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 |
|---|---|---|---|---|---|
2022-11-10T17:26:46.451Z
|
2022-11-02T00:00:00.000Z
|
253429215
|
s2ag/train
|
Cardiovascular complications of treatment for prostate cancer.
Prostate cancer, an androgen-dependent disease, is one of the leading causes of mortality in men. It can present as localised disease, locally advanced or distant metastatic disease. Treatment options for patients with prostate cancer include surgery, chemotherapy, brachytherapy, radiation therapy and hormonal therapy. There are multiple treatment options for each stage of the disease, but hormone therapy is usually reserved for advanced stages. Cardiovascular disease is the leading cause of death in patients with prostate cancer and both diseases share common risk factors. Hormone therapy improves prognosis in patients with more advanced disease, albeit at the cost of cardiovascular toxicity. Hormone therapy can be achieved with the use of agonists and antagonists of gonadotropin-releasing hormone receptors, androgen receptor blockers and enzyme inhibitors of androgen synthesis. Drug-specific cardiotoxicity caused by treatments for prostate cancer has not been fully elucidated. Cardiovascular disease in patients with prostate cancer is mainly managed via an ABCDE approach, a strategy to optimise common risk factors. With newer agents improving the prognosis for patients with prostate cancer, cardiovascular toxicity will have a greater impact on the outcomes of these patients. This article reviews cardiovascular risks associated with therapy for prostate cancer with a focus on hormonal therapy.
|
v2
|
2022-11-24T06:17:17.961Z
|
2022-11-02T00:00:00.000Z
|
253800292
|
s2ag/train
|
The triple negative breast cancer drugs graveyard: a review of failed clinical trials 2017-2022
ABSTRACT Introduction Triple-negative breast cancer (TNBC) accounts for 15–20% of breast cancers (BC) and has the worst prognosis. It is characterized by the absence of both hormone receptor (HR) and human epidermal growth factor receptor 2 (HER2). TNBC has more limited therapeutic options compared to other subtypes, meaning that there is still a long way to go to discover target treatments. Areas covered Our review aims to summarize phase II/III clinical trials enrolling patients with TNBC that have been published between 2017 and 2022 but failed to reach their primary endpoint. We here try to emphasize the limitations and weaknesses noted in negative studies and to point out unexpected results which might be useful to enhance the therapeutic approach to TNBC disease. Expert opinion A deeper understanding of the mechanisms behind TNBC heterogeneity allowed to enhance the knowledge of new prognostic and predictive biomarkers of response. However, it is also through several failed clinical trials that we were able to define new therapeutic approaches which improved TNBC patients’ clinical outcomes. Nowadays, we still need to overcome several difficulties to fully recognize different intracellular and extracellular pathways that crosstalk in TNBC and the mechanisms of resistance to identify novel tailored-patients’ therapies.
|
v2
|
2022-11-24T06:17:19.903Z
|
2022-11-02T00:00:00.000Z
|
253800506
|
s2orc/train
|
The impact of hospital volume on survival in patients with locally advanced colonic cancer
Abstract Background High hospital volume has been shown associated with improved survival in patients with several cancers. The aim of this nationwide cohort study was to investigate whether hospital volume affects survival in patients with locally advanced colonic cancer. Methods All patients with non-metastatic locally advanced colonic cancer diagnosed between 2007 and 2017 in Sweden were included. Tertiles of annual hospital volume of locally advanced colonic cancer were analysed and 5-year overall and colonic cancer-specific survival were calculated with the Kaplan–Meier method. HRs comparing all-cause and colonic cancer-specific mortality rates were estimated using Cox models adjusted for potential confounders (age, sex, year of diagnosis, co-morbidity, elective/emergency resection, and university hospital) and mediators (preoperative multidisciplinary team assessment, neoadjuvant chemotherapy, radical resection, and surgical experience). Results A total of 5241 patients were included with a mean follow-up of 2.7–2.8 years for low- and high-volume hospitals. The number of patients older than 79 years were 569 (32.3 per cent), 495 (29.9 per cent), and 482 (26.4 per cent) for low-, medium- and high-volume hospitals respectively. The 3-year overall survival was 68 per cent, 60 per cent and 58 per cent for high-, medium- and low-volume hospitals, respectively (P < 0.001 from log rank test). High volume hospitals were associated with reduced all-cause and colon cancer-specific mortality after adjustments for potential confounders (HR 0.76, 95 per cent CI 0.62 to 0.93 and HR 0.73, 95 per cent CI 0.59 to 0.91, respectively). The effect remained after inclusion of potential mediators. Conclusions High hospital volume is associated with reduced mortality in patients with locally advanced colonic cancer.
Introduction
Substantial attention has been paid to hospital volume as a quality marker of cancer surgery during the past decades. It may be a marker of surgical experience, and a proxy for important structural factors related to multidisciplinary management.
Several studies have presented an association between high-volume hospitals and improved short-and long-term survival in patients with pancreatic cancer and oesophageal cancer 1,2 . Surgery of these tumours is complex and postoperative intensive care is often necessary.
The importance of hospital volume in more common tumours, such as colonic cancer demanding less complex surgery, is controversial. Surgical experience and access to advanced perioperative care may not be as important in these patients.
Although, there is evidence of improved overall survival in patients with breast cancer resection in high-volume hospitals, studies on colonic cancer and survival have not found the same association [3][4][5][6] ; however, subgroups of patients with colonic cancer, requiring more complicated treatments, may benefit from centralization to high-volume hospitals. These subgroups include patients with metastatic disease or locally advanced tumours.
Most locally advanced tumours can be treated curatively, but they require access to neoadjuvant treatments, advanced surgery, and sometimes postoperative intensive care.
The impact of hospital volume on survival in patients with locally advanced colonic cancer has been seldom explored. The primary aim of this study was to evaluate whether hospital volume affects survival of patients with locally advanced colonic cancer. The secondary aim was to evaluate the association between overall annual hospital volume of colonic cancer surgery, regardless of tumour stage, and survival in patients with locally advanced colonic cancer.
National registers
The Swedish Colorectal Cancer Registry (SCRCR) collects data on all patients diagnosed with colorectal cancer and includes data on patient characteristics, and surgical and postoperative data, as well as recurrence and survival. Patients with colonic cancer have been registered since 2007 and the coverage has improved from 94 per cent in 2007 to 97 per cent in 2017 7 . The National Patient Register (NPR) includes data on inpatient care from 1964 and on outpatient doctor visits, except for primary care, from 2001 8 . The Swedish Cancer Register (SCR) was founded in 1958 and it is compulsory for all healthcare providers to report all newly detected cancer cases to the register 9 . All deaths in Sweden are registered in the Swedish Cause of Death Register (CDR) 10 .
Patients
Data on all patients resected for colonic cancer between 2007 and 2017 in Sweden were retrieved from the SCRCR (35 799). Patients with appendiceal cancer (560), metastatic disease (6775), or pT1-3 tumours (22 333) were excluded (Fig. 1). Tumours registered as pT4 in the SCRCR were defined as locally advanced colonic cancer. The registers did not differ between tumours penetrating the serosa (pT4a) and tumour overgrowth to surrounding tissues (pT4b). Patients with a previous history of primary colonic cancer before the study start were excluded.
Exposures and outcomes
The primary exposure was annual hospital volume of surgery for locally advanced colonic cancer. Data on hospitals performing this surgery were retrieved from the NPR. The hospitals were categorized into volume tertiles, yielding three mutually exclusive categories: less than 10, 11-19, and more than 19 resections per year.
The primary outcome under investigation was 5-year all-cause mortality, and as secondary outcome, colonic cancer-specific mortality was investigated.
Potential confounders and mediators
Age at the time of diagnosis was categorized in three groups (less than 65, 65-79, and more than 79 years old). The level of co-morbidity at diagnosis was measured using the Charlson co-morbidity index (CCI) together with ASA grade. Data from the NPR and SCR were used to calculate CCI, which was further divided into three groups: 0, 1, and more than 2 points, whereas ASA grade was divided into 1-2 and 3-5 [11][12][13] . No points were given for colonic cancer. Preoperative staging of the tumour (cT) and local lymph nodes (cN) were presented according to the TNM classification. The tumour location was defined as right-sided if the tumour was in the right or transverse colon and left-sided if located from the splenic flexure to the sigmoid. Emergency resection was defined as surgery performed due to an emergent medical condition. A resection was considered radical if it was registered as microscopically radical (R0) in the SCRCR. Surgical experience was categorized into colorectal surgeon, general surgeon, and resident. Preoperative multidisciplinary team (MDT) assessment was registered as yes or no in the SCRCR. Potential confounders and mediators were illustrated in a directed acyclic graph (Fig. S3).
