No significant differences in the protein synthesis of CTGF was detected between the normoxia and H/R groups with the presence or absence of MG132. Changes in CTGF stability following H/R was then examined using CHX, a protein translation blocker (Fig. 5). Notably, CTGF stability was significantly higher in the H/R group compared with that in the normoxic control group at 30 and 60 min (Fig. 5).
Discussion
The tumor microenvironment can contain fibroblasts, immune cells, blood vessels and the ECM (8). Fibroblasts are typically quiescent but can be activated into myofibroblasts during the wound-healing response (6). In addition, CAFs can directly regulate cancer cell proliferation, tumor immunity, angiogenesis, ECM remodeling and metastasis, suggesting that they can be a target for anti-cancer therapy (8). Several preclinical studies have reported CAFs to be possible targets for anti-cancer therapy in lymphoma, Lewis lung cancer, melanoma and gastrointestinal cancer (6)(7)(8). α-SMA is a marker that can be used to reflect the myofibroblast population of CAFs, such that docetaxel-conjugate nanoparticles have been shown to target α-SMA + stromal-suppressed metastases in a mouse model of breast cancer (24). Furthermore, selective depletion of myofibroblasts has been documented to attenuate angiogenesis in a pancreatic ductal adenocarcinoma mouse model. However, depletion of α-SMA + myofibroblasts in mouse pancreatic cancer can also increase the population of immunosuppressive CD4 + Foxp3 + Tregs, leading to more invasive tumors (25). In the present study, CM from HUVECs under H/R conditions, in addition to that from CTGF-overexpressing HUVECs, was found to activate fibroblasts into α-SMA + myofibroblasts. This suggests that blood vessels can promote neighboring fibroblasts into differentiating into myofibroblasts. A clinical trial of bevacizumab targeting endothelial cell precursors with CAFs has previously been conducted; the addition of bevacizumab to the standard of care significantly improved overall survival in malignant pleural mesothelioma (26). Although targeting α-SMA + myofibroblasts therapeutically remains to be a controversial topic (24,25), targeting only CAFs, CAFs with endothelial cells or other types of cells is a promising strategy (26). However, future studies are required to define this strategy more precisely.
CTGF is mainly secreted from endothelial cells and can modulate complex biological processes during normal embryonic development and tissue repair (10). Abnormal CTGF expression profiles have been observed in several diseases, including tissue fibrosis (of the lung, heart and liver), systemic sclerosis and tumors in major organs (23,27). It has been previously reported that >30 types of human cancers are associated with the dysregulated aberrant expression of CTGF (27). Higher CTGF expression is associated with more aggressive inflammatory colorectal cancer, whilst CTGF expression has also been found to be increased in breast cancer, chondrosarcoma and glioma (10,27). By contrast, CTGF can also function as a tumor suppressor. CTGF expression has been observed to be reduced in non-small cell lung cancer cells, where decreased CTGF expression may be involved in lung tumorigenesis (10,28). In the present study, CTGF from HUVECs stimulated fibroblast differentiation, suggesting a possible association of HUVECs with CAF generation. Further investigations on the functional roles of CTGF-expressing endothelial cells and myofibroblasts in tumors or ischemic diseases are required.
CTGF can bind to several types of receptors, such as integrins, heparan sulfate proteoglycans, lipoprotein receptor-related proteins and tyrosine kinase receptors (29). However, integrins are known to be the principal CTGF receptors (29). CTGF mediates downstream effects through α 6 , β 1 , α V and β 3 integrins (15). Physiologically, CTGF enhances the lactogenic differentiation of mammary epithelial cells by binding to integrins α 6 and β 1 , and to a lesser degree β 3 integrin (15). In addition, CTGF can activate β 1 integrin signaling in primary skin fibroblasts (30) and pancreatic stellate cells through α V β 3 (13,31). In the present study, CM from CTGF-overexpressing endothelial cells stimulated fibroblast differentiation, which was inhibited by a function-blocking antibody against integrin α V β 3 , but not by a function-blocking antibody against integrin α 6 .
The present study has a number of limitations. Although it was demonstrated that H/R increased CTGF expression in HUVECs and had a direct effect on the differentiation of fibroblasts to myofibroblasts through integrin α V β 3 , the mechanism of integrin-mediated fibroblast differentiation by H/R endothelial CM was not explored. Integrin-linked kinase (ILK) is a key mediator of integrin signaling that interacts with the cytoplasmic domain of β-integrins (32). Therefore, ILK may be a downstream candidate in this case. It has previously been reported that integrin α V β 3 is involved in the stress fiber formation through signaling molecules, such as focal adhesion kinase (FAK), PKCα and RhoA (33,34). In human aortic smooth muscle cells, osteoprotegerin, a ligand for integrin α V β 3 , mediated the phosphorylation of FAK and actin cytoskeleton reorganization (34). Integrin α V β 3 triggers the formation of focal adhesions and stress fibers through the activation of the transforming protein RhoA in astrocytes (33). Another limitation of the present study is that the binding of CTGF onto integrin α V β 3 on fibroblasts was not confirmed. CTGF-overexpressing HUVEC CM facilitated the differentiation of fibroblasts into myofibroblasts, which was inhibited by a functional blocking antibody against integrin α V β 3 . Therefore, it is highly likely that CTGF will bind to α V β 3 . However, more specific assays, such as a CTGF adhesion assay (15), are required to investigate the possible interaction between CTGF and integrin α V β 3 . The previous report by Morrison et al (15) demonstrated the interaction between CTGF and integrin complexes using an adhesion assay. HC11 epithelial cells adhered onto CTGF-coated wells, where function-blocking antibodies against both α 6 and β 1 interrupted this CTGF-mediated epithelial cell adhesion (15). Further studies are required to determine the underlying mechanism(s) of CTGF/integrin α V β 3 -mediated fibroblast differentiation, especially the stress fiber formation in fibroblasts.
Furthermore, another limitation of the present study is that the proteasomal degradation and stability of CTGF was analyzed using cellular homogenates. Since CTGF is a secreted protein, it is not sufficient to analyze proteasomal degradation and stability using cell lysates. Interestingly, whilst CTGF is a secreted protein, CTGF can also strongly bind to heparin and other matrix components, rendering it detectable in the supernatants or in cellular homogenates, depending on the cell type investigated (35,36). Although CTGF (38 kDa) was readily detected in cell lysates, it was not detectable in the conditioned medium (35). Upon stimulation of CTGF with serotonin, enhanced levels of CTGF protein were detected in the cellular homogenates, whereas no protein was detectable in cell culture supernatants (37). Therefore, the regulatory mechanisms associated with proteasomal degradation or stability of CTGF, especially secreted CTGF, under H/R and additional conditions require additional in-depth investigations.
To conclude, the supernatants of CTGF-overexpressing HUVECs stimulated fibroblast differentiation, which was significantly inhibited by a function-blocking antibody against integrin α V β 3 . These findings suggest that communication between CTGF-secreting endothelial cells and neighboring fibroblasts can lead to the development of myofibroblastassociated diseases, which may be a potential therapeutic target for the treatment of cancer or ischemic diseases.
|
v2
|
2022-11-26T17:28:33.936Z
|
2022-11-24T00:00:00.000Z
|
253903116
|
s2ag/train
|
No evidence that spice consumption is a cancer prevention mechanism in human populations
Why humans historically began to incorporate spices into their diets is still a matter of unresolved debate. For example, a recent study (Bromham et al. 2021, Nat Hum Behav) did not support the most popular hypothesis that spice consumption was a practice favoured by selection in certain environments to reduce food poisoning, parasitic infections, and foodborne diseases. Because several spices are known to have anticancer effects, we explored, using the same dataset, the hypothesis that natural selection and/or cultural evolution may have favoured spice consumption as an adaptive prophylactic response to reduce the burden of cancerous pathologies. Patterns of spice use in 36 countries, however, are not consistent with a cancer mitigation mechanism: the age-standardised rate of almost all gastrointestinal cancers was not related to spice consumption. Thus, directions other than foodborne pathogens and cancers should be explored to understand the health reasons, if any, why our ancestors developed a taste for spices.
|
v2
|
2022-11-27T16:49:23.671Z
|
2022-11-24T00:00:00.000Z
|
254000441
|
s2orc/train
|
Prognostic Hematologic Biomarkers Following Immune Checkpoint Inhibition in Metastatic Uveal Melanoma
Simple Summary Uveal melanoma is a rare form of melanoma but is the most common tumor in the eye. Despite having effective treatments for the initial tumor, many patients experience the spread of their cancer to distant body sites. There is no uniform way of treating metastatic disease, but physicians often use therapies that harness the patient’s immune system because these same treatments have been very effective in other types of melanoma. Not all patients respond to this therapy though, and some develop toxicity related to the treatment. The goal of this paper was to identify features or blood markers that may help determine response to treatment early. Specifically, we analyzed a molecule called lactate dehydrogenase (LDH) and the ratio of different white blood cells at the start of therapy and 2 months after treatment was started. We found that these non-invasive blood markers could be useful in determining which patients are responding to treatment. Abstract Background: There is no standardized treatment for metastatic uveal melanoma (MUM) but immune checkpoint inhibitors (ICI) are increasingly used. While ICI has transformed the survival of metastatic cutaneous melanoma, MUM patients do not equally benefit. Factors known to affect ICI response include the hematologic markers, lactate dehydrogenase (LDH) and neutrophil:lymphocyte ratio (NLR). We evaluated the prognostic value of LDH and NLR at the start of ICI and on treatment in MUM. Methods: MUM patients were treated between August 2006 and May 2022 with combination ipilimumab/nivolumab or ipilimumab/nivolumab/pembrolizumab single-agent therapy. Univariable (UVA) and multivariable (MVA) analyses were used to assess the prognostic value of predefined baseline factors on progression-free (PFS) and overall survival (OS). Results: In forty-six patients with MUM treated with ICI, elevated baseline and on-treatment LDH was prognostic for OS (start of ICI, HR (95% CI): 3.6 (1.9–7.0), p < 0.01; on-treatment, HR (95% CI): 3.7 (1.6–8.8), p < 0.01) and PFS (start of ICI, (HR (95% CI): 2.8 (1.5–5.4), p < 0.0001); on-treatment LDH (HR (95% CI): 2.2 (1.1–4.3), p < 0.01). On-treatment NLR was prognostic for PFS (HR (95% CI): 1.9 (1.0–3.9), p < 0.01). On-treatment LDH remained an important contributor to survival on MVA (OS: HR (95% CI): 1.001 (1.00–1.002), p < 0.05); PFS: HR (95% CI): 1.001 (1.00–1.002), p < 0.01). Conclusions: This study demonstrates that LDH and NLR could be useful in the prognostication of MUM patients treated with ICI. Additional studies are needed to confirm the importance of these and other prognostic biomarkers.
Introduction
Uveal melanoma (UM) is the most common intraocular malignancy [1]. The current primary tumor treatments-plaque radiotherapy and enucleation-provide 95-99% local tumor control. Despite good local control, systemic prognosis remains poor, as nearly half of patients die from metastatic disease within 10 years of their original diagnosis [2][3][4] To date there is no standardized treatment algorithm for metastatic UM (MUM). Approximately 90% of UMs harbor driver mutations in GNAQ/GNA11, which are genes that code for Gα proteins that mediate multiple downstream signaling cascades including the mitogen activated kinase pathway (MAPK). The MAPK pathway and its components have been the focus in the development of targeted therapies in cutaneous melanoma but when applied to patients with MUM, no survival benefit has yet been shown [5,6]. For example, when compared to chemotherapy, selumetinib, a MAP/ERK kinase (MEK) inhibitor, did not show a statistically significant improvement in overall survival (OS) (11.8 vs. 9.1 months) [7]. This drug has also been studied in combination with the cytotoxic agent, dacarbazine, in a phase III randomized controlled trial. Similarly, no significant difference in progression-free survival (PFS) or objective response rate (3 vs. 0%) was seen [8].
Based on their effectiveness in advanced stage cutaneous melanoma, immune checkpoint inhibitors (ICI), such as those blocking anti-cytotoxic T-lymphocyte-associated protein 4 (anti-CTLA-4) and/or programmed cell death-1 (PD-1)/programmed cell death ligand-1 PD-L1) are widely used in the treatment of MUM [9][10][11]. While these agents have revolutionized patient outcomes for advanced stage cutaneous melanoma as well as improved outcomes in mucosal melanoma, patients with MUM do not derive equal benefit. One study prospectively evaluated the use of pembrolizumab as a first-line therapy. Overall survival (OS) for patients who derived objective clinical benefit was 12.8 months, which is similar to other agents [12]. A small phase II study evaluating combination ipilimumab 3 mg/kg and nivolumab 1 mg/kg reported an overall response rate (ORR) of 18% including one complete response (CR) and a median OS of 19.1 months [13]. Forty percent of patients in this trial experienced a grade 3-4 treatment-related adverse event, which is expected for this treatment regimen and is consistent with other reports [14][15][16]. These high toxicity rates highlight the importance of identifying patients who will benefit from ICI.
Features thought to positively influence response to ICI include high tumor mutational burden, high expression of PD-1, low lactate dehydrogenase (LDH), and the absence of liver metastases (LM)-features not typically present in uveal melanoma patients [17][18][19][20][21][22][23][24][25]. In fact, the first and most common site of metastasis for patients with UM is the liver with >90% of patients developing tumors in this anatomic location [26]. Liver-directed therapies are commonly used in the treatment of MUM patients with limited LM. For patients with resectable disease, liver resection has shown a survival benefit. One retrospective study reported a post operative median OS of 14 months for all patients; if complete resection was possible the median OS increased to 27 months [27]. Other regional liver directed therapies like transarterial chemoembolization (TACE), hepatic artery infusion (HAI), or selective internal radiotherapy result in a broad range of survival outcomes [28]. A separate single institution retrospective study reported that local therapy (surgery or intrahepatic chemotherapy) correlated with prolonged survival on univariate and multivariate analyses (median OS 32.4 months) [29].
Another feature thought to be prognostic and predictive of ICI response across a variety of tumor types, including metastatic melanoma, is an elevated ratio of neutrophils to lymphocytes (NLR) at the start of therapy [30][31][32][33]. The change in NLR in response to ICI has also been linked to early response to therapy [34,35]. Despite its proven utility in other cancer types, there is only one published study that assessed NLR in patients with metastatic uveal melanoma and, importantly, this was not in the context of immunotherapy [36]. While there has been some effort to understand and identify features that impact survival in MUM, there are few studies that assess these factors in patients receiving ICI [37]. Furthermore, even fewer or these studies were conducted in the first-line setting, as most evaluated ICI agents as a salvage therapy.
To gain insight into factors that may be prognostic for response to ICI in MUM we conducted a real-world review and retrospective analysis of all MUM patients who were treated with either single-agent or combination ICI. We first evaluated the established patient demographics, primary tumor features, and treatment characteristics in the prognostication of MUM patients and extended this analysis to the hematologic biomarkers, LDH and NLR, at baseline and again while on treatment. We determined the objective tumor response for each patient and examined distinguishing characteristics of patients who benefited from ICI. Furthermore, we describe our real-world institutional experience treating patients with this complicated disease.
Patient Population and Data Sources
We performed a single-center, retrospective analysis of 46 metastatic uveal melanoma patients who received immunotherapy between September 2012 and May 2022. All these patients were treated with either single agent ipilimumab, nivolumab, or pembrolizumab therapy or ipilimumab/nivolumab combination therapy. Patients and data were collected via the University of Michigan electronic medical record (EMR) system. Patient data was extracted manually. All clinical records were obtained with the approval of Institutional Review Boards and patients' consents were waived following Institutional Review Board protocol review (HUM00163915, HUM00046408).
Data Collection and Treatment Outcomes
Patient demographic information and tumor characteristics included age, sex, Eastern Cooperative Oncology Group (ECOG) performance status, ocular location of the primary tumor (iris, ciliary body, choroid), ciliary body involvement, extraocular extension, tumor thickness, longest basal diameter, gene expression profile (GEP) class, PRAME status, histopathology, time from primary diagnosis to metastatic relapse and metastatic disease sites at the start of ICI (baseline). Hematologic markers collected included LDH at the following time points: metastatic disease diagnosis, baseline, and 8 weeks post ICI-start (on treatment), as well as absolute lymphocyte count (ALC), absolute neutrophil count (ANC), and absolute eosinophil count (AEC) at baseline and on treatment. A hard cut-off of 60 days following therapy start was used for all "on treatment" assessments. Patients who did not have this time-point were excluded from the analysis. NLR was defined as the ratio of the ANC to the ALC in each peripheral blood sample. dNLR was calculated using a formula previously demonstrated [32]:dNLR = ANC/(white blood cell count -ANC). Delta NLR was calculated as the difference between on treatment and baseline NLR. Treatment characteristics collected included treatment of the primary (enucleation vs. plaque RT), liver-directed therapy type, lines of prior therapy, immunotherapeutic agent(s), cycles of ICI completed, reason for ICI discontinuation, immune related adverse event (IRAE) grade, and objective response using the RECIST criteria (version 1.1) [38]. For the purposes of this study, responders were designated as those who derived clinical benefit from therapy, i.e., those who had complete response (CR), partial response (PR), or stable disease (SD). Non-responders were patients who had progressive disease (PD).
The endpoints analyzed included OS or PFS. Date of radiographic progression was determined by manual review of radiological reports and date of death was determined by manual review of the EMR. Survival time was measured from the start of immunotherapy to the date of death (for OS) or to disease progression (for PFS). Death certificates and hospital encounters at the end of life were reviewed. All patients had either (1) progressive disease radiographically, (2) progressive symptoms felt related to their cancer, or (3) cancer listed as a primary or secondary cause of death. Patients were censored at the date of last known follow-up, defined as the most recent encounter with a documented provider.
Patients with missing values pertinent to the specific analysis were excluded. Two of 46 patients (4.3%) were lost to follow up.
Statistical Analysis
OS and PFS estimates were generated with Kaplan-Meier method. Comparison of survival outcomes between groups and hazard ratios (HR) were generated using Gehan-Breslow-Wilcoxon and Mantel-Haenszel tests in univariable analyses. LDH was split into two groups based on the "normal" cutoff used at the University of Michigan: below 240 mg/dL and greater than or equal to 240 mg/dL. NLR was also split into two groups at the median NLR value for this cohort at each specified time-point. Multivariable Cox regression was performed to estimate the effect of each measure on survival. Continuous variables included longest basal diameter, tumor thickness, baseline and on treatment NLR as well as LDH at metastatic diagnosis, baseline, and while on treatment. The categorical variables, concurrent ipilimumab/nivolumab (vs. single agent ICB) and ECOG performance status (0 vs. >0), were also included in the model. HRs and 95% confidence intervals (CI) for each measure were generated for the overall cohort as well as subgroup analyses from the interaction term by the Wald method. For investigation of possible differences in baseline characteristics between responders (designated as CR, PR, SD) versus non-responders (designated as PD), Fischer-exact test was used for categorical variables and independent t-test for continuous variables. In all cases, two-tailed p-values were calculated with a significance cut-off of p < 0.05. All analyses were conducted using SPSS statistical software (IBM Corp, version 28
Survival Outcomes
The median OS for the entire cohort was 11.4 months (95% CI: 7.5-21.5 months) with a 49% survival rate at one year. Most patients had progressed by 6 months after the start of ICI (median PFS: 3.2 months, 95% CI: 2.6-5.1 months) (Figure 1). To identify factors that may be prognostic for survival or response to ICI we completed univariable analyses on patient demographic factors, primary tumor characteristics, treatment characteristics and hematologic lab values at baseline and while on treatment. Of the patient demographic and primary tumor features, the only factor that significantly impacted OS was primary
Survival Outcomes
The median OS for the entire cohort was 11.4 months (95% CI: 7.5-21.5 months) with a 49% survival rate at one year. Most patients had progressed by 6 months after the start of ICI (median PFS: 3.2 months, 95% CI: 2.6-5.1 months) (Figure 1). To identify factors that may be prognostic for survival or response to ICI we completed univariable analyses on patient demographic factors, primary tumor characteristics, treatment characteristics and hematologic lab values at baseline and while on treatment. Of the patient demographic and primary tumor features, the only factor that significantly impacted OS was primary
Patient Factors Associated with Clinical Benefit from ICI
To objectively evaluate treatment response in this cohort, we used RECIST criteria to quantitatively measure each patient's tumor at the start of immunotherapy and again at their next imaging assessment (9-12 weeks after therapy initiation). Of the 46 patients included in this study, 2 did not have available imaging and 5 had clinical and/or radiographic progression that was not quantifiable. In total, 34 (73.9%) patients had progressive disease (PD) while 10 (21.7%) patients derived a clinical benefit from therapy (i.e., CR, PR, or SD). There was one patient who achieved a CR (2.2%), one with PR (2.2%), and eight with SD (17.4%). The average time to progression for those with SD was 24.5 months after treatment initiation (IQR: 42 months) ( Figure 4A).
We sought to better understand if any specific patient, tumor, treatment or hematologic factors were related to clinical benefit from treatment. A greater proportion of patients in the responders group received combination ipilimumab/nivolumab as their first-line agent (40% vs. 20%, p = 0.24) ( Figure 4B, Table S2-S4). Moreover, patients who received combination or sequential anti-CTLA-4 and anti-PD-1 therapy at any point in
Patient Factors Associated with Clinical Benefit from ICI
To objectively evaluate treatment response in this cohort, we used RECIST criteria to quantitatively measure each patient's tumor at the start of immunotherapy and again at their next imaging assessment (9-12 weeks after therapy initiation). Of the 46 patients included in this study, 2 did not have available imaging and 5 had clinical and/or radiographic progression that was not quantifiable. In total, 34 (73.9%) patients had progressive disease (PD) while 10 (21.7%) patients derived a clinical benefit from therapy (i.e., CR, PR, or SD). There was one patient who achieved a CR (2.2%), one with PR (2.2%), and eight with SD (17.4%). The average time to progression for those with SD was 24.5 months after treatment initiation (IQR: 42 months) ( Figure 4A). Fisher-exact test was used for group comparisons. Responders (n = 10) are comprised of patients who derived a clinical benefit from treatment, i.e., complete response (CR (n = 1)), partial response (PR (n = 1)), and those with stable disease (SD (n = 8)). Non-responders (n = 34) are comprised of all Fisher-exact test was used for group comparisons. Responders (n = 10) are comprised of patients who derived a clinical benefit from treatment, i.e., complete response (CR (n = 1)), partial response (PR (n = 1)), and those with stable disease (SD (n = 8)). Non-responders (n = 34) are comprised of all those who had progressive disease (PD). Objective response was determined using RECIST criteria. LDH = lactate dehydrogenase, NLR = neutrophil:lymphocyte ratio.
We sought to better understand if any specific patient, tumor, treatment or hematologic factors were related to clinical benefit from treatment. A greater proportion of patients in the responders group received combination ipilimumab/nivolumab as their first-line agent (40% vs. 20%, p = 0.24) ( Figure 4B, Table S2, Table S3, Table S4). Moreover, patients who received combination or sequential anti-CTLA-4 and anti-PD-1 therapy at any point in their treatment course had significantly better OS (HR (95% CI): 2.3 (1.1-4.5), p = 0.0012). Median survival was 23.6 months in this group compared to 5.6 months for patients who only received monotherapy. Median PFS for patients who received combination therapy was 3.9 months compared to 2.7 months, however, this difference did not reach significance (HR (95% CI): 1.7 (0.9-3.1), p = 0.09) (Supplementary Figure S1).
Of the demographic variables analyzed there were more people with poor performance status in the non-responder (p = 0.046). Primary tumor characteristics including tumor thickness and longest basal diameter were not different between the groups (p = 0.78 and p = 0.75, respectively). Patients who benefited from therapy had a longer latent period between the diagnosis of their primary lesion and the development of metastases (64 vs. 22 months, p = 0.32) and fewer patients (70% vs. 94.1%, p = 0.069) had liver metastases at the start of immunotherapy. Responders also had lower median LDH at stage IV diagnosis (178 vs. 211 mg/dL, p = 0.14)) and while on treatment (202 vs. 258 mg/dL, p = 0.24) however, none of these reached statistical significance. Similarly, while median on treatment NLR values were higher for the non-responder group, the mean difference was not significant (Baseline, 2.6 vs. 2.6, p = 0.85; On treatment, 2.0 vs. 3.3, p = 0.47).
Discussion
In this study of 46 metastatic uveal melanoma patients, the median OS was 11.4 months. We describe factors, including the hematologic markers LDH and NLR, that may be important for prognostication in patients who receive ICI. We also describe differences in clinicopathologic and treatment characteristics between patients who had an objective benefit from immunotherapy and those who did not.
LDH has long been shown to be an important prognostic marker in cancer due to its facilitation of glycolysis [18,39,40] and has also been show to modify the immune microenvironment, potentially impacting the efficacy of ICI [41,42]. Its prognostic value in cutaneous melanoma is well established and included in the AJCC staging system [43]. Recently published studies of MUM patients, including a phase 2 clinical trial evaluating the efficacy of pembrolizumab, also demonstrate LDH to be prognostic [14,44]. In our study we confirmed the prognostic importance of LDH in MUM at various points in the clinical disease course. Furthermore, patients who benefited from ICI had lower median on treatment LDH values.
Another prognostically valuable hematologic marker that has been identified for solid tumors, including metastatic cutaneous melanoma, is NLR [31,32]. While NLR has been shown to be prognostically important in other contexts, such as COVID-19 infection [45], recent studies showed independent value in cancer patients with NSCLC treated with adjuvant immunotherapy [34]. In our review of the literature very few studies have evaluated the role of NLR in UM patients. A recent study by Meijer et. al, examined the association between NLR and other systemic inflammatory markers, including erythrocyte sedimentation rate (ESR) and c-reactive protein (CRP), and metastasis-free survival at the time patients were treated for their primary tumor. While NLR was not a prognostic marker in this study, the authors showed that high CRP levels were associated with a longer metastasis-free survival [46]. To our knowledge, there is only one published report evaluating NLR in patients with MUM. This study examined NLR as it related to first-line treatment response, time-to-relapse after receiving first-line therapy, and OS [36]. NLR was not prognostic for treatment response but was significant in time-to-relapse and OS. Importantly, only 6.7% of patients included in this study received ICI, making it difficult to extrapolate these conclusions to patients receiving immunotherapy [36]. Our study demonstrated that NLR is a significant prognostic factor for PFS, particularly while patients are on ICI treatment, and therefore may be a useful aid in clinical decision making and patient stratification. Other hematologic markers including absolute cell counts, dNLR, and on treatment delta NLR were not significant in our study. While also not reaching significance in our split analysis, the median NLR was consistently lower in responders than in non-responders at both time points.
Liver metastases, which are the most common metastatic site in UM, have also been shown to impact response to ICI via systemic loss of antigen-specific T cells and possibly by other unknown mechanisms [22,47]. While not prognostic in our study, a greater percentage of non-responders had liver involvement. A separate study in MUM patients receiving ICI showed that patients who lived longer had extrahepatic metastases in addition to liver metastases [48]. While this finding may be due to biased selection for patients with more indolent disease, it is also possible that extrahepatic metastases facilitate tumor recognition by circulating immune cells, bypassing the immunosuppressive environment of the liver. This study also showed that patients who received liver-directed therapy had longer survival [48]. This finding was not reproduced in our patient cohort, likely due to a limited sample size. Currently, there is a phase I clinical trial investigating the feasibility of hepatic ablation of melanoma metastases, in conjunction with ipilimumab and nivolumab, to enhance immunotherapy efficacy (HAMMER trial; NCT05169957) [49]. Besides an increased proportion of patients with LM, UM has a significantly lower tumor mutational burden and less PD-1/PD-L1 expression when compared to cutaneous melanoma which may limit response to ICI [50,51].
Our institutional experience in treating MUM patients with ICI is consistent with other published reports at an overall one-year survival rate of approximately 49%. As the treatment strategy for metastatic melanoma has evolved over time, our approach to agent selection has also changed. In our cohort, a total of 9 patients were treated with initial anti-CTLA-4 therapy; 7 of which were treated prior to 2015. All but one of these patients progressed on therapy and were eventually switched to a PD-1 inhibitor (either pembrolizumab or nivolumab). Twenty-four of our patients received single-agent pembrolizumab as their first-line therapy, with mixed response. Consistent with other reports, patients who received combination or sequential anti-CTLA-4 and anti-PD-I therapy at any point in their treatment course tended to live longer with a median OS of 23.6 months vs. 5.6 months for patients who received life-time single-agent therapy ( Figure S1) [14,[52][53][54]. Though biased towards patients with indolent enough disease to receive additional therapies, these findings suggest that dual-agent therapy may be beneficial in this patient population.
Limitations
The major limitations of this study are its small sample size, retrospective design, and lack of control or validation cohorts. When compared to other prospective and retrospective studies, the overall survival is similar at approximately one year. While the overall sample size was limited due to the rarity of the disease, the vast majority of patients included in this study received ICI as a front-line agent (93.6%), which is an improvement over many other published reports. Because GEP was not commercially available until late 2009, many of the patients included in this study did not have any available genetic data, including GEP class or PRAME status which together have shown prognostic value [55]. Recently, mutation of the methyl-binding domain 4 (MDB4) gene was identified as a predictive marker for immunotherapy response [56]. Encoding a glycosylase integral to DNA repair, it is suggested that defective MBD4 leads to increased tumor mutational burden, which in turn enhances the efficacy of immune checkpoint inhibitors. MDB4 mutational status, therefore, is an important confounder that we were unable to control for in this study. The prognostic and predictive value of CRP as a single pre-treatment measurement [57,58] and as a longitudinal metric [59] during ICI has been demonstrated for other tumor types. It would be interesting to evaluate CRP before and during treatment in metastatic uveal melanoma, however, our patient cohort did not have CRP consistently measured precluding us from inclusion of this important metric in our study. Furthermore, other potential prognostic biomarkers, tumor mutational burden, tumoral PD-L1, and HLA alleles were not available.
Conclusions
The findings of our study suggest that hematologic markers, such as LDH and NLR, may be helpful for patient prognostication and clinical decision making in patients with MUM being treated with immunotherapy. These agents are increasingly used in the treatment of MUM despite only modest improvement in survival outcomes. In this realworld analysis, we also demonstrated that patients who received initial combination ipilimumab/nivolumab tended to have better OS compared with patients who received initial monotherapy suggesting that combination therapy may be beneficial in the treatment of this deadly disease. Because immunomodulatory agents are not benign with respect to toxicity, more studies are needed to identify additional prognostic and predictive biomarkers that can aid in the identification of patients who will benefit from these therapies.
|
v2
|
2022-11-27T16:49:23.707Z
|
2022-11-24T00:00:00.000Z
|
253991431
|
s2orc/train
|
Antibacterial Spirotetronate Polyketides from an Actinomadura sp. Strain A30804
Large scale cultivation and chemical investigation of an extract obtained from Actimonadura sp. resulted in the identification of six previously undescribed spirotetronates (pyrrolosporin B and decatromicins C–G; 7–12), along with six known congeners, namely decatromicins A–B (1–2), BE-45722B–D (3–5), and pyrrolosporin A (6). The chemical structures of compounds 1–12 were characterized via comparison with previously reported data and analysis of 1D/2D NMR and MS data. The structures of all new compounds were highly related to the spirotetronate type compounds, decatromicin and pyrrolosporin, with variations in the substituents on the pyrrole and aglycone moieties. All compounds were evaluated for antibacterial activity against the Gram-negative bacteria, Acinetobacter baumannii and Gram-positive bacteria, Staphylococcus aureus and were investigated for their cytotoxicity against the human cancer cell line A549. Of these, decatromicin B (2), BE-45722B (3), and pyrrolosporin B (7) exhibited potent antibacterial activities against both Gram-positive (MIC90 between 1–3 μM) and Gram-negative bacteria (MIC90 values ranging from 12–36 μM) with weak or no cytotoxic activity against A549 cells.
Introduction
Antimicrobial resistance is one of the leading threats to human health globally [1]. Recently, it was estimated that 1.2 million people died from antibiotic-resistant bacterial infections, which was more than that caused by HIV/AIDS or malaria [2]. Thus, there is clearly an urgent need for new and effective antimicrobials. For many years, our group has engaged in a screening effort to detect secondary metabolites from Nature that can inhibit pathogenic microorganism [3][4][5][6], such as Staphylococcus aureus. Notably, S. aureus is one of the leading pathogens (the second after Escherichia coli) for fatalities associated with resistance [2]. Microbes have been one of the most prolific sources of small molecules where two third of known microbial secondary metabolites are derived from actinobacteria [7,8]. Actinobacteria is a large group of morphologically and physiologically diverse bacteria well known for their production of natural products and biotechnologically relevant compounds [9]. It is known that 70% of them are produced by the genus Streptomyces. Among the non-Streptomyces species, Actinomadura has been reported as a producer of chemically and biologically unique polyketides possessing antitumor, antimicrobial and anticoccidial activities. Up to date, several spirotetronate-class polyketides have been isolated from Actinomadura. For instance, nomimicin, a polyketide isolated from an Actinomadura strain TP-A0878 in a compost sample collected at Nomi, Ishikawa, Japan. Subsequently nomimicins B-D, new tetronate-class polyketides were also isolated from a marine-derived actinomycete of the genus Actinomadura. These compounds showed antimicrobial activities [10,11]. 2 of 18 During our screening assays, a few MeOH extracts from our in-house Actinobacteria strains library [12] exhibited antibacterial activity against S. aureus. Chemical dereplication of the active extracts suggested the presence of several spirotetronate polyketides, a group of compounds that was known for their significant pharmacological potentials [13]. The spirotetronate polyketides feature an unusual aglycone core containing a typical tetronic acid spiro-connected to a six-membered ring and linked to a trans-decalin moiety. Further chemical analysis of the active extracts shortlisted one extract from Actinomadura sp. strain A30804 that produced higher yields of potentially new spirotetronate analogs for large-scale cultivation. Bioassay guided purification of the extract yielded twelve spirotetronate natural products, including six known compounds, decatromicin A-B (1-2) [14,15], BE-45722B-D (3)(4)(5) [16][17][18], and pyrrolosporin A (6) [19,20], as well as six new analogs, pyrrolosporin B (7) and decatromicin C-G (8-12) (Figure 1). In this report, we describe the purification and structure elucidation of these natural products and further demonstrate the antimicrobial potential of the spirotetronates. Moreover, the discovery of these structurally intricate compounds enriches the chemical diversity of the spirotetronates and potentially leads to a better understanding of preliminary structure-activity relationships.
During our screening assays, a few MeOH extracts from our in-house Actinobacteria strains library [12] exhibited antibacterial activity against S. aureus. Chemical dereplication of the active extracts suggested the presence of several spirotetronate polyketides, a group of compounds that was known for their significant pharmacological potentials [13]. The spirotetronate polyketides feature an unusual aglycone core containing a typical tetronic acid spiro-connected to a six-membered ring and linked to a trans-decalin moiety. Further chemical analysis of the active extracts shortlisted one extract from Actinomadura sp. strain A30804 that produced higher yields of potentially new spirotetronate analogs for large-scale cultivation. Bioassay guided purification of the extract yielded twelve spirotetronate natural products, including six known compounds, decatromicin A-B (1-2) [14,15], BE-45722B-D (3)(4)(5) [16][17][18], and pyrrolosporin A (6) [19,20], as well as six new analogs, pyrrolosporin B (7) and decatromicin C-G (8-12) (Figure 1). In this report, we describe the purification and structure elucidation of these natural products and further demonstrate the antimicrobial potential of the spirotetronates. Moreover, the discovery of these structurally intricate compounds enriches the chemical diversity of the spirotetronates and potentially leads to a better understanding of preliminary structure-activity relationships.
General Experimental Procedures
Specific rotations were recorded using JASCO P-2000 digital polarimeter. Bruker DRX-400 NMR spectrometer with cryoprobe was used to record NMR spectra. The NMR spectrometer was equipped with 5 mm BBI ( 1 H, COSY, edited HSQC, and HMBC) or BBO ( 13 C) probe head with z-gradients. The 1 H and 13 C NMR chemical shifts were referenced to the residual solvent peaks for MeOH-d 4 at δ H 3.31 and δ C 49.0 ppm; or DMSO-d 6 at δ H 2.50 and δ C 39.5 ppm. Preparative RP-HPLC was performed using XTerra MS C 18 Prep column (19 × 300 mm, 10 µm) on an Agilent 1260 Infinity Preparative-Scale LCMS Purification System hyphenated with Agilent 6130B single quadrupole MS as a detector. LCMS data were recorded using Agilent UHPLC 1290 Infinity coupled to Agilent 6540 accurate-mass quadrupole time-of-flight (QTOF)-ESIMS. Standard gradient conditions of 98% H 2 O (0.1% FA) to 100% CH 3 CN (0.1% FA) were run over 8.6 min using an Acquity UPLC BEH C 18 (2.1 × 50 mm, 1.7 µm) column, all at a flow rate of 0.5 mL/min. The QTOF were set using the same parameter as previously reported [4].