Statistics
Time since diagnosis was used as the underlying timescale throughout, but patients started being at risk at the time of the tumour resection (delayed entry). Follow-up ended on the date of death, date of migration, or at the end of follow-up (31 December , whichever came first. The Kaplan-Meier method was used to estimate survival proportions, and differences in survival between patients in different hospital volume groups were tested using the log rank test. Study population averaged survival proportions (standardized survival) were predicted from a multivariable flexible parametric survival model adjusted for sex, age at diagnosis, year of diagnosis, ASA score, CCI, emergency resection, and university hospital 14 .
Cox regression models were fitted to estimate HRs with 95 per cent confidence intervals (c.i.) contrasting all-cause and colonic cancer-specific mortality by hospital volume tertiles. Both univariable and multivariable models were fitted, where the latter were adjusted for potential confounders (sex, age at diagnosis, year of diagnosis, ASA score, CCI, emergency resection, and university hospital). Both main effects models and interaction models, including type of hospital (university hospital/non-university hospital) as an effect modifier, were fitted.
Additionally, potential mediators (preoperative MDT assessment, microscopically radical resection, neoadjuvant chemotherapy, and surgical competence) were included in the main effects models to assess the direct effect of hospital volume. All models used the clustered sandwich estimator of the standard errors, with hospital as cluster, to allow for the correlation between observations within cluster. The assumption of proportional hazards was formally tested using Schoenfeld's residuals.
A number of sensitivity analyses were performed. First, to exclude postoperative mortality, conditional survival analyses restricted to patients who survived for more than 90 days after the tumour resection, with resection date plus 90 days as the start of follow-up, were conducted. Second, to further explore the combined effect of hospital volume and preoperative MDT assessment, a model including an interaction term between these two variables was fitted. In this analysis, hospital volume was dichotomized into high and low annual volume (1-14 and more than 14 resections per year). Third, an analysis with hospital volume categorized using annual hospital volume of all colonic cancer resections (and divided into quartiles) was performed.
Additionally, to further account for the correlation between patients treated within the same hospital, we fitted frailty models (instead of fixed-effects models) where patients treated at the same hospital shared frailty. This did not alter the results and hence will not be explored further.
Missing information on any of the variables in the model was handled using the missing indicator approach. All analyses were
Patient and surgical characteristics
A total of 5241 patients underwent resection for locally advanced colonic cancer between 2007 and 2017 after exclusions (Fig. 1). Among these, 1760 patients (33.6 per cent) were treated in low-volume hospitals, 1657 patients (31.6 per cent) in medium-volume hospitals, and 1824 patients (34.8 per cent) in high-volume hospitals. Mean follow-up time was 2.7 years in low-volume hospitals and 2.8 years in medium-and high-volume hospitals. Patient and tumour characteristics are presented in Table 1, stratified by annual hospital volume of locally advanced colonic cancers.
There were no significant differences in ASA score or CCI. Preoperative MDT was performed in 808 patients (45.9 per cent) Resected organs are presented in Table S1. Adherent bowel resection was the most common multivisceral resection in all three groups (n = 71 (4.0 per cent), n = 225 (13.6 per cent), and n = 237 (13.0 per cent) in low-, medium-and high-volume hospitals respectively). Resection of three or more organs was performed in 50 patients (2.8 per cent) in low-volume hospitals, compared with 74 (4.5 per cent) and 102 (5.6 per cent) patients in medium-and high-volume hospitals respectively. The number of hospitals in each volume group per year is illustrated in Fig. S1.
Survival analyses
The 90-day mortality rate was 4.6 per cent in high-volume hospitals, 6.5 per cent in medium-, and 6.3 per cent in low-volume hospitals ( Table S2). The proportion of recurrent disease was 19.0 per cent, 24.8 per cent, and 26.9 per cent respectively.
There were significant differences in both overall and colonic cancer-specific survival by hospital volume (Figs. 2 and 3). Threeyear overall survival was higher in high volume hospitals (68 per cent) than in medium volume and low volume hospitals (60 per cent and 58 per cent, P < 0.001 from log rank test). Similar differences were seen in the 3-year colon cancer specific survival (79 per cent, 71 per cent and 69 per cent for high, medium and low volume hospitals, P < 0.001). The study population mean survival proportions showed similar associations between high annual hospital volume and survival in locally advanced colonic cancer (Fig. 4).
On the basis of multivariable models, high-volume hospitals were associated with reduced all-cause mortality after adjustments for potential confounders (HR 0.76, 95 per cent c.i. 0.62 to 0.93; Table 3) and additional inclusion of mediators (HR 0.81, 95 per cent c.i. 0.68 to 0.97). The effect of hospital volume differed significantly between university and non-university hospitals (P < 0.001 from Wald test of interaction). High-versus low-volume hospitals was HR 0.70 (95 per cent c.i. 0.50 to 0.98) at university hospitals, whereas for non-university hospitals the relative rate was HR 0.77 (95 per cent c.i. 0.60 to 1.00). When investigating the combination of hospital volume and preoperative MDT assessment, low volume was associated with an increased mortality without preoperative MDT assessment (HR 1.48, 95 per cent c.i. 1.09 to 1.99) compared with patients assessed before surgery in an MDT conference in high-volume hospitals (Table S3). In the analyses with delayed entry, the adjusted HRs remained unchanged (data not shown).
The 5-year all-cause mortality by hospital volume based on the annual volume of all colonic cancers is illustrated in Fig. S2. There was no significant association between hospital volume and all-cause mortality after adjustments for potential confounders and mediators in the multivariable model (Table S4).
Discussion
In this nationwide study, high hospital volume was associated with an improved long-term outcome in patients with locally advanced colonic cancer and a resection rate of at least 20 resections per year was associated with decreased overall mortality. The positive effect of hospital volume remained after adjustments for potential confounders, mediators, and effect modifiers. This further emphasizes that the total effect of hospital volume cannot be explained by known mediators, such as preoperative MDT assessment, neoadjuvant chemotherapy, or radical surgery. To our knowledge this is the first study focusing on the effect of hospital volume restricted to patients with locally advanced colonic cancer.
Colorectal surgery in Sweden has gradually been centralized during the past 15 years. This has mainly affected rectal cancer, both complicated and standard cases, which today are treated in medium-and high-volume hospitals. The evidence for improved survival in patients with rectal cancer managed at high-volume hospitals is diverging [15][16][17][18] . Colonic cancer surgery has been considered as a more basic surgery that can be performed in smaller units; however, surgery of locally advanced colonic tumours should be considered as an exception. Despite the need for complex surgery and perioperative care, these patients can have the same prognosis as standard patients with colonic cancer if offered the appropriate treatment 19,20 . The spectrum of multivisceral resections is very broad. Less-complicated resections of the abdominal wall or small bowel have a postoperative recovery similar to that after standard resections. More advanced resections of the great vessels or duodenum, including the biliary tract, can be very demanding for the surgeon and are associated with severe postoperative complications. In the present study, resection of two or more organs were more common in high-volume hospitals. This can be due to selection bias if patients with more complicated locally advanced tumours are denied surgery in low-volume hospitals. On the other hand, the differences in multivisceral resections could also be explained by referral of the more complicated cases from low-volume hospitals to medium-and high-volume hospitals. High-volume hospitals were associated with improved overall survival despite more complicated multivisceral resections being conducted in this group.
There are several studies on hospital volume and colonic cancer but none focusing on patients with locally advanced colonic cancer. A US study from 2000 showed an association between hospital volume and overall survival, with the effect concentrated on colonic cancer stage II-III disease, but the external validity was low given that the cohort was restricted to Medicare-enrolled patients aged 65 years and older 21 . Another study of patients with stage I-IV colonic cancer presented a significant difference in 5-year overall survival between medium-and high-volume hospitals (52 per cent versus 56 per cent, P < 0.0118) but no association with low-volume hospitals 22 . All included hospitals participated in a quality assurance programme on a voluntary basis, which complicates the generalizability. A Cochrane analysis from 2012 presented a significant association between high-volume hospitals and improved 5-year overall survival for rectal cancer (HR 0.85, 95 per cent c.i. 0.77 to 0.93), but no such association could be found for colonic cancer 23 .
In the present study of locally advanced colonic cancer, high-volume hospitals were clearly associated with both decreased 5-year all-cause and colonic cancer-specific mortality after adjustments for potential confounders.