Molecular Identification and Phylogenetic Analysis of the Bacteria Isolate A30804
The DNA of bacterial strain A30804 was extracted from the plate using the DNeasy PowerSoil Pro Kit (Qiagen, Hilden, Germany) according to the manufacturer's protocol. Upon extraction, DNA purity and yield of A30804 was measured with NanoDrop2000 spectroscopy system (ThermoFisher Scientific, Waltham, MA, USA). Bacterial 16S rRNA genes of interest were amplified from the extracted DNA using the universal 16S primers 27F (5 -AGA GTT TGA TCC TGG CTC AG-3 ) and 1492R (5 -TAC GGY TAC CTT GTT ACG ACT T-3 ) [21,22]. The PCR amplification reactions were performed using Applied Biosystems ProFlex Thermocycler (ThermoFisher Scientific, Waltham, MA, USA) with a total reaction volume of 20 µL that comprised of 2.0 µL of 10× PCR buffer with 20 mM MgCl 2 , 2.0 µL of 2 mM dNTPs, 1 unit of Taq polymerase (ThermoFisher Scientific, USA), 1.0 µL of 10 µM of each primer and 1.0 µL of purified DNA templates. The thermal cycling profile used was, 5 min initial denaturation at 95 • C; further denaturation at 95 • C (30 cycles, 30 s each), annealing at 60 • C for 50 s, followed by 1 min extension at 72 • C, and a final extension at 72 • C for 5 min. A negative control and non-template were included in the run. This was followed by Sanger Sequencing of the PCR amplified DNA fragments (1st BASE, Singapore). Alignment of the sequences were carried out using Benchling and followed by analysis using BLAST [National Center for Biotechnology Information (NCBI)]. A30804 was aligned using ClustalW with respective closely related actinobacteria strains obtained from the GenBank databases with I6S rRNA region. Neighbor-joining tree algorithm method was used to establish the genetic relationship between the strains. The phylogenetic tree was constructed with a bootstrapped database containing 1000 replicates in MEGA 11.0 software (Mega, US). The I6S rRNA gene sequence of strain A30804 has been deposited in GenBank database of NCBI under the accession number OP225395.
Fermentation and Extraction of Bacterial Crude Extract
Actinomadura sp. strain A30804 were cultured in 5 mL SV2 media, (For 1 L, add 15 g glucose (1st BASE, Singapore), 15 g glycerol (VWR, Radnor, PA, USA), 15 g soya peptone (Oxoid, Basingstoke, Hampshire, UK), and 1 g calcium carbonate (Sigma Aldrich, St. Louis, MO, USA), pH adjusted to 7.0) for 3 days at 28 • C under constant shaking at 200 rpm to generate a seed culture. The seed culture was inoculated into fresh CA09LB media (10 g meat extract (Sigma-Aldrich, St. Louis, MO, USA), 4 g yeast extract (BD Biosciences, Franklin Lakes, NJ, USA), 20 g glucose (1st BASE, Singapore), and glycerol 3 g (VWR, Radnor, PA, USA) in 1 L of H 2 O, pH adjusted to 7.0) in a 1:20 volume ratio and incubated at 28 • C in the dark with shaking at 200 rpm. After 9 days of incubation, the cultures were centrifuged to separate biomass and supernatant, followed by lyophilization. The dried cultures were extracted by MeOH and filtered through filter paper (Whatman Grade 4, Maidstone, Kent, UK). MeOH was removed under reduced pressure to give a crude extract of a combined weight of 40.02 g.
Chemical Structural Data
The UV spectra and HRESIMS spectra of 7-12, and 1D and 2D NMR spectra of 1-12 are provided in Supplementary Table 1.
Biological Assays
Antimicrobial effect of compounds 1-12 were tested against a panel of Gram-negative bacteria, including Acinetobacter baumannii (ATCC ® 19606™), Pseudomonas aeruginosa (ATCC ® 9027™), and Klebsiella aerogenes (ATCC ® 13048™), the Gram-positive Staphylococcus aureus Rosenbach (ATCC ® 25923™) and the fungal strain Aspergillus fumigatus (ATCC ® 46645™). The minimum inhibitory concentration (MIC) and minimum bactericidal/fungicidal concentration (MBC/MFC) determination were carried out using the microbroth dilution method, based on the Clinical Laboratory Standards Institute (CLSI) guidelines, with minor modifications. Antibacterial assays were tested with cells seeded at 5.5 × 10 5 cells/mL, whereas the antifungal assay was carried out at 2.5 × 10 4 spores/mL. The tested compounds were incubated together with the bacterial cells at 37 • C for 24 h, and at 25 • C for 72 h for the fungal spores for MIC testing. OD 600 measurement was subsequentially performed on the plates to determine the inhibitory effect of the compounds. MBC/MFC was determined by transferring 5 µL of the treated culture into new media microplates, where the plates were incubated under the same condition, followed by OD 600 measurement. Isolated compounds 1-12 were also tested against A549 human lung carcinoma cells (ATCC ® CCL-185™) for cytotoxicity assessment, where cells were seeded at 3.3 × 10 4 cells/mL. Following that, compounds were added to the cells and further incubated for 72 h at 37 • C in the presence of 5% CO 2 . The cytotoxic effect of the compounds was detected using PrestoBlue™ cell viability reagent (ThermoFisher Scientific, Waltham, MA, USA). After the addition of the reagent, cells were further incubated for 2 h before fluorescence measurement at excitation 560 nm and emission 590 nm. All assays were performed in triplicates to ensure reproducibility. Standard inhibitors gentamicin (Gibco, Waltham, MA, USA) was used as the assay control for antibacterial, amphotericin (Sigma Aldrich, St. Louis, MO, USA) for antifungal and puromycin (Sigma Aldrich, St. Louis, MO, USA) for cytotoxicity testing. The dose-response curves for IC 90 and IC 50 values determination were plotted using the GraphPad Prism 8 software (GraphPad, San Diego, CA, USA).
Results and Discussion
Molecular identification was used to identify the closely related species of interest to the isolated strain A30804. The sequence obtained from the 16S rRNA gene sequence of A30804 was aligned and further analyzed where a nucleotide BLAST search was performed against the NCBI 16S ribosomal RNA database. The phylogenetic relatedness using the neighbor-joining analysis method of isolated strain and its closely related species obtained from the Genbank database is shown in Figure 2. Phylogenetic analysis based on 16S rRNA gene sequences revealed that strain A30804, accession number OP225395 belonged to the genus Actinomadura ( Figure 3). The isolate formed a subcluster with Actinomadura sp. 2EPS, Actinomadura chibensis strain IFM 10266 and Actinomadura sp. strain SS19 recovered by neighbor-joining analysis. Previous studies have shown that Actinomadura sp. have been reported to produce spirotetronate polyketides compounds such as decatromicin A and B [15,17].
Compound 7, white amorphous powders, had a molecular formula of C44H53Cl3N2O10, as revealed by (-)-HRESIMS. The 1 H, 13 C, and HSQC NMR data (Table 1) indicated the presence of five methyl, eight methylene, seventeen methine, and fourteen non-protonated carbons. Detailed analysis of 2D NMR spectra suggested that 7 belonged to the spirotetronate polyketide structural class (Figure 4). For example, 1 H-1 H COSY spectrum readily assigned the fragment H-5 along the chain to H-19; HMBC cross-peaks from H3-30 to C-19, C-21, and C-25, together with HMBC correlations from H-21 to C-22, C-23, and C-31 as well as 1 H-1 H COSY between H-23 and H2-24 and HMBC correlation from H-24 to C-25 allowed the identification of a cyclohexene ring (linked to C-19) decorated with a carboxylic acid at C-22. 1 H-1 H COSY between H2-32 and H-23 and HMBC correlation from H-33 to C-23 established an ethyl fragment attached to C-23. Further HMBC correlations from H2-27 to C-4, C-5, and C-13; and H3-28 to C-4 assigned the position of CH3-CH2-fragment on C-4. The presence of a sugar moiety attached to C-9 was evident by 1 H-1 H COSY between H-1'/H2-2'/H-3'/H-4'/H-5'/H3-6', and HMBC correlation from H-9 to an anomeric carbon at δC 102.7 ppm. The features of the 1 H and 13 C spectra of 7 were very similar to those for pyrrolosporin A (6) ( Supplementary Information Table S3), except that the -CH resonance from the pyrrole moiety in 6 was absent in 7. Analysis of MS data and the isotopic pattern of the molecular ion peak suggested the presence of three chlorine atoms in 7. These observations together with MS data analysis revealed that 7 had a unique trichlorinated pyrrole moiety. Consistently, MS/MS fragmentation analysis showed a fragment with an m/z value of 687.3651, which suggested a loss of the trichlorinated pyrrole carbonyl fragment ( Figure 6) [23]. Similar coupling constant values and NOESY correlations in 7 to those of pyrrolosporin A (6) indicated the same relative configurations, for instance, a coupling constant of 8. Compound 8, white amorphous powders, had a molecular formula of C45H58N2O10 assigned based on (-)-HRESIMS data. Comparison of 1 H NMR data of 8 with those of 7 showed an additional singlet methyl at δH 1.79 and three additional proton resonances at δH 6.16, 6.82, and 6.90, characteristic 1 H resonances of a pyrrole moiety (Table 1). The additional methyl was positioned at C-18 based on HMBC correlations from δH 1.79 to C-17, C-18, and C-19 ( Figure 4). The other features of 1 H and 13 C of 8 were similar to those of decatromicin A (1). Based on these observations, compound 8 was a dechlorinated version of decatromicin A (1). The relative configurations of 8 were deemed to be the same as in 7 following analyses of J-coupling and NOESY data ( Figure 5). Thus, the structure of 8 was assigned and named decatromicin C. It is worth mentioning that compound 8 was previously obtained from a dechlorination reaction of 1 using tri-n-butylstannane [14]. However, spectroscopic data of 8 was not reported. We have also included HRMS and full NMR spectroscopic data of 8, which were not reported in the original paper. This is the first report of the identification of 8 from a natural source.
Compound 9 had a molecular formula of C34H44O7 following HRMS data analysis. 1 H and 13 C NMR spectra of 9 were similar to those of 8. However, the 1 H NMR signals for the pyrrole group and amino sugar in 8 were missing in 9 (Table 2). Furthermore, the 1 H NMR resonance of H-9 was shifted downfield from δH 3.39 in 8 to δH 3.41 in 9. Detailed 2D NMR (Figure 7) data analysis indicated that 9 was the aglycone core of 8, and named decatromicin D. The relative configuration of 9 was deemed to be identical as 8 following NOESY data interpretation ( Figure 8). Previously, aglycone 9 was obtained from a degradation study of decatromicin A (1) [14]. Here, we report the isolation and identification of 9 from a natural source.
Compound 10, white amorphous powders, was assigned the molecular formula C34H46O5. The 1 H and 13 C NMR spectra of 10 were almost identical to those of 9, indicating that compound 10 had the same spirotetronate aglycone core structure. 1 H NMR data showed that compound 10 had an extra methyl compared to 9 ( Table 2). The 13 C carbonyl NMR resonance at δC 171.4 in 9 was not observed in 10, suggesting that the carboxylic acid group in 9 was replaced by a methyl group. This was supported by HMBC correlations from the singlet methyl at δH 1.79 to C-21, C-22, and C-23 (Figure 7). The same relative configuration previously assigned for 9 was also determined for 10 following analyses of the NOESY data and 1 H-1 H coupling constants (Figure 8). Thus, the structure of 10 was established and named decatromicin E.
The molecular formula C35H48O5 was assigned for compound 11 following (-)-HRESIMS spectrum analysis. MS and NMR data comparison between 10 and 11 revealed that the latter compound had an additional -CH2moiety (Table 2). 1 H-1 H COSY spectrum established a fragment of H2-32/H2-33/H-34 which was attached to C-23 since HMBC correlation from H-23 to C-32 was observed (Figure 7). The relative configuration of 11 (Table 1). The additional methyl was positioned at C-18 based on HMBC correlations from δ H 1.79 to C-17, C-18, and C-19 ( Figure 4). The other features of 1 H and 13 C of 8 were similar to those of decatromicin A (1). Based on these observations, compound 8 was a dechlorinated version of decatromicin A (1). The relative configurations of 8 were deemed to be the same as in 7 following analyses of J-coupling and NOESY data ( Figure 6). Thus, the structure of 8 was assigned and named decatromicin C. It is worth mentioning that compound 8 was previously obtained from a dechlorination reaction of 1 using tri-n-butylstannane [14]. However, spectroscopic data of 8 was not reported. We have also included HRMS and full NMR spectroscopic data of 8, which were not reported in the original paper. This is the first report of the identification of 8 from a natural source.
Compound 9 had a molecular formula of C 34 H 44 O 7 following HRMS data analysis. 1 H and 13 C NMR spectra of 9 were similar to those of 8. However, the 1 H NMR signals for the pyrrole group and amino sugar in 8 were missing in 9 (Table 2). Furthermore, the 1 H NMR resonance of H-9 was shifted downfield from δ H 3.39 in 8 to δ H 3.41 in 9. Detailed 2D NMR (Figure 7) data analysis indicated that 9 was the aglycone core of 8, and named decatromicin D. The relative configuration of 9 was deemed to be identical as 8 following NOESY data interpretation (Figure 8). Previously, aglycone 9 was obtained from a degradation study of decatromicin A (1) [14]. Here, we report the isolation and identification of 9 from a natural source. was judged to be identical to 10 following the NOESY data comparison (Figure 8). Therefore, the structure of 11 was elucidated and named decatromicin F. The minor compound 12 had the same molecular formula C35H48O5 as that of 11, indicating that these compounds were structural isomers. However, the methyl doublet at δH 1.00 in 11 was absent in 12. This observation together with HMBC correlation from H3-30 to C-8 located an ethyl group on C-8 ( Figure 7). 1 H-1 H coupling constants and the NOESY spectrum analysis assigned the relative configurations in 12 to be the same as in 10 ( Figure 8). Hence, the structure of 12 was elucidated and named decatromicin G. Due to the minute amount of decatromicin F (11) and G (12), several of 13 C NMR resonances for these molecules were not observed in the 13 C NMR spectra (Table 2). However, their 1 H/2D NMR, UV, and MS spectra were comparable with other isolated compounds (1-10), indicating they belonged to spirotetronate polyketide type of compounds. During the isolation of compounds 7-12, six known spirotetronate polyketides, including decatromicin A-B (1-2), BE-45722B-D (3)(4)(5), and pyrrolosporin A (6) were also isolated and identified. Decatromicins A (1) and B (2) were initially discovered from Actinomadura sp. MK73-NF4 and showed potent antimicrobial activity against Grampositive methicillin-resistant Staphylococcus aureus (MRSA) [14,15], while BE-45722 series of compounds (3)(4)(5) were analogs of the decatromicins with an altered decoration of was judged to be identical to 10 following the NOESY data comparison (Figure 8). Therefore, the structure of 11 was elucidated and named decatromicin F. The minor compound 12 had the same molecular formula C35H48O5 as that of 11, indicating that these compounds were structural isomers. However, the methyl doublet at δH 1.00 in 11 was absent in 12. This observation together with HMBC correlation from H3-30 to C-8 located an ethyl group on C-8 ( Figure 7). 1 H-1 H coupling constants and the NOESY spectrum analysis assigned the relative configurations in 12 to be the same as in 10 ( Figure 8). Hence, the structure of 12 was elucidated and named decatromicin G. Due to the minute amount of decatromicin F (11) and G (12), several of 13 C NMR resonances for these molecules were not observed in the 13 C NMR spectra (Table 2). However, their 1 H/2D NMR, UV, and MS spectra were comparable with other isolated compounds (1-10), indicating they belonged to spirotetronate polyketide type of compounds. During the isolation of compounds 7-12, six known spirotetronate polyketides, including decatromicin A-B (1-2), BE-45722B-D (3)(4)(5), and pyrrolosporin A (6) were also isolated and identified. Decatromicins A (1) and B (2) were initially discovered from Actinomadura sp. MK73-NF4 and showed potent antimicrobial activity against Grampositive methicillin-resistant Staphylococcus aureus (MRSA) [14,15], while BE-45722 series of compounds (3)(4)(5) were analogs of the decatromicins with an altered decoration of Compound 10, white amorphous powders, was assigned the molecular formula C 34 H 46 O 5 . The 1 H and 13 C NMR spectra of 10 were almost identical to those of 9, indicating that compound 10 had the same spirotetronate aglycone core structure. 1 H NMR data showed that compound 10 had an extra methyl compared to 9 ( Table 2). The 13 C carbonyl NMR resonance at δ C 171.4 in 9 was not observed in 10, suggesting that the carboxylic acid group in 9 was replaced by a methyl group. This was supported by HMBC correlations from the singlet methyl at δ H 1.79 to C-21, C-22, and C-23 (Figure 7). The same relative configuration previously assigned for 9 was also determined for 10 following analyses of the NOESY data and 1 H-1 H coupling constants (Figure 8). Thus, the structure of 10 was established and named decatromicin E.
The molecular formula C 35 H 48 O 5 was assigned for compound 11 following (-)-HRESIMS spectrum analysis. MS and NMR data comparison between 10 and 11 revealed that the latter compound had an additional -CH 2 -moiety ( Table 2). 1 H-1 H COSY spectrum established a fragment of H 2 -32/H 2 -33/H-34 which was attached to C-23 since HMBC correlation from H-23 to C-32 was observed (Figure 7). The relative configuration of 11 was judged to be identical to 10 following the NOESY data comparison (Figure 8). Therefore, the structure of 11 was elucidated and named decatromicin F. The minor compound 12 had the same molecular formula C 35 H 48 O 5 as that of 11, indicating that these compounds were structural isomers. However, the methyl doublet at δ H 1.00 in 11 was absent in 12. This observation together with HMBC correlation from H 3 -30 to C-8 located an ethyl group on C-8 ( Figure 7). 1 H-1 H coupling constants and the NOESY spectrum analysis assigned the relative configurations in 12 to be the same as in 10 ( Figure 8). Hence, the structure of 12 was elucidated and named decatromicin G. Due to the minute amount of decatromicin F (11) and G (12), several of 13 C NMR resonances for these molecules were not observed in the 13 C NMR spectra (Table 2). However, their 1 H/2D NMR, UV, and MS spectra were comparable with other isolated compounds (1-10), indicating they belonged to spirotetronate polyketide type of compounds.
The antibacterial activity of spirotetronate polyketides 1-12 was evaluated against a panel of bacterial strains, namely A. baumannii, K. aerogenes, P. aeruginosa, and S. aureus Rosenbach. All compounds were inactive against the Gram-negative bacteria P. aeruginosa and K. aerogenes ( Figure S56). Compounds 1-8, 10, 12 demonstrated activities against S. aureus Rosenbach, the only Gram-positive bacteria strain tested (Table 3 and Figure 9). Interestingly, compounds 2, 3 and 7 exhibited inhibitory activity against Gram negative strain A. baumannii (Table 3 and Figure 10), which was not reported previously for this class of compounds. In addition, the antifungal activities of 1-12 was also evaluated against A. fumigatus; no antifungal activity was observed ( Figure S57). Furthermore, all the compounds were investigated for their cytotoxicity against the human lung carcinoma cell line; and compounds 3-5 showed weak cytotoxic activity towards A549 cells ( Figure S58). Notably, it has been well documented that decatromicin compounds exhibited a wide range of biological properties; and are strong antibiotics against Gram-positive bacteria, including contemporary strains of methicillin-resistant Staphylococcus aureus (MRSA) [14,15]. Compounds 1-5 and 8 possess the same aglycone core structure. In this series of compounds, at least two chloro substitutions and a free NH on the pyrrole moiety are important for antibacterial activity against A. baumannii as seen in compounds 2 and 3 (i.e., MIC 90 between 12 µM and 30 µM). In addition, compound 3 was 2-fold more active against A. baumannii (MIC 90 of 12.3 µM and MBC 90 of 30.4 µM) as compared to compound 2 (i.e., MIC 90 of 28.3 µM and MBC 90 >100 µM) as the number of chloro substituent on the pyrrole moiety increases (Table 3).
Similar observation was made in compounds 6 and 7 which possess the same aglyone structure. For instance, compound 7 which was trichloro substituted at the pyrrole moiety showed antibacterial activity against A. baumannii (i.e., MIC 90 of 36.3 µM) while compound 6 (i.e., dichloro substituted) was inactive against A. baumannii. Moreover, a 3-fold increase in antibacterial activity was observed in compound 3 as compared to compound 7 when the C-18 position in the aglycone moiety was substituted with a methyl group (Table 3).
Compounds 1-8, 10, 12 were active against S. aureus Rosenbach. Among these compounds, compounds 2, 3 and 7 were the most potent (MIC 90 between 1 µM and 3 µM). Our studies showed that the potency of these compounds against S. aureus Rosenbach was likely attributed to at least two chloro-substitutions on the pyrrole moiety. On the other hand, a decrease in antibacterial activity against S. aureus Rosenbach was observed in compounds 4 and 5 (i.e., MIC 90 between 5 µM and 7 µM) when NH group in pyrrole moiety was substituted with a methyl group even though their pyrrole moieties were substituted with at least two chlorine. In addition, the absence of any chlorine and methyl substitutions on the pyrrole moiety in compound 8 led to 3 to 10-fold decrease in antibacterial activity against S. aureus Rosenbach (i.e., MIC 90 of 15.7 µM) as compared to compounds 2-5.
Interestingly, a replacement of the carboxylic acid group on the C-22 position in the aglycone moiety in compound 9 with a methyl group in compound 10 led to antibacterial activity against S. aureus Rosenbach as shown in Table 3 and Figure 9 (i.e., 9: MIC 90 > 100 µM; 10: MIC 90 of 17.5 µM). This could be due to the poor cell membrane permeability of the carboxylic acid group, thus resulting in no antibacterial activity in compound 9.
For compounds 10 and 11, a replacement of the ethyl group with a propyl group on the C-23 position in the aglycone moiety resulted in weaker activity against S. aureus Rosenbach with an MIC 90 value of 17.5 µM in 10, whilst 11 did not exhibit any antibacterial activity. In addition, a replacement of the methyl group with an ethyl group on the C-8 position in the aglycone moiety as seen in compounds 10 and 12, respectively, did not lead to any significant change in activity against S. aureus Rosenbach, with a MIC 90 value of 15.6 µM in compound 12. Of all the compounds tested, compounds 2 and 3 (decatromicin B and BE-45722B, respectively) and 7 (pyrrolosporin B) showed the most potent antibacterial activities against both Gram-negative and Gram-positive bacteria, as well as weak or no cytotoxic activity against A549 cells.
Conclusions
Actinomadura sp. strain A30804 was found to produce spirotetronate polyketides 1-12. Compounds 2 and 3 (decatromicin B and BE-45722B, respectively) and 7 (pyrrolosporin B) not only exhibited weak to no cytotoxicity to laboratory human cell line A549, but also demonstrated potent antibacterial activity against both Gram-positive bacteria, S. aureus Rosenbach (ATCC ® 25923™) with MIC 90 values ranging from 1 µM to 3 µM and Gramnegative bacteria, A. baumannii (ATCC ® 19606™) with MIC 90 values ranging from 12 µM to 36 µM. Preliminary structure-activity relationship studies showed the presence of at least two chloro-substitutions on the pyrrole moieties was important for antibacterial activity in this series of spirotetronate compounds. This report further supported the potential of these spirotetronate polyketides as potent antibacterial agents.
|
v2
|
2022-11-27T17:10:00.123Z
|
2022-11-24T00:00:00.000Z
|
253997347
|
s2orc/train
|
A Clinical Guide to Peptide Receptor Radionuclide Therapy with 177Lu-DOTATATE in Neuroendocrine Tumor Patients
Simple Summary Peptide receptor radionuclide therapy (PRRT) is one of the treatment options for locally advanced or metastatic gastroenteropancreatic (GEP) neuroendocrine tumors (NETs). This treatment makes use of radioactive labeled somatostatin analogues, with 177Lu-DOTATATE as its established standard. PRRT has positive effects in tumor control and it lowers the risk of disease progression or death. It also improves the quality of life of the patient. Unfortunately, important risk factors for a minority of patients include renal and hematological toxicities. NET is a rare disease and treating patients with PRRT requires clinical expertise. This guide gives an overview of the background of PRRT and the current results in NET patient care. Abstract Peptide receptor radionuclide therapy (PRRT) with [177Lu]Lu-[DOTA0,Tyr3]-octreotate (177Lu-DOTATATE) has become an established second- or third-line treatment option for patients with somatostatin receptor (SSTR)-positive advanced well-differentiated gastroenteropancreatic (GEP) neuroendocrine tumors (NETs). Clinical evidence of the efficacy of PRRT in tumor control has been proven and lower risks of disease progression or death are seen combined with an improved quality of life. When appropriate patient selection is performed, PRRT is accompanied by limited risks for renal and hematological toxicities. Treatment of NET patients with PRRT requires dedicated clinical expertise due to the biological characteristics of PRRT and specific characteristics of NET patients. This review provides an overview for clinicians dealing with NET on the history, molecular characteristics, efficacy, toxicity and relevant clinical specifics of PRRT.
Introduction
Peptide receptor radionuclide therapy (PRRT) with radiolabeled somatostatin analogs (SSAs) has become an established second-or third-line treatment option for patients with progressive well-differentiated (grade 1-2) gastroenteropancreatic (GEP) neuroendocrine tumors (NETs). Due to the increasing incidence and prevalence of GEP-NETs over recent decades [1] and the development of NET-specific treatments and treatment protocols, there is a growing utilization of such systemic treatment for this advanced patient group. PRRT with [ 177 Lu]Lu-[DOTA 0 ,Tyr 3 ]octreotate ( 177 Lu-DOTATATE) is the first registered theranostic application in the field of NETs [2,3]. With this therapy, radiolabeled SSAs target the somatostatin receptor (SSTR) subtype 2 that is over-expressed on the cancer cell surface [4]. Treatment of NET patients with PRRT requires dedicated clinical expertise due to the biological characteristics of PRRT (for example, binding to SSTR, DNA damage induction 177 Lu-DOTATATE leads to tumor cell binding via SSTR2. After internalization of the radiopharmaceutical-SSTR2 complex, local radiation by beta particles can lead to cell death through the induction of DNA damage (image created with BioRender.com, accessed on 1 September 2022).
The Choice for 177 Lu-DOTATATE
Clinical development of PRRT commenced with SSAs labeled with indium-111, yttrium-90 and lutetium-177 [12]. PRRT with 177 Lu-DOTATATE is approved by the Food and Drug Administration (FDA) and European Medicine Agency (EMA) and consists of the SSA octreotate [13] and 1,4,7,10-tetraazacyclotetradecane-1,4,7,10-tetraacetic acid (DOTA), which is the chelator for stable binding of lutetium-177 to octreotate [14]. Lutetium-177 is a medium-energy beta-emitting radionuclide and a low-energy gamma-emitting radionuclide. The gamma emission allows imaging and dosimetry where the beta emission is used for therapeutic purposes. The beta emission of lutetium-177 shows tissue penetration with a maximum range of 2 mm. Detailed information on chemical structure, production and quality control of 177 Lu-DOTATATE can be found elsewhere [15]. Indium-111 is much less suitable for therapy compared to lutetium-177, because its auger electrons will not reach the nucleus and will therefore not be able to induce sufficient DNA damage 177 Lu-DOTATATE leads to tumor cell binding via SSTR 2 . After internalization of the radiopharmaceutical-SSTR 2 complex, local radiation by beta particles can lead to cell death through the induction of DNA damage (image created with BioRender.com, accessed on 1 September 2022).
The Choice for 177 Lu-DOTATATE
Clinical development of PRRT commenced with SSAs labeled with indium-111, yttrium-90 and lutetium-177 [12]. PRRT with 177 Lu-DOTATATE is approved by the Food and Drug Administration (FDA) and European Medicine Agency (EMA) and consists of the SSA octreotate [13] and 1,4,7,10-tetraazacyclotetradecane-1,4,7,10-tetraacetic acid (DOTA), which is the chelator for stable binding of lutetium-177 to octreotate [14]. Lutetium-177 is a medium-energy beta-emitting radionuclide and a low-energy gamma-emitting radionuclide. The gamma emission allows imaging and dosimetry where the beta emission is used for therapeutic purposes. The beta emission of lutetium-177 shows tissue penetration with a maximum range of 2 mm. Detailed information on chemical structure, production and quality control of 177 Lu-DOTATATE can be found elsewhere [15]. Indium-111 is much less suitable for therapy compared to lutetium-177, because its auger electrons will not reach the nucleus and will therefore not be able to induce sufficient DNA damage [16]. Another important part of the potential therapeutic effect is the choice of SSA. Octreotate shows a higher affinity to SSTR 2 than the earlier developed octreotide. This can explain why the uptake of 177 Lu-DOTATATE in the tumor was increased three to four times in comparison to 111 In-diethylenetriaminepentaacetic acid (DTPA)-octreotide [17]. Therapeutic activity of 111 In-DTPA-octreotide, known for its diagnostic use as OctreoScan ® , originally showed promising preliminary results on symptom control in NETs, but radiological tumor regression was rare and high levels of myelotoxicity were observed [18,19]. Given the specific characteristics of the different radionuclides, not only is the variance in efficacy important to take into account when selecting the preferred radionuclide, but also the risk of toxicity has to be taken into account. Less renal toxicity was seen in patients treated with 177 Lu-DOTATATE compared to patients treated with [ 90 Y]Y-[DOTA 0 ,Tyr 3 ]octreotide ( 90 Y-DOTATOC). This can be explained by the lower beta-energy and tissue penetration of lutetium-177. Due to the lack of gamma emission, yttrium-90 is also not suitable for scintigraphy with a gamma camera [20], however, positron emission tomography (PET) imaging is possible [21]. 177 Lu-DOTATATE can also be used for tumor dosimetry to obtain more knowledge about the tumor-absorbed dose. An indication of the absorbed dose can be obtained by dosimetric calculation on the total body scintigraphy. Quantitative single photon emission computed tomography/computerized tomography (SPECT/CT) can be more precise in the estimation of the tumor-absorbed dose per cycle. However, these scans are more time consuming, especially for dosimetry at multiple time points. A reduction of uptake in the tumor in every subsequent cycle might be an indication of a good tumor response. On the other hand, there is still a large variety between cycles and different tumor sides within the patient [22]. A significant dose-response correlation was found in pancreatic NETs (panNETs) but not in small intestinal NETs (SI-NETs) [23,24]. The different range of the beta particles from lutetium-177 and yttrium-90 could potentially be an advantage in patients with tumors of different sizes, for instance in an induction setting, although further research is needed [25,26].
Tumor Control
Clinical evidence of the efficacy of PRRT in tumor control was proven by the phase III NETTER-1 study and several phase II clinical trials [27][28][29]. The NETTER-1 clinical trial compared four cycles of 7.4 GBq (800 mCi total) 177 Lu-DOTATATE plus intramuscular longacting octreotide 30 mg every 4 weeks to a control group which received octreotide 60 mg every 4 weeks in grade 1 and 2 midgut NET patients, progressive on a long-acting SSA. The rate of progression-free survival (PFS) after 20 months was 65% of patients treated with 177 Lu-DOTATATE compared to 11% of patients in the control group. The risk of progressive disease (PD) or death was 79% lower in the 177 Lu-DOTATATE group and the objective response rate (ORR) was significantly higher in the 177 Lu-DOTATATE group (18%) than in the control group (3%). Sixteen percent of the patients treated with 177 Lu-DOTATATE showed PD compared to 56% of the control group [27]. After a follow-up of 76 months in the intention-totreat population, the median overall survival (OS) was 48 months in the 177 Lu-DOTATATE group, which did not significantly differ from the median OS of 36 months in the control group. The similarity in OS between both groups may be explained by the fact that 36% of the patients in the control group crossed over to the PRRT arm during follow-up [30]. In the largest phase II prospective trial to date, Brabander et al. included patients treated with PRRT with 177 Lu-DOTATATE for GEP-NETs, bronchial NET and NET of unknown primary origin. A median PFS of 29 months and a median OS of 63 months were observed in 443 patients with bronchial and GEP-NETs who received a cumulative activity of 22.2-29.6 GBq (600-800 mCi) 177 Lu-DOTATATE. An ORR of 39% after PRRT was found, whereas stable disease (SD) was observed in 43% of the patients [29]. In other phase II trials, retrospective studies and meta-analyses of PRRT with 177 Lu-DOTATATE, similar survival and response data were found [31][32][33][34][35][36]. An overview of these results is given in Table 1. Although PRRT with 177 Lu-DOTATATE is the best option for treatment of patients with advanced GEP-NETs at this moment, complete response rates are still rare (1-2%) [27,29].
Symptom Control
PRRT is not only effective in reducing tumor growth, but it can also lead to symptomatic improvement in NET patients, including associated hormonal syndromes [37]. In the NETTER-1 trial, PRRT with 177 Lu-DOTATATE provided a significant benefit in quality of life (QoL) compared to octreotide LAR. Time to deterioration (TTD) was significantly longer in the 177 Lu-DOTATATE group, with positive outcomes in the domains of global health, physical functioning, diarrhea, pain, body image, disease-related worries and fatigue [38]. In a prospective series of 265 GEP-and bronchial NET patients, QoL, performance status and symptoms (particularly insomnia, appetite loss and diarrhea) improved significantly after PRRT [39]. Importantly, QoL in asymptomatic NET patients did not decline during therapy. In a single center study of 144 patients, a symptomatic response to PRRT with regard to diarrhea, abdominal pain, flushing and fatigue was observed in 70%, 63%, 64% and 53%, respectively [33]. Results are summarized in Table 1. In patients with a functioning NET, PRRT has shown to be an effective treatment regarding symptom control and circulating hormone levels [40,41]. Carcinoid syndrome (CS) is the most prevalent hormonal NET syndrome and is caused by excretion of hormones and amines such as serotonin, histamine, catecholamines, prostaglandins and tachykinins. In a study involving 22 patients with refractory CS who received PRRT for symptomatic control, flushing and bowel movement frequency improved significantly [42]. Two-thirds of the patients who had at least two episodes of flushing per day had a minimal decrease of 50% of these episodes. Of patients with bowel movement frequency of at least four times a day, 47% experienced > 30% decrease, while 29% experienced > 50% decrease [37]. In a systematic review, symptomatic improvement after PRRT was observed in 74% of patients with diarrhea and in 6% of patients with flushing [43]. Together, these results have positioned PRRT with 177 Lu-DOTATATE as a viable option for refractory CS [44]. In addition, case reports have shown improvement of symptoms and echocardiographic parameters of carcinoid heart disease after PRRT, which is a severe symptom of CS [45]. Similar positive effects of PRRT on hormonal levels and symptoms have been observed in patients with functioning panNET syndromes, such as insulinoma, gastrinoma, glucagonoma and VIPoma [40].
Prediction of Response to PRRT
There are a variety of NET subtypes with differences in prognosis and response to PRRT. GEP-NETs show higher PFS (25-30 months) and OS (39-71 months) after PRRT than non GEP-NETs (12-29 months PFS and 39-53 months OS) [28,29]. Within the group of GEP-NETs, panNETs and small intestinal NETs are the most prevalent. After PRRT, a median PFS of 30 months for both groups was observed with a median OS of 71 months in panNETs and 60 months in midgut NETs [29]. The outcome of PRRT has also been associated with the extent of tumor burden and in particular the hepatic tumor load [46]. An extensive tumor burden correlated with an increased rate of PD after PRRT and a significantly shorter OS [29,46]. Patients of whom the liver had over 50% tumor burden showed significantly lower OS and PFS in phase II trials [47,48]. Conversely, in the NETTER-1 trial, total liver tumor burden was found not to correlate with PFS, whereas the presence of at least one large target lesion of > 3 cm was associated negatively with PFS [49]. Additionally, higher uptake on SSTR imaging correlated significantly with better response rates after PRRT [50]. Recently, the cumulative activity of 177 Lu-DOTATATE was found to be an independent predictive factor for survival after PRRT. In 130 patients who received a reduced activity of 177 Lu-DOTATATE because of disease-unrelated causes (such as bone marrow and renal toxicity), PFS, OS and response rates were all lower as compared to 350 control patients who received 800 mCi 177 Lu-DOTATATE [51]. The association between PFS and OS and reduced administered activity of 177 Lu-DOTATATE persisted in multivariable Cox regression analyses in which relevant confounders were included. 18 F-FDG PET/CT has limited added value in the assessment for PRRT in well-differentiated NETs. Although neuroendocrine neoplasm (NEN) patients with positive uptake on 18 F-FDG PET/CT had a significantly shorter PFS than patients with a negative scan [52], this modality is likely more a prognostic than predictive marker [53]. A grading score has been developed to integrate the results of both 18 F-FDG PET/CT and SSA PET imaging ranging from non-aggressive somatostatin receptor imaging (SRI)-positive, FDG-negative to aggressive FDG-positive, SRI-negative NETs. This NETPET score was found to be a significant prognostic marker of OS in patients with GEP-NETs or bronchial NETs, but its predictive role in PRRT is unknown [54,55].
Predictive Markers
The biochemical response to treatment in NETs has often been evaluated by serum chromogranin A (CgA) levels. An increased level is associated with hepatic tumor burden, rapid tumor progression and a shorter OS [56]. In patients who received PRRT, a decrease in CgA of >50% was observed in 25-52% of patients [33,36,57]. Patients with a significant decline in CgA levels after PRRT displayed prolonged PFS and OS [57]. Changes in CgA levels during PRRT should be interpreted with caution though, as an increase in CgA can be caused by both tumor progression and cell damage or lysis by PRRT [58]. The PRRT predictive quotient (PPQ) is a circulating transcript assay that can predict the response of GEP-NETs and bronchopulmonary NETs. A positive PPQ measured before the start of PRRT accurately predicted the outcome of disease control after PRRT in 97% of cases, whereas a negative PPQ was associated with PD after PRRT in 61% of cases at 2 months and 89% at 6-9 months after PRRT [59]. Levels of the NETest, a circulating transcript analysis of 51 genes, during and after PRRT were found to be associated with response according to RECIST on imaging [60]. A lack of decrease in NETest levels at the fourth PRRT cycle signified the advent of PD in advance of imaging, with an overall diagnostic accuracy of 93% to predict response [61].
PRRT Protocol
The eligibility of a patient to receive PRRT should be discussed in a multidisciplinary team in a specialized NET center for each individual patient. Although multiple protocols for PRRT with 177 Lu-DOTATATE have been proposed, the most commonly used treatment schedule entails four cycles of 7.4 GBq 177 Lu-DOTATATE, which is based on the Rotterdam Erasmus MC protocol and also used in the NETTER-1 study [27,29]. The interval between cycles is 6 to 10 weeks. In case of toxicity, this interval can be extended up to 16 weeks [2,3,27]. In parallel with the administration of 177 Lu-DOTATATE, an amino acid solution of 2.5% arginine and lysine in 1 L saline is co-infused in order to prevent renal toxicity. This infusion starts approximately 30-60 min before the administration of 177 Lu-DOTATATE with a total infusion time of 4 h. The potential volume overload of the amino acid infusion in decompensated carcinoid heart disease patients can generally be managed by a prolonged infusion period or the administration of loop diuretics [62]. The amino acid infusion can cause nausea for which an antiemetic, typically ondansetron or granisetron, should be given prophylactically before the start of the infusion. 177 Lu-DOTATATE allows for post-therapy scintigraphy with planar imaging or SPECT/CT. At patient discharge, the radiation exposure should be measured and patients should receive tailored advice on the duration of radiation safety precautions at home, to avoid a high radiation exposure to other people, particularly children and pregnant women. Patients with NET-associated hormonal syndromes who have an indication for continuation of SSA use should adjust the moment of the injections to the PRRT cycles. Long-acting SSA should not be given within 4-6 weeks before a cycle of PRRT because of interference with the radiolabeled SSA. Although there is conflicting evidence from two limited single center studies whether continuation of SSA treatment is beneficial in non-functioning NETs [63,64], this practice is often adopted. If the patient suffers from severe hormonal symptoms, short-acting SSA can be used to bridge this period up till 24 h before PRRT. Radiopharmaceuticals such as 177 Lu-DOTATATE need to be administered at specialized facilities by medical personnel trained in radiation safety. These facilities should adhere to national and international regulations on the use of radiopharmaceuticals and be licensed by the regulatory authorities. Depending on local protocol and exposure regulations, PRRT with 177 Lu-DOTATATE can be given in an in-patient as well as an out-patient setting. In between cycles, patients should be reviewed for adverse effects, including full blood count and renal and liver function. Response evaluation by cross-sectional imaging is usually performed 2-3 and 6 months after the last cycle of PRRT. Long-term follow-up is determined on an individual basis taking into account the tumor biology and therapeutic response [65]. Pseudo-progression is a phenomenon that should be considered in the response evaluation when an increase in tumor size is seen during or short after treatment with PRRT. Pseudo-progression is probably based on localized, temporarily edema caused by inflammation as a response to PRRT and does not show the actual tumor response to the therapy [58]. When pseudo-progression is suspected, functional imaging (for example, PET/CT) can help differentiate between true progression and pseudo-progression [31].
Salvage PRRT
In NET patients who showed tumor response at least 18 months after the first cycle of 177 Lu-DOTATATE, re-treatment with PRRT (R-PRRT) with two additional cycles of 7.4 GBq each after renewed PD has shown antitumoral effects. In a meta-analysis on the effect of R-PRRT, the pooled median PFS was 14 months with a pooled median OS of 27 months. Similarly, the pooled ORR was 17% with a disease control rate of 77%. Response rates, PFS and OS were lower than for initial PRRT [66], nonetheless R-PRRT remains a potential option for GEP-NET patients when other systemic treatment options are limited. The limited efficacy of R-PRRT as compared to initial PRRT might be explained by the administration of lower cumulative activity (i.e., generally half of the initial PRRT dosage) [51], the increase in tumor bulk at baseline before R-PRRT and potential changes in the tumor biology, such as a longitudinal increase in Ki-67. In the largest study to date by van der Zwan et al., no difference in toxicity after R-PRRT as compared with initial PRRT was observed, particularly no increased occurrence of nephrotoxicity or significant hematological disease [67]. In cases where R-PRRT has provided additional benefit on tumor response and prolonged PFS, further re-treatment at the time of progression can be considered [67].
Patient Selection
PRRT with 177 Lu-DOTATATE is registered for patients with GEP-NETs that are progressive on SSA treatment. Non-radiolabeled or 'cold' long-acting SSAs such as octreotide LAR and lanreotide are generally considered the first-line systemic treatment option for advanced and metastatic SSTR-positive GEP-NETs, particularly for tumors with a Ki-67 index below 10% [68]. Among the different second-line options in well-differentiated panNETs, 177 Lu-DOTATATE seems to compare favorably to targeted therapy (everolimus or sunitinib) or chemotherapy [69]. In the randomized phase II OCLURANDUM trial, PRRT with 177 Lu-DOTATATE was associated with a higher median PFS of 20.7 months compared to 11.0 months for sunitinib treatment in patients with advanced, SSTR-positive, progressive panNET [70]. Currently, the randomized phase III COMPETE trial is comparing PRRT with 177 Lu-edotreotide with everolimus in advanced, progressive GEP-NET patients. A subset of GEP-NET patients present with extensive tumor bulk or high proliferative rate (Ki-67 index of 10-55%). In these cases, treatment with octreotide LAR or lanreotide has questionable antiproliferative effects [71]. Given its ORR of 39%, which increases to 55% in panNETs [29], 177 Lu-DOTATATE can be considered as first-line therapy if response is clinically necessitated [72]. These response rates compare favorably to targeted therapy [73][74][75] and for panNETs appear similar to capecitabine-temozolomide chemotherapy [76]. Poorly differentiated neuroendocrine carcinomas (NECs) and well-differentiated grade 3 NETs are high-grade NENs that display a more aggressive biological behavior than the more common grade 1 and 2 NETs [77]. PRRT is currently not considered a standard treatment option for high-grade NENs [68,78]. The rate of SSTR 2 expression in grade 3 NETs ranges from 67-92% and in NECs from 32-50%, compared to a positive expression rate in grade 1 and 2 NETs ranging from 67-96% [5,[79][80][81][82]. In a meta-analysis of PRRT comprising four studies, grade 3 NET patients had a median PFS of 19 months and median OS of 44 months after PRRT. The median PFS was 11 and 4 months and the median OS was 22 and 9 months for NEC with a Ki-67 of 21-55% and NEC with a Ki-67 above 55%, respectively [83]. Recent studies implicated that PRRT could be considered in grade 3 GEP-NETs and GEP-NEC with a Ki-67 of 21-55%. Importantly, to qualify for PRRT, uptake in all lesions is required on somatostatin receptor imaging [83].
Eligibility Criteria for PRRT
There are several inclusion and exclusion criteria to decide if a patient is eligible for treatment with PRRT. A key criterion for PRRT is the degree of uptake on SSTR imaging which is scored by the Krenning score based on planar 111 In-DTPA-octereotide imaging. It was reported that 68 Ga-DOTA-SSA PET/CT results in higher Krenning scores than 111 In-DTPA-octereotide imaging [84]. The uptake of all tumor lesions should minimally be equal to the physiological uptake in the liver (Krenning score grade 2) on 111 In-DTPAoctreotide scintigraphy or higher than the physiological uptake in the liver on 68 Ga-DOTA-SSA PET/CT for a patient to be eligible for PRRT. The latter functional imaging is more accurate for detecting SSTR-positive primary tumors and metastases and therefore superior for assessing the total extent of disease [85][86][87]. PRRT is only applicable for patients with a Karnofsky Performance Scale of at least 60 [27]. PRRT is contra-indicated for patients who are pregnant or breastfeeding, patients with severe cardiac impairment (NYHA III or IV) and those with a life expectancy less than 3 months [31]. Since PRRT can induce toxicity, the following pre-treatment laboratory values are required: creatinine clearance > 40 mL/min, hemoglobin levels ≥ 6 mmol/L, leukocytes > 2 × 10 9 /L, platelet count > 75 × 10 9 /L, bilirubin, alanine aminotransferase (ALAT) and aspartate aminotransferase (ASAT) < 3 times the upper limit of normal and albumin > 3 g/dL [2,3].
PRRT-Related Toxicity
Adverse events related to PRRT are frequently mild and include nausea, abdominal pain and asthenia [27,50]. Increased hair loss is observed in up to 60% of patients treated with 177 Lu-DOTATATE, but this is temporary and seldom leads to baldness [88]. Besides these mild adverse events, PRRT can induce more severe toxicities which can be dose-limiting and adjustments to the treatment schedule might be required [89]. When the toxicity has subsided within 16 weeks after the last dose, guidelines advise to administer half of the original activity of 177 Lu-DOTATATE during the next cycle. PRRT should be discontinued when the toxicity persists after 16 weeks or recurs after the half dose [3]. In the NETTER-1 trial, 7% of the patients received a reduced dose because of dose-limiting toxicities [27]. The kidneys and the bone marrow are the critical organs for dose-limiting toxicities.
Hematological Toxicity
PRRT can induce hematological toxicity through bone marrow radiation. The vast majority of patients only have mild and reversible hematological toxicity with a nadir at 4-6 weeks after administration of PRRT [27,36,50]. However, grade 3 or 4 neutropenia, thrombocytopenia or leukopenia have been observed in, respectively, 1%, 2% and 1% of the patients treated with 177 Lu-DOTATATE in the NETTER-1 trial [27]. PRRT-induced severe lymphopenia is the most common hematological toxicity [27,29], but it has not been associated with increased susceptibility for infections [90]. Thrombocytopenia is the most common cause of dose reduction in PRRT, whereas bleeding complications are rare [27]. Caution should be taken in patients with widespread bone metastases (Figure 2A), due to the risk of persistent cytopenia. In the absence of alternative treatment options, PRRT should preferably be initiated at half the regular activity (3.7 GBq) 177 Lu-DOTATATE in these patients. Additionally, there is a relevant long-term risk of 2% for the development of myelodysplastic syndrome (MDS) and 1% for acute myeloid leukemia (AML) after PRRT [27,29,91]. Little is known about the pathophysiology of persistent hematological toxicity, but a role for clonal hematopoiesis has been postulated [92]. Known risk factors for severe hematological toxicity include decreased renal function, pre-existent cytopenias, extensive tumor mass, age over 70, extensive bone metastases and pre-treatment with myelotoxic chemotherapy [91,93,94]. Additionally, women are at higher risk for developing subacute grade ≥ 2 thrombocytopenia than men, which was independent from other risk factors in a multivariable analysis [89].
Nephrotoxicity
Due to SSTR expression in the kidneys and the renal excretion of radiolabeled SSAs, the kidneys receive a high radiation dose during PRRT. Following glomerular filtration, SSAs are reabsorbed in the proximal tubuli of the renal cortex because of active transport mechanisms [95]. By saturating this re-uptake mechanism through the use of lysine and arginine, re-absorption of radiolabeled peptides can be significantly reduced. This results in less radiation-induced nephrotoxicity by a reduction of the absorbed kidney dose up to 40% [96,97]. However, nephrotoxicity after PRRT still occurs such as tubulointerstitial scarring, atrophy and thrombotic microangiopathy [98]. In the NETTER-1 study, grade ≥ 3 renal toxicity was observed in 5% of the 177 Lu-DOTATATE group and in 4% of the control group [30]. Bergsma et al. reported an overall creatinine clearance loss of 3.4% 1 year after PRRT. No subacute grade ≥ 3 renal toxicity was seen and 1.5% of the patients showed grade 3 renal toxicity in the long term. However, all these patients had a creatinine clearance of <60 mL/min at baseline [99]. A reduced kidney function can lead to a delayed renal excretion of 177 Lu-DOTATATE and this has also been associated with a higher risk of hematological toxicity [95]. Risk factors associated with renal toxicity include age > 60 years, hypertension, diabetes mellitus, pre-existing renal disease, cumulative radiation dose to the kidneys, previous nephrotoxic chemotherapy, tumor or metastases close to the kidney and previous PRRT with 90 Y-DOTATOC [100]. Post-renal obstruction can be observed in some GEP-NET patients, particularly in those with retroperitoneal or pelvic metastases, but this can also be caused by the primary tumor, nephrolithiasis and abdominal or retroperitoneal fibrosis [101]. Post-renal obstruction can lead to reduced renal excretion of 177 Lu-DOTATATE and thereby increase the exposure of radiation to the kidneys, risking (permanent) deterioration of renal function ( Figure 2B) [102,103]. In these patients, it is necessary to correct the hydronephrosis of a functional kidney before the start of PRRT to lower the risk of radiation-induced toxicity, for instance with a double J urethral stent or percutaneous nephrostomy [101,104].
Hepatotoxicity
Mild and severe hepatotoxicity has been observed in, respectively, 12% and 0.4-2.5% of patients after PRRT [28,33]. Based on clinical experience, liver failure after PRRT can
Hepatotoxicity
Mild and severe hepatotoxicity has been observed in, respectively, 12% and 0.4-2.5% of patients after PRRT [28,33]. Based on clinical experience, liver failure after PRRT can occur in cases of severe (>90%) liver involvement of NET metastases (example of a patient with severe liver involvement with an enlarged risk of liver failure in Figure 2C). Moreover, there is a concern about the cumulative hepatotoxicity in patients treated with a combination of PRRT and radioembolization. In the HEPAR phase I trial, one patient died of liver failure after treatment with a dose of 80 Gy 166 Ho-radioembolization [105]. No liver failure was observed in 67 patients undergoing 90 Y-radioembolization after PRRT [106,107]. Adherence to eligibility criteria is crucial and in borderline cases a first trial infusion with 3.7 GBq 177 Lu-DOTATATE is advised. There was no relevant hepatotoxicity reported in the NETTER-1 trial [27]. Grade 3 and 4 toxicity of aminotransferases was observed in 3% in the phase II trial by Brabander et al. Three of these patients had persisting toxicity after 3 months, but no hepatic failure was observed [29].
Hormonal Crisis Due to PRRT
A hormonal crisis can be triggered by PRRT due to excessive release of metabolically active amines or peptides from the tumor. A PRRT-induced hormonal crisis was observed in 1% of patients and can occur during the infusion of the radiolabeled SSA up to 48 h after infusion [108]. Risk factors for a hormonal crisis include CS, elevated 5-hydroxyindoleacetic acid (5-HIAA) and CgA levels, metastatic disease, high tumor burden, higher age and histamine release due to medication as β2 agonist bronchodilators. Prophylactic measures consist of controlling the CS before initiation of PRRT, having a good nutritional status and reducing the time without non-radiolabeled SSAs. When a hormonal crisis occurs, symptom control is necessary. The treatment should be hormone-specific and may also include general measures such as monitoring of vital parameters, saline infusion, correction of electrolyte disturbances and inpatient observation if needed. SSAs can be restarted safely 1 h after infusion of 177 Lu-DOTATATE and should be considered in severe cases of SSA-responsive CS, gastrinoma, insulinoma and VIPoma. Post-infusion administration of corticosteroids can be considered in patients at risk for carcinoid crisis or those presenting with hormonal crisis [109].
Nervous System Location or Compression
GEP-NETs rarely metastasize to the central nervous system, with an incidence of brain metastases of 1.5-5% in NET patients [110,111]. More frequently, vertebral bone metastases can compress the spinal cord or its nerve roots ( Figure 2D) [112]. PRRT can reduce tumor volume and metabolic activity of central nervous system metastases, thereby reducing pain and spinal compression [113]. The radiation-induced damage by PRRT can be accompanied by temporary edema, leading to increased compression of surrounding vital structures and potentially (worsening of) neurological symptoms. In cases of brain metastases or for those patients at risk of spinal cord or nerve compression, prophylactic therapy with corticosteroids can be considered to prevent or minimize treatment-induced edema. Some expert centers have used a one-week regimen of 4 mg dexamethasone two times a day. Caution is required with this treatment since it is possible that the use of corticosteroids negatively influences the uptake of SSA [114].
Mesenteric Fibrosis and Mesenteric or Peritoneal Metastases
Another midgut NET-related complication is the presence of mesenteric fibrosis, which occurs in up to half of the patients with a small intestinal NET metastatic to the mesenteric lymph nodes. Mesenteric metastases can induce local fibrosis by the secretion of serotonin and other mediators, inducing a desmoplastic reaction, which subsequently can lead to intestinal obstruction and ischemia ( Figure 2E) [115,116]. Peritoneal metastases can lead to adhesions between the abdominal wall and the bowel which can cause intestinal immobility, bowel obstruction or 'frozen abdomen' (Figure 2F) [117]. The proportions of patients treated with PRRT who had mesenteric and/or peritoneal metastases ranged from 6-51% [27,117,118]. In a retrospective study of 132 NET patients with a mesenteric mass, 177 Lu-DOTATATE resulted in an ORR of this mass in only 4% of the patients [119]. In an analysis of NET patients with mesenteric or peritoneal disease, 23% of patients at high risk for complications experienced at least one episode of bowel obstruction within three months of PRRT. These patients were treated with corticosteroids, surgery and total parenteral nutrition [117]. Although a direct causal relationship has not been established, treating physicians should be aware that PRRT seems to be ineffective in reducing the size of a mesenteric mass while there may be an increased risk of abdominal complications in patients with extensive peritoneal disease or desmoplastic changes associated with mesenteric metastases. A post-infusion trial of corticosteroids can be considered in patients at high risk of these complications [117].
Conclusions and Future Directions
PRRT with 177 Lu-DOTATATE is a widely applicable therapy for patients with advanced and metastatic NETs. It is clear that PRRT contributes to reduction of tumor growth and stabilization of the disease in a significant proportion of these patients. Next, PRRT improves QOL and reduces symptoms [27,29]. Despite the successes, it is important that clinicians treating NET patients are aware of the toxicities (mainly renal and hematological) associated with PRRT and discuss every patient in a multidisciplinary team where careful selection, preparation and management of these patients before and during therapy with this unique radiopharmaceutical can be discussed. Despites its successes, there still is room for improvement of the efficacy and safety of PRRT. Ongoing research projects are exploring how to improve the current strategies. A combination therapy with radiosensitizers is a way to increase the effect of PRRT. This can be conducted by the inhibition of poly ADP-ribose polymerase (PARP)-1 which is essential for the repair of SSBs [120] or inhibition of heat shock protein 90 (HSP90), a chaperone molecule that plays a role in cell protection and maintenance [121]. Other radionuclides are being studied for potential use in PRRT, such as actinium-225, lead-212 and bismuth-213. These alpha-emitting radionuclides can potentially cause more damage to the tumor cells compared to PRRT with 177 Lu-DOTATATE and thus increase treatment efficacy [122][123][124]. Altogether, PRRT with 177 Lu-DOTATATE is a very potent treatment modality for patients with progressive well-differentiated (grade 1-2) GEP-NET and there are many different strategies currently being explored that can contribute even to better patient care.
|
v2
|
2022-11-27T17:31:25.882Z
|
2022-11-24T00:00:00.000Z
|
253986262
|
s2ag/train
|
Health risks of environmental exposure to microplastics
Plastics are materials widely used in all sectors. The subject of interest in recent years has become so-called microplastics, whose composition and structure are causing new environmental hazards. The presence and accumulation of microplastics in the environment threaten the ecological balance, the water environment, food sustainability and safety, and ultimately human health. Human exposure to microplastics is primarily through the oral route, so the main source of human exposure to microplastics is diet. Despite many studies focusing on microplastic contamination in seafood, fish, and shellfish, estimating total human exposure to microplastics via the oral route is difficult, due to the lack of research on other foods in this area. The risks to human health from inhaling microplastics remain unclear. According to the WHO, there is no reliable evidence of the harmful effects of microplastic on the human body, but the phenomenon requires further research. Likely health effects of human exposure to microplastic include respiratory and gastrointestinal effects, oxidative stress, and cancer. There is a need to raise public awareness about environmental exposure to microplastics and effective waste management.
|
v2
|
2022-11-28T06:42:08.600Z
|
2022-11-24T00:00:00.000Z
|
254018209
|
s2orc/train
|
Intensity modulated proton arc therapy via geometry-based energy selection for ependymoma
We developed a novel method of creating intensity modulated proton arc therapy (IMPAT) plans that uses computing resources efficiently and may offer a dosimetric benefit for patients with ependymoma or similar tumor geometries. Our IMPAT planning method consists of a geometry-based energy selection step with major scanning spot contributions as inputs computed using ray-tracing and single-Gaussian approximation of lateral spot profiles. Based on the geometric relation of scanning spots and dose voxels, our energy selection module selects a minimum set of energy layers at each gantry angle such that each target voxel is covered by sufficient scanning spots as specified by the planner, with dose contributions above the specified threshold. Finally, IMPAT plans are generated by robustly optimizing scanning spots of the selected energy layers using a commercial proton treatment planning system. The IMPAT plan quality was assessed for four ependymoma patients. Reference three-field IMPT plans were created with similar planning objective functions and compared with the IMPAT plans. In all plans, the prescribed dose covered 95% of the clinical target volume (CTV) while maintaining similar maximum doses for the brainstem. While IMPAT and IMPT achieved comparable plan robustness, the IMPAT plans achieved better homogeneity and conformity than the IMPT plans. The IMPAT plans also exhibited higher relative biological effectiveness (RBE) enhancement than did the corresponding reference IMPT plans for the CTV in all four patients and brainstem in three of them. The proposed method demonstrated potential as an efficient technique for IMPAT planning and may offer a dosimetric benefit for patients with ependymoma or tumors in close proximity to critical organs. IMPAT plans created using this method had elevated RBE enhancement associated with increased linear energy transfer.
| INTRODUCTION
Proton arc therapy with spot scanning has been shown to improve the treatment plan conformity and surrounding organ at risk (OAR) sparing [1][2][3][4] . However, it is limited by a high computational cost for treatment planning. Optimization of all scanning spots from a relatively large number of beam angles may require high memory usage and long computation time. Energy-switching time is another limiting factor; for a synchrotron-type proton accelerator in particular, energy switching may take up to seconds. [5][6][7] Sanchez-Parcerisa et al. 8 compared different methods of proton arc treatment planning that select one or two energies per beam angle based on water-equivalent range relative to the distal and proximal limits of the target region. However, this oversimplified energy selection strategy may not be sufficient for tumors with complex shape and heterogeneous anatomical surroundings in terms of clinically acceptable dose coverage. Ding et al 1 iteratively increased the sampling frequency of beam angles, during which process energy layers were redistributed to newly created angles. Then a filtration process was used to remove low-weighted energy layers. Robust optimization was integrated in the planning iterations. They reported on clinical applications of their proton arc therapy planning approach 9 .
Brain tumors such as ependymoma often involve treatment targets located near or overlapping critical organs, such as the brainstem. Authors have reported that proton therapy may be associated with increased radiation necrosis 10, 11 . Further analysis of differential damage induced by protons and photons using imaging changes as biomarkers suggested that increased linear energy transfer (LET) contributes to a higher incidence of radiation damage to brain tissue in the proton therapy cohort 12,13 . One of the strategies to mitigate the relative biological effectiveness (RBE) of high proton LET in treating brain tumors is to include an increasing number of beam angles in the treatment plan, with the hope of spreading out high-LET protons that are mostly located at the distal ends of the beams. This may lead one to intuitively expect proton arc therapy to naturally have this advantage to a great extent. Recent studies 14,15 have shown that proton arc therapy may offer greater flexibilities than conventional IMPT in limiting high-LET proton irradiation within the target by adopting an energy selection method that stops proton beams mostly at the interior target.
As part of the effort to develop the proton arc therapy capability at our center, we aimed to achieve a simplified workflow for proton arc therapy planning to encourage further works on this topic. Our hypothesis here is that proton arc therapy plans can be generated to potentially offer clinical benefits, with a simplified workflow, and reduced computing resource usage compared to the existing methodologies. Herein we report on our novel method of proton arc treatment planning based on an energy selection process, the only added step to the regular intensity-modulated proton therapy (IMPT) planning workflow, that we developed to create high-quality, robust proton arc therapy plans, potentially with limited computing resources. Here we refer to spot-scanning-based proton arc therapy as intensity modulated proton arc therapy or IMPAT. We also report on the in silico application of our method to treatment of ependymoma in four patients.
| Treatment Planning Optimizer
Our IMPAT planning method was developed and preliminarily implemented with in-house program modules, and the beam-line calculation and optimizer of the commercial proton treatment planning system (TPS) RayStation® (RaySearch Laboratories AB, Stockholm, Sweden). Our energy selection process relied on RayStation providing the initial full spectrum of energy layers that longitudinally cover the whole treatment target with proper spacings at each gantry angle. After energy selection, the RayStation optimizer produced reasonably acceptable solutions for our treatment planning tasks with a relatively large number of gantry angles but with a greatly reduced number of energy layers at each angle. Robust proton plan optimizations were applied with a setup uncertainty ±2 mm in x, y and z directions and range uncertainty ±2.5% for all clinical cases in this study. Robustness objectives were used for targets for robust coverage of targets.
| Beam arrangement
The proton arc was formed by consecutively placing gantry angles at 5° intervals to form a proper partial arc. For ependymoma treatment, start and end gantry angles of the partial arc were selected to avoid passing beams through optic organs, nasal and oral cavities, and cochleae. This is to achieve the best protection possible to those normal tissues and organs. To facilitate the assessment of IMPAT plan quality, the plans were compared side by side with reference three-field IMPT plans with the similar objective functions and optimization processes. IMPT beam angles were selected based on the similar principles to minimize passing beams through critical organs as well as to follow the target shape.
| Geometry-based energy selection
The water-equivalent path length was calculated for all dose voxels within the treatment target plus margins at each gantry angle by ray-tracing with computed tomography (CT) data. The resolution of the dose grid was set to 2 mm. The calibration data that map CT Hounsfield units to relative stopping power ratios to water were obtained from beam configurations of the TPS. For each dose voxel, our energy selection process collects all scanning spots with dose contributions above a specified fraction of the maximum spot dose at the Bragg peak ( Figure 1). In the present study, a fractional threshold of 0.8 allowed for a sufficiently large pool of scanning spots with reasonable memory usage (<500 MB per patient for all four patients). The above data can be denoted as the dose deposition matrix D with its element Dij representing the relative dose received by dose voxel i from scanning spot j. Specifically, we call D the major dose deposition matrix to reflect the fact that only dose contributions above the threshold are included in D. Because only the major dose contributions from scanning spots had to be counted, dose contributions to dose voxels were estimated using lateral spot profiles modeled by the single Gaussian function according to the lateral distance between the voxel and the central axis of a scanning spot 16,17 . The sigma of the single Gaussian function was quantified based on in-air spot sizes, and in-medium multiple Coulomb scattering. The multiple Coulomb scattering was calculated using the generalized Highland approximation. 18, 19 One may consider to take the proton plan robustness into consideration by including a set of beam perturbation scenarios during ray tracing and scanning-spot searching. However, this did not result in a noticeable improvement in our preliminary implementation and therefore was not applied in this study.
From the initial set of energy layers provided by the TPS , our energy selection module picks a subset of energy layers at each gantry angle such that the total number of scanning spots with dose contributions above a threshold fraction to each dose voxel meets or exceeds a specified minimum number, or minimum peak repetition. A minimum peak repetition of 7 was used in generating IMPAT plans for all four patients. With collections of scanning spots above the dose threshold, our energy selection is carried out in iterations. During each iteration, energy layers from all gantry angles are sorted according to the number of dose voxels that receive dose contributions above the fractional threshold from at least one scanning spot on the examined layer, and the energy layer with dose contributions above the threshold to the most dose voxels is selected. When sorting energy layers at each iteration, the energy layers selected in previous iterations are excluded from the sorting, as are the dose voxels that already satisfy the minimum peak repetition. The energy selection stops when all dose voxels satisfy the minimum peak repetition. Figure 2 illustrates the workflow of the above process.
| Patient study
Four ependymoma patients were selected to demonstrate IMPAT planning results using the proposed energy selection method. The prescription dose was 5400 cGy to be delivered in 30 fractions.
Raystation distinguishes between objective functions and constraints such that constraints must not be violated, whereas objective functions are to be minimized. In comparing the quality of the IMPAT and regular IMPT plans, the aim was to achieve optimal clinical target volume (CTV) coverage without violating the same maximum dose constraint on the brainstem. For this study, a constraint value of 5450 cGy was applied in plan optimization for all four patients. By maintaining the same brainstem constraint, the doses to other OARs were decreased to as much as possible without dramatically compromising the CTV coverage. When possible, CTV coverage was normalized such that the prescription dose covered 95% of the CTV while maintaining the similar brainstem dose.
Treatment plan quality was reported using dose-volume histograms (DVHs), plan robustness, the homogeneity index, and the conformity index. Plan robustness is shown qualitatively in band DVH graphs. The homogeneity index was calculated using the following formula described in Wu, et al. 20 : where and are the minimum doses received by 98% and 2% of the target volume, respectively, and is the prescription dose. The conformity index or conformation number is calculated as approximately as follows: where is the physical dose and the LET dependency of the RBE factor is approximated as 1 + .
With the constant set to 0.04 µm/keV, the RBE factor equals 1.1, a constant RBE factor adopted by many proton facilities, at the center of a spread-out Bragg peak of 5-cm modulation and 10-cm range where the dose-averaged LET is about 2.5 keV/µm. The part of the RBE dose enhanced by variable proton LET can be represented by term ; the RBE dose enhancement was compared for the IMPAT and IMPT plans.
| RESULTS
The CTV volumes, CTV and brainstem overlaps, and field arrangements of IMPAT and IMPT plans for the patients in this study are listed in Table 1
| DISCUSSION
Proton treatment plan optimization is often a resource-intensive process in terms of computing memory and speed. IMPAT considerably magnifies this demand. Our novel IMPAT planning method first determines a minimum set of energy layers per gantry angle based on ray tracing, and single Gaussian lateral spot profiles, which is followed by proton treatment plan optimization. Therefore, it can be used The initial full spectrum of energy layers at each gantry angle is crucial for the success of the energy selection. Although this was found not necessary for each proton arc planning task, the convergence of the energy selection iterations can be checked prior to the selection by counting for each target voxel how many scanning spots from the initial pool of energy layers are contributing spot doses above the specified dose threshold. The count of major contributing spots per target voxel should be greater (normally much greater) than the specified minimum peak repetition, which guarantees the energy selection terminates after finite iterations. If the count is small relative to the specified minimum peak repetition, the initial full spectrum of energy layers should be examined on whether they completely cover the entire treatment target, or spot spacings are too coarse.
The parameter choices of the minimum peak repetition and the fractional dose threshold were made by trial and error with a limited number of clinical cases tested. As more and more cases of various treatment sites are being tested, we expect these parameters can be optimized for specific treatment sites and scenarios. Generally speaking, as the minimum peak repetition is decreased, the decreased plan quality can be expected. As it is increased, however, the energy layers per gantry angle may increase rapidly and the delivery efficiency will be negatively impacted. The choice of the fractional dose threshold considered its impact on the energy selection efficiency, as well as memory usage. If the value is much smaller than 1.0 (e.g., 0.5 or 0.6), the energy selection suffers from too much noise from scanning spots contributing doses from the proximal parts of their single-spot dose profiles too far away from the Bragg peaks. If the value is too close to 1.0, a large number of energy layers may be selected for each field angle, or worse, the energy selection may fail to converge. Although the plan quality may improve to a certain degree with increased energy layers and scanning spots, it may start deteriorating due to the effect of "MU starvation"28 or the optimizer becoming inefficient for the enormously large dimension of the problem. It is possible that at each iteration of the energy selection, multiple energy layers could make major spot dose contributions to the same count of target voxels. With our currently implementation, one layer is selected randomly among those equally contributing layers. Further improvement of our algorithm could be made by considering additional selection criteria, for example, to further improve OAR sparing.
The conformity that a proton treatment plan can achieve depends on multiple factors, such as the target volume and complexity of the target shape. For the patients in the present study, overlap of the CTV and brainstem may have limited the plan conformity. The conformity indices for both IMPAT and IMPT were highest for the patient with the least overlap. The difference in conformity between IMPAT and IMPT also was the smallest for that patient. Although it met clinical dose objectives generally adopted, sparing of cochleae was noticeably worse in the proton arc plan than in the reference IMPT plan for patient 3.
Upon analyzing cochlea doses contributed from each field angle, the added field angles in the proton arc plan appeared to increase the chance of scanning spots being placed close to the cochleae for this patient. In terms of sparing an anatomic region, a proton optimizer generally does not work as effectively as not placing spots in or near that region at all.
In a regular IMPT plan, the optimizer often assigns heavy weights to energy layers that are of or close to the highest energy when each beam is examined individually. Those layers are commonly located at the edge of or even outside the targets. Hence, high-LET regions at the distal ends of heavy-weighted layers are often located outside the targets. Our energy selection process aims to cover targets with sufficient scanning spots with a minimum number of energy layers per beam angle. As a result, energies with ranges reaching around the middle of target depths often have a large number of scanning spots covering a relatively large cross-section of a tumor and tend to be selected with high priority. This is attributed to the fact that energy layers near distal and proximal depths normally have smaller numbers of scanning spots than do those at the middle depths for naturally shaped tumors. This likely contributed to the IMPAT plans having higher LET-associated RBE enhancement in the CTVs than did the corresponding reference IMPT plans. This effect was reported previously by Bertolet et al 14 . Due to overlap of the CTV and brainstem, the RBE enhancement was also elevated in the brainstem for all patients except patient 1, whose overlap was much smaller than that of the other three patients. We chose a gantry angle interval of 5° to sufficiently demonstrate the proposed planning method. No fundamental difficulty should prevent the method from working with finer or coarser gantry angle arrangements.
| CONCLUSIONS
The proposed IMPAT planning method is capable of creating plans of clinically acceptable quality for ependymoma patients. It has potential benefits of enhancing the conformity and homogeneity of target coverage, especially in close proximity to critical OARs. Furthermore, it allows for plan generation that is not much different than that using regular IMPT optimization processes by applying the proposed energy selection step beforehand with cost-efficient computing resource consumption. Further research is warranted to distribute the RBE enhancement, that is potentially associated with our IMPAT planning method, in targets and critical organs in a desirable manner. Ratio of RBE enhancements for the IMPAT to reference IMPT plans in the CTV (ratio of mean enhancements in blue triangles) and brainstem (ratio of D1% in orange circles) in the four study patients.
FIGURE 8
The proton energies (as shown by dots in various colors) picked by the geometry-based energy selection process congregated in narrow energy ranges for majority of field angles; proton energies also varied relatively continuously between adjacent fields. These characteristics may potentially improve delivery efficiency of proton arc plans.
|
v2
|
2022-11-25T14:14:37.774Z
|
2022-11-25T00:00:00.000Z
|
253841777
|
s2orc/train
|
Immunogenomic correlates of immune-related adverse events for anti–programmed cell death 1 therapy
Despite impressive antitumor efficacy of programmed cell death 1 (PD-1) inhibitors, this inhibition can induce mild to severe autoimmune toxicities, termed immune-related adverse events (irAEs). Yet, predictive pretreatment biomarkers for irAEs development across cancer types remain elusive. We first assessed cellular and molecular factors. To determine factors predicting the risk of irAEs for anti–PD-1 immunotherapy across multiple cancer types, an integrative analysis of cellular and molecular factors from 9104 patients across 21 cancer types and 4865522 postmarketing adverse event reports retrieved from adverse event reporting system was then performed. Accuracy of predictions was quantified as Pearson correlation coefficient determined using leave-one-out cross-validation. Independent validation sets included small cell lung cancer and melanoma cohorts. Out of 4865522 eligible adverse events reports, 10412 cases received anti–PD-1 monotherapy, of which, 2997 (28.78%) exhibited at least one irAE. Among established immunogenomic factors, dendritic cells (DC) abundance showed the strongest correlation with irAEs risk, followed by tumor mutational burden (TMB). Further predictive accuracy was achieved by DC and TMB in combination with CD4+ naive T-cells abundance, and then validated in the small cell lung cancer cohort. Additionally, global screening of multiomics data identified 11 novel predictors of irAEs. Of these, IRF4 showed the highest correlation. Best predictive performance was observed in the IRF4 – TCL1A – SHC-pY317 trivariate model. Associations of IRF4 and TCL1A expression with irAEs development were verified in the melanoma cohort receiving immune checkpoint inhibitors. Collectively, pretreatment cellular and molecular irAEs-associated features as well as their combinations are identified regardless of cancer types. These findings may deepen our knowledge of irAEs pathogenesis and, ultimately, aid in early detection of high-risk patients and management of irAEs.
Introduction
Immune checkpoint inhibitors (ICIs) targeting programmed cell death 1 (PD-1) pathway have brought remarkable clinical benefits in diverse cancers (1). The ICIs work by blocking the PD-1 from binding with its partner proteins, resulting in immune activation in the tumor microenvironment (2). Nevertheless, ICI use is commonly associated with autoimmune toxicities, known as immune-related adverse events (irAEs) (3). The most common irAEs are observed in skin, colon, endocrine organs, liver, and lungs, but any organ can be affected and some infrequent irAEs may be serious and fatal, such as encephalitis and myocarditis (3,4). Preexisting autoantibodies (5), gut microbiome (6), tissue-resident Tcells (7), microRNAs (8), and cytokines (6) have all been involved in irAEs in single cancer type. The pathogenesis of irAEs remains poorly characterized and no biomarkers are routinely used in standard clinical practice to recognize patients at high risk for irAEs development.
Although high tumor mutational burden (TMB) has recently been reported to correlate with elevated irAEs risk across cancer types (9), large proportion (> 50%) of variation in irAEs risk has not yet been accounted for during anti-PD-1 therapy, indicating the role of other factors in leading to irAEs. Herein we systematically study this hypothesis, aiming to identify additional pretreatment immunogenomic factors that contribute to irAEs development regardless of cancer types. To this end, we analyzed cleaned large-scale pharmacovigilance data of irAEs from The US Food and Drug Administration's Adverse Event Reporting System (FAERS) and The Cancer Genome Atlas (TCGA) multiomics data from whole-exome sequencing, mRNA sequencing, microRNA sequencing, and reverse phase protein array across multiple cancer types, and lastly, validated our hypothesis in independent cohorts.
Materials and methods
Details about the methods are provided in Supplementary Methods, and a flow chart illustrating main analyses conducted is presented as Supplementary Figure S1.