In the sub-analysis on hospital volume based on total annual colonic cancer volume, there was no significant association between high hospital volume and survival after adjustments for potential confounders and mediators. This indicates that volume of more complicated cases is more important for improved survival in patients with locally advanced colonic cancer than the overall volume of colonic cancer surgery.
The association between hospital volume and survival most likely depends on several factors that are important in the multimodality treatment of locally advanced tumours. Preoperative MDT assessment, correct preoperative staging, neoadjuvant chemotherapy, and surgical competence are among them. Earlier studies have shown that preoperative MDT assessment is associated with improved survival in patients with oesophageal, lung, and rectal cancer but also in locally advanced colonic cancer [24][25][26][27] . Preoperative MDT assessment was more common with increasing hospital volume. Interestingly, further adjustment for preoperative MDT assessment did not change the results. To further explore this association, a combined exposure of hospital volume and preoperative MDT assessment was computed, which showed that low hospital volume was significantly associated with increased mortality without MDT assessment. Furthermore, high hospital volume was important for long-term survival, regardless of preoperative MDT assessment and other known mediators, which further enhances the importance of centralizing patients with locally advanced disease to high-volume units.
It is likely that high-volume hospitals have better access to surgical competence specialized in colorectal surgery; however, the results remain after adjustments for surgical competence as a mediator.
University hospitals are commonly referral units for more advanced cancer disease, such as locally advanced colonic cancer. High hospital volume is not the only advantage related to an academic setting. These hospitals often have high patient volumes for different procedures, such as thoracic surgery and vascular surgery, which makes them more experienced in the perioperative management related to complex surgery. Research activities with ongoing clinical trials are also more common in university hospitals and associated with improved outcome in patients with colorectal cancer 28 ; however, in the present study, the type of hospital modified the effect on overall mortality only in high-volume hospitals, and it is likely that the benefit of treatment in high-volume units cannot be explained by the type of hospital.
The major strength of this study is the large and nationwide setting, which enables the generalizability to other countries with comparable populations. The completeness of colonic cancer registration in the SCRCR was 98.5 per cent between 2008 and 2015 29 . By linkage to national patient registers the data set was completed with data on co-morbidities, previous cancer disease, and survival. Furthermore, this is the only study, to our knowledge, focusing on hospital volume in locally advanced colonic cancer surgery.
One limitation is that register information and treatment details, such as surgical quality as well as type and length of adjuvant chemotherapy, were missing. As in all studies of an observational nature there is a risk of residual confounding. For example, patients living in remote areas are more likely to be managed in low-volume hospitals than patients in urban areas. These patients are often older and have more co-morbidities. The potential differences in co-morbidity were handled by adjustments for both ASA score and CCI in the multivariable models. Lymph node harvest is an indicator of the quality of the surgery. Other possible measurements, such as the quality of the mesocolic excision, were not available in the register.
Another limitation is that the registers did not differ between pT4a and pT4b tumours. This study focused on the complex surgery of pT4b tumours, but the inclusion of pT4a tumours may have diluted the results.
In this study of patients having surgery for locally advanced colonic cancer, high hospital volume was associated with decreased mortality, and the association cannot be explained by known mediators. This knowledge should be considered in the discussion of centralization of patients with locally advanced colonic cancers.
Funding
This work was supported by grants provided by Region Stockholm (ALF project), Cancerföreningen in Stockholm and Stiftelsen Konung Gustav V: s Jubileumsfond.
|
v2
|
2022-11-25T16:44:10.055Z
|
2022-11-02T00:00:00.000Z
|
253865638
|
s2ag/train
|
Robotic-assisted laparoscopic radical prostatectomy: Initial outcomes of 500 cases.
INTRODUCTION
We aimed to present our experience of robot-assisted laparoscopic radical prostatectomy (RARP).
MATERIAL AND METHODS
The study was a retrospective review of 500 patients who underwent RARP between March 2015 and July 2021 in our clinic. A transperitoneal approach was used in all patients. All patients had clinically organ-confined prostate cancer (≤ cT2c).
RESULTS
The mean age of the patients was 64.6 ± 5.7 years. The median PSA was 11.4 ng/dL (range 0.3-92.7). The mean operative time was 183.5 min. Positive surgical margin rate was 19.4%. During a mean follow-up of 23.5 months, 96 patients (19.2%) received adjuvant radiotherapy due to the biochemical recurrence and 28 patients (16%) with lymph node positivity received early adjuvant hormone therapy. Considering the continence rates, 69% of the patients were total continence in the 3rd month, while this rate increased to 83 in the 6th month and 91% in the 12th month.
CONCLUSION
RARP is a safe and feasible method for experienced centers with patient comfort, surgeon comfort, and successful oncological and functional results.
|
v2
|
2022-11-26T06:16:12.726Z
|
2022-11-02T00:00:00.000Z
|
253880062
|
s2ag/train
|
What’s the latest with investigational drugs for soft tissue sarcoma?
ABSTRACT Introduction Despite extensive research undertaken in the past 20–30 years, the treatment for soft tissue sarcoma (STS) has remained largely the same, with anthracycline-based chemotherapy remaining the first choice for treating advanced or metastatic STS. Areas covered This review focuses on newly approved drugs for STS and current research directions, including recent results of late-phase trials in patients with STS. We cover several different histological subtypes, and we discuss the role of adoptive cell transfer (ACT) therapies for the treatment of synovial and myxoid/round cell (high-grade myxoid) liposarcoma, one of the most promising areas of treatment development to date. We searched clinicaltrials.gov and pubmed.ncbi.nih.gov, as well as recent year proceedings from the annual conferences of the American Society of Clinical Oncology (ASCO), European Society for Medical Oncology (ESMO), and Connective Tissue Oncology Society (CTOS). Expert opinion Immune-oncology drugs (IOs) show promise in certain subtypes of STS, but it is recognized that PD-1/PD-L1 axis inhibition is not enough on its own. Better trial stratifications based on the molecular categorization of different subtypes of STS are needed, and more evidence suggests that ‘one size fits all’ treatment is no longer sustainable in this heterogeneous and aggressive group of tumors.
|
v2
|
2022-11-03T17:45:55.160Z
|
2022-11-03T00:00:00.000Z
|
253260743
|
s2orc/train
|
A novel senescence-related lncRNA signature that predicts prognosis and the tumor microenvironment in patients with lung adenocarcinoma
Background: Cellular senescence has recently been considered a new cancer hallmark. However, the factors regulating cellular senescence have not been well characterized. The aim of this study is to identify long non-coding RNAs (lncRNAs) associated with senescence and prognosis in patients with lung adenocarcinoma (LUAD). Methods: Using RNA sequence data from the Cancer Genome Atlas Lung Adenocarcinoma (TCGA-LUAD) and senescence genes from the CellAge database, a subset of senescence-related lncRNAs was first identified. Then, using univariate and multivariate Cox regression analyses, a senescence lncRNA signature (LUADSenLncSig) associated with LUAD prognosis was developed. Based on the median LUADSenLncSig risk score, LUAD patients were divided into high-risk and low-risk groups. Kaplan-Meier analysis was used to compare the overall survival (OS) in the high- and low-risk score subgroups. Differences in Gene Set Enrichment Analysis (GSEA), immune infiltration, tumor mutation burden (TMB), tumor immune dysfunction and exclusion (TIDE) module score, chemotherapy, and targeted therapy selection were also compared between the high-risk and low-risk groups. Results: A prognostic risk model was obtained consisting of the following nine senescence-related lncRNAs: LINC01116, AC005838.2, SH3PXD2A-AS1, VIMS-AS1, SH3BP5-AS1, AC092279.1, AC026355.1, AC027020.2, and LINC00996. The LUADSenLncSig high-risk group was associated with poor OS (hazard ratio = 1.17, 95% confidence interval = 1.102–1.242; p < 0.001). The accuracy of the model was further supported based on receiver operating characteristic (ROC), principal component analysis (PCA), and internal validation cohorts. In addition, a nomogram was developed consisting of LUADSenLncSig for LUAD prognosis, which is consistent with the actual probability of OS. Furthermore, immune infiltration analysis showed the low-risk group had a stronger anti-tumor immune response in the tumor microenvironment. Notably, the levels of immune checkpoint genes such as CTLA-4, PDCD-1, and CD274, and the TIDE scores were significantly higher in the low-risk subgroups than in high-risk subgroups (p < 0.001). This finding indicates the LUADSenLncSig can potentially predict immunotherapy efficacy. Conclusion: In this study, a lncRNA signature, LUADSenLncSig, that has dual functions of senescence phenotype identification and prognostic prediction as well as the potential to predict the LUAD response to immunotherapy was developed.