IrAEs risk evaluation
FAERS is a database of spontaneously gathered adverse event reports, containing great collection of reports of irAEs on anti-PD-1 agents in a real-world situation. A FAERS search engine named OpenVigil (version 2.1) was used to retrieve cleaned adverse event reports (10). Only reports with nivolumab or pembrolizumab as the suspected cause of adverse events were considered. Further, given that overall prevalence of irAEs and severity was higher with combined PD-1 and cytotoxic T-lymphocyte associated protein 4 (CTLA-4) antibodies as compared with monotherapy (11,12), anti-PD-1 agents plus ipilimumab combination therapy was excluded. The irAEs reported in FAERS were defined as 106 preferred terms in the Medical Dictionary for Regulatory Activities according to previously published irAEs management guidelines during ICI therapy (3,4,13,14), and listed in Supplementary Table S1. Lastly, reporting odds ratio (ROR) was calculated for each cancer type to evaluate the risk of a cancer type developing any irAE, which represents standard practice for quantitative analyses of data in spontaneous reporting systems such as FAERS (9,15,16).
Molecular and cellular data sources from TCGA
Datasets of somatic mutations, mRNA, microRNA, and protein expression for 9104 samples across 21 cancer types (Supplementary Figure S1) with available irAEs ROR data were downloaded from the TCGA Pan-Cancer Atlas project hosted in the UCSC Xena Hubs (17). TMB was then calculated as the count of nonsynonymous mutations for each patient based on somatic mutations, and log-transformed. On the basis of the mRNA expression dataset, several tumor immune microenvironmentrelated signatures were generated, including cytolytic index to assess intratumoral cytolytic activity (18), interferon (IFN)gamma and expanded immune signatures (19), and a transcriptional signature reflecting CD8 + T-cells exhaustion (20). Proportion of PD-1-high samples for each cancer type was also determined, with percentile 80 th of PD-1 mRNA expression in the entire TCGA cohort as the cutoff (21).
Other immunogenomic factors, including T cell receptor (TCR) diversity, intratumor heterogeneity, and neoantigen load, were obtained from Genomic Data Commons Pan-Cancer Atlas panimmune data portal (22).
Abundance data of 30 immune cell types in the tumor microenvironment for the TCGA samples were inferred using xCell (23) and downloaded from the xCell website. The abundance was defined as an enrichment score which showed resemblance to the fraction of specific cell type in the tumor microenvironment (23).
Lastly, median values of individual aforementioned immunogenomic factors except the PD-1-high samples proportion were calculated for each cancer type. Raw data of mRNAs, microRNAs, proteins, and phosphoproteins were preprocessed separately, and then their median expression levels per cancer type were determined for further analyses.
Objective response rates across cancer types
Objective response rate (ORR) for PD-1 or its ligand PD-L1 inhibitor monotherapy across 19 types of cancers (Supplementary Figure S1) was compiled from Yarchoan et al. (24). To evaluate the correlation of tumor response with irAEs risk, Pearson correlation coefficient (R) between the ORR and the corresponding irAEs ROR across these cancer types was calculated.
Small cell lung cancer and melanoma cohorts
Given that molecular data for small cell lung cancer is lacking in TCGA but its irAEs ROR could be calculated in our study, we focused on an independent cohort encompassing 71 small cell lung cancer patients with both somatic mutations and mRNA sequencing data of pretreatment tumors available (25). TMB was calculated as previously described. To estimate abundances of immune cells, gene expression data in fragments per kilobase of exon per million reads mapped units was fed to the R package xCell (23). The ICIs-treated cohort in our study consisted of 60 patients with metastatic melanoma, which received either anti-PD-1 blockade (nivolumab or pembrolizumab) or nivolumab plus ipilimumab therapy (26). All irAEs were classified according to the United States Health and Human Services Common Terminology Criteria for Adverse Events v.5.0. Grade 0 reflected no toxicity, and irAEs occurrence was defined as grade 1+. RNA sequencing was performed for peripheral blood mononuclear cell samples obtained at baseline, and then read counts were normalized to gene-level transcripts per million (TPM) for further differential gene expression analyses against irAEs status.
Statistical analysis
To examine the relationships of molecular and cellular factors with irAEs risk, Pearson correlation analysis was used to calculate the Rs between their respective medians and the ROR across the 21 cancer types above. For combinations of irAEs risk-associated factors, a multivariable linear regression analysis with leave-one-out cross-validation in predicting ROR across cancer types was performed using the R package caret. Prediction performance of linear models was determined as R and unexplained variance (1 − R 2 ). Multicollinearity among variables in a multivariable linear model was quantified as variance inflation factor (VIF) calculated using the R package rms; a VIF > 4 was considered as an indicator of multicollinearity. Log-likelihood ratio test was applied to comparing the goodness-of-fit between different models using the R package lmtest. Specifically, the log-likelihood ratio test was conducted between the bivariate model and corresponding single variable models to determine the bivariate model fitness; for the trivariate model fitness comparison, the log-likelihood ratio test was conducted between the trivariate model and corresponding bivariate models. Multiple testing correction was performed to control the false discovery rate (FDR) by the Benjamini-Hochberg method. All P values were 2-sided and statistical significance was expected at FDR <.05 unless stated otherwise.
In the melanoma cohort, Mann-Whitney U test was used to compare the difference in gene expression between irAEs status. To eliminate the possibility that the associaton between gene expression and irAEs status was skewed by ICIs therapy type, a logistic regression model was adopted to control for different therapy classes. All statistical analyses were done in R statistical software v.3.5.2.
Association of established immunogenomic factors with irAEs risk
A total of 4865522 reports were identified in FAERS, including 10412 cases that received the anti-PD-1 monotherapy for 22 cancer types. Of those 10412, 2997 cases (28.78%) exhibited at least one irAE. As shown in Supplementary Figure S2, the irAEs RORs varied between cancer types, ranging from the lowest 0.94 in cholangiocarcinoma to the highest 5.87 in melanoma.
Given that the relationship between irAEs onset and survival advantage of patients on ICIs has been shown in large case series studies in multiple cancers (27), we first examined the correlation of irAEs ROR with ORR. Our analysis demonstrated a significant positive correlation between them (R = 0.51; P = .03) (Supplementary Figure S3). Next, we investigated whether established immunogenomic correlates of response to ICI therapy may associate with irAEs risk. Strong association signals were identified for 15 factors (P <.05 for all), with 12 passing the correction for multiple testing (FDR <.05 for all): 5 immune cells, TMB, 3 immune expression signatures, and 3 checkpoint-related factors. ( Figure 1A). Specifically, the strongest correlation between dendritic cells (DC) abundance and ROR was observed (R = 0.69; FDR = .02) ( Figure 1B . Consistent with the previous study (9), elevated TMB was demonstrated to correlate with increased risk of irAEs (R = 0.63; FDR = .04) ( Figure 1C). Additionally, 3 immune expression signaturesexpanded immune signature, IFN-gamma signature, and cytolytic indexwhich are related to IFN-gamma signaling and activated T cell biology (18,19), displayed significant correlations with ROR ( Supplementary Figures S4E-G). We also found that checkpoint-
Combination of DC, TMB, and CD4 + naive T-cells for irAEs risk prediction
We further investigated whether certain combinations of those 12 correlates of irAEs ROR could provide additional accuracy in predicting irAEs risk. The performances of 66 bivariate models were first evaluated. Of these 31 models showed significant statistical differences compared with their respective univariate models in terms of the fitness (loglikelihood ratio test, P <.05 for all) and no signs of collinearity were detected (VIF < 4 for all) (Figure 2A; Supplementary Table S2). However, only TMB and CD4 + naive T-cells or DC bivariate models outperformed the DC-based univariate model (TMB -CD4 + naive T-cells model, R = 0.73; TMB -DC model, R = 0.71; both FDR = .01) ( Figures 2B, C). Of note, some cancer types, which had RORs higher than would be predicted by the TMB -CD4 + naive T-cells model, exhibited higher DC abundance (e.g., lung adenocarcinoma), and some with lower-than-predicted RORs showed lower DC abundance (e.g., glioma) ( Figure 2B). The same was true for CD4 + naive T-cells abundance in the TMB -DC bivariate model ( Figure 2C). Thus, we next examined whether inclusion of the third variable would aid in contributing additional information beyond the bivariate model. Indeed, of the resulting trivariate models, combined DC, TMB, and CD4 + naive T-cells model achieved the best predictive accuracy (R = 0.81; FDR = 1.1×10 -4 ), and exhibited pronounced model promotion in comparison with their corresponding bivariate models (log-likelihood ratio test, P = 8.7×10 -4 relative to TMB -DC model; P = 2.8×10 -4 relative to TMB -CD4 + naive T-cells model) (Figure 3; Supplementary Table S3). Likewise, there was no sign of collinearity for this trivariate model (Supplementary Table S3). Collectively, these results emphasized the importance of integrating multiple factors in determining irAEs risk.
External validation of DC -TMB -CD4 + naive T-cells model
Having identified candidate composite models of irAEs risk, we next attempted to verify our findings in an independent cohort of small cell lung cancer, a cancer type not depicted in TCGA and known to have high TMB but low response rates to ICIs. As shown in Figure 3B, estimated ROR by univariate TMB model markedly deviated from the actual ROR of 2.71. This striking deviation was also observed in previous work showing substantially lower-than-anticipated ROR for small cell lung cancer (9). However, strong improvements were seen after incorporating DC and/or CD4 + naive T-cells into our prediction models, further supporting the validity of synergistic combination of DC, TMB, and CD4 + naive T-cells.
Given these results, we asked whether a composite model integrating the 11 novel molecular determinants with the 12 prior immunogenomic ones could outperform the IRF4 -TCL1A -SHC-pY317 model. In contrast, none of constructed bivariate models outperformed it (Supplementary Table S6). Moreover, the combination of DC, TMB, CD4 + naive T-cells, IRF4, TCL1A, and SHC-pY317 did not improve our ability to predict irAEs risk (R = 0.81).
Validation of IRF4 and TCL1A associated with irAEs in ICI-treated melanoma cohort
Lastly, we examined associations of IRF4 and TCL1A genes with irAEs development in patients with melanoma receiving ICI treatment. As shown in Figure 6A, IRF4 mRNA level was significantly elevated in patients developing irAEs compared with those without any irAEs (median expression, 4.36 vs 3.98; Mann-Whitney U test, P = .04). We next wondered whether this association was skewed by ICI therapy type. After correcting for therapy classes, IRF4 remained associated with irAEs development (logistic regression, P = .03). In contrast, the difference in TCL1A mRNA level stratified by irAEs status was not statistically significant, although there was a trend toward high TCL1A expression in irAEs-experiencing patients subgroup (median expression was 1.32 for irAEs-experiencing patients vs 0.75 for irAEs-free ones; Mann-Whitney U test, P >.05). Further analysis revealed that patients with grade 3, 1, or no irAEs had higher TCL1A expression than those experiencing the most severe irAEs (median expressions for grade 4, 3, 1, and 0 were 4.7, 1.61, 1.32, and 0.81, respectively; Mann-Whitney U test, P <.05 for all) ( Figure 6B).
Discussion
Our integrative analyses of large-scale cleaned pharmacovigilance data and multiomics profile offer a valuable collection of baseline predictors for irAEs development regardless of caner types, as In our work, we first investigated the relationships between the 41 established immunogenomic factors and irAEs risk, then found 12 irAEs-related factros, including 5 immune cells, TMB, 3 immune expression signatures, and 3 checkpoint-related factors. Of these, 4 cell types (DC, Mast cells, CD4 + T-cells, and CD4 + naive T-cells ) have not yet been demonstrated to be associated with ICI efficacy. Next, we investigated the relationships between genome-wide mRNA, microRNA, and protein profiles and irAEs risk, and found 11 de novo generated irAEs-related molecular factors, all of which were not reported to be implicated in ICI efficacy. Actually, there was only moderate correlation between irAEs risk and ICI efficacy, implying that immunological mechanisms underlying irAEs development and efficacy were not completely shared. Thus we indeed identified some factors predictive of irAEs risk but not efficacy, although further experimental study is warranted to classify the biological significance of novel features identified in our study.
Based on aforementioned predictors, a trivariate model combining DC, TMB, and CD4 + naive T-cells, which considerably reduced the unexplained variance in predicting irAEs risk from 0.60 (1 -0.63 2 ) utilizing TMB alone to 0.34 (1 -0.81 2 ), was generated, suggesting that all these factors may get involved in the mediation of irAEs development. We hypothesized that high TMB may contribute to increased irAEs risk due to consequent increment in immunogenic neoantigens, which could resemble peptides in normal tissues and be recognized as non-self antigens by the immune system (41), thus eliciting irAEs in target tissues as cross-reactive immune responses in the ICI therapy setting. This hypothesis invokes the theory of molecular mimicry that has been involved in the pathogenesis of autoimmune diseases, and where antibodies or TCRs recognizing pathogenic antigens could also cross-react against self-antigens (42). Evidences supporting the validity of neoantigenic molecular mimicry in the onset of irAEs come from observations in the cancer context that (1) TCRs reactive to certain neoantigens exhibited crossreactivity to the wild-type peptides (43), and (2) shared T-cell clones were identified between tumors and target tissues of irAEs from ICI-treated patients in whom irAEs developed (44,45).
Nonetheless, as suggested in our prediction model, it was not sufficient for immunogenic neoantigens alone to trigger irAEs, but abundant DC were required. As the most potent antigenpresenting cell type, DC are critical for priming naive T-cells by presenting antigens via major histocompatibility complex molecules and providing costimulatory signal (46). The engagement of DC in triggering autoimmunity has been documented via various mechanisms (47). For instance, deficient apoptosis of DC may increase DC numbers and lead to the onset of systemic autoimmunity (48). Additionally, previous studies suggest that DC may transfer the majority of tumor antigens from tumors to draining lymph nodes for the purpose of efficient priming of T-cells (49)(50)(51). Thus, a possible mechanism whereby neoantigenic mimicry may be implemented is that, intratumoral DC locally capture immunogenic neoantigens in tumor microenvironment, and subsequently migrate to draining lymph nodes where they disseminate neoantigens and stimulate resident T-cells. Upon being educated by DC, these T-cells would circulate systemically to induce neoantigen-specific immunopathologies such as irAEs against the cross-reactive self-antigen at distal sites.
Given the similarity between the irAEs and that of a chronic graft-versus-host-disease (GVHD) reaction following allogenic hematopoietic stem cell transplantation, there is a new theory for ICIs-induced irAEs (52). It was hypothesized that ICIs induced a graft-versus-malignancy effect, which eradicated metastatic cancer in a minority of patients, but also involved an auto-GVHD reaction that leaded to widespread autoimmunity in the majority. Based on this theory, an off-label low-dose nivolumab plus ipilimumab regimen was developed and tested in 131 unselected stage IV cancer patients (53). The irAEs profile of this combined low-dose treatment was significantly safer than that of the established protocols without compromising efficacy. Our finding that DC abundance showed the strongest correlation with irAEs risk supports the auto-GVHD reaction theory as host-derived DC are also important to elicit allogeneic T cell responses (54).
Our model also highlights the potential role of CD4 + naive T-cells in tumor microenvironment in promoting irAEs developement. The recruitment of CD4 + naive T-cells into non-lymphoid tissues, including tumors, has been reported (55,56). although their biological significances remain uncertain. It is notable that CD4 + T-cells are of fundamental importance in mediating autoimmunity. And this role is achieved via the differentiation of CD4 + naive T-cells into various lineages of T helper cells, depending on external A B FIGURE 6 Association of IRF4 cytokine microenvironment and transcription factors they induce (57). Further performance enhancement (unexplained variance = 0.24) was seen in the composite model comprising 3 novel molecular predictors (IRF4, TCL1A, and SHC-pY317). All these features have been implicated in immune regulation (28)(29)(30)(31)(32)(33)38). Importantly, we observed elevated expression level of IRF4 in ICI-treated melanoma patients with irAEs. IRF4 is a member of the interferon regulatory factor family of transcription factors and selectively expressed in lymphoid and myeloid cells. IRF4 deletion in mice may induce transplant acceptance by establishing CD4 + T-cells dysfunction (58) and render mice resistant to several autoimmune diseases (28), such as ulcerative colitis and experimental autoimmune encephalomyelitis. Intriguingly, a MEK1/2 inhibitor trametinib was capable of inhibiting IRF4 expression in activated CD4 + Tcells (58). Collectively, these evidences suggest the therapeutic potential of targeting IRF4 expression for abrogating inflammatory toxicities from immune checkpoint blockade.
MicroRNAs are critical posttranscriptional regulators of target genes expression, and the number of microRNAs implicated in immune disorders like autoimmunity has increased dramatically (40). A recent study has shown that microRNA-146a may regulate irAEs by ICIs in mice (8). Of note, we found 6 microRNAs predictive of irAEs risk. Given that miRNAs act by targeting multiple genes within a pathway, thus causing a broader yet specific response (59), our finding may further spark the possibility of using microRNAs as therapeutics for irAEs with multifactorial origin.
We also noted a study profiled for serum cytokines/ chemokines in 47 cancer patients with ICIs treatment (60). It revealed that patients with irAEs had lower baseline levels and higher posttreatment elevation in serum IFN-gamma-inducible small cytokines (CXCL9 and CXCL10). In our work, the IFNgamma signature in tumor microenvironment showed positive correlation with irAEs risk. This observation may be associated with the difference between circulating and tumor immune microenvironment, and deserve further investigation.
This study has several limitations. First, FAERS is a spontaneous reporting database which may include reporting bias and inaccurate reports, although it has previously been used to determine risk factors linked to the development of irAEs (9,61). Second, cancer patients with more responsive tumor immune microenvironment may remain on ICI treatment longer. However, the majority of irAEs reported during anti-PD-1 therapy occur within the first few months of commencing treatment (62), which implys that treatment duration may not bias our results. Given the identification of markers (e.g., TMB, immune signatures, and PD-L1 expression) predictive of both ICI response and irAEs risk in our study, we propose that the association between response and irAEs risk may be partially linked via high tumor immunogenicity and immunoresponsive microenvironment represented by these predictors. Therefore, it is necessary to discern markers able to discriminate anti-tumor efficacy from the risk of irAEs in patients with ICI treatment. Notably, all 11 novel molecular features in our study have not been reported to be associated with anti-tumor efficacy. Third, in addition to cancer-associated immunogenomic features reported in our study, host features, such as age, genetic susceptibility to autoimmunity, pre-existing autoimmune disease, and gut microbiome, may influence the development of irAEs (6). Fourth, further experimental study is required to classify the biological significance of novel features identified in our study.
In conclusion, our approach allowed us to identify cellular and molecular candidates as well as their optimal combinations for identifying patients with the risk of irAEs development during anti-PD-1 therapy, irrespective of cancer types. These findings may advance our understanding of mechanisms that drive irAEs development and tailoring personalized surveillance strategies.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding authors.
Ethics statement
Ethical review and approval was not required for the study on human participants in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.
Author contributions
YS, XH, and LZ conceived the concept and designed the study. LZ conducted statistical analysis. LZ drafted the manuscript. YS and XH performed the critical revision of the manuscript for important intellectual content. YS and XH obtained funding and supervised the study. All authors contributed to the article and approved the submitted version.
Funding
This work was supported by China National Major Project for New Drug Innovation (2017ZX09304015, and 2019ZX09201-002).
|
v2
|
2022-11-25T14:43:54.643Z
|
2022-11-25T00:00:00.000Z
|
253841928
|
s2orc/train
|
Transcriptome analysis revealed the role of mTOR and MAPK signaling pathways in the white strain of Hypsizygus marmoreus extracts-induced cell death of human hepatoma Hep3B cells
The aim of this study was to investigate the anticancer mechanisms of white genius mushroom (WGM). WGM is a popular edible mushroom in Taiwan and has been demonstrated to mediate potent antiproliferation effects against human Hep3B liver cancer cells in our previous study. According to next generation sequencing technology and KEGG pathway enrichment analysis, mTOR and MAPK signaling pathways were markedly changed during treatment with WGM extracts in Hep3B cells. Therefore, this study examined the effects of WGM extracts on the expression of mTOR and MAPK signaling pathway-related proteins, such as PI3K, Akt, mTOR, Ras, Raf, MEK, ERK, p38 and JNK in Hep3B cells. According to the results of immunoblotting, we demonstrated that the protein expression of the members of PI3K/Akt/mTOR and MAPK signaling pathways were involved in WGM extracts-induced cell death. Furthermore, the inhibitors of PI3K/Akt/mTOR and MAPK signaling pathways such as rapamycin, MK2206, LY3214996 and SB202190, blocked the induction of cell death and vacuoles formation induced by WGM extracts. This study also demonstrated that WGM extracts is able to inhibit Hep3B cell migration and colony formation in a dose-dependent manner. In addition to being a very popular food, WGM should be a pharmacologically safe natural agent for cancer treatment. Therefore, WGM might be designed to develop into a dietary chemopreventive agent for the cancer treatment.
Introduction
Cancer is a major public health concern and the leading cause of death in Taiwan. According to the report provided by the Ministry of Health and Welfare in Taiwan, liver cancer is always the second leading cause of cancer death for the past 20 years. Drug resistance continues to be the principal reason for achieving cures in patients with liver cancer. In recent years, an advent of cancer cells develop resistance to anticancer drugs has led researchers to expedite and put in more effort in the development of new and more effective anticancer drugs. Cancer chemoprevention is defined as the use of food supplements or synthetic compounds to suppress, prevent or delay cancer development and progression. Natural products, such as fruits and vegetables, were also receiving renewed attention as the discovery of cancer chemopreventive agents. For prevention the development of the liver cancer by blocking the initiation stage of tumorigenesis, the development of an effective cancer chemopreventive agent derived from the daily intake of food is urgently needed.
Mushrooms have been shown to have numerous biological activities including anticancer activity, antimicrobial effect and immunomodulating effects Kim et al., 2011;Kwak et al., 2015). Based on the consideration of anticancer activity and immune-promoting effect, mushrooms are people's favourite food. Anticancer activity of mushrooms against human cancer cells, such as breast, prostate and colorectal cancer cells, involves apoptosis, cell cycle arrest and inhibition of cell proliferation (Hu et al., 2002;Stanley et al., 2005;Hsu et al., 2008). Transcriptome profiling is an effective tool in large-scale investigation of gene expression patterns in distinct cellular states. Next generation sequencing (NGS) refers to massive scale RNA sequencing technology that allows investigation of a transcriptome profiling. Therefore, NGS has been widely used in the detection of changes in gene expression in different groups or treatments. KEGG (Kyoto Encyclopedia of Genes and Genomes) is a collection of databases dealing with metabolism, genetic information processing, environmental information processing, cellular processes, organismal systems, human diseases and drug development (Kanehisa and Goto, 2000). KEGG pathway enrichment facilitates to determine the differential expressed genes involved in the most important signal transduction pathways and metabolic pathways.
Phosphatidylinositol 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR) signaling is one of the most important intracellular signaling pathways that is related to cell proliferation, motility, survival and apoptosis of cancer cells (Katso et al., 2001;Engelman et al., 2006). Many reports have indicated that inhibition of PI3K/Akt/mTOR signaling pathway triggers apoptosis and inhibits the proliferation of tumor cells, which have elevated Akt levels (Hennessy et al., 2005;Mandal et al., 2005). Activation of the PI3K/Akt/mTOR pathway leads to tumor development and anticancer drugs resistance (Hennessy et al., 2005;Martini et al., 2014). The Ras/Raf/ERK signaling and PI3K/Akt signaling pathway are highly interconnected (Castellano and Downward, 2011). MAPK (mitogen-activated protein kinase) cascades plays a key role in the cellular response to extracellular stimuli (Raingeaud et al., 1995;Liu et al., 1996). MAPK signaling pathway has been found to be involved in the proliferation, differentiation and apoptosis of human cancers Guo et al., 2016). Three well-defined subgroups of mammalian MAPKs are extracellular signal-regulated kinase (ERK1/2), jun N-terminal kinase/stress activated protein kinase (JNK/SAPK) and p38. Among all MAPK signal transduction pathways, the Ras/Raf/ERK signaling pathway is the most important signaling cascade and plays an important role in the proliferation and migration of tumor cells (Roberts and Der, 2007;Sun et al., 2015). It has been reported that inhibition of ERK signaling pathway can result in both decreased cellular proliferation and increased cellular death (Liu et al., 2018;Li et al., 2020).
White genius mushroom (WGM) is white strain of H. marmoreus and is one of the most important edible mushrooms in Taiwan. In our previous study, we have identified the cytotoxicity of white genius mushroom extracts on Hep3B cells which is partially dependent on the production of ROS in Hep3B cells (Li et al., 2022). In order to get a better understanding of the underlying mechanisms of WGM extracts-produced anticancer activity of Hep3B cells, the differential expressed genes and pathways involved in the WGM extracts-induced cell death of Hep3B cells were identified using the NGS technology and KEGG tool. According to the KEGG pathway enrichment analysis, autophagy, mitophagy and apoptosis pathways were markedly changed by WGM extracts in human Hep3B liver cancer cells (Li et al., 2022). Although WGM extracts were found to have anticancer activities, the exact mechanism of the anticancer effect of WGM extracts is substantially unknown. In this study, WGM extracts was examined for its anticancer activities and exact mechanisms in hepatocellular carcinoma Hep3B cells, and expected to develop into a dietary chemopreventive agent in the future.
Preparation of white genius mushroom
White genius mushroom used in this study was harvested in September and purchased form 8329 Farm (Changhua, Taiwan). The voucher specimens of white genius mushroom (CMU-RX-HM-2020021) were deposited in School of Pharmacy, China Medical University, Taichung, Taiwan. The air-dried WGM (122.2 g) was soaked thrice with 1 L of 95% ethanol at room temperature for 3 days. The combined filtrate was then concentrated under reduced pressure at 40°C. The yield of dry extract of WGM was about 5.6%. WGM extracts are dissolved in 100% dimethylsulfoxide (DMSO) and stored at -20°C until use. In our previous study, we have demonstrated that the IC 50 (half maximal inhibitory concentration) of WGM extracts was about 175 μg/ml (Li et al., 2022). Since the concentration of IC 50 of WGM was used to treat the cells in the inhibitor experiment, inhibitors had no significant protective effect on the WGMinduced vacuoles formation and cell death. Therefore, IC 40 , which was about 150 μg/ml WGM, was chosen in this study.
Total RNA extraction
Total RNA was extracted from Hep3B cells using AllPure Total RNA Isolation Kit (AllBio Science Inc., Taiwan) following the manufacturer's protocol. RNA concentration was detected using a spectrophotometer with wavelength at 260 nm.
Differential expressed genes analysis
Gene expression is measured using read density, with higher read density indicating higher gene expression levels. Gene expression was calculated using Cuffdiff (v2.2.1) software that calculates FPKM (fragments per kilo bases per million reads). The formula is: FPKM total exon f ragments mapped read Millions ( )× exon length KB ( ) Comparison of the expression levels of all genes under different experimental conditions by FPKM profiling. Gene differential analysis was performed using the Cuffdiff (v2.2.1). Based on the criteria of fold change greater than 2 and p-value less than 0.05, the results of the Cuffdiff analysis were further analyzed to identify genes with significantly differential expression. Differential expressed genes (DEGs) of WGM extracts-treated samples compared to controls were analyzed using DESeq (v1.18.0) Bioconductor package, which is a model based on a negative binomial distribution. After adjustment by the Benjamini-Hochberg method for controlling the false discovery rate, p-values for genes were set at < 0.05 to detect differential expressed genes. The raw sequencing data were uploaded to the NCBI Sequence Read Archive (SRA) with the accession ID PRJNA813700.
Cluster analysis of differential expressed genes
Cluster analysis is the calculation and classification of data based on similarity, thereby grouping samples or genes with similar expression patterns into one group. This can predict the function of unknown genes and predict whether they are involved in the same cellular pathway or metabolic process. The FPKM value of different genes under different experimental conditions was taken as the expression level and used for hierarchical clustering. The most obvious feature of this method is the generation of dendrogram. Different clusters are represented by different colored regions. Gene expression patterns are similar within the same cluster and close to each other, and they may have similar biological functions. The gplots in R software were used for cluster analysis, and the data of union_for_cluster were clustered by the log relative expression Frontiers in Pharmacology frontiersin.org value of genes. We used algorithms to obtain the distance between genes, and then calculated the relative distance between the genes through repeated operations. Finally, clustering was performed by dividing genes into different subclusters based on their relative distances.
Kyoto encyclopedia of genes and genomes enrichment analysis of differential expressed genes KEGG is a collection of databases dealing with metabolism, genetic information processing, environmental information processing, cellular processes, organismal systems, human diseases and drug development (http://en.wikipedia.org/wiki/ KEGG). We used scripts in house to enrich significant differential expressed genes in KEGG pathways. KEGG pathway units and a hypergeometric test were used to perform the pathway enrichment analysis and find the pathways of the differentially express genes that are significantly enriched in the transcriptome data. Below is the formula: N is the number of genes with pathway annotations, n is the number of differential expressed genes in N, M is the number of genes annotated for a particular pathway in all genes, and m is the number of differential expressed genes annotated for that pathway. The threshold Q value is ≤ 0.05.
Assessment of morphological changes
Cells were seeded at a density of 5 × 10 4 cells per well onto 12well plate 48 h before being treated with drugs. Hep3B cells were incubated without or with indicated various concentrations WGM extracts for 24 h. The morphology of the cells was photographed with an Olympus IX 73 phase contrast microscope at objective ×10 magnification. A field was chosen in the center of each well at approximately the same location for photography.
Wound healing assay
The migratory activity of Hep3B cells was assessed using a wounded migration assay. Wounded migration assay was performed as previously described (Liao et al., 2015). As cell reached confluence or near confluence, a linear wound was scratched across each well by a sterile 200 μl pipette tip. After washing, cells were treated with vehicle alone (control) or with 75, 125 or 150 μg/ml WGM extracts for 24 h. Images of wounds at 0 h and 24 h after scratching were obtained with an Olympus IX 73 phase-contrast microscope (Olympus Optical Co., Tokyo, Japan) at objective ×10 magnification.
Transwell migration experiments
In vitro cell migration was also performed using 24-well transwell inserts with a pore size of 8 μm (Corning Life Sciences, NY, United States). The transwell chambers were used according to the manufacturer's protocol. The cell pellets were resuspended in DMEM supplemented with 0.1% FBS at a cell density of 1 × 10 6 cells/ml. Aliquots of 100 μl cell suspension (1 × 10 5 cells per well) containing DMSO or WGM extracts were loaded into transwell inserts that were subsequently placed into the 24well plate, and 600 μl medium supplemented with 10% FBS was used as a chemoattractant in the lower chamber. After incubated for 24 h at 37°C, the cells remaining on the upper surface of the membrane were removed with a cotton swab, and the cells on the lower surface of the membrane are the migrated cells. Filters were fixed with 4% paraformaldehyde solution for 20 min and stained with 0.1% crystal violet for 10 min, then the cells that passed through the filter were photographed by Frontiers in Pharmacology frontiersin.org 04
FIGURE 1
Cluster analysis and KEGG pathway enrichment of differential expressed genes between WGM extracts-treated and control groups in Hep3B cells. Cells were treated with vehicle alone or with 150 μg/ml WGM extracts for 24 h. (A) Cluster analysis of differential expressed genes between the WGM extracts-treated and control groups. The regions of different colors represent different clusters. The color scale of the heatmap illustrates the log 2 of fold change of the WGM extracts-treated/control samples shown in the heatmap. (B) The KEGG pathway enrichments of the DEGs between WGM extracts-treated and control samples. The y-axis shows the names of the enriched pathways. The results were analyzed to determine genes with significant differential expression according to the criteria of fold change greater than 2 and FDR less than 0.05.
Colony formation assay
To assess the effects of WGM extracts on cell proliferation, the colony formation assay was carried out in vitro. Hep3B cells were seeded at a density of 5 × 10 3 cells per well onto 6-well plate 48 h before drug treatment. Cells were treated with vehicle alone or with 125 or 150 μg/ml WGM extracts for 24 h. After treatment, the media was replaced with fresh complete growth media without the WGM extracts. Treated cells were washed by phosphate-buffered saline and then cultured in culture medium for 3 and 4 days with medium replacement every 2 days. After incubated at 37°C for 3 and 4 days, colonies were fixed with 4% paraformaldehyde solution and then stained with 0.5% crystal violet. Finally, the colonies were photographed.
Data analysis and statistics
One-way ANOVA and Bonferroni post-hoc test were used to analyze differences between each experimental group. A p-value less than 0.05 was considered significant.
Results
Processing of next generation sequencing data and differential expressed genes analysis Since the exact mechanism of the anticancer effect of WGM extracts is substantially unknown, this study takes advantage of NGS technology to compare differences in the transcriptome profiles between control and WGM extractstreated cells. The software of Cuffdiff (v2.2.1) and HTSEQ (v0.6.1) were used for gene expression calculation. Based on the criteria of fold change greater than 2 and p-value less than 0.05, the results of the Cuffdiff analysis were further analyzed to identify genes with significantly differential expression. To directly assess differential expressed genes (DEGs) expression in these samples and confirm the classification, a DEG heat map was drawn using the R package pheatmap (version 1.0.8). Gene expression patterns are similar within the same cluster and close to each other, and they may have similar biological functions. Cluster analysis of differentially expressed genes log 10 (FPKM +1) values are used for clustering. Genes of low expression are in blue, and high expressed in red. The heat map clearly indicated the distinct separation of the WGM extracts-treated and control groups ( Figure 1A). Frontiers in Pharmacology frontiersin.org mTOR and mitogen-activated protein kinase signaling pathways were involved in white genius mushroom extracts-induced cell death of Hep3B cells After screening differential expressed genes, KEGG pathway enrichment analysis was subsequently performed to investigate the function of the identified genes. Pathway functional enrichment facilitates to determine the DEGs involved in the most important signal transduction pathways and metabolic pathways. The top 30 KEGG pathways are presented in Figure 1B. KEGG pathway analysis showed that the DEGs induced by WGM extracts were found to be involved in molecular pathways for organismal systems, metabolism, human diseases, genetic information processing, environmental information processing and cellular processes ( Figure 1B). In the cellular processes of KEGG pathway analysis, autophagy, mitophagy and apoptosis pathways were found to be significantly changed in WGM extracts-treated cells ( Figure 1B). However, the involvement of autophagy, mitophagy and apoptosis pathways in WGM extracts-induced cell death was investigated in our previously study. Hippo, mTOR, FoxO, MAPK and TGF-beta signaling pathways were significantly altered in the environmental information processing ( Figure 1B). The mTOR and MAPK signaling pathways are necessary, however, to promote growth and proliferation in many mammalian cell types. We focused our attention on the investigation of the effects of WGM extracts on the expression of mTOR and MAPK signaling pathways in Hep3B cells in this study. As shown in Figure 2A, 112 genes (87 upregulated and 25 downregulated) were related to mTOR signaling pathway and 85 genes (76 upregulated and Effects of WGM extracts on the protein levels of the members of mTOR and MAP signaling pathways in Hep3B cells. The effects of WGM extracts on mTOR and MAPK signaling pathway-related proteins were analyzed by Western blotting. Cells were treated with vehicle alone or with 75, 125 or 175 μg/ml WGM extracts for 24 h. Protein samples were analyzed by SDS-PAGE (5% for PI3K and mTOR, 8% for Raf, 10% for Akt and Aktp (S473), 12% for β-actin, ERK, ERK1p(Thr202/Tyr204)/ERK2p(Thr185/Tyr187), JNK, JNKp(Thr183/Tyr185), MEK, p38 and p38p (Thr180/Tyr182) and 14% for Ras), and then probed with primary antibodies followed by secondary antibodies. (A) Representative blots. (B) The blots were quantified by Lane 1D Gel imaging analysis software. Protein expression was normalized using β-actin. Fold change = normalized signal treated/normalized signal control. Results are expressed as the mean ± S.D. of three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001 compared to the control values.
Frontiers in Pharmacology frontiersin.org Frontiers in Pharmacology frontiersin.org 9 downregulated) to MAPK signaling pathway. The differential expressed genes involved in mTOR and MAPK signaling pathways were shown in Supplementary Tables S1, S2, respectively. The results of a hierarchical clustering analysis of DEG involved in mTOR and MAPK signaling pathways are displayed in a heat map as a dendrogram ( Figures 2B,C).
The effects of white genius mushroom extracts on the expression of mTOR and mitogen-activated protein kinase signaling pathways-related proteins According to the KEGG pathway enrichment analysis, the present study examined the effects of WGM extracts on the expression of mTOR and MAPK signaling pathway-related proteins, such as PI3K, Akt, mTOR, Ras, Raf, MEK, ERK, p38 and JNK, in Hep3B cells. The expression of mTOR and MAPK signaling pathway-related proteins during WGM extracts-induced cell death was performed by Western blotting techniques. After treating the Hep3B cells with WGM extracts for 24 h, the protein levels of PI3K and mTOR were increased after treatment with WGM extracts for 24 h ( Figure 3A). Cells treated for 24 h with WGM extracts showed a marked dose-dependent decrease in the expression of the ERK and JNK protein, and an upregulation of the Ras, Raf and MEK protein ( Figure 3A). The present study also investigated the expression of pERK protein in WGM extracts-induced Hep3B cell death. Treatment with WGM extracts increased pERK protein in Hep3B cells up to 125 μg/ml, but pERK level was decreased after treatment with 175 μg/ml WGM extracts. It is interesting to note that WGM extracts induced a significant increase in the protein expression of Ras, but did not induce ERK activity, in cells exposed to 175 μg/ml WGM extracts for 24 h ( Figure 3A). These results indicate that WGM extracts-mediated ERK activation occurs through a Ras-independent pathway. As shown in Figure 3A, the present study also demonstrated that the expression patterns of pERK protein are similar to those seen in Akt and pAkt expression after treatment with WGM extracts for 24 h ( Figure 3A). In our study, not only the expression of ERK, which is associated with cell proliferation, but also p38, which is regulated by various environmental stresses, is regulated by WGM extracts. As shown by immunoblotting, WGM extracts caused a marked increase in the protein levels of pp38 in Hep3B cells ( Figure 3A). However, the protein levels of p38 were decreased after treatment with WGM extracts ( Figure 3A). As shown in Figure 3B, the Western blot results of protein were quantified by Lane 1D Gel imaging analysis software. These results indicate that changes in the expression of PI3K/mTOR and MAPK signaling pathwaysrelated proteins in WGM extracts-treated cells is associated with WGM extracts-induced Hep3B cell death.