Introduction
Lung cancer is a common malignant tumor, with the second highest incidence and highest mortality rate in the world (Ferlay et al., 2021). More than 1.3 million people die from lung cancer worldwide each year (Siegel et al., 2021). The 5-year survival rate for non-small cell lung cancer (NSCLC) is 23%, however, for small cell lung cancer is approximately only 6% (Siegel et al., 2021). Lung adenocarcinoma (LUAD) is the most common type of NSCLC, accounting for roughly 40% of all cases (Seguin et al., 2022). LUAD is more difficult to prevent and treat than lung squamous cell carcinoma, which is mostly associated with smoking (Seguin et al., 2022). Despite recent advances in genomics and targeted therapies for lung cancer, the disease is susceptible to drug resistance, resulting in a poor prognosis. In recent years, the development of immunotherapy has led to a new treatment direction for lung cancer. In 2015, CheckMate-063, a phase II single-arm trial, was the first large study in which the efficacy of immunotherapy in NSCLC was reported (Rizvi et al., 2015). In two phase III studies, CheckMate-017 (Brahmer et al., 2015) and CheckMate-057 (Borghaei et al., 2015), patients with NSCLC treated with nivolumab or docetaxel had significantly better overall survival (OS), response rate, and progression-free survival (PFS) when treated with nivolumab than docetaxel. Due to these impressive results, the FDA approved nivolumab as a second-line treatment for NSCLC in March 2015. Unfortunately, most LUAD patients do not respond to immune checkpoint inhibitor (ICI) therapies due to a lack of appropriate patient-selective biomarkers. To improve patient outcomes, new effective immunotherapy markers need to be developed. Cellular senescence is a program of stable cell cycle arrest in response to various intrinsic and extrinsic stimuli to remove senescent cells to maintain body homeostasis (Kumari and Jat, 2021). Cellular senescence is primarily caused by progressive telomere shortening, telomere structure change, mitosis, carcinogenic activation, ionizing radiation, oxidation, genotoxic stress, epigenetic changes, chromatin disorder, protein steady-state disorder, mitochondrial dysfunction, inflammation, tissue damage signal, radiation therapy, or chemotherapy (Pazolli et al., 2012;García-Prat et al., 2016;Mikuła-Pietrasik et al., 2020;d'Adda di Fagagna et al., 2003;Passos et al., 2010). Both cancer and senescence are caused by the accumulation of cell damage. Previous research has shown that senescence is both beneficial and detrimental in the process of tumorigenesis and development, and viewed as an example of the dichotomy of multiple effects in the evolutionary process (Ohtani et al., 2012;Schosserer et al., 2017). Conversely, senescence causes senescent cells to enter a permanent cell stagnation cycle to maintain tissue homeostasis and prevent tumor formation (Wang et al., 2020a). However, when senescent cells are not cleared by the immune system and accumulate, cell senescence may have harmful results, promoting the occurrence, development, invasion, and metastasis of tumors through multiple pathways (Wang et al., 2020a).
Phenotypic changes during cell senescence are controlled by changes in the specific proteins expressed. These processes are mainly regulated by proteins linking DNA and RNA as well as various non-coding RNAs, including long non-coding RNAs (lncRNAs) (Grammatikakis et al., 2014). LncRNA-XIST expression was shown downregulated in senescent cells and inhibited NSCLC cell proliferation and promoted apoptosis by triggering cell necrosis mediated by the miR-335/ SOD2/ROS signaling pathway, thereby inhibiting NSCLC progression (Liu et al., 2019). LncRNA H19 regulates the imprinting of gene clusters containing H19 and insulin-like growth factor 2 (IGF2) (Monnier et al., 2013). Both IGF2 (Issa et al., 1996) and H19 (Gabory et al., 2010) are associated with growth, proliferation, cell circumference, apoptosis, and senescence. LncRNA H19 is also highly expressed in lung cancer, and by inhibiting the function of miR-200a, upregulates the expression of ZEB1 and ZEB2, promoting the epithelial-mesenchymal transition and enhancing lung cancer cell proliferation and metastasis (Zhao et al., 2019). Taken together, lncRNAs can characterize cellular senescence and serve as an important tool in determining LUAD patient prognosis.
In the present study, a senescence-related lncRNA signature was constructed and shown to predict the Frontiers in Genetics frontiersin.org 02 prognosis of LUAD patients. In addition, a nomogram was developed that included a LUAD senescence-related lncRNA signature (LUADSenLncSig). Clinical factors, gene enrichment, mutations, immune cell infiltration, and potential response to targeted therapy and immunotherapy were further compared in the LUADSenLncSig high-risk and low-risk groups. In the present study, the regulatory network of cellular senescence was investigated and hypothesized to improve the efficacy of individualized treatment for LUAD.
Dataset and sample extraction
The LUAD RNA sequencing data, clinical, and mutations were downloaded from The Cancer Genome Atlas Lung Adenocarcinoma (TCGA-LUAD) database (https://portal.gdc. cancer.gov/). Data from 539 LUAD patients were initially collected. Patients with incomplete follow-up information, survival <30 days, or lack of complete clinicopathological data were excluded from the follow-up analysis and finally, 448 patients were retained. The online database CellAge (http://genomics.senescence.info/cells) was used to download 279 senescence-related genes identified from gene manipulation experiments (Avelar et al., 2020).
Differentially expressed genes (DEGs) of cellular senescence in LUAD and normal tissue
The threshold of log 2 fold change absolute value >1 and a false detection rate (FDR) < 0.05 were used to detect the differentially expressed genes (DEGs) between 539 tumor tissues and 59 normal tissues. The R package limma was used to visualize DEGs. Enrichment analysis of DEGs was performed using the Kyoto Encyclopedia of Gene and Genome (KEGG) and Gene Ontology (GO).
Identification of a lncRNAs senescencerelated signature (LUADSenLncSig) associated with LUAD prognosis
The senescence-related lncRNA-mRNA co-expression gene expression network was developed using Pearson correlation coefficient absolute value >0.3 and p < 0.001 as thresholds to identify senescence-related lncRNAs. The Sankey diagram generated using the R package ggalluvial and Cytoscape software (version 3.7.2) were used to visualize the lncRNA-mRNA co-expression network. First, univariate Cox regression analysis was used to identify lncRNAs associated with LUAD prognosis, then the lncRNAs were further incorporated into multivariate Cox regression analysis to construct the prognosis model of LUAD; this lncRNA senescence-related signature was termed LUADSenLncSig. The risk score of the prognostic model was calculated using the following formula and LUAD patients were divided into high-and low-risk groups based on the median risk score: risk score = explncRNA1 × coef lncRNA1 + explncRNA2 × coef lncRNA2 + explncRNAi × coef lncRNAi. To determine whether the LUADSenLncSig risk score was an independent prognostic index, the clinicopathological variables were included in univariate and multivariate Cox regression analyses. To investigate the distribution of living conditions, the risk score level was used. Furthermore, the accuracy of the risk model was assessed using the receiver operating characteristic (ROC) curve. The R package pheatmap was used to display the clinicopathological variables of the highand low-risk groups. The distribution of patients with different risk scores was assessed using principal component analysis (PCA) and the visualization was performed using R package scatterplot3d.
Construction of the nomogram
A nomogram was constructed by combining risk scores and clinicopathological features and using the R package rms to predict survival at 1, 3, and 5 years in patients with LUAD. Calibration curves were used to determine if the predicted survival matched the actual survival.
Gene Set Enrichment Analysis (GSEA) of the senescence-related lncRNA predictive signature Gene Set Enrichment Analysis (GSEA) was performed for the high-and low-risk groups to identify different functional enrichment of the two groups. Significant biological processes and pathways were enriched at the threshold of nominal (NOM) p < 0.05 and the FDR q-value < 0.25. The R package ggplot2 was used to visualize the results.
Analysis of somatic mutations and tumor mutation burden (TMB)
The number of somatic non-synonymous mutations in each sample was calculated using the R package maftools (Mayakonda et al., 2018). The tumor mutation burden (TMB) uses nonsynonymous and code-shifting indels and a 5% detection limit to calculate the number of somatic, coding, base replacement, and insert-deletion mutations found in the genome database per Frontiers in Genetics frontiersin.org megabase. The TMB was also compared between the high-risk and low-risk groups and survival curves and risk scores for TMB were plotted.
Estimation of immune infiltration
The immune cell infiltration fraction and immune-related pathway activity in high-and low-risk LUAD samples were assessed using single-sample GSEA (ssGSEA) (Rooney et al., 2015) and the CIBERSORT algorithm as cross-validation (Newman et al., 2015). The Wilcoxon rank-sum test was used to determine whether a significant difference existed in immune cell proportions between the low-risk and highrisk groups.
Potential relationship between the LUADSenLncSig and immunotherapy, chemotherapy, and target therapy First, the differential expression of 40 immune checkpoints in high-risk and low-risk groups was compared. Furthermore, the tumor immune dysfunction and exclusion (TIDE, http://tide.dfci.harvard.edu/) module was used to distinguish potential immunotherapy responses between high-risk and low-risk groups. This module predicts anti-PD1 and anti-CTLA4 treatment responses based on the transcriptional expression profiles of patient genomes before treatment. In addition, to further evaluate the role of LUADSenLncSig in predicting the therapeutic response of LUAD, the half-maximal inhibitory concentration (IC 50 ) of commonly used chemotherapeutic drugs and targeted therapeutic drugs was calculated. Wilcoxon signed-rank test and R package pRRophetic were used to compare and visualize the IC 50 values in the high-risk and low-risk groups.
Statistical analysis
The Wilcoxon test was used to compare the expression of senescence-related DEGs in cancer tissues and normal tissues. The Kaplan-Meier method and log-rank test were used to compare the OS rate between the high-risk and low-risk groups. The survivalROC package was used to generate ROC curves and calculate the area under the curve (AUC). The Kruskal-Wallis test was used to compare the differences between groups. The clinical data were analyzed using the chi-square test or the Fisher's exact test. A Pearson correlation coefficient was used to evaluate the relationship between lncRNA expression and immune infiltration and immune checkpoint gene expression. All statistical analyses were performed using R software (version 4.1.2).
Enrichment analysis of senescencerelated DEGs in LUAD
The study flow chart is shown in Figure 1. First, the expression levels of 279 senescence-related genes in tumor and normal tissues were compared to determine if the genes were abnormally expressed in tumor tissues (Supplementary Table S1). In LUAD tumor tissues, 23 of 62 differentially expressed genes were downregulated and 39 were upregulated (Figures 2A,B, Supplementary Table S2). KEGG pathway analysis revealed the five most enriched pathways were human T-cell leukemia virus one infection, cellular senescence, cell cycle, Kaposi sarcoma-associated herpesvirus infection, and p53 signaling pathway ( Figure 2C). Conversely, GO analysis showed the five most enriched categories were cell aging, aging, mitotic cell cycle phase transition, regulation of mitotic cell cycle phase transition and cellular response to chemical stress ( Figure 2D). These findings indicate the DEGs are primarily involved in cell senescence, cell cycle, virus infection, cell stress, and cell apoptosis.
Correlation between LUADSenLncSig and prognosis of LUAD patients
The risk score of LUADSenLncSig was calculated as follows: risk score = (0.348 × LINC01116 expression) + (−0.889 × AC005838. The formula was used to calculate the risk score of each patient and the patients were divided into two groups based on the median risk score: high-risk group (n = 228) and low-risk group (n = 220) ( Figure 4A). Kaplan-Meier analysis showed the OS was significantly shorter in the high-risk group than in the low-risk group ( Figure 4A). Figures 4B,C show risk scores and survival statistics for individual patients, with a greater number of deaths with increasing risk scores. Univariate and multifactorial Cox regression analyses showed the LUADSenLncSig risk score was an independent prognostic factor for LUAD (Figures 4D,E) with an AUC of 0.764 and the best predictor of LUAD prognosis compared with other clinicopathological variables ( Figure 4F). The AUC of the 1-year, 3-year, and 5-year ROC curves were 0.707, 0.677, and 0.772, respectively, indicating that LUADSenLncSig performed well in LUAD prognosis ( Figure 4G). Figure 5A depicts the expression levels of the nine lncRNAs and clinicopathological factors in the LUADSenLncSig model. To differentiate between high-risk and low-risk patients, PCA with genome-wide, senescencerelated DEGs, senescence-related lncRNAs, and LUADSenLncSig, was performed ( Figures 5B-E). The LUADSenLncSig model clearly distinguished low-risk and high-risk populations, as shown in Figure 5E, demonstrating the accuracy of the model. Furthermore, Figure 6A-Q shows the relationship between clinical parameters and risk score. Significant correlations were observed between the risk score and age (> 65 years and ≤ 65 years, Figure 6A and B), sex (female and male, Figure 6C,D), M0 stage ( Figure 6E), N0 and N1 stage ( Figure 6G,H), overall TNM stage 2 and 3 ( Figure 6K,L), and T2, T3, and T4 stage ( Figures 6O-Q). The LUADSenLncSig risk score was proven an independent prognostic risk factor for patients with LUAD.
Construction of the nomogram
A nomogram clinical prognostic assessment plot was established using the LUADSenLncSig risk score and other clinicopathological factors to assess the probability of survival at 1, 3, and 5 years for patients with LUAD ( Figure 7A). Based on the three calibration plots, the nomogram-estimated mortality was similar to the actual mortality ( Figures 7B-D).
Internal validation of the LUADSenLncSig
The TCGA-LUAD patients (n = 448) were randomly assigned to two internal validation cohorts (n = 224 in the first internal cohort and n = 224 in the second internal cohort) to determine if LUADSenLncSig is universally applicable to the OS predictive value of LUAD. The clinical characteristics of the sample are detailed in Table 1. Patients in the high-risk group had a shorter OS than subjects in the low-risk group in the first and second internal cohorts ( Figures 8A,C), which was consistent with the overall TCGA-LUAD dataset results. Furthermore, the AUC of 1-year, 3-year, and 5-year survival in the first internal cohort was 0.787, 0.683, and 0.79, respectively ( Figure 8B), and in the second internal cohort was 0.629, 0.679, and 0.734, respectively ( Figure 8D). These results showed that LUADSenLncSig performs adequately in both internal validation cohorts, indicating the robustness of the prediction model. The relationship between LUADSenLncSig and TMB Somatic mutations were separately examined in 223 high-risk patients and 216 low-risk patients ( Figures 10A,B). Higher mutation rates were observed in the high-risk group than in the low-risk group for TTN (53% vs. 40%), MUC16 (44% vs. 36%), CSMD3 (42% vs. 36%), and RYR2 (39% vs. 33%). In addition, although difference in TMB was not found between the high-and low-risk groups ( Figure 10C), the combination of high TMB with LUADSenLncSig risk scores in the low-risk group led to better prognosis ( Figures 10D,E), indicating a synergistic effect between these two indicators. In theory, a higher TMB increases tumor neoantigen production, thus, increased immune recognition and tumor cell killing are likely. In addition, an association between high TMB and ICI response rates and PFS has been found in previous studies (Sholl et al., 2020). Thus, the immune infiltration and potential immunotherapy response between the two groups were assessed.
The profile of immune infiltration in the high-and low-risk LUAD subgroups
The results of ssGSEA analysis showed the number of anti-tumor immune cells, such as CD8 + T cells, macrophages, and T helper cells, was significantly higher in the low-risk group than in the high-risk group ( Figure 11A). Immune function scores, such as antigen-presenting cell (APC) coinhibition, C-C chemokine receptor (CCR), checkpoint, human leukocyte antigen (HLA), T cell co-inhibition, T cell co-stimulation, and type II interferon (IFN) response, were significantly higher in the low-risk group than in the high-risk group ( Figure 11B). These results indicate a stronger anti-tumor immune response in the tumor microenvironment in the low-risk group. In addition, the CIBERSORT algorithm showed the number of CD8 + T cells and macrophages M2 was significantly higher in the low-risk group than in the high-risk group ( Figure 11C), which confirmed the ssGSEA analysis results. These differences based on the theory that immunotherapy must rely on a pre-existing immune-hot microenvironment
LUADSenLncSig has a potential relationship with LUAD immunotherapy, chemotherapy, and target therapy The expression levels of 37 immune checkpoint genes differed between the high-and low-risk groups ( Figure 12A). Immunotherapy markers such as CD274 (PD-L1 protein coding gene), PDCD-1 (PD-1 protein coding gene), and CTLA-4, which are now widely used in clinical trials, were found significantly higher in the low-risk group ( Figure 12A), indicating potential immunotherapeutic responses in lowrisk patients. Furthermore, as shown in Figure 12B, when the online software TIDE was used to predict the efficacy of anti-PD1 or anti-CTLA4 treatment for LUAD patients, TIDE scores were significantly higher in the low-risk subgroups than in high-risk subgroups (p < 0.001). This finding demonstrates the LUADSenLncSig has the potential to predict immunotherapy efficacy.