The effects of the inhibitors of PI3K/Akt/ mTOR signaling pathway on white genius mushroom extracts-induced cell death and membrane-enclosed vacuoles of Hep3B cells As shown by the Western blotting analysis, we guess that the cell death of WGM extracts treated Hep3B cells is related to PI3K/Akt/mTOR signaling pathway. Therefore, we used Akt inhibitor MK2206 and mTOR inhibitor rapamycin to demonstrate that PI3K/Akt/mTOR signaling pathway is an important intracellular signaling pathway in the WGM extracts-induced cell death. MK2206, a well-known Akt inhibitor, was used to demonstrate whether Akt activity can promote the WGM extracts-induced cell death of Hep3B cells. The present study demonstrated that MK2206 (0.3 μM) was only slightly restore the cell death induced by WGM extracts ( Figure 4A). MK2206 (0.3 and 3 μM, pretreatment 1 h) had a significant preventive effect, however, on the WGM extracts (150 μg/ml)-induced membrane-enclosed vacuoles of Hep3B cells ( Figure 4B). The mTOR inhibitor rapamycin, an autophagy inducer, was also used in this study. As shown in Figures 5A,B, a significant preventive effect on the WGM extracts (150 μg/ml)-induced cell death and vacuoles formation was observed in Hep3B cells pretreatment with 0.5 or 1 μM rapamycin. Hep3B cells did not show any cytoplasmic vacuolization after 24 h treatment with 0.5 or 1 μM rapamycin alone, but the proliferation of Hep3B cells were slightly inhibited by the presence of rapamycin ( Figure 5A). Our previous study has demonstrated that 3-MA, PI3K inhibitor, did not recover the membrane-enclosed vacuoles and cell death induced by 150 μg/ml WGM extracts (data not shown). These results showed that PI3K/Akt/mTOR signaling pathway was involved in the WGM extracts-induced cell death in Hep3B cells. The lysosomotropic agent chloroquine, an autophagy inhibitor, was also used in this study. The WGM extracts (150 μg/ml)-induced vacuoles formation and cell death of Hep3B cells were partially blocked by pretreatment with 10 or 50 μM chloroquine ( Figure 6A). It is interesting to note that chloroquine alone had a slight effect on the vacuoles formation in Hep3B cells ( Figure 6B).
The effects of extracellular signalregulated kinase inhibitor LY3214996 and p38 inhibitor SB202190 on white genius mushroom extracts-induced cell death and membrane-enclosed vacuoles of Hep3B cells
It is well known that MAPK signaling pathway was an important intracellular signaling pathway that is related to cell Frontiers in Pharmacology frontiersin.org proliferation, motility, survival and apoptosis of cancer cells.
According to the results of immunoblotting, WGM extracts caused a marked increase in the protein levels of pERK in Hep3B cells. The ERK inhibitor LY3214996 was used in this study. Pretreatment with LY3214996 (5 and 10 nM) significantly inhibited the WGM extracts (150 μg/ml)induced cell death and membrane-enclosed vacuoles of Hep3B cells (Figure 7). In this study, not only the expression of phosphorylation of ERK but also phosphorylation of p38 is regulated by WGM extracts according to the results of immunoblotting. This study determined the effect of p38 inhibitor SB202190 on the WGM extracts-induced cell death of Hep3B cells. Pretreatment with SB202190 (0.5 and 1 μM) for 1 h had a significant preventive effect on the WGM extracts (150 μg/ ml)-induced cell death and membrane-enclosed vacuoles of Hep3B cells (Figure 8). Based on the above data, these results indicate that WGM extracts-mediated the activation of ERK and p38 is involved in the WGM extracts-induced membraneenclosed vacuoles and even cell death.
The effect of white genius mushroom extracts on the migration potential of Hep3B cells
Controlling the invasion and metastasis of cancer cells has been recognized as a new strategy for cancer prevention and treatment. The MAPK and mTOR signaling pathway are most important signaling cascade and play key role in the proliferation and migration of tumor cells. Since cell migration is an essential step in the cancer metastatic process, the present study investigated the effect of WGM extracts on the migration ability of Hep3B cells. For examination of the ability of Hep3B cell migration, a wound-healing assay was performed to examine whether WGM extracts can inhibit Hep3B cell migration. Wound healing experiment was performed on cells treated with 75, 125 and 150 μg/ml WGM extracts for 24 h. As shown in Figure 9A, an increase in the distance of the wound edge indicates that the speed of cell migration is significantly reduced after treatment with WGM extracts. To obtain further support for the inhibition of cell migration by WGM extracts in Hep3B cells, the transwell migration assay, which is used to examine the migratory response of cancer cells to treatments, were performed in this study. The results from that assay showed that WGM extracts had a significant inhibition of the Hep3B cell migration ability in a dose-dependent manner through the transwell membrane compared with those in the control cells ( Figure 9B). Based on the above data, we suggested that WGM extracts be able to inhibit Hep3B cell migration in a dose-dependent manner.
White genius mushroom extracts inhibits the proliferation of Hep3B cells in vitro
Colony formation assay was used to mimic cancer growth from a single cell to grow into a cancer cell colonies or tumor mass. Plate colony formation assay was used to examine the effect of WGM extracts on the colony-forming ability of Hep3B cells. Cells were plated onto 6-well plates and incubated with vehicle alone or with 125 or 150 μg/ml WGM extracts and treated cultures were maintained in culture for an additional 3 or 4 days to allow formation of colonies. As shown in Figure 10A, WGM extracts significantly inhibited the colony-forming ability of Hep3B cells in a dose-and time-dependent manner, as fewer colonies were observed in cells treated with higher concentrations of WGM extracts. In other words, higher doses inhibited colony formation and resulted in a significant decrease in colony numbers in the WGM extracts-treated group. In control group, Hep3B cells grow and form colonies within 3 or 4 days of incubation ( Figure 10A). In our previous study, we have confirmed the cytotoxicity of white genius mushroom extracts on human Hep3B liver cancer cells. WGM extracts induced a significant concentration-dependent inhibition of Hep3B cell growth with IC 50 (half maximal inhibitory concentration) value of 175 μg/ml. We further investigated whether WGM extracts selectively causes cytotoxicity in different cancer cell lines. Therefore, this study evaluated the effects of WGM extracts on cell growth of hepatocellular carcinoma HepG2 cells, human lung squamous carcinoma CH27 cells, breast cancer MCF-7 cells and prostatic adenocarcinoma PC-3 cells. WGM extracts had no significant cytotoxic effect on the HepG2, CH27 and MCF-7 cells after treatment with WGM extracts for 24 h ( Figure 10B). Furthermore, WGM extracts treatment leads to an increase in cell viability of MCF-7 cells ( Figure 10B). The treatment of PC-3 cells with WGM extracts for 24 h resulted in a slight cytotoxic effect ( Figure 10B). The concentration of inducing about 50% cell death by WGM extracts is more than 200 μg/ml for PC-3 cells ( Figure 10B). Based on the above data, we demonstrated that the cytotoxic effect of WGM extracts on Hep3B cells was significantly higher than that in PC-3, HepG2, CH27 and MCF-7 cells. Therefore, Hep3B cells were chosen for investigation of the anticancer mechanisms of WGM extracts in this study.
Discussion
Hepatocellular carcinoma (HCC) is one of the most common cancers in Taiwan, with a high incidence and mortality rate. However, the drugs used to treat for liver cancer may appear high toxicity, many side effects and drug resistance, leading to poor quality of life after treated for cancer. Therefore, the search for new drugs or chemoprevention agents with few side effects and
FIGURE 7
The effects of LY3214996 on WGM extracts-induced cell death and membrane-enclosed vacuoles of Hep3B cells. Cells were pretreated with 5 or 10 nM LY3214996 (LY) for 1 h and then treated with 0.1% DMSO or 150 μg/ml WGM extracts (W) for 24 h. (A) The effects of LY3214996 on WGM extracts-induced cell death of Hep3B cells. After treatment, the viable cells were measured by MTT assay. All determinations are expressed as the mean % control ±S.D. of duplicate from three independent experiments. **p < 0.01, ***p < 0.001 compared to the control values. † † † p < 0.001 compared to the WGM extracts alone. (B) The effects of LY3214996 on WGM extracts-induced membrane-enclosed vacuoles of Hep3B cells. After treatment, the cells were immediately photographed with an Olympus IX 73 phase-contrast microscope. All results are representative of three independent experiments.
Frontiers in Pharmacology frontiersin.org high efficacy is in demand in order to treat liver cancer patients. White genius mushroom (WGM) is a popular edible mushroom whose anticancer activity was demonstrated to be partially dependent on the production of ROS in human Hep3B liver cancer cells in our previous study (Li et al., 2022). Furthermore, autophagy, mitophagy and apoptosis pathways were identified as significant in WGM extracts-treated Hep3B liver cancer cells according to the KEGG pathway enrichment analysis (Li et al., 2022). Although WGM extracts were found to have anticancer activities, the exact mechanisms of the anticancer effect of WGM extracts are not fully understood. In this study, WGM extracts was evaluated for its anticancer activities and exact mechanisms in hepatocellular carcinoma Hep3B cells. Since WGM is a pharmacologically safe natural agent, WGM might be a dietary chemopreventive agent for cancer treatment with an excellent safety profile in vivo. RNA expression profiles produced by NGS technology allow comprehensive investigation of transcribed sequences in distinct cellular states (Wu et al., 2019;Shi et al., 2020). The KEGG pathways database is the comprehensive coverage of a wide range of different biochemical pathways. In general, KEGG pathway enrichment analysis was conducted to identify differential expressed genes (DEGs) into significantly enriched metabolic pathways or signaling pathways. In this study, the KEGG pathway analysis was used to screen the enrichment of dysfunctional signaling pathways of the DEGs in WGM extracts-treated cells. KEGG pathway enrichment analysis showed that the DEGs induced by WGM were found to be involved in organismal systems, metabolism, human diseases, genetic information processing, environmental information processing and cellular process. In the environmental information processing of KEGG pathway analysis, Hippo, mTOR, FoxO, MAPK and TGF-beta signaling pathways were significantly changed. The Hippo signaling pathway is a highly conserved tumor suppressor pathway that exerts a critical role in modulating cell division, cell proliferation and apoptosis. It has been suggested that Hippo signaling pathway is a potent growth and tumor suppressor pathway in the mammalian liver (Lu et al., 2010;Hong et al., 2015). FoxOs (forkhead box class O proteins) are also considered to be tumor suppressors and known to be implicated in the progression of several human cancers (Su et al., 2014;Jiramongkol and Lam, 2020). The mTOR and MAPK signaling pathways are necessary to promote growth and proliferation in many cancer cell types. PI3K/Akt/mTOR and MAPK signaling pathways had been demonstrated to be involve in cell proliferation, motility, survival and apoptosis of cancer cells (Chen, 2010;Ma et al., 2013;Zhang et al., 2016). Recently, PI3K/Akt/mTOR and Ras/Raf/MEK/ERK signaling pathways have also been recognized as a new strategy for cancer therapy (Asati et al., 2016). Therefore, we focused our attention on the investigation of the effects of WGM extracts on the expression of mTOR and MAPK signaling pathways in Hep3B cells in this study.
FIGURE 8
The effects of SB202190 on WGM extracts-induced cell death and membrane-enclosed vacuoles of Hep3B cells. Cells were pretreated with 0.5 or 1 μM SB202190 (SB) for 1 h and then treated with 0.1% DMSO or 150 μg/ml WGM extracts (W) for 24 h. (A) The effects of SB202190 on WGM extracts-induced cell death of Hep3B cells. After treatment, the viable cells were measured by MTT assay. All determinations are expressed as the mean % control ±S.D. of duplicate from three independent experiments. ***p < 0.001 compared to the control values. † p < 0.05 compared to the WGM extracts alone. (B) The effects of SB202190 on WGM extracts-induced membrane-enclosed vacuoles of Hep3B cells. After treatment, the cells were immediately photographed with an Olympus IX 73 phasecontrast microscope. All results are representative of three independent experiments.
Frontiers in Pharmacology frontiersin.org
This study demonstrated that PI3K/Akt/mTOR signaling pathway was involved in the WGM extracts-induced cell death in Hep3B cells according to the results of the Western blotting analysis. Wu et al. (2010) demonstrated that PI3K inhibitor 3-MA can block autophagosome formation by inhibiting PI3K, thereby inhibiting autophagy. However, 3-MA did not recover the membrane-enclosed vacuoles and cell death induced by WGM extracts in our previous study (Li et al., 2022). In this study, we demonstrated that Akt inhibitor MK2206 was only slightly restore the cell death induced by WGM extracts. MK2206 had a significant preventive effect, however, on the WGM extracts-induced membrane-enclosed vacuoles of Hep3B cells. Furthermore, pretreatment with the mTOR inhibitor rapamycin abolished the WGM extractsinduced cell death and membrane-enclosed vacuoles of Hep3B cells in this study. It is interesting to note that MK2206 did not have as much effect on the WGM extracts-induced cell death as rapamycin did. In addition to 3-MA, we demonstrated that the inhibitors of PI3K/Akt/ mTOR signaling pathway are able to prevent the WGMinduced vacuoles formation and cell death, indicating PI3K/Akt/mTOR signaling pathway is an important intracellular signaling pathway in the WGM extractsinduced cell death. In this study, the Akt and mTOR inhibitor can reverse the WGM extracts-mediated Hep3B cell death of Hep3B cells which are consistent with those of other studies reporting that the dependence of PI3K/ Akt/mTOR signaling on the growth and survival of certain tumors has broad implications in cancer therapy (Guertin and Sabatini, 2007;Fasolo and Sessa, 2008;Zhou et al., 2011;Hassan et al., 2013). It is worthy of note that mTOR inhibitor is significantly more effective in preventing the membrane-enclosed vacuoles formation and cell death induced by WGM extracts than PI3K or Akt inhibitors.
FIGURE 9
The effect of WGM extracts on cell migration of Hep3B cells. (A) The migratory activity of Hep3B cells was assessed using a wounded migration assay. A linear wound was made with a sterile 200 μl pipette tip. Cells were treated with vehicle alone (Control) or with 75, 125 or 150 μg/ml WGM extracts for 24 h. Images of wounds at 0 h and 24 h after scratching were obtained with an Olympus IX 73 phase-contrast microscope at objective ×10 magnification. All results are representative of three independent experiments. (B) The effects of WGM extracts on cell migration of Hep3B cells were also measured by Transwell assay in the study. Cell migration was also detected using Transwell assay with a pore size of 8 μm the Transwell inserts for 24-well plates. Cells were treated with vehicle alone or with 125 or 150 μg/ml WGM extracts for 24 h. Filters were fixed with 4% paraformaldehyde solution for 20 min and stained with 0.1% crystal violet for 10 min, then the cells that passed through the filter were photographed by Olympus IX 73 phase-contrast microscope at objective ×20 magnification. Left: Representative images of Transwell migration assay. Right: Cells were counted from 4 random microscope fields for each sample in three independent experiments. Results are expressed as the mean ± S.D. *p < 0.05, ***p < 0.001 compared to the control values.
Frontiers in Pharmacology frontiersin.org In this study, we demonstrated that mTOR inhibitor rapamycin actually reversed WGM-induced cell death and vacuole formation. However, rapamycin is also a well-known autophagy inducer, which activates autophagy by repressing the mechanistic target of mTOR which is associated with lysosomal biogenesis and autophagosome-lysosome fusion (Huang et al., 2021). This study also investigated whether chloroquine could inhibit the induction of cell death and vacuole formation induced by WGM of Hep3B cells. Chloroquine, which is lysosomotropic agent, is a classic inhibitor of autophagy that blocks autophagic processes by inhibiting the binding of autophagosomes to lysosomes (Mushtaque and Shahjahan, 2015). Furthermore, lysosomotropic agents are one of the triggers for cytoplasmic vacuolation caused by lysosomal dysfunction (Zou et al., 2020). The present study showed that chloroquine had a significant preventive effect on the WGM extracts-induced cell death of Hep3B cells. It is interesting to note that both the autophagy inducer rapamycin and the autophagy inhibitor chloroquine prevented WGM-induced membrane-enclosed vacuoles and cell death in Hep3B cells. We also demonstrated that chloroquine was more effective than rapamycin in preventing WGM extract-induced cell death, however the prevention of vacuole formation had the opposite effect. According to the results of immunoblotting, WGM extracts induced a marked increase in the expression of the mTOR protein. It indicated that WGM might attempt to inhibit autophagy process by activating mTOR protein expression. Therefore, we guessed that rapamycin may inhibit WGMinduced cell death through induction autophagy, which is essential for maintaining cell survival. While autophagy is considered as cell survival mechanism, the occurrence of autophagy is also thought to promote apoptosis and even accelerate cell death (Gump and Thorburn, 2011;Rubinstein et al., 2011;Mukhopadhyay et al., 2014;Wei et al., 2014). Since chloroquine prevents WGM extract-induced cell death more efficiently than rapamycin, we further confirmed that Hep3B cells should attempt to survive by inducing autophagy during WGM extracts-induced death cells in this study.
The most important pathway interacting with PI3K/Akt in different types of cancers is the Ras/Raf/ERK pathway (Castellano and Downward, 2011). MAPK signaling pathway signaling pathway was found to be a crucial intracellular signaling pathway, which is closely related to proliferation, apoptosis and metastasis of cancer cells Guo et al., 2016). According to the results of immunoblotting, WGM extracts induced a marked dosedependent decrease in the expression of the ERK protein, and an upregulation of the Ras, Raf and MEK protein in this study. The expression patterns of pERK protein are similar to those seen in Akt and pAkt expression after treatment with WGM extracts for 24 h in this study. There is abundant evidence that PI3K/Akt has surprisingly extensive cross talk with ERK1/2 (Yang et al., 2019;Duff et al., 2021). Furthermore, PI3K/Akt and ERK1/2 pathways are downstream of EGF-EGFR signaling in many cell types (Zhang et al., 2012;Fan et al., 2015;Li et al., 2019). We also demonstrated that treatment with WGM extracts increased pERK protein in Hep3B cells up to 125 μg/ml, but pERK level was decreased after treatment with 175 μg/ ml WGM extracts. It is interesting to note that WGM extracts induced a significant increase in the protein expression of Ras in cells exposed to 175 μg/ml WGM extracts for 24 h. These results indicate that WGM extracts-mediated ERK activation occurs through a Ras-independent pathway. It had also been reported that the crosstalk between ERK and Akt is mediated Frontiers in Pharmacology frontiersin.org by EGFR and independent of Ras or Raf mutation (Klinger et al., 2013). In addition to the ERK, p38 is also an important intracellular signaling pathway in the MAPK signaling pathway. In our study, not only the expression of ERK, which is associated with cell proliferation, but also p38, which is regulated by various environmental stresses, is regulated by WGM extracts. As shown by immunoblotting, WGM extracts caused a marked increase in the protein levels of pp38, but p38 levels decrease in Hep3B cells. According to the results of immunoblotting, the ERK inhibitor LY3214996 and p38 inhibitor SB202190 were used in this study. The present study showed that both inhibitors partially abolished the WGM extracts-induced cell death and membrane-enclosed vacuoles of Hep3B cells. It indicate that WGM extracts-mediated the activation of ERK and p38 is involved in the WGM extracts-induced membraneenclosed vacuoles and even cell death. Based on the above data, we also demonstrated that MAPK pathway is an important intracellular signaling pathway in the WGM extracts-induced cell death.
Migration and invasion are important factors that accelerate the occurrence and progression of malignant tumors. Many cellular signaling pathways are thought to be involved in the proliferation and migration of cancer cells (Guo et al., 2016;Luo et al., 2016;Yang et al., 2016). PI3K/Akt and MAPK signaling pathways have been reported to be important signaling pathways involved in the network regulation on cell migration in many different kinds of cancer (Guo et al., 2016;Luo et al., 2016;Yang et al., 2016). According to the results of immunoblotting and experiments of inhibitors, this study examined whether WGM extracts will yield an inhibitory effect on cell migration of Hep3B cells. To obtain further support for the inhibition of cell migration by WGM extracts in Hep3B cells, the transwell migration and wound healing assays, which are used to examine the migratory response of cancer cells to treatments, were performed in this study. The results from those assays showed that WGM extracts had a significant inhibition of the Hep3B cell migration ability in a dosedependent manner. Colony formation assay was used to mimic cancer growth from a single cell to grow into a cancer cell colonies or tumor mass. Plate colony formation assay was used to examine the effect of WGM extracts on the colony-forming ability of Hep3B cells. We demonstrated that WGM extracts had a significant inhibitory effect on the colony forming ability of Hep3B cells in a concentrationand time-dependent manner. Based on the above data, we suggest that PI3K/Akt/mTOR and MAPK signaling pathways be involved in the WGM extracts-mediated inhibition of cell colony formation and migration of Hep3B cells. It is worthy to note that WGM extracts selectively induced cytotoxicity in different cancer cell lines. We demonstrated that the cytotoxic effect of WGM extracts on Hep3B cells was significantly higher than that in prostate cancer PC-3 cells, liver cancer HepG2, lung cancer CH27 and breast cancer MCF-7 cells.
Conclusion
According to the KEGG pathway enrichment analysis, the results of immunoblotting and experiments of inhibitors, we demonstrated that PI3K/Akt/mTOR and MAPK signaling pathways were involved in WGM extracts-induced vacuoles formation and cell death. Our results are the first findings to indicate that WGM extracts induces cell death in Hep3B cancer cells and that the induction of vacuoles formation and cell death coincides with the PI3K/Akt/mTOR and MAPK signaling pathways. This study also demonstrated that WGM extracts had a significant inhibition of the Hep3B cell colony formation and migration ability in a dose-dependent manner. The present findings indicate that WGM should be a pharmacologically safe natural agent and might be a dietary chemopreventive agent for the cancer treatment.
Frontiers in Pharmacology frontiersin.org
|
v2
|
2022-11-26T06:16:13.956Z
|
2022-11-25T00:00:00.000Z
|
253879494
|
s2ag/train
|
Dysregulation of HNF1B/Clusterin Axis Enhances Disease Progression in a Highly Aggressive Subset of Pancreatic Cancer Patients.
Pancreatic ductal adenocarcinoma (PDAC) is a lethal malignancy and is largely refractory to available treatments. Identifying key pathways associated with disease aggressiveness and therapeutic resistance may characterize candidate targets to improve patient outcome. We used a strategy of examining the tumors from subset of PDAC patient cohorts with worst survival to understand the underlying mechanisms of aggressive disease progression and to identify candidate molecular targets with potential therapeutic significance. Non-negative matrix factorization (NMF) clustering, using gene expression profile, revealed three patient subsets. A 142-gene signature specific to the subset with the worst patient survival, predicted prognosis and stratified patients with significantly different survival in the test and validation cohorts. Gene-network and pathway analysis of the 142-gene signature revealed dysregulation of Clusterin (CLU) in the most aggressive patient subset in our patient cohort. HNF1B positively regulated CLU, and a lower expression of HNF1B and CLU associated with poor patient survival. Mechanistic and functional analyses revealed that CLU inhibits proliferation, 3D spheroid growth, invasiveness and epithelial-to-mesenchymal transition in pancreatic cancer cell lines. Mechanistically, CLU enhanced proteasomal degradation of EMT-regulator, ZEB1. In addition, orthotopic transplant of CLU-expressing pancreatic cancer cells reduced tumor growth in mice. Furthermore, CLU enhanced sensitivity of pancreatic cancer cells representing aggressive patient subset, to the chemotherapeutic drug gemcitabine. Taken together, HNF1B/CLU axis negatively regulate pancreatic cancer progression and may potentially be useful in designing novel strategies to attenuate disease progression in PDAC patients.
|
v2
|
2022-11-26T06:16:14.008Z
|
2022-11-25T00:00:00.000Z
|
253880167
|
s2ag/train
|
A review of the cost-effectiveness of genetic testing for germline variants in familial cancer.
Background: Targeted germline testing is recommended for those with or at risk of breast, ovarian and colorectal cancer. The affordability of genetic sequencing has improved over the past decade, therefore the cost-effectiveness of testing for these cancers is worthy of reassessment.Objective: To systematically review economic evaluations on cost-effectiveness of germline testing in breast, ovarian or colorectal cancer.Methods: A search of PubMed and Embase databases for cost-effectiveness studies on germline testing in breast, ovarian or colorectal cancer, published between 1999 to May 2022. Synthesis of methodology, cost-effectiveness, and reporting (CHEERS checklist) was performed.Results: The incremental cost-effectiveness ratios (ICERs; in 2021-adjusted US$) for germline testing vs. the standard care option, in hereditary breast or ovarian cancer (HBOC) across target settings were as follows: 1) population-wide testing: 344-2.5 million/QALY 2) women with high-risk: dominant-78,118/QALY, 8,337-59,708/LYG 3) existing breast or ovarian cancer: 3,012-72,566/QALY, 39,835/LYG 4) metastatic breast cancer: 158,630/QALY. Likewise, ICERs of germline testing for colorectal cancer across settings were: 1) population-wide testing: 132,200/QALY, 1.1 million/LYG 2) people with high-risk: 32,322-76,750/QALY, dominant-353/LYG 3) patients with existing colorectal cancer: dominant - 54,122/QALY, 98,790-6.3 million/LYG. Key areas of underreporting were inclusion of a health economic analysis plan (100% of HBOC and colorectal studies), engagement of patients and stakeholders (95.4% of HBOC, 100% of colorectal studies) and measurement of outcomes (18.2% HBOC, 38.9% of colorectal studies).Conclusion: Germline testing for HBOC was likely to be cost-effective across most settings, except when used as a co-dependent technology with the PARP inhibitor, olaparib in metastatic breast cancer. In colorectal cancer studies, testing was cost-effective in those with high-risk, but inconclusive in other settings. Cost-effectiveness was sensitive to the prevalence of tested variants, cost of testing, uptake and benefits of prophylactic measures. Policy advice on germline testing should emphasize the importance of these factors in their recommendations.
|
v2
|
2022-11-26T06:16:14.071Z
|
2022-11-25T00:00:00.000Z
|
253879274
|
s2ag/train
|
Appropriate use of tapentadol: focus on the optimal tapering strategy
Abstract Objective Due to its opioid and non-opioid mechanism of action, tapentadol is considered an atypical opioid with improved gastrointestinal tolerability versus traditional opioids. As for all opioid analgesics it is important to understand how to discontinue a treatment when it is not needed anymore. The aim of this article was to provide an overview of opioid therapy in non-cancer pain, with a specific focus on tapering of tapentadol in patients with chronic non-cancer pain, and suggestions on how to achieve tapering. Methods Studies for this narrative review were identified via PubMed using a structured search strategy, focusing on management of chronic non-cancer pain with opioids, and the efficacy, tolerability, and pharmacology of tapentadol prolonged release. Publications were limited to English-language articles published within the last ∼10 years. Results The review discusses the use and discontinuation of opioids in general, as well clinical data on discontinuation of tapentadol specifically. We provide a flow chart, which can be used by clinicians in the context of their own clinical experience to appropriately taper tapentadol in patients with chronic non-cancer pain. The flow chart can be easily tailored to individual patient characteristics, duration of tapentadol treatment, response to progressive dosage reduction, and likelihood of withdrawal symptom occurrence. Conclusions While tapentadol is associated with a low frequency of opioid withdrawal symptoms after abrupt discontinuation, use of a tapering strategy is prudent. Tapering strategies developed for opioids in general can potentially be safely individualized in tapentadol-treated patients, although research on tapering strategies for tapentadol is required.
|
v2
|
2022-11-26T06:16:14.125Z
|
2022-11-25T00:00:00.000Z
|
253879874
|
s2ag/train
|
PERFORM: a non-interventional study assessing the patients' treatment starting with 1L palbociclib in HR+/HER2- ABC.
The prospective, non-interventional PERFORM study describes and analyzes the effectiveness of palbociclib in combination with endocrine therapy (aromatase inhibitor or fulvestrant) as first-line treatment for patients with locally advanced or metastatic HR+/HER2- breast cancer in the real-world setting in Germany and Austria. PERFORM will reflect current patient characteristics and routine treatment patterns including treatment sequences and time to subsequent (chemo)therapy. Besides, second-line treatment effectiveness and patient-relevant end points such as longitudinal patient-reported outcome measurements beyond disease progression will be analyzed. Accounting for the heterogenous real-world patient population, data on clinicopathologic subgroups underrepresented in clinical trials such as elderly or male will be analyzed. Taken together, PERFORM will close knowledge gaps from clinical trials in real world.
|
v2
|
2022-11-26T14:04:46.684Z
|
2022-11-25T00:00:00.000Z
|
253882275
|
s2orc/train
|
CXCL1 promotes colon cancer progression through activation of NF-κB/P300 signaling pathway
Background The upregulated expression of CXCL1 has been validated in colorectal cancer patients. As a potential biotherapeutic target for colorectal cancer, the mechanism by which CXCL1 affects the development of colorectal cancer is not clear. Methods Expression data of CXCL1 in colorectal cancer were obtained from the GEO database and verified using the GEPIA database and the TIMER 2.0 database. Knockout and overexpression of CXCL1 in colorectal cancer cells by CRISPR/Cas and "Sleeping Beauty" transposon-mediated gene editing techniques. Cell biological function was demonstrated by CCK-8, transwell chamber and Colony formation assay. RT-qPCR and Western Blot assays measured RNA and protein expression. Protein localization and expression were measured by immunohistochemistry and immunofluorescence. Results Bioinformatics analysis showed significant overexpression of CXCL1 in the colorectal cancer tissues compared to normal human tissues, and identified CXCL1 as a potential therapeutic target for colorectal cancer. We demonstrate that CXCL1 promotes the proliferation and migration of colon cancer cells and has a facilitative effect on tumor angiogenesis. Furthermore, CXCL1 elevation promoted the migration of M2-tumor associated macrophages (TAMs) while disrupting the aggregation of CD4+ and CD8+ T cells at tumor sites. Mechanistic studies suggested that CXCL1 activates the NF-κB pathway. In the in vivo colon cancer transplantation tumor model, treatment with the P300 inhibitor C646 significantly inhibited the growth of CXCL1-overexpressing colon cancer. Conclusion CXCL1 promotes colon cancer development through activation of NF-κB/P300, and that CXCL1-based therapy is a potential novel strategy to prevent colon cancer development.
Introduction
Colorectal cancer is one of the most common malignancies. Globally, there were approximately 1.15 million new cases of colorectal cancer in 2020, accounting for 6.0% of all new cancer cases. In the same year, there were approximately 580,000 deaths due to colorectal cancer, accounting for 5.8% of all cancer-related deaths [1]. Colorectal cancer is the third most common and fatal tumor among both males and females in the United States. According to the latest epidemiological survey, there were an estimated 149,500 new cases and 52,980 deaths in 2021 [2]. The main cause of colon cancer-related mortality is resistance to the current standard chemotherapy regimens, leading to a high incidence of metastatic recurrence [3]. Therefore, identifying effective therapeutic targets for colorectal cancer and improving the sensitivity of chemotherapy are key imperatives to increase the survival rate of patients.
The Nuclear Factor "kappa-light-chain-enhancer" of activated B-cells (NF-κB) is an important immune defense transcription factor that mediates cytoplasmic/ nuclear signaling pathways [4] and regulates the gene expression of various cytokines and cytokine receptors and adhesion molecules in the inflammatory nuclear immune response [5]. It is also associated with apoptotic pathways, cell proliferation and differentiation, and resistance to chemotherapy/radiotherapy in tumors [6,7]. Therefore, targeting the NF-κB pathway may be an effective means of preventing cancer progression and improving chemotherapy sensitivity.
The chemokine C-X-C motif ligand 1 (CXCL1), also known as GROα, belongs to the CXC chemokine family and is secreted by macrophages, neutrophils, and epithelial cells. CXCL1 signaling is through binding to the G protein-coupled chemokine receptor CXCR2 [8], which is predominantly expressed on myeloid cell populations (neutrophils, monocytes, and macrophages) [9]. This receptor directs the efflux of myeloid cells from the bone marrow and their migration to tumor sites with high CXCL1 expression, where they promote tumor immune escape by inhibiting the proliferation, activation, and motility of effector T cells [10,11] and stimulating the expansion of Treg [12].
Studies have demonstrated high expression of CXCL1 in a variety of cancers and its association with cancer progression and inflammation. For example, several studies have found elevated levels of CXCL1 in plasma, serum, ascitic fluid, and tumor tissue of patients with ovarian cancer [13,14]. High CXCL1 expression showed a significant association with lymph node metastasis and poor overall survival in patients with breast cancer [15]. In our previous study, we found high expression of CXCL1 in colorectal cancer and its close correlation with the clinicopathological features; in addition, CXCL1 overexpression was closely associated with tumor diameter, stage, degree of infiltration, and lymph node metastasis [16]. Wang et al. suggested that CXCL1 may enhance the metastasis of colorectal cancer by interacting with CXCR2 [17]. However, the specific mechanism of action of CXCL1 in colorectal cancer is not well characterized.
The current study was designed to study the molecular mechanism by which CXCL1 promotes the progression of colorectal cancer. We hypothesized that CXCL1 promotes the development of colon cancer through the NF-κB/P300 pathway. We first verified the differential expression of CXCL1 in colorectal cancer, and found significantly higher expression of CXCL1 in colorectal cancer compared to that in normal tissues. Knockout of CXCL1 in colon cancer cells significantly inhibited the growth of colon cancer, while overexpression of CXCL1 showed the opposite result. In the mouse model, treatment with P300 inhibitor, C646, led to significant inhibition of tumor growth after overexpression of CXCL1. Bioinformatics analysis using the TCGA database highlighted the clinical relevance of our findings. These data not only suggest that CXCL1 may be a potential therapeutic biomarker of colorectal cancer, but also reveal the potential mechanism by which CXCL1 promotes the progression of colorectal cancer. Our results may help inform novel strategies for targeted therapy of colorectal cancer.
The sequencing dataset and patient parameters for colorectal cancer were obtained from TCGA, containing 698 samples of 51 normal and 647 tumor types. The expression of CXCL1 in tumors and paracancerous tissues of colorectal cancer patients were compared to evaluate its diagnostic value. The staging correlation of colorectal cancer phenotypes with high and low CXCL1 expression was analyzed according to the median expression level.
Identification of the CXCL1 expression profile
The raw expression data from the above five GEO datasets were preprocessed into expression matrices using R software and Microsoft Excel 2019. One of the R commands, NormizeBetween Arrays, was executed to normalize the raw expression data. Differentially expressed genes were analyzed by running the LIMMA package, which is a collection of R commands. By using ImageGP (http:// www. ehbio. com/ Image GP/ index. php/ Home/ Index/ index. html), a web tool for visualizing clustering of multivariate data, the common differentially expressed genes (DEGs) of the above five datasets were identified and displayed in a Venn diagram.
Expression of CXCL1 in colorectal cancer and normal tissues
Human Protein Atlas (HPA: https:// www. prote inatl as. org/) was used to reveal the distribution of CXCL1 in normal human tissues. Gene Expression Profiling Interactive Analysis (GEPIA: http:// gepia. cancer-pku. cn/ index. html) and Tumor Immune Estimation Resource (TIMER 2.0: http:// timer. cistr ome. org/) were used to explore the distribution of CXCL1 in colorectal cancer and normal colon tissues.
Cell lines and culture conditions
Mouse colon cancer cell line MC38 was obtained from the Shanghai Institute of Digestive Surgery (Shanghai, China). The MC38 CXCL1−/− cell line was generated using the CRISPR/Cas 9 technique (see below for details); the MC38 CXCL1+/+ cell line was obtained by transducing CXCL1 to MC38 using the "Sleeping Beauty" transposon (see below for details). Cell lines were cultured in highsugar DMEM medium containing 1% penicillin/streptomycin and 10% fetal bovine serum at 37 °C with 5% CO 2 humidification.
RNA extraction and real-time fluorescence quantitative PCR
Total RNA was extracted from colon cancer lines using Trizol reagent (Magen, R4801-02, Shanghai, China) and 1 µg of total RNA was reverse transcribed using the Pri-meScript RT reagent kit (Vazyme, Nanjing, China) to detect relevant mRNAs. Real-time fluorescence quantitative PCR was performed using SYBR Premix Ex Taq II (Vazyme, Nanjing, China) at 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s, 55 °C for 30 s, and 72 °C for 30 s. The primer sequences used for PCR are shown in Table 1. Changes in relative expression of different samples were calculated by the ABI Q6 Fast real-time PCR system (Applied Biosystems, Foster City, CA, USA) using the 2 −∆∆CT method. GAPDH was used as an internal reference gene.
CRISPR/Cas9-mediated gene knockout
The designed gRNA oligonucleotides were annealed and cloned into the vector pX459 (the primer sequences: m-CXCL1-sg-F: GCT GCT GGC CAC CAG CCG CC; m-CXCL1-sg-R: GGC GGC TGG TGG CCA GCA GC). To knock out the target gene, we introduced the pX459 eukaryotic expression vector carrying the corresponding gRNA into MC38 cells using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA) and screened with 1 µg/mL of puromycin for 3 days after 1 week. Subsequently, the surviving cells were transferred to fresh culture medium without puromycin for 2 days. Plates were spread at an ultralow density to ensure clone formation from individual cells, and then clones were selected and expanded for gene knockout validation by sequencing and Western Blot assay of the target genomic region.