Finally, the relationship between the LUADSenLncSig risk score and the efficacy of LUAD chemotherapy and target therapy was analyzed. As shown in Figures 12C-F, the traditional chemotherapy drug methotrexate ( Figure 12C) as well as novel targeted therapies such as the P21-activated kinase 1 (PAK1) inhibitor IPA.3 ( Figure 12D), the altered Akt inhibitor MK.2206 ( Figure 12E), and the CDK4/6 inhibitor PD.0332,991 (palbociclib, Figure 12F), were among the more sensitive drugs in the high-risk group. Conversely, paclitaxel ( Figure 12G) and docetaxel ( Figure 12H) as well as tyrosine kinase inhibitors erlotinib ( Figure 12I) and Frontiers in Genetics frontiersin.org
FIGURE 9
LUADSenLncSig-based GSEA of different risk groups. (A) Based on the GSEA results, the KEGG genes were differentially enriched for senescence-related lncRNA expression. Five KEGG items, the cell cycle, p53 signaling pathway, oocyte meiosis, glycosphingolipid biosynthesis-lacto and neolacto series, and adherens junction were enriched in the high-risk group. Asthma, intestinal immune network for IgA production, hematopoietic cell lineage, autoimmune thyroid disease, and T cell receptor signaling pathway were enriched in the low-risk group based on the NES, NOM p-value, and FDR q-value (B) Differential enrichment of genes in GO with senescence-related lncRNAs. Five GO items, spindle localization, establishment of spindle orientation, establishment of mitotic spindle localization, cadherin binding, and microtubule cytoskeleton organization involved in mitosis showed a significant differential enrichment in the high expression phenotype. The other five GO terms, negative regulation of adaptive immune response, mast cell activation involved in immune response, T cell activation involved in immune response, T cell differentiation involved in immune response, and negative regulation of B cell mediated immunity, were found significantly enriched in the low expression phenotype based on the NES, NOM p-value, and FDR q-value. The reference files from the (MSigDB) were c2. cp.kegg.v7.4. symbols.gmt and c5. go.v7.4 symbols. gmt (http://software.broadinstitute.org/gsea/msigdb/index.jsp). LncRNAs, long non-coding RNAs; LUAD, lung adenocarcinoma; LUADSenLncSig, LUAD senescence lncRNA signature; GSEA, Gene Set Enrichment Analysis; KEGG, Kyoto Encyclopedia of Gene and Genome; NES, normalized enrichment score; NOM p-value, nominal p-value; FDR, false detection rate; GO, Gene Ontology; MSigDB, Molecular Signatures Database.
Frontiers in Genetics frontiersin.org BIBW2992 (alphatinib, Figure 12J), were the more sensitive drugs in the low-risk group based on the IC 50 .
Discussion
Lung cancer is the leading cause of cancer-related mortality worldwide. The recent FDA approval of ICIs for LUAD has changed the therapeutic landscape. However, the overall response rate of ICI is <20% (Brahmer et al., 2015) and only a subset of LUAD patients benefit from ICI treatment. Consequently, fully assessing whether patients benefit from ICI treatment is important. Biomarkers provide information for this treatment decision, however, effective biomarkers to predict efficacy in clinical applications do not currently exist (Brueckl et al., 2020). Tumor cell immunohistochemical PD-L1, plasma soluble PD-L1 (sPD-L1), TMB/blood TMB (bTMB), mismatch repair defect (dMMR)/ microsatellite instability-high (MSI-H), and other biomarkers such as KRAS, STK11, KEAP1, and DNA damage response (DDR) gene variation are potential biomarkers (Doroshow et al., 2019). PD-L1 detected using tumor cell immunohistochemistry (IHC) is a standard biomarker to predict the effectiveness of ICI therapy. Pembrolizumab has been used in first-line therapy with a PD-L1 threshold of 50% in patients with advanced NSCLC (Arbour and Riely, 2019). However, PD-L1 has disadvantages of low expression and false negatives in some patients as well as spatial differences within and between tumors (Costantini et al., 2019). In the KEYNOTE-010 (Herbst et al., 2016) and KEYNOTE-042 (Mok
FIGURE 10
The relationship between LUADSenLncSig risk scores and somatic mutation and TMB in LUAD tissues. Waterfall plot showing the somatic mutations between high-(A) and low-risk (B) LUAD patients. (C) Difference of TMB between patients from the low-and high-risk score subgroups. (D) Kaplan-Meier curves for the high-and low-TMB groups. (E) Kaplan-Meier curves for patients stratified based on both TMB and risk scores. The p-value represents the ANOVA test between the subgroups. TMB, tumor mutation burden; LUAD, lung adenocarcinoma; lnc, long non-coding.
Frontiers in Genetics frontiersin.org et al., 2019) studies, patients with high TMB levels in tumor tissues who received pembrolizumab had better objective response rate (ORR), PFS, and OS than patients in the chemotherapy group (Herbst et al., 2019). However, consensus is currently lacking regarding which mutations should be applied to the calculation of TMB. In addition, because the incidence of dMMR/MSI-H in lung cancer is very low (Takamochi et al., 2017), the predictive value of immunotherapy for lung cancer must be further confirmed. Compared with a single clinical biomarker, combining multiple biomarkers into a single model can improve prediction accuracy and contribute to accurately personalized treatment planning (Guo et al., 2021). Senescent cells (regardless of cell type) have recently been identified as important functional cell types in tumor microenvironments, including LUAD (Hanahan, 2022). Pituitary tumor transforming gene 1 (PTTG1) and Holliday cross recognition protein (HJURP) are senescent cell-specific genes. PTTG1 is associated with the progression of NSCLC and a poor prognostic factor for NSCLC patients (Wang et al., 2016). HJURP is a histone
FIGURE 11
Immune cell infiltration and immune-related functions in different risk groups. The ssGSEA algorithm was used to compute the score levels of infiltration of 16 immune cells. (A) and 13 immune-related functions. (B) in the high-and low-risk groups. (C) The Wilcoxon rank-sum test was used to determine differences between the 22 types of immune cells between high-and low-risk groups. SsGSEA, single-sample Gene Set Enrichment Analysis; aDCs, activated dendritic cells; iDCs, immature dendritic cells; NK, natural killer; pDCs, plasmacytoid dendritic cells; Tfh, T follicular helper; Th1, T helper type 1; Th2, T helper type 2; TIL, tumor-infiltrating lymphocyte; Treg, T regulatory cell; APC, antigen-presenting cell; CCR, C-C chemokine receptor; HLA, human leukocyte antigen; MHC, major histocompatibility complex; IFN, interferon; *p < 0.05; **p < 0.01; ***p < 0.001; ns, non-significant; p < 0.05 indicates statistical significance.
Frontiers in Genetics frontiersin.org H3 chaperone protein that influences mitosis cell cycle progression, DNA repair, and chromosome segregation (Wang et al., 2020b). HJURP is overexpressed in NSCLC and promotes NSCLC cell proliferation, migration, and invasion by activating the Wnt/βcatenin signal pathway (Wei et al., 2019). High HJURP expression is associated with shorter OS and disease-free survival (Wei et al., 2019). The detection of senescent cells remains controversial at present and the majority are linked to SASP expression, DNA damage, and β-galactosidase activity, none of which are specific or universal (Lee et al., 2006;van Deursen, 2014;Georgakilas et al., 2017;Kastenhuber and Lowe, 2017;Casella et al., 2019). Several lncRNAs have recently been found involved in the regulation of senescence in LUAD cells (Liu et al., 2019;Wei et al., 2019;Wan et al., 2021), however, the overall pattern of the LUAD cell senescence regulatory network remains largely unknown. Therefore, we developed a LUADSenLncSig signature that represents both senescence and prognosis of LUAD. The LUADSenLncSig model developed in this study shows good predictive performance for the OS of TCGA-LUAD samples. Furthermore, the nomogram, which includes the
FIGURE 12
Comparison of immune checkpoints, TIDE scores, sensitivity of chemotherapy and targeted therapy drugs in the high-and low-risk groups. (A) Expression of 37 immune checkpoint genes differed between the high-risk and low-risk groups. Red boxes represent high-risk patients and blue boxes represent low-risk patients (B). The online software TIDE prediction of the efficacy score in subgroups of LUAD patients treated with anti-PD1 or anti-CTLA4. The IC 50 values for (C) methotrexate (D) IPA.3 (E) MK.2206 (F) PD.0332,991 (palbociclib) (G) paclitaxel (H) docetaxel (I) erlotinib, and (J) BIBW2992 (alphatinib) in the high-risk and low-risk groups. TIDE, tumor immune dysfunction and exclusion module; IC 50 , half-maximal inhibitory concentration; *p < 0.05; **p < 0.01; ***p < 0.001; ns, non-significant; p < 0.05 indicates statistical significance.
Frontiers in Genetics frontiersin.org LUADSenLncSig risk score, has the potential to guide clinical decisions. Notably, LUADSenLncSig can stratify LUAD patients based on immune checkpoint gene expression levels and TIDE score, which is important for the selection of patients who may benefit from immunotherapy. Among the LUADSenLncSig, LINC01116 has been shown to promote LUAD progression by affecting the p-Akt signaling pathway (Zeng et al., 2020) and epithelial mesenchymal transformation (Shang et al., 2021), and is a lncRNA associated with multiple LUAD prognostic models (Geng et al., 2021;Gong et al., 2022a;Yang et al., 2022).
The notable aspect of this study is the important relationship between cell senescence and the prognosis and microenvironment of LUAD. At present, the role of the senescent microenvironment in tumors is often ignored in preclinical studies, which are usually designed for young mice rather than old mice (Fane and Weeraratna, 2020) and may help explain why many successful preclinical responses are not reproduced after entering real clinical trials. When considering comprehensive treatment for tumors based on the role and mechanism of genes involved in the regulatory network in LUAD tissue, senescence should be included as a parameter. In a clinical retrospective study, patients >60 years of age responded better to PD-1 than patients <60 years of age (Kugel et al., 2018). Furthermore, when compared with anti-CTLA-4 and anti-PD-L1, anti-PD-1 was ineffective against melanoma in aged mice (Padrón et al., 2018). When senescent cells (expressing p16INK4a) were pharmacologically eliminated in aging mice, the incidence of spontaneous tumorigenesis and cancer-related death was reduced (Baker et al., 2016). Similarly, in the present study, the relationship between the LUADSenLncSig model derived from senescence genes and the potential efficacy of LUAD was determined. Based on LUADSenLncSig stratification, the expression of most immune checkpoints, activation of immune pathways, infiltration of antitumor immune cells, and TIDE score in the low-risk group were higher than in the high-risk group (Figures 11, 12), indicating that low-risk patients may benefit more from immunotherapy. In addition, compared with other studies using the TCGA database to develop lncRNA prognostic models Deng et al., 2020;Zhao et al., 2020;Chen et al., 2021;Feng et al., 2021;Tang et al., 2021), our LUADSenLncSig model is very comparable. This shows that senescent cells in TME are helpful and as valuable as other biomarkers in predicting prognosis and estimating the efficacy of immunotherapy of LUAD. Interestingly, other studies have also focused on the contribution of cellular senescence-associated LncRNAs to tumor prognosis in tumors including liver (Huang et al., 2022a), colorectal (Huang et al., 2022b), and gastric cancers. These studies have similar AUC values to the time-dependent ROC curves of our model and importantly illustrate the contribution of senescent cells to the tumor microenvironment.
The present study had several limitations. First, more external data should be considered to assess whether the LUADSenLncSig model fully matches the additional dataset. Second, several key data points that affect patient prognosis, such as who received second-line treatment, were lacking and could not be included in the nomogram, which may affect the model's accuracy. Third, functional studies are needed to better understand the molecular mechanism of the lncRNA effect associated with senescence.
In conclusion, we developed a LUADSenLncSig lncRNA signature that can be used to predict LUAD prognosis. Notably, LUADSenLncSig was associated with the level of immune infiltration and the potential efficacy of tumor immunotherapy. These findings highlight the potential future direction of tumor immunotherapy with an emphasis on cellular senescence therapy. In addition, because the technology is more readily available and is becoming less expensive, bulk-sequencing is a technique that may more easily be introduced into the clinic as a standard management method. We believe that we can expand the bulk-sequencing-generated lncRNA model to the standard care of LUAD patients if sufficient external data is available, which can validate the predictive efficacy of senescence-related lncRNAs.
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.
Author contributions
MS and XF conceived the study and design and provided administrative support. XF and EH contributed to data analyses, wrote, reviewed, and edited the manuscript. XX, KY, SW, and XH contributed to data analysis and reviewed the manuscript. All authors read and approved the final manuscript. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-03T17:51:37.972Z
|
2022-11-03T00:00:00.000Z
|
253260795
|
s2orc/train
|
Identifying tumor antigens and immune subtypes of renal cell carcinoma for immunotherapy development
Renal cell carcinoma (RCC) is one of the leading causes of death in men. Messenger ribonucleic acid (mRNA) vaccines may be an attractive means to achieve satisfactory results. Cancer immunotherapy is a promising cancer treatment strategy. However, immunotherapy is not widely used in renal cell carcinoma, as only a few patients show a positive response. The present study aimed to identify potential antigens associated with renal cell carcinoma to develop an anti-renal cell carcinoma mRNA vaccine. Moreover, the immune subtypes of renal cell carcinoma cells were determined. The Cancer Genome Atlas (TCGA) analysis revealed gene expression profiles and clinical information. Antigen-presenting cells infiltrated the immune system using Tumor Immune Estimation Resource (TIMER) tool (http://timer.cistrome.org/). GDSC (Genomics of Drug Sensitivity in Cancer) database were used to estimate drug sensitivity. The 13 immune-related genes discovery could be targets for immunotherapy in renal cell carcinoma patients, as they were associated with a better prognosis and a higher level of antigen-presenting cells. These immune subtypes have significant relationships with immunological checkpoints, immunogenic cell death regulators, and RCC prognostic variables. Furthermore, DBH-AS1 was identified as a potential antigen for developing an mRNA vaccine. The CCK8 assay demonstrated that the proliferative capacity of 786-O and Caki-1 cells overexpressing DBH-AS1 was higher than in the control group. In addition, transwell assay revealed that 786-O and Caki-1 cells overexpressing DBH-AS1 showed higher invasion capacity compared with control. This study provides a theoretical basis for the development of mRNA vaccines. Our findings suggest that DBH-AS1 could be potential antigens for developing RCC mRNA vaccines.
Introduction
Men are more likely than women to develop renal cell carcinoma (RCC), which is one of the world's 10 most common types of cancer (1). Most deaths from renal cell carcinomas are caused by clear cell renal cell carcinomas (ccRCCs). The treatment depends on the diagnosis of tumor characteristics and staging. Currently, most patients with RCC present with metastatic disease, and a 30%-40% chance of developing metastases is expected in the remaining patients (2). There are many places where metastatic RCC can occur, such as the lungs, lymph nodes, bones, liver, and brain. RCC has a broad metastatic potential and is very common even after radical nephrectomy (3). According to literature reports, most metastatic lesions occur within the first 5 years (4). A previous study reported that age, race, and family history were significant risk factors for RCC patients (5).
Cancer immunotherapy can be divided into passive immunotherapy and active immunotherapy (6). Active immunotherapy stimulates a patient's immune system to activate natural killer cells or cytotoxic T cells or to produce antibodies against tumor-specific antigens (7). One of the functions of immune checkpoint inhibitors is to block programmed death 1 (PD-1), restoring T cells to target cancer cells (8). Various drugs have been developed to inhibit certain solid tumor progression. Nivolumab and pembrolizumab are emerging antitumor drugs that work by blocking the immune checkpoint protein PD-1, hence promoting T-cell recovery to target cancer cells. Furthermore, PD-L1 inhibitors are currently used in some cancers and are being explored for other cancers (9).
In the last two decades, treating and managing metastatic RCC have undergone substantial improvements. Initially, cytokines were utilized in first-generation immunotherapies. Interleukins or interferons were the standard approaches, with poor results (10). Immune checkpoint inhibitors (ICIs) alone or in combination have shown better results than traditional immunosuppressants (11 ). Targeted immunotherapy can be used in place of antiangiogenic drugs because ccRCC is also considered an immunogenic tumor with many immune cells like tumor-infiltrating lymphocytes (TILs) (12). The tumor microenvironment (TME) is complex and evolving. In addition to stromal cells, fibroblasts, and endothelial cells, the TME also includes innate and adaptive immune cells. TME plays a critical role in drug resistance, according to studies combining antiangiogenic and targeted immunotherapies, which are currently available as a first-line treatment option (13). It is possible to develop resistance to ICIs through primary, congenital, secondary, or acquired mechanisms. These are neoantigen loss, deficient antigen presentation, alternative immune checkpoints, and deficient interferon signaling. Interferon-g is key in enhancing the PD-L1 expression and inducing immunosuppressive molecules (14). TIGIT, LAG-3, TIM-3, and other immune checkpoints suppress antitumor immunity, contributing to drug resistance.