"Sleeping Beauty" transposon-mediated gene transfer CXCL1 was constructed into vector PT2/SVNeo to establish PT2/SVNeo-CXCL1 recombinant transposon vector. To overexpress the target gene, we mixed the recombinant transposon vector and SB100 in a homogeneous ratio of 1:2, introduced into MC38 cells using Lipofectamine 2000 (Thermo Fisher Scientific, Waltham, MA, USA), and spread the plate at ultralow density after 48 h to ensure clone formation from individual cells. Then the clones were selected and expanded. The relative mRNA expression of the target gene was verified by real-time fluorescence quantitative PCR.
Western Blot analysis
Different cells were washed separately with ice-cold PBS, gently scraped with a cell spatula, and lysed using radioimmunoprecipitation assay (RIPA) cell lysis buffer (Beyotime, P0013B, Shanghai, China). Quantification was performed using the BCA protein concentration assay kit (Beyotime, P0012S, Shanghai, China). 60 µg of total protein was passed through 12.5% SDS polyacrylamide gels, and then transferred to PVDF membranes (Bio-Rad, Hercules, CA). The membranes were closed in 5% BSA for 2 h and then incubated overnight with the corresponding primary antibodies at 4 °C, followed by incubation with the corresponding IRDye 800CW or 680 LT secondary antibodies (1:10,000, Abcam, China) in dark for 1 h at room temperature. The fluorescence signals of membranes were detected using Odyssey CLx Western blot detection system (Westburg, Leusden, the Netherlands). The images were analyzed using ImageJ 1.43 software. The expression of β-actin was Table 1 Primer sequences used for qRT-PCR analyses Table 2.
Cell proliferation assay
MC38 WT, MC38 CXCL1−/− , and MC38 CXCL1+/+ cells were inoculated onto 96-well plates at a density of 5 × 10 3 cells/well, respectively. After apposition of cell, cell viability was assayed using a cell counting kit (CKK8, Beyotime, China) according to the manufacturer's instructions to ensure consistent numbers of various cell inoculations. Cell proliferation was subsequently assayed every 24 h for a total of 4 times.
Transwell migration assay
Transwell chambers were used to assess cell migration. A culture medium containing 10% fetal bovine serum was added to the lower chamber. 1 × 10 5 colon cancer cells (MC38 WT, MC38 CXCL1−/− , and MC38 CXCL1+/+ ) were suspended in serum-free culture medium and inoculated in the upper chamber, respectively. After 24 h incubation, the colon cancer cells were removed from the upper chamber with a cotton swab. The cancer cells that penetrated and adhered to the bottom of the filter membrane were fixed with 4% paraformaldehyde in PBS for 15 min, then stained with 0.5% crystal violet for 20 min, and then imaged under 20 × magnification. For statistical analysis, the number of invading cells was calculated from three independent experiments, with an average from four image fields per experiment. The colonies were observed under the microscope.
Colony formation assay
The groups of cells (MC38 WT, MC38 CXCL1−/− , MC38 CXCL1+/+ ) were inoculated at a density of 6 × 10 2 cells/well in six-well plates. After 10 days, cells were washed 3 times with PBS and fixed with 4% paraformaldehyde for 30 min, followed by staining with 0.5% crystalline violet staining solution.
ELISA analysis
The culture supernatant derived from cell medium was collected and the concentration of the CXCL1 secretory protein levels was determined using mouse CXCL1 ELISA kit (Solarbio, Beijing, China) according to the manufacturer's method. Absorbance was measured at 450 nm.
Animal experiments
All animal experiments were conducted according to the Guide for the Care and Use of Laboratory Animals approved by the Fujian Provincial Office for Managing Laboratory Animals and were guided by the Animal Care and Use Committee of Fujian Medical University. Six-week-old C57BL/6 male mice were purchased from the Shanghai SLAC Laboratory Animal Co. (Shanghai, China), and housed at a density of five to six mice per cage in a controlled environment (12 h daylight/ dark schedule) and provided ad libitum access to food and water. MC38, MC38 CXCL1−/− , and MC38 CXCL1+/+ cells were suspended in PBS (1 × 10 6 cells, 100 µL) and injected subcutaneously into the right flank of mice to establish mouse colon cancer tumor models. Mice (MC38, MC38 CXCL1−/− , and MC38 CXCL1+/+ ) were divided into NC group (n = 5) and C646 group (n = 5). Mice in the c646 groups were administered intraperitoneal injection of c646 every 3 days (MCE, 30 nmol/g) for 2 weeks. Mice in the NC groups were administered the same volume of PBS placebo. The tumors of mice were measured every 3 days with vernier calipers and the tumor volume was calculated using the following formula: V = (length) × (width) 2 /2.
Immunohistochemistry
Tumor specimens were fixed in 10% neutral buffered formalin for 24 h, followed by standard tissue processing and embedding. Sections of paraffin-embedded tumor specimens were cut at 4 µm and dried overnight at 37 °C. Sections were then dewaxed twice in xylene for 10 min each and rehydrated by passage through a graded ethanol series. Endogenous peroxidase was inactivated using 3% hydrogen peroxide for 30 min at room temperature. The slides were heated in citrate buffer for antigen repair. The slides were Table 2.
Immunofluorescence
The first half of the immunofluorescence step for tumor tissues was similar to immunohistochemistry. Primary antibodies were removed and washed three times with PBS for 10 min each, and incubated with secondary fluorescent labeling antibodies at room temperature for 2 h. Finally samples were incubated with 4,6-diamidino-2-phenylindole for nuclear staining. Tumor tissues were imaged using an Axiolmage2 ortho-fluorescence phase contrast microscope (Zeiss, Oberkochen, Germany) for section acquisition. Primary antibodies were diluted at 1:1000 using immunostaining primary antibody dilution (Beyotime, P0103, Shanghai, China) and fluorescent secondary antibody was diluted at 1:1000 using immunostaining secondary antibody dilution (Beyotime, P0108, Shanghai, China).
Statistical analysis
Data were expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed using GraphPad Prism 8.0 software. Between-group differences were assessed using Student's t test or one-way analysis of variance (ANOVA). For all analyses, P values < 0.05 were considered indicative of statistical significance. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.
CXCL1 identified as an upregulated gene in GEO datasets
To explore the key genes in colorectal cancer, we simultaneously performed differential gene expression analysis in the five selected GEO datasets. The results showed 30 differentially expressed genes that were common to these five datasets (Fig. 1a, b). Among these 30 overlapping genes, 18 genes were down-regulated genes and 12 genes were up-regulated genes, and CXCL1 was one of the 12 up-regulated genes. The other 11 up-regulated genes, such as FOXQ1, MMP7, MMP11, andCDH3, have been reported in previous cancer-related studies [18][19][20][21][22]. The function and mechanism of action of CXCL1 have seldom been researched in the context of colorectal cancer. Therefore, we selected CXCL1 as the main object of interest. As shown in Fig. 1c, CXCL1 expression was significantly higher in colorectal cancer tissues than in the paratumoral normal tissues.
To further validate CXCL1 expression, we first analyzed CXCL1 expression in human normal tissues. CXCL1 Fig. 1 Analysis of the CXCL1 expression profile in the five GEO databases (GSE41328, GSE106582, GSE25070, GSE156355, and GSE113513). a Venn diagram showing the commonly differentially expressed genes in colorectal cancer; b total symbols of the overlapping down-regulated and up-regulated genes; c CXCL1 mRNA expression between normal tissues and colorectal cancer tissues. **P < 0.01, ***P < 0.001 expression in human normal tissues was obtained from Human Protein Atlas, and the data were based on HPA RNA-seq, Genotype-Tissue Expression (GTEx), and Functional Mammalian Genomes 5 dataset (FANTOM5) (Fig. 2). CXCL1 was found to be highly expressed in human normal colon tissues. Subsequently, we analyzed CXCL1 expression profiles in the GEPIA database and TIMER 2.0 database (Fig. 3a, d), and based on expression data in the databases, we determined that CXCL1 expression was significantly higher in human colorectal cancer tissues than in paratumoral normal tissues. These findings indicated the potential involvement of overexpression of CXCL1 in the genesis and progression of colorectal cancer.
Colorectal cancer data analysis in TCGA database
To further investigate the effect of CXCL1 on colorectal cancer, we compared the expression of CXCL1 in colorectal cancer tissues and paratumoral normal tissues of patients with colorectal cancer using the TCGA database. In addition, we evaluated its diagnostic value, and staged colorectal cancer phenotypes with high and low CXCL1 expression according to the median expression level. A total of 644 colorectal cancer patients were included in this analysis, and RNA sequencing was performed on 647 tumor tissues and 51 paratumoral normal colorectal tissue specimens from these 644 patients. The analysis showed a decrease in CXCL1 expression with the progression of tumor stage (Fig. 3b). Receiver operating characteristic (ROC) curve (Fig. 3c) showed that CXCL1 had a high accuracy for the diagnosis of colorectal cancer [area under the curve (AUC) (95% CI): 0.928 (0.893-0.963)]. The above analysis suggests that CXCL1 is pathologically and clinically associated with the development and progression of colorectal cancer.
CXCL1 promotes the proliferation and migration of colon cancer cells
In order to test the hypothesis that CXCL1 promotes colorectal cancer progression, we used CRISPR/Cas9 technology to knockout CXCL1 in MC38 colon cancer cells to construct MC38 CXCL1−/− cell line; in addition, we also used "Sleeping Beauty" transposon to transduce CXCL1 into MC38 cells to construct MC38 CXCL1+/+ Fig. 3 a CXCL1 expression profile in colorectal cancer based on data from the GEPIA database. b Correlation of CXCL1 expression with clinical characteristics of patients with colorectal cancer; c the ROC curves of CXCL1 gene mRNA expression in tumor tissues and paratumoral normal colorectal tissues of TCGA cohort. d CXCL1 expression profile in colorectal cancer based on data from TIMER 2.0 database. ns no significant, *P < 0.05, **P < 0.01, ***P < 0.001 cell line. The knockout and overexpression efficiencies were assessed by Western Blotting and qPCR (Fig. 4a, b). The results showed successful construction of the MC38 CXCL1−/− and MC38 CXCL1+/+ cell lines. The cell biological behavior was then examined by in vitro assays. In this part, we studied the effect of CXCL1 on the proliferation of colon cancer cells by CCK-8 and clone formation assays. CCK-8 results showed that MC38CXCL1 + / + cells had the fastest proliferation rate, followed by MC38 WT cells, while MC38 CXCL1−/− cells had the slowest proliferation rate (Fig. 4c). Similar results were obtained in cell clonal formation experiments (Fig. 4e). In other words, the number of MC38 cells overexpressing CXCL1 was significantly higher than that of wild-type cells, while the number of MC38 cells with CXCL1 knockout was significantly lower than that of wild-type cells. This result suggested that CXCL1 can promote the proliferation of colon cancer cells. We examined the effect of CXCL1 on migration of colon cancer cells using Transwell assay. The results showed that CXCL1 overexpression significantly promoted the migration of colon cancer cells, while CXCL1 knockout significantly inhibited the migration of colon cancer cells (Fig. 4d). These findings suggest that CXCL1 exerts its role as an oncogene to regulate cell proliferation and migration in colon cancer.
CXCL1 promotes colon cancer growth and angiogenesis
To further validate the promoting effect of CXCL1 on colon cancer growth in vivo, we transplanted three cell lines (MC38 WT, MC38 CXCL1−/− , MC38 CXCL1+/+ ) percutaneously at the right flank of C57BL/6J male mice and started assessing the tumor volume 1 week later (Fig. 5a). The result showed that overexpression of CXCL1 in colon cancer cells significantly promoted colon cancer growth, while knockout of CXCL1 in colon cancer cells inhibited colon cancer growth (Fig. 5b-d).
In addition, dissection of tumor tissues revealed significant vascular infiltration in tumors of mice overexpressing CXCL1 (Fig. 6b). Therefore, we hypothesized 4). ns no significant, **P < 0.01, ***P < 0.001, ****P < 0.0001 Fig. 6 Positive correlation between CXCL1 and the number of tumor vessels. a, c Immunofluorescence assay revealed that CXCL1 overexpression significantly elevated CD31 expression in the colon cancer models, whereas CXCL1 knockout in MC38 resulted in an inhibition of CD31 expression (× 200). b MC38. CXCL1+/+ tumors surrounded by abundant blood vessels. **P < 0.01 that CXCL1 promotes angiogenesis and colon cancer. To verify the relationship between CXCL1 and angiogenesis, we examined CD31 in the tumors of each group of mice using immunofluorescence, and the results showed a significant increase in CD31 expression after CXCL1 overexpression, while the opposite result was observed after CXCL1 knockout (Fig. 6a, c). These data suggest a positive correlation between CXCL1 and CD31 expression. However, further studies are required to identify the pathway by which CXCL1 leads to increased vascularity in colon cancer.
To determine whether the autocrine mechanism is involved in the CXCL1-mediated malignant process in colon cancer, we measured the CXCL1 secretion of each group of cells (MC38 WT, MC38 CXCL1−/− , MC38 CXCL1+/+ ) by ELISA, and the results showed that the CXCL1 protein secreted in CXCL1 knockout cells decreased significantly, while the CXCL1 protein secreted in overexpressing CXCL1 cells was significantly increased (Fig. 7b). The function of CXCL1 is mainly mediated by G protein-coupled receptor CXCR2 binding, and TCGA database analysis shows that there is a high positive correlation between CXCL1 and CXCR2 RNA expression in colorectal cancer (Fig. 7a). Later, we analyzed the expression of CXCR2 in various groups of mouse tumors by immunohistochemistry, and the results showed that CXCR2 was localized on the colon cancer cell membrane and cytoplasm (Fig. 7d), which provided the possibility of binding to CXCL1, but the protein expression level of CXCR2 was negligible compared to the high level of CXCL1 expression in colon cancer cell lines. The results in the HPA database also showed that CXCR2 was almost not expressed in normal colon tissue and colon cancer tumor tissue (Fig. 7c). Based on the above results, we believe that CXCL1 cannot mediate the development of colorectal cancer through autocrine mechanisms.
CXCL1 expression is correlated with immune cell infiltration in colon cancer
The tumor microenvironment plays an indispensable role in tumorigenesis and tumor progression. The interaction between tumor microenvironment and colon cancer is mainly mediated by multiple immune cells and their secreted cytokines, which affect tumor proliferation and metastasis. To detect the effect of CXCL1 on the tumor microenvironment, we examined CD4, CD8, and CD163 expression in tumor tissues of each group of mice (Fig. 8).
The results showed that CXCL1 overexpression in colon cancer cells increased the infiltration of tumor-associated macrophages (TAMs), while decreasing the CD4 and CD8 cells around the tumor. The infiltration of CD4, CD8, and CD163 in tumor tissues after CXCL1 knockout was not significantly different from that in wild-type colon cancer tissues. These data suggest that CXCL1 overexpression reduces T-cell aggregation and increases infiltration of M2-TAMs in colon cancer tissues.
CXCL1-mediated colorectal cancer development is NF-κB/ P300-dependent
CXCL1, one of the most important chemokines, is involved in the development of several inflammatory diseases and shows elevated expression in the inflammatory response [23]. After carcinogen-induced expression, CXCL1 can lead to chronic inflammation by recruiting neutrophils, resulting in tumor formation [23][24][25]. CXCL1 has been reported to activate NF-κB in other cancer types; however, whether CXCL1 can activate NF-κB in colon cancer has rarely been reported. Therefore, we examined the activity of the NF-κB pathway. Firstly, we analyzed the correlation between CXCL1 and several key genes related to the NF-κB pathway by bioinformatics analyses, and the results showed a significant moderate to strong positive correlation of CXCL1 with IL-6, IL-1β, TNF-α, and IFNγ (Fig. 9a-d). To further verify the results, we examined the expression of these genes in MC38, MC38 CXCL1−/− , and MC38 CXCL1+/+ cells by real-time fluorescence quantitative PCR. The results showed that the mRNA expression levels of IL-6 remained consistent with CXCL1 expression levels; IL-1β expression was reduced after CXCL1 knockdown and no significant changes were observed after CXCL1 overexpression; while TNF-α expression was reduced in both ( Fig. 9e-g). This result suggested a significant association between CXCL1 and NF-κB pathway.
Further, we detected the activation of NF-κB pathway by Western Blotting (Fig. 10a-d). The result showed that the phosphorylation level of IκBα was significantly decreased after CXCL1 deletion, while there was no significant change in the phosphorylation level of IκBα after CXCL1 overexpression. However, CXCL1 deletion and overexpression did not significantly affect the phosphorylation of total p65. Since NF-κB needs to enter the nucleus to regulate transcription, we also examined the expression of p65 protein level in the nucleus to assess whether NF-κB was activated. The results showed that CXCL1 overexpression led to a significant increase in the expression of p65 protein level in the nucleus. The above results suggested that CXCL1 is an upstream factor of the NF-κB pathway and its activation of NF-κB pathway may not depend on phosphorylation of p65, whereas NF-κB activation can be achieved by acetylation in addition to phosphorylation.
P300, an important histone acetylase, has been reported to directly acetylate the non-histone transcription factor p65, thereby affecting cell growth and differentiation. To verify the role of P300 in mediating CXCL1 activation of the NF-κB pathway, first, we examined the mRNA expression of P300 in MC38, MC38 CXCL1−/− , and MC38 CXCL1+/+ cells by real-time fluorescence quantitative PCR. The results showed significantly increased expression level of P300 in the MC38 cells overexpressing CXCL1 (Fig. 10e). Then, we added the P300 inhibitor C646 to the cell culture system. After adding different doses of C646 to MC38 CXCL1+/+ cells, C646 inhibited the expression of CXCL1 and IL-6 to different degrees, and the expression trends of CXCL1 and IL-6 remained consistent (Fig. 10h, i). Subsequent Western Blotting results showed significantly decreased expression level of p65 in the nuclei of MC38 CXCL1+/+ cells treated with C646 (Fig. 10f, g). The above results suggested an inhibitory effect of C646 on the activation of NF-κB pathway. In the animal experiments, we administered C646 intraperitoneally to each group of mice. Fig. 8 Relationship between CD163, CD4, and CD8 infiltration and expression of CXCL1 in colon tumor tissues. Immunohistochemistry assay revealed that CXCL1 overexpression significantly elevated CD163 infiltration in colon tumor tissues, whereas CXCL1 knockout in MC38 led to increased CD4 and CD8 cell infiltration (× 200). *P < 0.05, **P < 0.01, ***P < 0.001 In the MC38 CXCL1+/+ group, the tumor volume significantly decreased after C646 treatment, and was restored to MC38 WT level after treatment, while the tumors of mice in the MC38 CXCL1−/− group showed no significant changes (Figs. 11, 12). These results suggest a critical role of P300 in mediating CXCL1-induced activation of the NF-κB pathway.
Discussion
In the present study, we applied bioinformatics analysis and identified increased expression of CXCL1 in colorectal cancer. Previous analyses of clinical colorectal cancer tissue samples have also shown consistent results [16]. Subsequently, we explored the role of CXCL1 in colorectal cancer from the perspective of cell biological behavior. We found that CXCL1 promotes the proliferation, migration, and invasion of colorectal cancer cells. In the subsequent mouse model experiments, CXCL1 was found to promote the progression of colon cancer. These results provide the rationale for further studies on the role of CXCL1 in colorectal cancer.
Angiogenesis is a key factor in carcinogenesis and is regulated by multiple molecular pathways. We observed a positive correlation between CXCL1 and the number Fig. 10 CXCL1 activates the NF-κB pathway via P300. a-d CXCL1 activated NF-κB signaling in MC38 colon cancer cells, presenting as increased expression levels of p-IκBα/IκBα, nuclear-p65/p65. e CXCL1 overexpression in MC38 colon cancer cells promoted P300 mRNA expression. f, g C646 treatment significantly inhibited the nuclear-p65 expression. h, i mRNA expression levels of CXCL1 and IL6 decreased after C646 treatment of MC38 CXCL1+/+ cells, and the decreased level of IL6 remained consistent with CXCL1. ns no significant, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 of tumor vessels in colon cancer. In colorectal cancer patients, CXCL1 also showed a positive correlation with VEGF expression [26]. It has been shown that primary colorectal cancer cells secrete VEGF-A and stimulate TAMs to produce CXCL1 in primary tumors [17], while in tumors of mice treated with chemotherapy, TAMs accumulate around the blood vessels and promote tumor revascularization and recurrence by releasing VEGF-A [27]. In contrast, our data suggest that CXCL1 overexpression increases the expression of CD31 and the infiltration of TAMs. These findings suggest a close association of CXCL1 with TAMs and tumor angiogenesis. However, further studies are required for in-depth characterization of the underlying mechanism of this association.
In colorectal cancer, there are multiple tumor microenvironments, and the complex interactions between tumors and their microenvironments influence the development of tumors and the metastatic spread of cancer cells [28]. T cell infiltration in tumor tissue is required for tumor regression [29,30]. Studies have shown that CXCL1 tends to be highly expressed in tumor cells with low T cell clones, that CXCL1 production promotes the recruitment of MDSCs into the tumor, thereby inhibiting CD8 + T cell infiltration, and that CXCL1 production by tumor cells is required for an immunosuppressive phenotype [31]. Studies have also shown that RIP3 increases CXCL1 expression thereby promoting myeloid cellinduced adaptive immunosuppression in tumors [32,33]. In addition, CXCL1 produced in primary colorectal cancer cells was shown to form a pre-migratory ecology by recruiting MDSCs to eventually promote liver metastasis [17]. In contrast, the CXCL1 receptor CXCR2 antagonist, SB225002, was found to reduce the accumulation of PMN-MDSCs and increase CD8+ T cell infiltration in tumors [34]. Our bioinformatics results suggested an association between CXCL1 and multiple immune cells, while overexpression of CXCL1 in a subsequent mouse model showed increased infiltration of CD4+ and CD8+ T cells. These findings suggested that CXCL1 increases infiltration of T cells, causes myeloid cell-mediated immunosuppression, and thus promotes tumor progression.
A variety of inducible transcription factors, including NF-κB, can be activated by interaction with cellular coactivators [35]. The transcriptional activity of NF-κB can be optimized by interaction with P300/cyclic adenosine monophosphate response element binding protein (CBP) [36]. P300/CBP, a transcriptional co-activator, is believed to regulate transcription through its histone acetyltransferase (HAT) activity [37,38]. Through enzymatic activity localized to the HAT structural domain, P300/CBP activates transcription by transferring acetyl groups to the ε-amino groups of histone lysine residues, leading to elevated levels of histone 3 lysine 27 acetylation (H3K27ac), P53 [39,40], and NF-κB acetylation, thereby promoting expression of a variety of genes. In the present study, CXCL1 was found to activate the NF-κB pathway in colon cancer cells, which is consistent with previous reports. In the study by Kou et al., CXCL1 increased nuclear translocation of NF-κB and activated the NF-κB/ HDAC1 pathway to promote progression of prostate cancer [41]. In contrast, NF-κB-mediated CXCL1 production was found to contribute to the maintenance of bone cancer pain, suggesting some regulatory relationship between NF-κB and CXCL1. In the present study, treatment of CXCL1-overexpressing colon cancer cells with C646 (an inhibitor of P300) led to inhibition of CXCL1 expression along with decrease in the protein level of p65 in the nucleus. Knockout of CXCL1 in colon cancer cells significantly inhibited the growth of tumors in mice compared to that in the other groups. Upon subsequent treatment of mice with C646, the tumors could not be further reduced. In contrast, C646 treatment of mice with CXCL1-overexpressing colon cancer significantly inhibited tumor growth, suggesting the involvement of P300 in CXCL1-mediated bioactivity.
In conclusion, we demonstrated overexpression of CXCL1 in colon cancer tumors. CXCL1, an important cytokine that promotes colon cancer development, induces myeloid cell-mediated immunosuppression. Inhibition of P300 activity blocked NF-κB activation and CXCL1-induced pro-tumor growth effects. Our findings suggest the involvement of NF-κB/P300 in CXCL1 downstream signaling. Therefore, CXCL1 is a potential prognostic biomarker and therapeutic target in the context of colorectal cancer.
|
v2
|
2022-11-26T14:44:46.159Z
|
2022-11-25T00:00:00.000Z
|
253882425
|
s2orc/train
|
Noninvasive detection of pancreatic ductal adenocarcinoma using the methylation signature of circulating tumour DNA
Background Pancreatic ductal adenocarcinoma (PDAC) has the lowest overall survival rate primarily due to the late onset of symptoms and rapid progression. Reliable and accurate tests for early detection are lacking. We aimed to develop a noninvasive test for early PDAC detection by capturing the circulating tumour DNA (ctDNA) methylation signature in blood. Methods Genome-wide methylation profiles were generated from PDAC and nonmalignant tissues and plasma. Methylation haplotype blocks (MHBs) were examined to discover de novo PDAC markers. They were combined with multiple cancer markers and screened for PDAC classification accuracy. The most accurate markers were used to develop PDACatch, a targeted methylation sequencing assay. PDACatch was applied to additional PDAC and healthy plasma cohorts to train, validate and independently test a PDAC-discriminating classifier. Finally, the classifier was compared and integrated with carbohydrate antigen 19-9 (CA19-9) to evaluate and maximize its accuracy and utility. Results In total, 90 tissues and 546 plasma samples were collected from 232 PDAC patients, 25 chronic pancreatitis (CP) patients and 323 healthy controls. Among 223 PDAC cases with known stage information, 43/119/38/23 cases were of Stage I/II/III/IV. A total of 171 de novo PDAC-specific markers and 595 multicancer markers were screened for PDAC classification accuracy. The top 185 markers were included in PDACatch, from which a 56-marker classifier for PDAC plasma was trained, validated and independently tested. It achieved an area under the curve (AUC) of 0.91 in both the validation (31 PDAC, 26 healthy; sensitivity = 84%, specificity = 89%) and independent tests (74 PDAC, 65 healthy; sensitivity = 82%, specificity = 88%). Importantly, the PDACatch classifier detected CA19-9-negative PDAC plasma at sensitivities of 75 and 100% during the validation and independent tests, respectively. It was more sensitive than CA19-9 in detecting Stage I (sensitivity = 80 and 68%, respectively) and early-stage (Stage I-IIa) PDAC (sensitivity = 76 and 70%, respectively). A combinatorial classifier integrating PDACatch and CA19-9 outperformed (AUC=0.94) either PDACatch (0.91) or CA19-9 (0.89) alone (p < 0.001). Conclusions The PDACatch assay demonstrated high sensitivity for early PDAC plasma, providing potential utility for noninvasive detection of early PDAC and indicating the effectiveness of methylation haplotype analyses in discovering robust cancer markers. Graphic Abstract Supplementary Information The online version contains supplementary material available at 10.1186/s12916-022-02647-z.
Background
Pancreatic ductal adenocarcinoma (PDAC) is widely considered one of the most lethal diseases worldwide. One main reason for its high mortality rate is the lack of effective early detection methods. Early symptoms, which typically include abdominal and back pain, diarrhea, weight loss and jaundice, are nonspecific for PDAC and may be associated with other gastrointestinal diseases. This complication is particularly prominent in the differential diagnosis between chronic pancreatitis (CP) and PDAC [1].
Currently, carbohydrate antigen 19-9 (CA19-9) is the most widely used clinical serum marker to detect PDAC, and it can reach a sensitivity and specificity of 75-90% in symptomatic patients prior to resection [2,3]. However, the low sensitivity in early-stage disease has limited its use in early detection protocols. Its practical application in early cancer detection is also hampered by false negatives in Lewis-negative individuals (5-10% of the general population) [4]. Moreover, several large-population studies have demonstrated that CA19-9 is ineffective in detecting PDAC in asymptomatic populations due to its high false-positive rate in conditions of inflammation and nonpancreatic cancers [5,6]. Endoscopic ultrasoundguided fine needle aspiration (EUS-FNA) is a commonly used method to obtain pathological diagnosis; however, it is invasive and has been linked to bleeding and/ or tumour dissemination [7]. Therefore, noninvasive and more accurate methods to detect early PDAC are highly desirable for improving the clinical outcomes of PDAC patients.
In recent years, aberrant DNA methylation has been proposed as a promising marker for noninvasive cancer detection [8]. DNA methylation patterns are profoundly altered in the genome of malignant cells during tumourigenesis and progression [9]. The epigenetic signatures of cancer cells can be utilized to detect circulating tumour DNA (ctDNA) molecules in circulating cell-free DNA (cfDNA) samples [10], which have been explored to develop blood markers for the early screening of PDAC [11][12][13][14][15][16][17].
To further improve the accuracy and robustness of ctDNA methylation markers for early PDAC screening from previous studies, we first employed novel methylation haplotype blocks (MHBs) for marker discovery and validation. MHBs are discrete genomic regions that have tightly coupled CpG methylation sites (i.e., methylation haplotypes) [18] that have been shown to have superiority over methylation on individual CpG sites and were screened as PDAC markers in prior studies, with respect to both sensitivity and specificity in deconvoluting trace amounts of ctDNA from total cfDNA [19]. While MHB analyses have been utilized to identify markers for noninvasive cancer screening in several cancer types [18,19], we were the first to apply it for PDAC marker screening and testing.
Second, we systematically developed new metrics to comprehensively interrogate methylation haplotypes as well as unmethylation haplotypes within MHBs (see Additional file 1: Supplementary Methods [19][20][21][22][23][24][25][26][27][28] for details) to enlarge the pool of potential PDAC markers. These de novo markers were combined with literaturebased candidate markers to form a highly informative marker panel, which was then developed into PDACatch, an ultrasensitive targeted methylation sequencing assay for detecting ctDNA methylation signatures in blood. The PDACatch classifier was built and validated in PDAC and control plasma samples and was tested in a singleblinded manner to confirm its accuracy. It was further compared with CA19-9 in the ability to discriminate PDAC plasma from healthy controls. Our results showed that the PDACatch classifier was more sensitive than CA19-9 in detecting plasma of early-stage PDAC, suggesting its potential to be optimized into diagnostics to detect PDAC early in blood.
Participants
All PDAC plasma samples except the test cohort in Phase III ( Fig. 1) were collected prior to surgery (pancreaticoduodenectomy or distal pancreatectomy) at the Peking Union Medical College Hospital (PUMCH) and Changhai Hospital, Navy Medical University (CHNMU) from Oct. 2017 to Oct. 2019. All PDAC tissues and matched adjacent nontumour tissues were obtained from surgical resection. All PDAC patients were pathologically diagnosed using specimens obtained from surgical resection. The slides were reviewed by two experienced pathologists (H.W. and Z.L.), and the diagnosis was confirmed. CP plasma samples were collected from CHNMU from May 2019 to Sept. 2019. The inclusion criteria for CP patients followed the international consensus [29]. Plasma samples of healthy individuals were collected from PUMHC and CHNMU from May 2018 to Sept. 2019. All cohorts provided informed consent. Samples of the test cohort in Phase III were purchased from ProteoGenex (Inglewood, CA, USA). A few tissue and plasma samples were used in two phases (Additional file 2: Fig. S1). This project was approved by the PUMHC Ethics Committee (No. JS-1490) and CHNMU Ethics Committee (No. CHEC2020-113).
PDAC patients were staged according to the eighth edition of the American Joint Committee on Cancer TNM Staging System. We defined early PDAC as cases with TNM Stage IA, IB, or IIA, which refer to patients without lymph node involvement or distant metastasis [30,31].
Plasma sample preparation, DNA extraction, RRBS and targeted methylation sequencing
All blood samples were collected in cfDNA BCT tubes (Streck). Plasma separation and plasma DNA extraction were performed as previously described (23). Briefly, blood samples were centrifuged at 1600×g for 10 min at 4 °C. The supernatant was transferred and centrifuged again at 16,000×g for 10 min at 4 °C, and plasma was collected and aliquoted into nuclease-free tubes at −80 °C. Circulating cfDNA was extracted from plasma using a QIAamp Circulating Nucleic Acid Kit (Qiagen, 55114) according to the manufacturer's instructions. Tissue samples were freshly obtained after surgical resection. Genomic DNA was extracted using a DNeasy Blood and Tissue Kit (Qiagen).
Reduced representation bisulfite sequencing (RRBS) or targeted methylation sequencing was performed as previously described [19,32]. For each sample, 50-100 ng of genomic DNA or 20 ng of plasma DNA was used. All libraries were paired-end sequenced on Illumina platforms for 150 cycles.
Methylation haplotype measurements
Candidate methylation haplotype blocks (MHBs) were constructed as described previously [19]. The CpG sites within each MHB tend to be tightly coregulated on the epigenetic status at the level of single DNA molecules. We evaluated multiple block-level quantitative metrics to identify the most informative measurement for each target region. Such metrics included AMF (average methylation fraction), MHL (methylation haplotype load), UMHL (unmethylation haplotype load), MHFm (fully methylated haplotype fraction) and MHFu (fully unmethylated haplotype fraction).
AMF
AMF was defined as the average methylation level for all CpG sites in a specific target region. All detected CpG alleles were divided by all methylated CpG alleles of the target region where i represents a CpG site in this target region, M is the total number of CpG sites in this target region, N T,i represents the number of thymines observed at CpG site i and N C,I represents the number of cytosines observed at CpG site i.
MHL, MHL3, UMHL and UMHL3
MHL was defined as in Guo et al. [19], which is the normalized fraction of methylated haplotypes at different lengths.
where l is the length of haplotypes and P(MH i ) is the fraction of fully successive methylated CpGs within i loci. w i is the weight for the i-locus haplotype. The options for weights are w i = i for MHL and w i = i 3 for MHL3. Similar to MHL and MHL3, UMHL and UMHL3 are the normalized fractions of unmethylated haplotypes at different lengths.
MHFm and MHFu
TheMHFm metric was computed for each fully methylated haplotype over each targeted region using the equation: where i is the current locus, h is the current haplotype, N i,h is the number of reads at the current locus containing the current haplotype, and Ni is the total number of reads covering the current locus. MHFu is the fraction of fully unmethylated haplotypes.
Discovery of de novo markers
Target MHBs of the Phase II panel were selected from multiple sources: PDAC tissues vs. healthy plasma, PDAC tissues vs. para-tumour tissues, PDAC plasma vs. healthy plasma and literature searches.
When comparing the methylation profiles of PDAC tissues vs. healthy plasma and PDAC tissues vs. paratumour tissues, the analytical process includes marker filtering and differentially methylated MHB selection. In a library, MHBs with a sequencing depth < 10 were set to NA. MHBs with an NA rate greater than 10% in all libraries were removed from the analyses. Furthermore, MHBs with variations less than 0.02 were removed. After filtering, MHBs with FDR-adjusted p values < 0.05 were selected.
In PDAC plasma vs. healthy plasma analysis, 20 healthy plasma samples were randomly selected to balance cases and controls, which was repeated 500 times. In each iteration, the selected 40 samples were randomly split into a marker discovery set (15 PDAC, 15 healthy) and a validation set (5 PDAC, 5 healthy). Differentially methylated MHBs were identified in the discovery set with the Wilcoxon rank sum test FDR < 0.05. A random forest model was built with these markers in the discovery set and validated in the validation set. If AUC ≥ 0.75, the identified MHBs were kept. MHBs identified over 300 times were selected for downstream analysis.
Regional measurement selection
Methylation markers with low sequencing depth were filtered out. The remaining methylation markers detected in ≥ 90% of the training samples were kept for downstream analyses.
First, we determined the most discriminative measurements for each marker. For every measurement, a logistic regression model was built with one measurement at a time for the training samples using the Python package statsmodels (0.11.1) as follows: h θ (x): phenotype;x: measurement; θ 0 : intercept; θ 1 : coefficient of the measurement.
The p value of θ 1 was returned when the model was built. Measurements of each individual marker were ranked based on the p values, and the measurement with the smallest p value was selected as a regional measurement to represent the methylation status of the marker.
Incremental feature selection and classifier construction
The number of markers was determined by incremental marker selection. After regional measurements of each target were selected, missing measurement values were imputed with the values of 5 nearest neighbours (KNN).
The selected values of a marker were then scaled based on the median value and the 25-75% interquartile range. Then, the training samples were randomly split into 10 fractions: a support vector machine (SVM) model for each marker was built using 9 fractions and tested by the remaining 1 fraction. This process was repeated 10 times, during which the area under the curve (AUC) was calculated for each test and averaged. Note that we started with the marker with the smallest p value.
For a new marker, if its average AUC of the 10 tests did not decrease below the average AUC of the previously tested marker(s), this marker was included for classifier training. After all markers were tested, an SVM classifier for PDAC plasma was built using all included markers by classifying all training samples and was validated using the validation samples (Additional file 2: Fig. S2). Plasma samples with CA19-9 level information were selected to build a combinatorial classifier integrating the methylation classifier and serum CA19-9 level.
Statistical analyses
Statistical analyses were performed in R 3.5.0. In Phase II, chi-square tests were utilized to test the difference in methylation level distribution between the case and control groups for each marker: values of each marker were assigned to 10 windows evenly distributed from 0 to 1 for the PDAC and healthy groups. A chi-square contingency test was performed to test whether the distribution of samples in each window was identical between the PDAC and healthy groups. Measurements with the smallest chisquared test p value of each target were selected as the methylation status of the corresponding target regions. Binomial confidence intervals for sensitivity and specificity were calculated using the Clopper-Pearson method. To assess whether the difference observed between AUCs from the combinatorial model (i.e., CA19-9+PDACatch) and the CA19-9 alone model was statistically significantly different from 0, we considered a test statistic T [(T = AUCCA19-9 − AUCCA19-9+PDACatch)2/(s2 CA19-9 + s2 PDACatch)] [10], which looks at the difference in AUC between the two models divided by the sum of the variances from the two models. The fact that this test statistic followed a χ 2 distribution with 1 degree of freedom under the null hypothesis was used to calculate a resulting p value. A bootstrap percentile confidence interval (CI) approach was used to estimate a 95% CI for the AUC (1000 times).
Study design and sample description
This study utilized a total of 90 tissues (52 PDAC tumours, 38 matched para-tumour tissues) and 546 plasma samples (198 PDAC, 25 CP and 323 healthy controls) to sequentially develop a PDACatch assay (Fig. 1, Additional file 2: Fig. S1, Tables 1 and 2). PDAC samples were collected from 232 PDAC patients (18 PDAC patients provided both tissue and plasma samples). Among 223 PDAC patients with known stage information, 43/119/38/23 cases were Stage I/II/III/IV (Additional file 2: Fig. S2A). In Phase I, we discovered de novo PDAC-specific markers by analysing the genomic DNA methylation profiles of PDAC tumours, normal tissues and plasma samples using the RRBS method [19]. In Phase II, markers were tested in additional tissue and plasma samples for their PDAC-discriminating accuracy. The most informative markers were selected to develop a targeted sequencing assay, PDACatch. In Phase III, PDAC classifiers were built and validated in 199 plasma samples to separate PDAC patients from healthy individuals or CP patients. Furthermore, we conducted a single-blinded test of the PDACatch classifier using an independent cohort of PDAC plasma and healthy controls. In Phase IV, we compared PDACatch and CA19-9's performances in classifying PDAC plasma and explored an integrated classifier to further improve the accuracy. Note that a small number of tissue and plasma samples were shared in multiple phases of this study (see details in the relevant "Results" sections).
Discovery of PDAC-specific methylation markers in tissue and plasma
We searched predefined methylation haplotype blocks (MHBs) for de novo PDAC-specific DNA methylation markers by first profiling the methylation patterns of 46 PDAC tumours, 28 para-tumour tissues and 143 plasma samples (20 PDAC, 123 healthy) using RRBS (Table 1). Multiple MHB-specific metrics (see the "Methylation haplotype measurements" section for details) were used to quantify methylation levels to identify MHBs containing PDAC-specific methylation haplotypes as markers via PDAC tissue group vs. para-tumour tissue group (T2T), PDAC tissue vs. healthy plasma (T2P) and PDAC plasma vs. healthy plasma (P2P) comparisons ( Fig. 2A). The first set of 76 markers was yielded by intersecting the 1870 T2P markers and 700 T2T markers (Wilcoxon rank sum test, Benjamini and Hochberg FDR <0.05). The second set comprised 42 T2T markers 15 located within −1500 to +1000 bp of the transcription start sites of 819 genes exhibiting aberrant methylation changes in PDAC tissues or plasma [21][22][23][24]. The third set of 53 markers was selected from P2P markers via a model-based cross-validation marker selection process using an AUC value of over 0.75 as the cut-off for qualified markers. In total, 171 de novo markers were compiled for downstream analysis. Gene set enrichment analysis (GSEA) of these de novo MHBs revealed that multiple cancer-related pathways or biological processes were enriched in their associated genes (FDR < 0.05, Fig. 3A, B and Additional file 3: Table S1) [33], including 7 MsigDB hallmark pathways known to be dysregulated in PDAC (Fig. 3B) [23,34]. Furthermore, the PDAC marker genes previously published in the literature were also highly enriched in our top marker-associated genes (hypergeometric test, Additional file 4: Table S2) [21][22][23][24]. These results strongly supported that the de novo markers we selected are involved in PDAC carcinogenesis.
Marker optimization and PDACatch assay development
In Phase II, we further reduced the number of PDAC markers to minimize model overfitting caused by the imbalance of hundreds of features and limited samples. To this end, all the selected markers were integrated with the PanSeer assay [32], and their separation power was validated. PanSeer is a targeted methylation sequencing assay that is highly sensitive for detecting early-stage cancer signals in blood. It is also readily customized for different sets of targets, making it versatile for investigating different types of cancers.
A combination of the PDAC de novo markers with the PanSeer markers formed a starting set of 750 markers for further testing (Fig. 2B). We filtered them by using them to discriminate PDAC tumours from para-tumour tissues (N = 27 and 17, respectively). Among them, 21 PDAC and 7 para-tumour tissues were previously used in Phase I. They were reused in Phase II to confirm that RRBSdiscovered markers can also be consistently detected by targeted methylation sequencing. The top 200 most discriminating markers (p < 0.05, Wilcoxon rank sum test) were selected and preliminarily filtered based on their ability to classify PDAC and healthy plasma (N = 29 and 55, respectively) via cross-validation (Fig. 2C) and the distribution of their methylation haplotype measurements in these samples (chi-square test, p < 0.01). The 185 most significant markers were chosen to develop the final version of the PDACatch assay to detect the PDAC marker signature in blood.
Model building and evaluation of the PDACatch classifier for early PDAC detection
We then sought to develop a PDAC early detection classifier to separate PDAC plasma from healthy controls by the PDACatch assay. To this end, 94 PDAC and 80 healthy samples, which were age-and sex-matched, were randomly split into a training set and a validation set at a 2:1 ratio ( Fig. 1 and Table 1). The training set included 19 PDAC plasma samples that were previously tested in Phase II and had sufficient remaining cfDNA. This was to increase the size of the training set to improve the trained classifier's robustness; however, no samples were reused in validation to prevent biasing the validation results. Using training samples, the 56 most discriminatory markers for PDAC were identified by 10-fold cross-validation incremental feature selection (Additional file 2: Fig. S3 and Additional file 5: Table S3) to build an SVM-based classifier with a high AUC of 0.93 in the training set (sensitivity = 71%, specificity = 91%) (Fig. 4A). The PDACatch classifier was then validated in the left-out validation set and achieved a similar AUC of 0.91 (sensitivity = 84%, specificity = 89%) using the same cut-off as in the training set, demonstrating its consistency and robustness (Fig. 4A, B). Covariant analysis also showed that the PDACatch classifier was independent of age, sex, tumour location and size (Fig. 4C-F).
Genes associated with the 56 markers of this classifier were annotated and a number of cancer-related genes (Fig. 3E). It may be worth exploring whether these transcription factors have regulatory roles in PDAC carcinogenesis. While testing CP samples, we found that the PDA-Catch classifier showed limited accuracy in stratifying CP from PDAC. We then rebuilt an SVM-based classifier to separate PDAC from CP plasma, which achieved an AUC of 0.85 for samples in the validation set (Additional file 2: Fig. S4A). This PDAC-CP classifier exhibited a consistent accuracy for PDAC across all stages (Additional file 2: Fig. S4B) with no significant covariate differences (Additional file 2: Fig. S4C-F). Although the limited CP samples likely have reduced performance during validation, the results did suggest potentially great feasibility in differentiating PDAC from CP using ctDNA methylation as markers to reduce misdiagnosis due to the lack of discriminatory symptoms.
Comparison of the PDACatch classifier with serum CA19-9 levels
As mentioned earlier, serum CA19-9 is commonly used as a blood marker to stratify PDAC risk. Thus, it is necessary to compare the performance of the PDACatch classifier with CA19-9 in all samples with available test results for CA19-9 to assess PDACatch's clinical utility and significance.
We compared the classification accuracy by PDACatch and CA19-9 on all 92 PDAC and 37 healthy cases with The same samples were also tested for CA19-9 levels. Orange dots show the CA19-9-positive cases (>37 U/ml), and blue dots show CA19-9-negative cases (≤37 U/ml) known CA19-9 levels from the training and validation samples of Phase III ( Fig. 1 and Table 2) and found that on balance, PDACatch was modestly more accurate than CA19-9, as demonstrated by the fact that PDACatch has a higher, or at least an equal, AUC score than CA19-9 for PDAC of each stage (Fig. 5A and Table 3). Importantly, PDACatch was more sensitive in detecting Stage I (sensitivity = 80 and 68% for PDACatch and CA19-9, respectively, Additional file 2: Fig. S5) or early-stage (I-IIa) PDAC plasma than CA19-9 (sensitivity = 76 and 70% for PDACatch and CA19-9, respectively). Note that in this comparison, PDACatch and CA19-9 had the same specificity of 89%. These results indicate that PDACatch may be more advantageous in detecting early PDAC cases than CA19-9.
Lastly, we explored integrating CA19-9 with the PDA-Catch classifier to potentially maximize the model accuracy. To this end, we used cases from the training and validation sets for PDACatch that had known CA19-9 levels for the combinatorial model's training (23 healthy, 62 PDAC) and validation (14 healthy, 31 PDAC), respectively. The combinatory classifier was trained by logistic regression and achieved an AUC of 0.93 (sensitivity = 82%, specificity = 87%); in validation, it achieved an AUC score of 0.96 (sensitivity = 94%, specificity = 93%), which was higher than either parental classifier (0.87 and 0.90 for CA19-9 and PDACatch, respectively) in classifying the same validation samples (Fig. 5, Table 3).
Because the combinatorial classifier had consistent performances in both training and validation cohorts, we further compared the combinatorial classifier's performances with CA19-9 in classifying all the cases of these 2 cohorts. Indeed, the combinatorial classifier had an AUC of 0.94, higher than CA19-9 (AUC = 0.89, Fig. 5A-E and Table 3). Additionally, it was more sensitive than CA19-9 when classifying Stage I (sensitivity = 92 and 68% for the combinatorial classifier and CA19-9, respectively, p < 0.05, McNemar's test, Additional file 2: Fig. S5) or earlystage PDAC plasma (I-IIa) (sensitivity = 88 and 70% for the combinatorial classifier and CA19-9, respectively, p < 0.05, McNemar's test, Additional file 2: Fig. S6). These results suggest that early detection of PDAC may be improved by combining PDACatch and CA19-9.
Independent test of the PDACatch classifier to distinguish PDAC and healthy plasma samples
To independently verify the PDACatch classifier's utility in classifying PDAC plasma, we conducted a singleblinded classification on a cohort of preoperative PDAC (N = 74) plasma samples and healthy controls (N = 65, Fig. 1 and Table 1) obtained from ProteoGenex, a biobank in the USA. The PDACatch assay was performed on these samples, and the same classifier and cut-off were applied to label these samples as PDAC or normal. The results showed that for the blind-test cohort, PDA-Catch achieved an AUC of 0.91 (sensitivity = 82%, specificity = 88%, Fig. 5C) in classifying PDAC cases, reaching a degree of accuracy that was essentially identical to that of the validation cohort (AUC = 0.91, sensitivity = 84%, specificity = 89%), further confirming its robustness and consistency. Stagewise, PDACatch detected early-stage PDAC (I-IIa) at a sensitivity of 80% and advanced-stage PDAC (IIb and above) at 83%, both of which were also consistent with the results of the validation cohort.
Importantly, PDACatch correctly identified all 7 CA19-9-negative PDAC samples in this cohort, achieving a sensitivity of 100% (Fig. 5D). While the number of such cases was relatively small in this cohort (22 of the 74 PDAC samples had serum CA19-9 levels measured), combined with the results of the same test on CA19-9-negative cases of the training and validation cohorts, it nonetheless demonstrated the PDACatch classifier's consistent accuracy at detecting CA19-9-negative PDAC cases. Taken together, we found that the PDACatch classifier performed consistently in classifying PDAC plasma in the independent blind-test cohort as it did in the training and validation cohorts, confirming its robustness and utility for the noninvasive detection of PDAC in blood.
Discussion
In this study, we investigated PDAC and control tissues and plasma to sequentially discover, develop, validate and test ctDNA methylation signatures for early PDAC detection. Methylation haplotype-based analyses were performed in the marker discovery phase to improve the specificity and robustness. PDACatch, a highly sensitive targeted methylation sequencing assay, was developed to integrate the most discriminating markers for PDAC. The 56-marker PDACatch classifier was built and performed better than CA19-9 in detecting early-stage PDAC and CA19-9-negative cases at high specificity. Most importantly, the PDACatch classifier was confirmed in an independent test to accurately classify PDAC plasma from healthy controls.
Taken together, these results are another step forward to achieving accurate detection of early PDAC using blood specimens, which is arguably the most cost-effective approach to reducing the high mortality rate of PDAC [40]. Neither imaging modalities nor non-CA19-9 serum markers have sufficient efficacy to detect early-stage PDAC [41]. Recently, ctDNA methylation has shown great potential as a blood marker to detect PDAC in its early stages [17]. Our results supported this notion, as both PDACatch-based classifiers achieved a high degree of sensitivity for early-stage PDAC plasma (76~82%), which is comparable to a recent multicancer early screening study's results that also used DNA methylation changes as blood markers for cancers [42]. We further demonstrated that methylation markers can be combined with CA19-9 to maximize the overall accuracy, especially for CA19-9-negative PDAC cases that lack Lewis antigens and/or have early-stage disease [43]. This finding is especially meaningful to improve the diagnostic and prognostic stratification of PDAC patients.
During de novo marker discovery, we analysed MHBs of tissues and plasma to identify candidate ctDNA markers for PDAC in addition to AMF. When the methylation haplotype was first introduced, only comethylation was quantified (i.e., by the MHL measurement) [19], which limits the number of identified methylation markers. In this study, to increase the size of potential PDAC plasma markers, we expanded the analysis to include co-unmethylation by the UMHL measurement. We also explored weighing the length of the haplotype with an exponent of 3 in the MHL3 and UMHL3 measurements instead of just 1 in MHL and UMH to utilize the density of CpG sites in marker discovery. Finally, we specifically analysed whether the fully methylated (MHFm) or unmethylated (MHFu) haplotype was differentially expressed between PDAC samples and their controls.
Admittedly, applying the new measurement did increase the complexity in marker selection and modeling. However, PDACatch performed with notably high accuracy and consistency in all study cohorts that included both Chinese and non-Chinese populations, proving that a model incorporating these new types of methylation haplotypes can be stable and robust for populations of diverse genetic and epigenetic backgrounds, in addition to having more potential markers to choose from for model development. Practically, because the new measurements such as UMHL were all developed using a principle or formula similar to that of MHL, they can be easily implemented by researchers who are reasonably well-versed with MHL.
In addition to diagnostic, monitoring and prognostic applications [44,45], ctDNA methylation has been investigated in recent years as a possible early detection biomarker in PDAC [17,[46][47][48][49]. Most of these studies were based on relatively small sample sizes and evaluated individual differentially methylated CpG sites or genes, which were typically selected from literature searches or by in silico prediction and might not be able to capture the complex biology of PDAC. In the present study, we conducted a comprehensive discovery and validation process in four sequential phases: (1) biomarker discovery; (2) assay development; (3) training, validation and testing; and (4) assay integration. Our classifier was trained and validated in a relatively large Chinese cohort and tested in an independent dataset of Caucasian individuals.
Early-stage and CA19-9-negative cases, two important target groups for the early detection of PDAC, were analysed separately. Moreover, MHBs rather than individual CpGs were analysed for marker discovery, during which novel metrics quantifying the methylation status of MHBs were used to obtain the most representative measurement to identify differentially methylated MHBs.
Several limitations should be acknowledged. Our preliminary results from using methylation markers to differentiate PDAC plasma from CP were encouraging. However, they need to be further validated using a larger number of CP samples. The sample sizes of several key categories of cases, namely, Stage I and CA19-9-negative PDAC, were also limited (Additional file 2: Fig. S2). In addition to the high cost and complexity, ctDNA tests likely suffer from the same problems of insufficient sensitivity and specificity as traditional biomarkers when applied to population screening and early cancer diagnosis, given that the fraction of ctDNA was extremely small in total plasma DNA in earlystage tumours [50]. Although our test demonstrated improved accuracy using PDACatch over CA19-9 for early-stage PDAC, its sensitivity and specificity are still not good enough for early detection or screening purposes in PDAC and need further improvement. Last, a considerable number of PDAC patients with advancedstage disease were included in our cohorts, which might overestimate PDACatch's sensitivity. The effectiveness of the PDACatch assay in early PDAC detection needs further evaluation in a large multicentre prospective study.
Conclusions
In summary, we conducted a de novo genome-wide screening using methylation haplotype-based analyses for PDAC-specific DNA methylation markers, and built a PDACatch classifier for early PDAC detection. Despite the limitations mentioned above, we believe that our study is an important step forward in reaching the goal of accurately noninvasively detecting early-stage PDAC to reduce the high mortality rate of PDAC. It will not only benefit early PDAC detection, but its methodology and analyses may play the foundation to develop DNA methylation-based diagnostics for other cancers.
|
v2
|
2022-11-27T06:17:05.945Z
|
2022-11-25T00:00:00.000Z
|
253967260
|
s2ag/train
|
Growing small solid nodules in lung cancer screening: safety and efficacy of a 200 mm3 minimum size threshold for multidisciplinary team referral
The optimal management of small but growing nodules remains unclear. The SUMMIT study nodule management algorithm uses a specific threshold volume of 200 mm3 before referral of growing solid nodules to the multidisciplinary team for further investigation is advised, with growing nodules below this threshold kept under observation within the screening programme. Malignancy risk of growing solid nodules of size >200 mm3 at initial 3-month interval scan was 58.3% at a per-nodule level, compared with 13.3% in growing nodules of size ≤200 mm3 (relative risk 4.4, 95% CI 2.17 to 8.83). The positive predictive value of a combination of nodule growth (defined as percentage volume change of ≥25%), and size >200 mm3 was 65.9% (29/44) at a cancer-per-nodule basis, or 60.5% (23/38) on a cancer-per-participant basis. False negative rate of the protocol was 1.9% (95% CI 0.33% to 9.94%). These findings support the use of a 200 mm3 minimum volume threshold for referral as effective at reducing unnecessary multidisciplinary team referrals for small growing nodules, while maintaining early-stage lung cancer diagnosis.
|
v2
|
2022-11-27T16:09:50.457Z
|
2022-11-25T00:00:00.000Z
|
254002570
|
s2orc/train
|
Immune-Related Gene Signatures to Predict the Effectiveness of Chemoimmunotherapy in Triple-Negative Breast Cancer Using Exploratory Subgroup Discovery
Simple Summary Chemoimmunotherapy combinations have transformed the treatment landscape for patients with triple-negative breast cancer (TNBC). However, the discovery of immune-related biomarkers is needed to optimally identify patients requiring the addition of immune-checkpoint inhibitors (ICIs) to chemotherapy. In this study, we identified immune-related gene signatures via exploratory subgroup discovery algorithm that substantially increase the odds of partial remission for TNBC patients on anti-PD-L1+chemotherapy regimen. We have also uncovered distinct cell populations for TNBC patients with various treatment outcomes. Our framework may result in better risk stratification for TNBC patients that undergo chemoimmunotherapy and lead to overall improvement of their health outcomes in the future. Abstract Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer with limited therapeutic options. Although immunotherapy has shown potential in TNBC patients, clinical studies have only demonstrated a modest response. Therefore, the exploration of immunotherapy in combination with chemotherapy is warranted. In this project we identified immune-related gene signatures for TNBC patients that may explain differences in patients’ outcomes after anti-PD-L1+chemotherapy treatment. First, we ran the exploratory subgroup discovery algorithm on the TNBC dataset comprised of 422 patients across 24 studies. Secondly, we narrowed down the search to twelve homogenous subgroups based on tumor mutational burden (TMB, low or high), relapse status (disease-free or recurred), tumor cellularity (high, low and moderate), menopausal status (pre- or post) and tumor stage (I, II and III). For each subgroup we identified a union of the top 10% of genotypic patterns. Furthermore, we employed a multinomial regression model to predict significant genotypic patterns that would be linked to partial remission after anti-PD-L1+chemotherapy treatment. Finally, we uncovered distinct immune cell populations (T-cells, B-cells, Myeloid, NK-cells) for TNBC patients with various treatment outcomes. CD4-Tn-LEF1 and CD4-CXCL13 T-cells were linked to partial remission on anti-PD-L1+chemotherapy treatment. Our informatics pipeline may help to select better responders to chemoimmunotherapy, as well as pinpoint the underlying mechanisms of drug resistance in TNBC patients at single-cell resolution.
Introduction
Triple-negative breast cancer (TNBC) occurs in about 10 to 20% of diagnosed breast cancers and defined by the absence or minimal expression of estrogen receptor (ER), progesterone receptor (PR) and epidermal growth factor receptor 2 (HER2) [1,2]. Due to its aggressive clinical phenotype and limited response to hormonal therapy, one in three TNBC patients will likely to relapse within the first three years of primary diagnosis [3]. Although numerous therapeutic agents have been evaluated for the treatment of early TNBC [4], only Olaparib has been approved for the treatment of the small group of patients with high-risk TNBC harboring germline BRCA1 or BRCA2 pathogenic variants in the adjuvant setting [5]. The emergence of cancer immunotherapy, however, is altering the paradigm in TNBC treatment.
TNBC, unlike other breast cancer subtypes, has high tumor mutational burden (TMB), which has been correlated with responsiveness to immune checkpoint inhibitors (ICIs) [6]. Indeed, checkpoint inhibition with the anti-PD1 antibody Pembrolizumab has been approved for advanced-stage, PD-L1 positive TNBC due to improved outcomes when combined with frontline chemotherapy [7]. Interestingly, ICIs are more effective in treating TNBC when given early in the course of the disease, which may be a result of immune escape mechanisms emerging as the condition progresses [8]. More recently, results from the KEYNOTE-522 trial indicated that adding checkpoint inhibition in the early stage setting does in fact improve long-term outcomes [9]. However, subgroup analyses did not pinpoint any strongly predictive biomarkers. For example, PD-L1 expression did not distinguish responders from non-responders in the early setting, with both PD-L1-negative and PD-L1-positive patients obtaining a benefit from Pembrolizumab. Moreover, the addition of immunotherapy increased adverse effects (AEs) [10]. In another study-IMPASSION131the combination of Paclitaxel with the PD-L1 inhibitor Atezolizumab failed to improve progression-free survival (PFS) or overall survival (OS) in TNBC patients [11]. These findings could be due to imbalances in prognostic features or accidental discoveries in a relatively small trial. Therefore, the exploration of immune-related biomarkers is needed to optimally identify patients requiring the addition of ICIs to chemotherapy [12,13].
In this work we determined homogenous TNBC subgroups based on both phenotypic and genotypic parameters using exploratory subgroup mining. We have also identified significant predictors that increase chances of partial remission in TNBC patients on chemoimmunotherapy treatment using multinomial regression model on TNBC scRNA-seq dataset. Lastly, we uncovered distinct immune cell populations (T-cells, B-cells, Myeloid, NK-cells) for TNBC patients with various treatment outcomes. We interpreted our results using biomedical knowledge, including findings from existing clinical trials, immunohistochemistry experiments and functional characterization of specific genes. The proposed informatics pipeline may assist health care professionals in the selection of chemoimmunotherapy responders, as well as determine the underlying causes of drug resistance in TNBC patients at a single-cell level and resolution.
Data Mapping
In this study we employed two datasets. Each dataset consisted of multiple phenotypic (either categorical or continuous) and genotypic (continuous only) variables. Each categorical variable was labeled based on the National Comprehensive Cancer Network (NCCN) Guidelines in Oncology [14]. For example, relapse-free status was categorized as (1) disease-free or (2) recurred. Continuous variables were converted into categoric variables by grouping values into several categories. For example, normalized gene expression values were categorized as (1) downregulated, (2) upregulated, or (3) non-differentially expressed.
The first TNBC dataset comprised of 422 patients. These patients were selected from 24 breast cancer studies available at the cBioPortal platform [15]. The final dataset included breast cancer patients based on the following immunohistochemical profile: ER-negative, PR-negative and HER2-negative. This dataset consisted of 12 phenotypic variables, in- cluding clinical-pathologic data (age at diagnosis, menopausal status, tumor type, tumor stage, tumor cellularity, histologic grade, TMB), treatment regimen (chemotherapy, radiotherapy, hormone therapy) and survival status (overall survival status, relapse-free status). There were 1067 genotypic variables in the form of normalized gene expression values derived from human immunome (immune-related genes) and human kinome (protein kinase genes).
The second TNBC dataset consisted of scRNA-seq profiles for 22 TNBC patients that underwent chemotherapy (Pactilaxel) or chemoimmunotherapy treatment (Paclitaxel with Atezolizumab) [16]. For this study we selected six phenotypic variables, including information about treatment timeline (pre-, post-treatment, progression), tissue type (tumor or blood), tumor site (brain, breast, chest wall, liver, lymph nodes), treatment type (anti-PD-L1+chemotherapy or chemotherapy only), treatment response (partial response (PR), stable disease (SD), progressive disease (PD) and cell cluster (T-cells, B-cells, Myeloid, NK-cells)). We used the same genotypic variables as in the TNBC subgroup discovery dataset.
The Informatics Pipeline
Our informatics pipeline has three modules: (1) exploratory subgroup discovery, (2) inference module based on multinomial regression model and (3) immune cell populations discovery. Our goal was two-fold: (1) to identify significant genes from exploratory subgroup discovery that increase odds of having partial remission after anti-PD-L1+chemotherapy treatment and (2) to uncover unique immune cell populations for TNBC subgroups with various treatment outcomes.
The main goal of exploratory subgroup discovery module was to determine homogenous patient subgroups based on expanatory phenotypic characteristics (Module A on Figure 1), where prevailing number of patients in that subgroup exemplify distinctive genotypic patterns (Module B on Figure 1). Each genotypic pattern had been represented as a combination of differentially expressed genes. For example, the genotypic pattern may consist of three genes: upregulated EGFR, downregulated MTOR and upregulated MAPK1 genes. On the first step, the algorithm determines the base subgroup (e.g., Chemotherapy = Yes) contingent on the most significant contrast against the rest of the population. On the next inclusion step it adds a new phenotypic variable, e.g., TMB = High, to the previous subgroup to generate a more focused subgroup (Chemotherapy = Yes and TMB = High). Subsequently, the exclusion step is employed to remove a less relevant inclusion move after each inclusion step. The exploratory search selects multiple paths that form multiple subgroups and have equally relevant genotypic patterns within each subgroup. When the algorithm reaches the most focused subgroup with the highest contrast score that cannot be further increased, the search would be terminated. Support [17] and growth rate [18] were used to measure the frequency for a specific genotypic pattern in the homogenous subgroup. We then applied a J-value [19] to prioritize each subgroup based on the relevance (contrasts) for all patterns in each subgroup [20].
To find significant predictors of partial remission on anti-PD-L1+chemotherapy regimen, we employed multinomial regression model (Module C on Figure 1) on the scRNA-seq TNBC dataset. The outcome variable was categorical and represented as a combination of treatment response, treatment timeline, and treatment type. For example, the level of outcome variable can be encoded as SD-Post_treatment-Chemo meaning that a fraction of TNBC patients achieved stable disease after treatment with chemotherapy only. Overall, there were ten levels of outcome variable. We set PD-Post_treatment-Chemo-progressive disease after chemotherapy-as a baseline for the model. The continuous covariates were encoded as genes with normalized gene expression values identified as a top 10% of genotypic patterns in the exploratory subgroup discovery stage. We used the multinom function from the nnet package [21] to estimate a multinomial logistic regression model. We computed p-values via two-tailed z-test to identify significant predictors of response to anti-PD-L1+chemotherapy treatment.
progressive disease after chemotherapy-as a baseline for the model. The continuous covariates were encoded as genes with normalized gene expression values identified as a top 10% of genotypic patterns in the exploratory subgroup discovery stage. We used the multinom function from the nnet package [21] to estimate a multinomial logistic regression model. We computed p-values via two-tailed z-test to identify significant predictors of response to anti-PD-L1+chemotherapy treatment. (2) chemotherapy, post treatment, PD. Using the top 10% of genotypic patterns from exploratory mining stage as an input, we generated heatmap plots for each condition in every TNBC subgroup of interest. For example, NME3 gene was represented as a geometric mean of NME3 expression values in CD4-Tcm-LMNA cells [22]. Finally, we compared immune cell populations in these two conditions to identify mutually exclusive cell populations that were associated with either partial remission after anti-PD-L1+chemotherapy treatment or progressive disease after chemotherapy treatment.
The Identification of Homogenous TNBC Subgroups
First, we ran the exploratory subgroup discovery algorithm on the TNBC dataset described in Section 2.1. The algorithm revealed 11944 subgroups. We focused our analysis of the 460 subgroups where TNBC patients had undergone chemotherapy. On the next step, we narrowed down the search to twelve homogenous subgroups based on TMB (low or high), relapse status (disease-free or recurred), tumor cellularity (high, low and moderate), menopausal status (pre-or post) and tumor stage (I, II and III). Since the lengths of genotypic patterns vary (up to 5 genes), we decided to make a union of top 10% of genotypic patterns for each subgroup of interest. Let us assume that each genotypic pattern is a set of elements, where each element is a unique differentially expressed gene (e.g., (2) chemotherapy, post treatment, PD. Using the top 10% of genotypic patterns from exploratory mining stage as an input, we generated heatmap plots for each condition in every TNBC subgroup of interest. For example, NME3 gene was represented as a geometric mean of NME3 expression values in CD4-Tcm-LMNA cells [22]. Finally, we compared immune cell populations in these two conditions to identify mutually exclusive cell populations that were associated with either partial remission after anti-PD-L1+chemotherapy treatment or progressive disease after chemotherapy treatment.
The Identification of Homogenous TNBC Subgroups
First, we ran the exploratory subgroup discovery algorithm on the TNBC dataset described in Section 2.1. The algorithm revealed 11,944 subgroups. We focused our analysis of the 460 subgroups where TNBC patients had undergone chemotherapy. On the next step, we narrowed down the search to twelve homogenous subgroups based on TMB (low or high), relapse status (disease-free or recurred), tumor cellularity (high, low and moderate), menopausal status (pre-or post) and tumor stage (I, II and III). Since the lengths of genotypic patterns vary (up to 5 genes), we decided to make a union of top 10% of genotypic patterns for each subgroup of interest. Let us assume that each genotypic pattern is a set of elements, where each element is a unique differentially expressed gene (e.g., upregulated MTOR gene). The union would represent a set of a collection of genotypic patterns, where each element would not be repetitive. These genotypic patterns were used as covariates for the multinomial regression model in the next section.
Significant Predictors of Partial Remission after Anti-PD-L1+Chemotherapy
The multinomial regression model on scRNA-seq TNBC dataset was able to identify significant predictors from exploratory subgroup discovery results that increase odds of having partial remission after anti-PD-L1+chemotherapy treatment versus progressive disease after chemotherapy (Table 1). Next, we highlight the importance of identified phenotypic features from Table 1 for TNBC patient outcomes. Using literature, high-TMB TNBC status may benefit specifically from ICIs in combination with chemotherapy [23] or ICIs alone [24]. TNBC patients have high TMB due to accumulation of genomic instability, which leads to the production neoantigens, thereby resulting in strong effector cell responses [25]. TNBC tumors have a "hot tumor phenotype", which characterized by a high degree of immune infiltration and associated with improved survival outcomes regardless of tumor stage, molecular subtype, PD-L1 status, age and treatment schedule [26]. The IMpassion130 trial tested immunotherapy agent Durvalumab in combination with chemotherapy or chemotherapy alone on 149 early stage TNBC patients. Median TMB was significantly higher in patients with pathologic complete response (pCR) (median 1.87 versus 1.39, p = 0.005), and odds ratios for pCR per mut/MB were 2.06 (95% CI 1.33-3.20) among all patients, 1.77 (95% CI 1.00-3.13) in the Durvalumab arm, and 2.82 (95% CI 1.21-6.54) in the chemotherapy arm. Interestingly, the association between pCR and TMB was more pronounced in patients treated with chemotherapy alone. The KEYNOTE-119 trial evaluated metastatic TNBC patients treated with Pembrolizumab monotherapy versus chemotherapy. The positive association was observed between TMB and clinical response to Pembrolizumab (ORR p = 0.154, PSF p = 0.014, OS p = 0.018) but not to chemotherapy (ORR p = 0.114, PFS p = 0.478, OS p = 0.906). ORR and hazard ratio (HR) for OS also suggested a trend towards increased benefit with Pembrolizumab versus chemotherapy in TNBC patients with high TMB. This clinical trial was constrained by the small sample size and low number of TMB-high cases.
In terms of relapse status, one study suggested that rapid versus late relapse in TNBC might be characterized by unique clinical and genomic features [27]. Both 'rapid relapse' (rrTNBC) and 'late relapse' (lrTNBC) groups had significantly lower expression of immune-related genes. Intriguingly, lrTNBCs were enriched for luminal signatures. There was no difference in TMB or percent genome altered across investigated subgroups of TNBC patients.
In connection to menopausal status, TNBC was observed primarily in postmenopausal patients [28]. The overexpression of the p53 protein, a significantly higher Ki-67 proliferation index value, and a higher nuclear grade was detected in TNBC premenopausal patients. A multivariate analysis estimated that menopausal status, nodal status, and tumor size were significant contributors for disease-free survival (DFS) in TNBC cases.
We had also discovered novel phenotypic features in TNBC subgroups, such as tumor cellularity and tumor stage. The evaluation of tumor cellularity, defined as the percentage of invasive tumor comprised of tumor cells, may represent an informative histologic measure of the differential response of TNBC to chemoimmunotherapy. To classify the severity of a malignant disease in a particular patient, the tumor staging system is employed during the course of disease. This system is essential in optimizing cancer patients treatment options and their risk stratification. Therefore, these features can be important in the design and analysis of intervention studies, including randomized clinical trials, to better assess their prognostic utility for TNBC patients.
Differences in Immune Cell Populations for Discovered TNBC Subgroups
This section described immune cell populations that were discovered in scRNA-seq TNBC data based on genotypic patterns from exploratory mining stage. We interpreted our results using biomedical knowledge, including findings from existing clinical trials, immunohistochemistry experiments and functional characterization of specific genes. The summary of our findings is presented in Table 2. Table 2. Immune cell populations that are linked to the specific TNBC outcome determined by our informatics pipeline.
Condition T-Cells B-Cells Myeloid Cells NK-Cells
anti-PD-L1 post treatment partial remission
T-Cells Global Cluster
The proliferative MKI67 + T-cells (Tprf-MKI67) were exclusively present in TNBC patients achieving progressive disease after chemotherapy. Based on literature findings, the expression of MKI67 gene was significantly correlated with lymph node metastases, tumor invasion and adverse survival outcome in TNBC [29]. In addition, more unfovourable survival outcomes in breast cancer patients with recurrent lesions were significantly correlated with high Ki-67 immunohistochemical expression levels (hazard ratio 2.307; 95% confidence interval 1.207-4.407, p-value = 0.011) [30]. Therefore, MKI67 may be an important biomarker of predictive and prognostic value in TNBC.
CD4-Tn-LEF1 and CD4-CXCL13 T-cells were linked to partial remission after anti-PD-L1+chemotherapy treatment. Importantly, these CD4 + T-cells express very high amounts of PD-1 and other co-stimulatory and inhibitory receptors. Therefore, they instrumental to Bcells for efficient antibody responses and their presence in tumor samples is often correlated with a better outcome in patients with solid tumors [31]. Based on biomedical literature, the presence of CD4-CXCL13 T-cells in TNBC tumors responsive to chemoimmunotherapy was detected through immunohistochemistry staining [16,32]. In addition to CXCL13 + T-cells, naïve LEF1 + T-cells (Tn-LEF1) were also linked to a favorable response to both anti-PD-L1+chemotherapy and chemotherapy. In a recent study, the magnitude of lymphocytic infiltration was assessed by a four-gene signature-HLF, CXCL13, SULT1E1 and GBP1, which was indicative of favourable outcome in TNBC after neoadjuvant therapy. This signature may help to identify early stage TNBC patients and being a novel prognostic biomarker of this aggressive disease [33].
The activated IFI6 + T-cells (Tact-IFI6) were linked to progressive disease after chemotherapy. The poor metastasis-free survival in breast cancer patients was linked to upregulation of mitochondrial antiapoptotic protein IFI6 that might be involved in regulation of mitochondrial ROS production [34]. Therefore, to improve clinical outcomes in breast cancer patients, the deactivation of mitochondrial functions of IFI6 is paramount.
B-Cells Global Cluster
The MKI67 + follicular B-cells (Bfoc-MKI67), NEIL1 + follicular B-cells (Bfoc-NEIL1) and MKI67 + memory B-cells (Bmem-MKI67) were exclusively present in TNBC patients with partial remission after anti-PD-L1+chemotherapy treatment. Based on biomedical literature, follicular B-cells was associated with favorable outcomes for TCGA patients with breast cancer [35]. The naïve B-cells, memory B-cells and follicular B-cells were present primarily in patients responsive to chemoimmunotherapy but not in patients responsive to chemotherapy treatment [16]. In regard with Bfoc-NEIL1 cell population, NEIL1 implicated in repair of oxidative damage associated with DNA replication or transcription [36]. Reduction in NEIL1 expression was associated with a poorer outcome in patients with breast invasive carcinoma [37]. Hence, NEIL1 could be a promising biomarker for TNBC patients that consider chemoimmunotherapy treatment.
Plasma IGHG1 + B-cells (pB-IGHG1) were linked to progressive disease after chemotherapy treatment. In TNBC, the expression of IGHG1 indicated the most significant prognostic value compared to trivial clinicopathological parameters [38]. Intriguingly, IGHG1 expression in B-cells and plasma cells could be associated with immune evasion and tumor cell proliferation in breast malignancies [39]. These data may imply that B cells or plasma cells could have pro-tumoral roles under particular conditions; however, the factors influencing the emergence of this pathologic phenotype and the roles played by B cells and plasma cells in these contexts remains unclear.
Myeloid Cells Global Cluster
The MMP9 + macrophages (macro-MMP9) were exclusively present in TNBC patients with partial remission after anti-PD-L1+chemotherapy treatment. The literature search revealed that MMPs have a intricate role in cancer progression and may exert both proand antitumorigenic activities [40]. Although MMP expression has been linked to tumor progression in various cancer types including breast cancer [41], clinical trials investigating the effect of broad-spectrum MMP inhibitors have failed, and in some cases, patients treated with these inhibitors even progressed after treatment comparing to control placebo group [42]. Indeed, the overexpression of MMP9 results in increased production of antiangiogenic fragments, decreased angiogenesis, and therapeutic effects of established breast cancer [43]. In another study, gene transfer of MMP-9 to ex vivo breast cancer tumors caused tumor regression via increased neutrophil infiltration and an activation of tumor-associated macrophages (TAMs) [44]. Therefore, MMP9 can serve as a biomarker for predicting tumor regression in TNBC.
The macro-CCL2, macro-CX3CR1, macro-IFI27, macro-IGFBP7, macro-IL1B9, macro-MGP and macro-SLC40A1 cells were exclusively present in TNBC patients achieving progressive disease after chemotherapy. Based on biomedical findings, CCL2 expression in breast carcinomas was highly associated with macrophage infiltration, and its expression was correlated with poor prognosis in breast cancer patients [45]. In another study, chemokine receptor CX3CR1 showed a role in angiogenic macrophage survival in the tumor microenvironment contributing to tumor metastasis [46]. In a similar fashion, IFI27 overexpression was shown to impair the tamoxifen-induced apoptosis in breast cancer cells [47]. Finally, IL1B signalling contributed to breast cancer metastasis by enhancing tumor cell motility and inhibiting cell proliferation [48]. These findings highlight the importance of CCL2, CX3CR1, IFI27 and IL1B expressed in macrophages in progression of TNBC.
The CLEC9A + dendritic cells (cDC1-CLEC9A), macro-CFD, macro-FOLR2, macro-MKI67, macro-SPP1, macro-TUBA1B, FCN1 + monocytes (mono-FCN1), mono-S100A89 and mono-SMIM25 cells were linked to progressive disease after chemotherapy treatment. Notably, CFD functioned as an enhancer of tumor proliferation and cancer stem cell properties in breast cancers [49]. In another study, SPP1-associated macrophages in the tumor-adipose microenvironment facilitate breast cancer progression [50]. Interestingly, S100A8/A9, which are calcium-binding proteins that are secreted primarily by granulocytes and monocytes, may be associated with the loss of estrogen receptor and may be involved in the poor prognosis of Her2 + /basal-like subtypes of breast cancer. Therefore, myeloid cell populations expressing CFD, SPP1 and S100A89 might be crucial biomarkers of poor treatment response in TNBC.
NK-Cells Global Cluster
The CNOT2 + group 2 innate lymphoid cells (ILC2-CNOT2) were exclusively present in TNBC patients achieving partial remission after anti-PD-L1+chemotherapy treatment. Indeed, ILC2s involved in both anti-tumor and pro-tumoral immunity in a variety of human cancers [51]. In terms of pro-tumoral immunity, the promotion of tumor growth and metastasis is achieved by crosstalk between ILC2s and tumor microenviroment (TME) [52]. In addition, the ILC2s trigger the apoptosis of tumor cells by recruiting and activating eosinophils [53], CXCL1L/CXCL2L molecules and macrophages with M1 profile [54].
The ZNF683 + group 1 innate lymphoid cells (ILC1-ZNF683) cells were exclusively present in TNBC patients achieving progressive disease after chemotherapy. The biomedical literature demonstrates that ILC1 cells involved in inhibiting the antitumoral immune response, enabling the differential tumor infiltration of ILC1 cells in patients to improve the levaraging of immunity in cancer therapies [55]. However, the role of ZNF683 gene in particular remains elusive.
ILC3-AREG and ILC3-IL7R cells were linked to partial remission after anti-PD-L1 +chemotherapy treatment. It had been shown that ILC3-IL7R could predict a favorable response to both treatment regimens, indicating its potential role in effective antitumor immunity [16]. In contrary, ILC1-VCAM1 cells were linked to progressive disease after chemotherapy treatment. Recent studies have shown that vascular cell adhesion molecule-1 (VCAM1) is aberrantly expressed in breast cancer cells and mediates prometastatic tumorstromal interactions [56]. Therefore, AREG + , IL7R + and VCAM1 + innate lymphoid cells can help determine prognosis for breast cancer patients.
Discussion
The analysis of the TNBC scRNA-seq data revealed distinct immune cell populations that are linked to either partial remission after anti-PD-L1+chemotherapy or progressive disease after chemotherapy only. In terms of T-cells, CD4-Tn-LEF1 and CD4-CXCL13 T-cells were linked to partial remission after anti-PD-L1+chemotherapy treatment, while Tact-IFI6 T-cells were linked to progressive disease after chemotherapy. The naïve B-cells, memory Bcells and follicular B-cells were mainly enriched in tumors responsive to chemoimmunotherapy but not in tumors responsive to chemotherapy treatment. The MMP9 + macrophages (macro-MMP9) were exclusively present in TNBC patients with partial remission after anti-PD-L1+chemotherapy treatment, while heterogenous population of macrophages, including macro-CCL2, macro-CX3CR1, macro-IFI27, macro-IGFBP7, macro-IL1B9, macro-MGP and macro-SLC40A1 cells were exclusively present in TNBC patients achieving progressive disease after chemotherapy. Finally, group 3 innate lymphoid cells (ILC3-AREG and ILC3-IL7R) were linked to partial remission after anti-PD-L1+chemotherapy treatment, while ZNF683 + group 1 innate lymphoid cells (ILC1-ZNF683) cells were exclusively present in TNBC patients achieving progressive disease after chemotherapy. Each of these cell populations have distinctive genetic markers that could be useful therapeutic targets for chemoimmunotherapy.
The role of T follicular helper and B-cell crosstalk in tumor immunity has been extensively studied over the last decade. Accumulating evidence suggests that tumor infiltrated lymphocyte (TIL) subpopulations (CD4, CD8, and CD19/20) constitute of both suppressive (pro-tumor) or effector (anti-tumor) phenotypes whose functions are influenced by the surrounding TME [57]. Natural or treatment-induced immune activation or suppression may determine the balance between pro-or anti-tumor immune cell crosstalk within a given tumor. Key anti-tumor effector activities include antibody-dependent cell cytotoxicity, complement activation, antibody-mediated tumor cell phagocytosis, antigen presentation, T cell activation, cytokine secretion, and direct tumor killing by TIL, including CD8, NK, B cells, and/or macrophages [58].
Despite of significant survival advantages that could be achieved after treatment with chemoimmunotherapy, most TNBC patients would not benefit. Therefore, more and more attention has been paid to the identification and development of biomarkers for the response of chemoimmunotherapy in recent years. Our informatics pipeline identified novel phenotypic and genotypic predictors in unsupervised manner that indicative of favorable outcome after chemoimmunotherapy. These predictors could be important biomarkers in the design and analysis of intervention studies and ultimately could help to optimize therapy decisions for TNBC patients. In addition, it may help to select better responders to chemoimmunotherapy, as well as pinpoint the underlying mechanisms of drug resistance in TNBC patients at single-cell resolution.
Conclusions
To tackle patient heterogeneity, chemoimmunotherapy combinations represent a feasible alternative for TNBC patients. However, matching patient subgroups to effective treatments that increase their chance of survival remains a challenging endeavor. In this work, we augmented our exploratory subgroup discovery algorithm to identify TNBC subpopulations that may benefit from chemoimmunotherapy. Specifically, we identified immune-related gene signatures that increased the likelihood of partial remission after anti-PD-L1+chemotherapy regimen versus progressive disease after chemotherapy in TNBC patients. Our novel informatics pipeline identified immune cell populations that associated with various treatment outcomes in TNBC. We also showed the importance of TMB and menopausal status among the investigated TNBC subgroups. The potential limitations include the usage of two disjoint datasets and the absence of outcome variable for immunotherapy outcomes in TCGA datasets. Further validation of our computational results in wet-lab studies would be a significant step toward improving survival outcomes for TNBC patients.
|
v2
|
2022-11-27T16:09:51.185Z
|
2022-11-25T00:00:00.000Z
|
254002585
|
s2ag/train
|
The surgical treatment of locally advanced angiosarcoma of the anterior mediastinum. A clinical case
Tumors of the mediastinum is a collective term that includes neoplasms of various degrees of malignancy, different morphogenesis, originating from heterogeneous tissues and combined into one nosological form according to anatomical localization. Angiosarcoma is a rare malignant tumor originating from a vessel wall. In the general structure of sarcomas, its frequency is estimated at 1%. The tumor is characterized by an aggressive course, local recurrence, hematogenous metastasis, poor sensitivity to chemotherapy and radiotherapy. Surgery remains the main treatment for angiosarcomas. The article presents a clinical case. A 55-year-old patient has multiple primary malignant neoplasms: cancer of the body of the uterus pT1aN0M0 IA stage. Surgical treatment from 12.15.2020 panhysterectomy; angiosarcoma of the anterior mediastinum with lesions of the manubrium, body of the sternum and the 1st rib on the left, invasion into the left brachiocephalic vein with the formation of a tumor thrombus in it, growing into the upper lobe of the left lung, T3N0M0, G1. Surgical treatment from 08.12.2021 resection of the manubrium and body of the sternum with resection of the anterior segments I, II, III ribs on both sides, resection of the left brachiocephalic vein, removal of an anterior mediastinal tumor, atypical resection of the upper lobe of the left lung en bloc, reconstruction of the sternum and ribs with an individual titanium 3D prosthesis. The patient was discharged from the hospital in a satisfactory condition.
|
v2
|
2022-11-27T16:10:59.711Z
|
2022-11-25T00:00:00.000Z
|
254005016
|
s2ag/train
|
The analysis of the relationship between transferrin receptor 1 (TfR1) and clinical, morphological and immunophenotypic characteristics of breast cancer: retrospective cohort study
Background. Transferrin receptor 1 (TfR1) expression has been identified in a number of malignant tumors. It is noted that its overexpression gives growth advantages to cancer cells. Estimation of transferrin receptor expression in breast cancer (BC) might be an important component in disease prognosis, choice of treatment, also might be an attractive target for targeted therapy.
Aim. To evaluate the expression of TfR1 by BC cells and to study its relationship with the clinical, morphological and immunophenotypic characteristics of the tumor.
Materials and methods. This study included 82 patients with BC who received treatment at the Blokhin National Medical Research Center of Oncology (Moscow). The expression of TfR1 on primary tumor cells was studied, the relationship of TfR1 with clinical, morphological and immunophenotypic characteristics of BC was analyzed. Immunophenotyping of the primary tumor was performed by the immunohistochemical method (immunofluorescent staining) on cryostat sections. Antibodies to CD71, CD95, CD54, CD29, MUC1, Pgp170 were used. The reaction was evaluated using a luminescent microscope (AXIOSKOP, Germany). The study was dominated by patients with stage IIB 54% and IIIB 21%. Infiltrative ductal BC was diagnosed in 67% (n=55) of patients, infiltrative-lobular in 22% (n=18) of cases, other types in 11.0% (n=9).
Results. BC cells expressed TfR1 in most cases (64.4%; n=61). A combination of TfR1 monomorphic expression with MUC1 monomorphic expression (74.4%; n=47) was noted. CD29 is presented both mosaic (38.7%) and monomorphic (51.6%). The Pgp170 antigen was monomorphically observed in 27.5% of cases. As the proportion of TfR+ cells increased, the expression frequency of the adhesion molecule CD54 increased from 10.5 to 33.3%, a positive correlation was established (r=0.293; p=0.008). In the group with TfR1 monomorphic expression, the frequency of tumors expressing the CD95 apoptosis molecule decreased: 25.0% vs 13% (p=0.042).
Conclusion. BC cells overexpress TfR1. TfR1 expression is associated with tumor immunophenotype.
|
v2
|
2022-11-27T17:14:04.140Z
|
2022-11-25T00:00:00.000Z
|
253990816
|
s2orc/train
|
An Elevated Neutrophil-to-Lymphocyte Ratio Predicts Poor Prognosis in Patients with Liver Cancer after Interventional Treatments
This study is aimed at examining the prognostic value of blood neutrophil-to-lymphocyte ratio (NLR) in patients with hepatocellular carcinoma (HCC). Demographic and clinical data of 543 HCC patients treated with interventional therapies were retrospectively analyzed. Preoperative NLRs were determined and receiver operating characteristic (ROC) curves were plotted for survival time in patients with high (NLR ≥3.8) and low (NLR<3.8) NLR. The median overall survival (OS) was 1241 days after interventional therapies and was significantly reduced in the high NLR group when compared to the low NLR group. The median progression-free survival time (PFST) of patients was also significantly shorter in the high NLR group than in the low NLR group. Univariate analysis revealed that tumor type, therapy method, maximum tumor size (>3 mm), and NLR (>3.8) were risk factors for OST and PFST (P < 0.05). Multivariate analysis indicated that tumor type, maximum tumor diameter, therapy method, and NLR (>3.8) were independent risk factors for PFST (P < 0.05). Our results demonstrate that preoperative NLR has prognostic value for patients with HCC undergoing interventional therapies, and high NLR is an indication of poor prognosis.
Introduction
Liver cancer is one of the leading malignant tumors in the world and ranks the fourth in the causes of cancer-related death [1,2]. For patients diagnosed with hepatocellular carcinoma (HCC) of all stages, the overall 5-year survival rate is estimated to be about 18%, and the incidence is increasing year by year. For instance, the incidence is about 18.3 per 1,000,000 persons in China, and the mortality rate is about 17.1/100000 [3]. Orthotopic liver transplantation (OLT) is one of the best treatment options for liver cirrhosis and HCC. However, due to insidious onset of HCC, a majority of patients are already at late stage once diagnosed, and only less than 20% can be treated with OLT or surgically [4,5]. It is therefore important to develop prognostic biomarker to better manage patients for this disease.
Inflammation-related prognostic indicators have been related to the survival and other prognostic parameters such as tumor aggressiveness [6]. They include a number of easily measurable indicators of inflammation that can be obtained in routine clinical blood-based tests, such as counts and levels of neutrophils, lymphocytes, monocytes, platelets, albumin, Creactive protein (CRP), and monocyte-to-lymphocyte ratios, among others [7][8][9]. Neutrophils in human peripheral blood have the functions of phagocytosis, chemotaxis, and bactericide, and lymphocytes are involved in the immune response [10,11]. Studies have shown that the normal NLR values in an adult, nongeriatric, population in good health are between 0.78 and 3.53 [12], neutrophil-to-lymphocyte ratio (NLR) has a potential as prognostic marker, and the elevated NLR is associated with poor prognosis of diseases and cancers such as breast cancer [13], gastric cancer [14,15], advanced melanoma treated with nivolumab [16], pancreatic cancer [17], in extensive-stage small cell lung cancer [18], and others [19,20]. However, it is unknown whether if it has prognostic value for HCC patients after interventional therapies.
In the present study, we retrospectively analyzed the relationship between the ratio and survival of HCC patients after interventional therapies.
2.1.
Patients. This is a single-center retrospective study. The medical records of patients who underwent interventional therapies for HCC between January 1, 2015 and December 31, 2019 at Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, China, were retrieved and analyzed. Patients were included if they fulfilled the following inclusion criteria: (1) had complete baseline clinical data for HCC, including CT scan findings on HCC size, (2) histologically proven HCC, and (3) received hepatectomy and other treatments for HCC. Patients were excluded if (1) distant metastasis was found at the first visit, (2) treated for other cancers within 6 weeks, (3) had severe infections or any hematology-related diseases, and (4) administered with any immunosuppressive medications within 6 months. Data retrieved from the hospital electronical medical data system included age, gender, history of smoking, maximum tumor diameter (MTD), pathological type, therapeutic method, and distant metastasis. This work was reported in line with the STROCSS criteria [21] and was approved by the Research Ethic Committee of Shandong Provincial Hospital Affiliated to Shandong First Medical University, Jinan, China (approval number: SPH-CR-2316, July 2020). Written informed consent was obtained from every patient.
Data Collection.
Peripheral venous blood samples (10 ml) were drawn from all patients within 2 days prior to the interventional therapies and used to assess neutrophil and lymphocyte counts. NLR was calculated and used to plot receiver operating characteristic (ROC) curve for postoperative median survival time. Since NLR was 3.8 at the maximum Youden's index, this value was used as optimal cut-point to group the patients into high and low NLR groups. The Youden index is a main summary statistic that measures the potential effectiveness of a biomarker based on the ROC curve [22]. The cutpoint that achieves this maximum is referred to as the optimal cut-point because it is the cut-point that optimizes the biomarker's differentiating ability when equal weight is given to sensitivity and specificity [23,24].
Statistical Analysis.
The Student t-test was used to compare normally distributed data. A chi-square test was used for categorical variables. Receiver operating characteristic (ROC) curves were constructed, and the areas under the curves (AUCs) were calculated to evaluate the predictive abilities of the NLR for discriminating patients with good and poor prognosis. Overall, progression-free survival rate of patients was estimated through Kaplan-Meier survival analysis. The log-rank test was used to compare survival rates for groups for high and low NLR. Prognostic factors were assessed using univariate and multivariate analyses (Cox proportional hazard regression model). Data were analyzed using IBM SPSS Statistics (v. 20.0, IBM, New York, USA). P < 0:05 was considered statistically significant.
Baseline Characteristics.
A total of 599 patients were found satisficing the inclusion criteria. Among them, 56 patients were later excluded due to various reasons, including the loss of follow-up, and the clinical data for the remaining 543 patients were collected and analyzed. The study population consisted of 287 males and 256 females with a median age of 55.8 years (range 32-82 years).
Among them, 151 were smokers, and most cancers had not distant metastasis. The MTD ranged from 3 to 7 mm. ROC curve for postoperative OS time revealed that NLR was 3.8 at the Youden index, and AUC was 0.891. For the prediction at this point, the sensitivity was 84.4% and specificity was 86.5% ( Figure 1). This value was used to group the patients into high (≥3.8, n = 256) and low (<3.8, n = 287) NLR groups. Analysis showed that the two groups were not statistically different in age, gender, history of smoking, MTD, tumor site and type, and therapeutic methods but their NLR values were different (3.8-5.4 in high vs 2.1-3.8 in low NLR groups (Table 1).
High NLR Reduces Survival Time.
After the interventional therapy, patients were followed-up for up to five years. By the end of this study, 309 patients died and 234 were alive. Taken all patients together, the median OS time was 1241 days, and 1-and 2-year OS rates were 64.10% and 32.80%, respectively. For patients with high NLR (≥3.8), the median OS time was 381 days, and 1-and 2-year OS rates were 33.10% and 12.30%, respectively, and for patients with low NLR (<3.8), the median OS time was 1465 days and 1-and 2-year OS rates were 85.40% and 43.60%, respectively. The difference in the survival time and rates were statistically significant between the high and low NLR groups (P < 0:01, Figure 2). The PFS time was 529 days in all patients. However, the PFS time was significantly shorter in high than in low NLR patients, (242 vs 761 days, P < 0:05, Figure 3).
Factors Affecting Prognosis of OS and PFS.
To analyze factors that affect OS and PFS after the therapy, we first performed univariate analysis, and the results indicated that the tumor type, therapy method, MTD, and NLR were significantly related to postoperative OS time and PFS time (P < 0:05, Tables 2 and 3). On other hand, other demographic and clinical features such as age, gender, smoking status, and tumor site were not significantly associated with the OS time and PFS time. These significantly related variables were then included in multivariate regression models for further analysis. The results revealed that therapy method, MTD, and NLR were the independent risk factors affecting OS time and PFS time (P < 0:05, Table 4).
Discussion
Most HCC patients are in the middle and late stages when diagnosed and may have missed the optimal surgery time [25]. Several treatment options are available for HCC patients, and among them, OLT and surgical resection are the mainstay treatments, although personalized therapies such as transarterial chemoembolization (TACE), transarterial radioembolization (TARE), and stereotactic body radiation (SBRT) as well as immunotherapy for HCC are being developed to improve overall survival [4,26]. However, the overall survival of patients is still not satisfactory after the therapeutic processes due to various reasons. Therefore, discovery of indicators that can predict the prognosis for HCC patients is highly demanded. Studies have shown that the occurrence and progression of HCC is related to inflammation over a long period [27,28]. NLR is an indicator of inflammation and is shown to be associated with prognosis of a variety of tumors [29]. Increased number of neutrophils in tumor and deceased lymphocyte count often indicate poor prognosis in the cancer patients [30,31] or patients after selective internal radiation therapy [32].
Since both lymphocytes and neutrophils mainly play roles in protecting the body from infections and are a part of the immune system, they are associated with prognosis of diseases, including cancers. For example, lymphocyte was shown to be able to predict the severity and prognosis in patients with HBV-related acute-on-chronic liver failure [33] and lung cancer [34]. Neutrophils increase as a result 3 BioMed Research International of detrimental outcome in several tumors [35]. However, in our patient data, no relationship between either absolute neutrophil or leucocyte count alone was found to be significantly associated with the prognosis.
NLR has been reported as a potential prognostic marker, and the elevated ratio is associated with poor prognosis of different cancers such as breast cancer [13], gastric cancer [14,15], advanced melanoma treated with nivolumab [16], pancreatic cancer [17], and in extensive-stage small cell lung cancer [18]. However, due to the heterogenicity of patient populations, the relationship needs to be analyzed for different patient populations to develop a cut-off that is best fit for specific patient groups (or at least large patient groups), because neutrophil and lymphocyte may vary in response to many factors, including treatment protocols, drugs, and methods. For HCC, NLR was found to accurately predict the probability of long survival after sorafenib treatment [36], and increased NLR was associated with poor survival after selective internal radiation therapy [32]. However, NLR variation after surgical section in HCC has not been well addressed.
In this study, we focused on HCC patients mainly after surgical sections and other interventional therapies such as liver transplant and radiation therapy and found that the NLR are related to OS and PFS times, and high NLR predicts shorter OS time and PFS time and has significant prognostic value. This is consistent with the previous results in lung cancer [37][38][39]. In addition, NLR is found to be related to the recurrence, metastasis, and prognosis of a number of solid tumors, such as esophageal cancer [40], prostatic cancer [41], and cervical cancer [42], and liver cancer [43] is useful in analyzing allergic conditions, inflammatory disorders, and infectious diseases [44,45]. Different from other clinical indicators such as tumor size and grading, which require use of relatively complex and invasive surgical procedure, neutrophil and lymphocyte counts are readily available in routine blood tests. Therefore, NLR is a convenient biomarker for predicting the prognosis of HCC patient and can be used to stratify patients before different surgical and interventional treatment options. For instance, patients with high NLR could be allocated to receive relatively less invasive surgery to reduce their postoperative risk. On other hand, patients with low NLR may be tolerant to liver transplant and section. In addition, NLR could be monitored over the therapeutic periods as an auxiliary index for the progress and outcome of HCC after treatment. However, since the lymphocyte and neutrophil counts are affected by many factors, especially infections [46,47] and drugs [48], and in a recent study, COVID-19 infection was also found to result in severe lymphopenia [49], therefore, cautions should be taken to interpret the changes of NLR in HCC patients, and additional data, particularly inflammation-related data, are needed to trace the therapeutic outcomes and to rule out other factors and diseases that may affect changes. For example, the pathogenesis of several diseases such as cardiovascular diseases [50], retinal artery occlusion [51], and spinal epidural abscess [52] have been found to result in high NLR, while treatment with 25-hydroxyvitamin D 3 and smoking cessation are associated with a reduced blood NLR [53,54], suggesting when NLR is used for individual patients, it should be evaluated along with other pathological conditions to obtain more reliable prediction. Mechanisms by which high NLR are associated with poor HCC prognosis may result from the interaction between tumor and inflammatory microenvironment [55,56]. Immune cells such as activated macrophages, stellate, and mast cells have the ability to infiltrate into tumors, leading to increased tumor growth [57]. The peritumor infiltration by neutrophils may trigger inflammatory response to release free radicals and angiogenic response to enhance tumor growth [58,59]. In addition, therapy method and MTD were also found to be related to the survival of HCC patients after treatments. This is consistent with early studies [38,60,61].
There are limitations in this study. This study is a singlecenter retrospective analysis; the sample size is relatively small. However, it may serve as starting point for multicenter and large prospective study in the future to further validate our conclusions for HCC patients.
Taken together, our study demonstrated that blood NLR may be used as prognostic marker to predict the prognosis 5 BioMed Research International of HCC for patients with middle and later stage HCC after interventional therapy. The preoperative NLR values may be used to stratify patients for different surgical and interventional treatment options before treatments and to monitor postoperatively the progress and outcomes of treatments.
Data Availability
The datasets used during the current study are available from the corresponding author on reasonable request.
Ethical Approval
This study was approved by the ethical committee of Shandong First Medical University. The work was carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans.
Consent
Written informed consent was obtained from all patients.
Disclosure
The sponsor did not have role in study design, experiment, manuscript writing, and publication.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
|
v2
|
2022-12-02T05:22:15.823Z
|
2022-11-25T00:00:00.000Z
|
254121913
|
s2orc/train
|
Durable response to low-dose pralsetinib in a renal insufficient patient with NSCLC harboring concurrent CCDC6-RET, LINCO1264-RET, and SEMA5A-RET fusions: A case report
Introduction: RET-rearranged fusions have been considered as oncogenic drivers in 1% to 2% of non-small cell lung cancers (NSCLC). ARROW study has demonstrated a new selective RET tyrosine kinase inhibitors (TKIs) shows remarkable and durable responses in RET-rearranged NCSLC. In this study mainly recruited patients with common fusion partners KIF5B and CCDC6. There is still a lack of definitive conclusions about effective of rare RET fusion variants to anti-RET therapies. Case report: A Chinese 58-year-old female renal insufficient patient with no history of smoking was diagnosed as stage IIIA (T2N2M0) lung adenocarcinoma. Next-generation sequencing targeting 520 cancer-related genes was performed on the pleural effusion samples and revealed 2 novel RET fusions LINCO1264-RET and SEMA5A-RET, concomitant with a common CCDC6-RET. Management and outcome: The patient was first treated with multiple lines of chemotherapy and switched to lenvatinib but failed to respond. Due to renal insufficiency, she subsequently received pralsetinib with gradually reduced dosages (400 mg-300 mg-200 mg-100 mg qd) and achieved a partial response (PR) lasting for more than 10 months, accompanied by the declined allele frequencies of all 3 RET fusions. Discussion/conclusions: We reported the first case of the pralsetinib efficacy in NSCLC with 3 concurrent RET fusions. Our case also indicates the sensitivity of the newly identified RET fusions to this RET selective inhibitor pralsetinib, and highlights the low-dose treatment option for patients with renal insufficient background.
Introduction
Mapping to chromosome 10q11.2, RET gene encodes a receptor tyrosine kinase comprised of an extracellular domain, a transmembrane region and an intracellular kinase domain. Genomic rearrangements of RET gene can form chimeric tyrosine kinase fusion proteins that often confer the constitutive oncogenic RET activation if the intact kinase domain is retained. [1] RET fusions occur in 1% to 2% of non-small cell lung cancers (NSCLC) and have been established as oncogenic drivers in this disease. [2] To date, a number of clinical studies have investigated a variety of multikinase inhibitors with anti-RET activity, such as vandetanib, cabozantinib and alectinib, in patients with RET-rearranged lung cancer. However, the objective response rate (ORR) (16%-47%) and median progressive-free survival (mPFS) (2.3-7.3 months) are inferior to that seen in other oncogene-addicted NSCLC with targeted therapies, such as EGFR, ALK and ROS1. [3] Moreover, it has been reported that the responsiveness of different RET fusion partners varies to the multi-target inhibitors. [4] More recently, a new generation of highly selective RET tyrosine kinase inhibitors (TKIs), namely selpercatinib and pralsetinib, has been developed and demonstrated remarkable and durable responses in RET-rearranged NCSLCs, driving the recent FDA approval of the drugs. [5] However, in these clinical trials, recruited patients mainly harbored the common RET fusion partners KIF5B and CCDC6. It remains elusive how other rare partners respond to these selective RET TKIs. Thus, there is still a lack of definitive conclusions on the potentially different responsiveness of diverse RET fusion variants to anti-RET therapies. Herein, we described a recurrent NSCLC patient who harbored 2 novel RET fusions LINCO1264-RET and SEMA5A-RET concomitant with a common CCDC6-RET. The refractory heavily-pretreated patient had chronic renal insufficiency and received pralsetinib with gradually reduced dosages. She achieved a partial response (PR) lasting for more than 9 months. We present the following case in accordance with the CARE reporting checklist.
Case presentation
A Chinese 58-year-old female never-smoker was referred to our clinic due to the tumor in the right lower lung field identified by a CT scan. Any other sign of metastasis did not discover, Eastern Cooperative Oncology Group (ECOG) performance status 1. The patient had no family history of cancer but had renal insufficiency due to the hydronephrosis (creatinine: 130 umol/L; creatinine clearance: 30 mL/min). In Jan 2017, she had a thoracoscopic lobectomy of the right lower lobe with systemic lymph node dissection. Postoperative histopathological and genetic tests indicated a stage IIIA (T2N2M0) lung adenocarcinoma with negative EGFR, ALK, and ROS1. Due to the history of renal insufficiency, the patient subsequently received adjuvant treatment with liposome paclitaxel for 6 cycles and local-regional radiotherapy in July 2017 (Fig. 1). After a disease-free survival of 19 months, the patient presented with right pleura thickening accompanied by increased FDG metabolism. She was treated with docetaxel plus carboplatin for 4 cycles and achieved a stable disease lasting for 8 months. In August 2018, the patient had shortness of breath, and was found with massive right pleural effusion and multiple nodules in the right pleura. She received monotherapy of docetaxel rechallenge for 2 cycles but the disease progressed. Then she was switched to a progressive disease-1 inhibitor toripalimab for 2 cycles but developed progressive disease again with the increased creatinine (up to 182 umol/L) (Fig. 1). Toripalimab was held on admission. The renal function of the patient improved significantly and achieved a baseline of serum creatinine (130 umol/L).
In December 2020, the patient signed informed consent form, pralsetinib (GAVRETO, Blueprint Medicines Corporation) was administrated with an initial dosage of 400 mg qd. Due to creatinine increased to 220 umol/L and a deficient creatinine clearance rate (20 mL/min), the pralsetinib dosage was gradually reduced as recommended (400 mg-300 mg-200 mg qd) and pralsetinib 200 mg was administrated intermittently from March 2021 to July 2021, since July 2021 The patient continued to take 100 mg pralsetinib orally the creatinine remained at 180 umol/L.
At the time of the first follow-up in Jan 2021 the patient had achieved a PR, CT scan showed disappearance of micronodules, reduction of right pleural effusion and re-expansion of the right lower lobe (Fig. 3B). The pralsetinib reduced to 200 mg the patient remained as PR as follow-up in April 2021 (Fig. 3C). Second next-generation sequencing was performed with her pleural effusion sample and revealed the retaining of the 3 RET fusions but with declined abundance: CCDC6-RET (AF: 2.30%), LINCO1264-RET (AF: 2.53%) and SEMA5A-RET (AF: 0.60%) (Fig. 1). Although the patient took pralsetinib 100 mg orally since July 2021, Chest CT scan was showed the disease was stable in Sep. 2021 (Fig. 3D). No distant metastasis of other organs was found, progressive disease free was 10 months.
The patient's creatinine was 182 umol/L on baseline, then increased to 220 umol/L after pralsetinib oral administration, required dose reductions from 400 to 300 to 200 to 100 mg. Adverse events decreased after dose reduction to 100 mg of pralsetinib, and were primarily grade 2 according to the CTCAE5.0. In addition, the patient had grade I proteinuria, hypertension and rash. Hypertension and rash are relieved after pralsetinib reduction, and proteinuria persists, which may be related to basic disease of renal insufficiency. In addition, patient didn't have increased count of eosinophils. Figure 1. Timeline of the patient's treatment history. BOR = best overall response, DFS = disease-free survival, NGS = next-generation sequencing, PD = progressive disease, PFS = progressive-free survival, PR = partial response.
Discussion
In NSCLC, more than ten RET fusion partners have been described, [3,6] among which KIF5B-RET occurs the most frequently with the prevalence ranging from 40% to 72%, [1,7] followed by CCDC6 (10-25%). Other identified partner genes include TRIM33, ZNF477P, ERCC1, HTR4, CLIP1 etc. [8] In the ARROW study, 13 cases had other or unknown types of RET fusion, accounting for 10.7%, but no specific fusion partner and its efficacy were reported. [9] In this case, we identified 2 novel RET fusions LINCO1264-RET and SEMA5A-RET co-occurring with the common CCDC6-RET. LINCO1264-RET (intergenic: R12) results in the exon 12 of RET gene 3'-juxtaposed with an intergenic region, while SEMA5A-RET (S5:R12) is predicted to produce an in-frame fusion of SEMA4 exon 5 with RET exon 12. Both novel fusions retain the intact RET kinase domain thus might be potential oncogenic drivers.
After the detection of RET fusions, the patient first received the treatment with a multikinase inhibitor lenvatinib, but the efficacy was limited, it similar with outcome of clinic trials, In phase 2 trials, cabozantinib and lenvatinib showed low response rates (ORR: 16%-28%, mPFS: 7.3 months). [9] A retrospective multicenter registry analysis showed response rates ranging from 18% to 37%. [10] Owing to the low activity and toxicity concerns with multikinase inhibitors, it is not recommended for the treatment of NSCLC with RET fusion.
Luckily with the recent approval of the new generation of selective RET TKIs, the patient switched to pralsetinib and showed an immediate response. Moreover, allele frequencies of all 3 fusions declined in the pleural effusion as the PR achieved, suggesting the responsiveness of all 3 of them to pralsetinib including the 2 novel ones. In our database, only 1 patient with 3 RET fusions was found, and the abundance decreased significantly after treatment. In the phase ARROW trial, pralsetinib demonstrated a promising ORR of 60% and DCR of 93% in RET-rearranged NSCLCs. [10] Most patients had a duration of response ≥ of 6 months. [9] So far, we have not found report about LINCO1264-RET and SEMA5A-RET co-occurring with the common RET was sensitive to pralsetinib or LOXO 292 (selpercatinib) in the NSCLC. Our case also showed that pleural effusion supernatant is an alternative liquid biopsy specimen for detecting rare RET fusion genes. [11,12] Of note, this case had a history of chronic renal insufficiency. The pralsetinib was administrated with gradually reduced dosages (down to 200 mg qd) and eventually discontinued due to the persistent increase in creatinine. Reducing the pralsetinib dosage to 100 mg qd administrated creatinine was maintained at grade 2. Pralsetinib is primarily metabolized by liver enzyme CYP3A4 and to a lesser extent by liver enzyme CYP2D6 and CYP1A2, in vitro. It was predominately excreted in feces (approximately 73%). In addition, mild and moderate renal impairment (CLcr 30-89 mL/min) are safe on the exposure of pralsetinib. It has revealed a tolerable toxicity with most treatment-related adverse events (TRAEs) being grade 1 to 2, consisting of increased aspartate aminotransferase (31%), anemia (22%), increased alanine aminotransferase (21%), constipation (21%) and hypertension (20%), elevated blood creatinane (13%), no at grate 3 to 4. In the ARROW trial, 15% and 60% of NSCLC patients required permanent discontinuation of pralsetinib and dosage interruptions, respectively, due to adverse reactions. Dosage reduction was required in 36% of patients. [9] In the last follow-up of our case, reducing the pralsetinib dosage to 100 mg qd the creatinine no further deterioration, and still have clinical benefits.
For the first time, we described a renal insufficient patient with NSCLC harboring 3 concomitant RET fusions, including 2 novel partners LINCO1264 and SEMA5A. The patient failed to respond to lenvatinib but achieved a PR to the reduced dosage of pralsetinib accompanied with declined frequencies of all 3 RET fusions. Our case also indicates the responsiveness of the newly identified RET fusions to pralsetinib and highlights the necessity of dosage interruption especially in treating patients with background diseases.
|
v2
|
2022-12-02T06:17:20.023Z
|
2022-11-25T00:00:00.000Z
|
254122501
|
s2ag/train
|
Applications of Indocyanine Green-Guided Near-Infrared Fluorescence Imaging in Pediatric Minimally Invasive Surgery Urology: A Narrative Review.
Background: Indocyanine green (ICG) is a fluorescent dye used for several indications in adult surgery, and, more recently, adopted also in the pediatric patients. This study aimed to review the literature published on the use of ICG near-infrared fluorescence (NIRF) in pediatric urology, to address its shortcomings and disadvantages and to detect the future perspectives. Materials and Methods: An electronic literature search of PubMed on all studies reporting use of ICG-NIRF in pediatrics was performed. We included only studies reporting ICG-NIRF application in minimally invasive surgery (MIS) for pediatric urology indications. Results: Forty-two articles reporting MIS procedures performed using ICG-NIRF in children were obtained, but only 15 studies that focused on urological applications of ICG-NIRF in children were included in this review. The included studies described use of ICG-NIRF for kidney malformations such as duplex system, kidney tumors, renal cysts, ureteral pathology, bladder malformations, varicocele, and lymph node sampling in tumors. The pediatric urological applications in which ICG-NIRF provided significant advantages included partial nephrectomy, lymphatics sparing varicocele repair, and oncological procedures. The ICG-NIRF use was clinically safe, without reported adverse systemic reactions in all pediatric series. The main drawback of this technology is the need of specific laparoscopic equipment such as camera system, light sources, and telescopes or the da Vinci Xi Robot, with the software for ICG-NIRF, Firefly®, already integrated within. Conclusions: ICG-enhanced fluorescence-guided surgery is gaining growing popularity among pediatric surgeons due to the excellent results that have been published until now. ICG-NIRF technology has proven to be safe, easy to use, not time-consuming, cheap, and very effective to improve intraoperative view and surgical ability. Nonetheless, further evidence, including larger series, longer follow-up, and more specific assessments, is necessary to confirm the preliminary results and enlarge the applications.
|
v2
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.