These resistance mechanisms are being overcome with new treatments currently being evaluated in ongoing clinical trials. Identifying biomarkers for metastatic kidney cancer is critical to select better treatments, lowering costs, and improving survival (15). Nevertheless, the limitations of the most studied biomarkers, PD-L1 immunohistochemistry, and TMB make it imperative to find more robust markers. New technologies may be able to provide this assistance.
The current study aimed to explore new RCC antigens, provide a basis for developing new immunotherapy drugs, and define immune subtypes that can be used to improve immunotherapy response in RCC patients. Furthermore, the study sought to provide insights into targeted therapy in RCC by conducting correlation analysis of antigen-presenting cells, prognostic-related genes, immune subtype analysis, immune checkpoints, and immune modulators.
Data sources
A key feature of The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/) is its representation of cancer gene expression, miRNA expression, copy number variation, DNA methylation, and single nucleotide polymorphisms (SNPs). TCGA database was used to retrieve mRNA expression data of the Kidney Renal Clear Cell Carcinoma (KIRC) cohort. cBioPortal (http://www.cbioportal. org,v3.2.11) provides open access to raw data from large-scale genome studies. In this study, cBioPortal was used to explore potential anti-tumor antigen gene changes in TCGA cohort. TCGA-KIRC on cBioPortal was used to retrieve complete expression profiles and survival information, and ImmPort retrieves immune-related genes for each patient.
Tumor Immune Estimation Resource analysis
An advanced tool for systematically analyzing immune infiltration in different cancer types is called TIMER (https:// cistrome.shinyapps.io/Timer/). A TIMER analysis was conducted to examine the relationship between immune cell infiltration and the expression level of the identified effective antigen.
Detection of the differential expression genes
We screened differentially expressed genes using the edgeR filter criteria [log2|fold change| > 2, false discovery rate (FDR) < 0.05]. The volcano plot displayed filtered results, as red indicates genes that have been upregulated, and blue indicates genes that have been downregulated.
Classification of immune subtypes
We quantified individual scores for each tumor case using the single-sample gene set enrichment analysis (ssGSEA) method. ssGSEA computes overexpression measures for a list of genes of interest relative to all other genes in the genome using a rank-based method. ssGSEA scores were calculated using logtransformed RNA-seq or microarray data. Our next step was to classify RCCs according to 29 immunobio signature enrichment levels (ssGSEA scores) and examine both tumor purity and the immune score for each RCC.
Drug sensitivity analysis
According to the Genomics of Drug Sensitivity in Cancer (GDSC) database, the chemotherapy sensitivity of each tumor sample was predicted with "pRRophetic" R package (https:// www.cancerrxgene.org/). IC 50 values for each chemotherapy drug were further determined by regression analysis. We performed cross-validation on thGDSC training set 1e 0 times to test the regression and prediction accuracy. Each parameter was set to its default value, including the "combat" parameter, which removes batch effects and averages repeated gene expressions.
Immune cell infiltration analysis
The relative proportions of 22 immune infiltrating cells were determined by analyzing RNA-seq data from RCC patients in different sub-groups using CIBERSORT algorithm. Following this, Spearman correlation analysis was performed to determine how gene expression relates to immune cell infiltration. It was considered statistically significant when p-value <0.05.
Gene set variation analysis
Gene set enrichment was evaluated using gene set variation analysis (GSVA), a non-parametric, unsupervised method. The gene-level changes in this analysis were transformed into pathway-level changes by scoring the gene set of interest, followed by determining the sample's biological function. The molecular signatures database (version 7.0) was used to retrieve the gene sets in the present study. A comprehensive evaluation of potential biological function changes in various samples was then conducted using the GSVA algorithm.
Construction of the prognostic prediction model
The univariate Cox regression, multivariate Cox regression, and LASSO regression were used to investigate the genes closely associated with RCC prognosis.
Weighted correlation network analysis
Weighted gene coexpression networks were constructed to evaluate co-expressed gene modules, genotype-phenotype relationships, and core genes in the network. Using weighted correlation network analysis (WGCNA) R package, we constructed a coexpression network using all genes in the dataset. We further analyzed genes with the highest variances using 3 as a threshold. Furthermore, a network connectivity estimation was conducted by transforming the weighted adjacency matrix into a topological overlap matrix (TOM). A hierarchical clustering tree was also constructed based on TOM matrix using hierarchical clustering. Different branches of the clustering tree represented different gene modules, whereas different colors represented different gene modules. A module was created by grouping genes whose expression patterns are similar. The weighted correlation coefficients of thousands of genes allowed the identification of multiple modules based on gene expression patterns.
GO and KEGG enrichment analysis
Key genes were annotated, and candidate genes' functions were explored using ClusterProfiler package. We explored related functional categories using gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). In GO and KEGG enrichment pathways, statistical significance was determined by p-values and q-values <0.05.
Gene set enrichment analysis
MSigDB (http://www.gsea-msigdb.org/gsea/downloads.jsp) was used to retrieve gene sets. GSEA was performed on the gene sets to identify enriched GO terms and KEGG pathways. The 50 best terms were selected from each subtype based on their significance. fetal bovine serum, 1% penicillin-streptomycin, and 1% glutamine in a humidified atmosphere with 5% CO 2 at 37°C. Overexpression plasmids and control vectors were designed and constructed by Shanghai Jikai Gene Co, Ltd. Transfection of the expression plasmid was performed according to the manufacturer's instructions.
CCK8 assay
CCK8 assay was conducted using the CCK8 kit (Dojindo, Shanghai, China). Briefly, 1,000 cells were planted into each well with 100 ml medium, and 10 ml of CCK8 was added to each well. After incubation at 37°C for 24 h in a humidified incubator with 5% CO 2 , the proliferative ability of the cells was measured at 450 nm.
Transwell assay
Transwell chambers with pore sizes of 8.0 mm were added to a 24-well plate to create upper and lower chambers. We then added 600 ml medium with 10% fetal bovine serum (FBS) to the lower chamber and 200 ml serum-free conditioned medium with 2 × 10 4 cells to the upper chamber. A cotton swab was used to wipe the upper chamber cells, and invasion cells were fixed and stained with crystal violet for counting after 24 h.
Statistical analysis
Multivariate data analysis was conducted using Cox proportional hazard model and Kaplan-Meier method for calculating survival curves and comparing them with the logrank test. R (version 3.6) software was used for all statistical analyses. Statistical significance was determined by a p-value <0.05 on both sides of the test.
Results
Immune subtype analysis showed the two immune groups of RCC In this study, 29 immune-related genomes representing various immune cells, pathways, and functions were assessed. ssGSEA method was used to cluster TCGA dataset containing RCC samples based on the immune cells' expression profiles ( Figure 1A). Dimensionality reduction algorithm t-SNE showed that the subtypes were highly consistent with two-dimensional t-SNE distribution patterns at a k-value of 2 ( Figure 1B). A group with high immunity was known as Immunity_H, while a group with low immunity was known as Immunity_L. Among Immunity_H subtypes, immune cell infiltration and immune pathways activation were reported, which indicates an immune hot phenotype. According to the Immunity_L subtype, there was low immune cell infiltration, indicating a cold immune phenotype.
Moreover, the immune subtype analysis demonstrated that each was associated with the different immune components in RCC patients. The Immunity_H subtype showed a higher stromal score, immune score, and estimate score than the Immunity_L subtype ( Figure 1C). The heatmap demonstrated that Immunity_L subtype was enriched with more tumor purity than Immunity_H subtype. Immunity_H subtype had greater infiltration and immune pathways activation than Immunity_L subtype in terms of immune cell infiltration and immune pathways activated ( Figure 1D). Differences in human leukocyte antigen genes, immune-related genes, immune cell infiltration, and activated immune pathways related to immune subtype Furthermore, we identified differences in gene expression between Immunity_H and Immunity_L. Based on the results, 2,819 genes were upregulated in the Immunity_H group as opposed to the Immunity_L group. Compared to the Immunity_L group, 1,847 genes were downregulated in the Immunity_H group (Figure 2A). Immune surveillance is facilitated by presenting tumor-associated antigens by MHC class I complexes. Immunotherapies that target immune checkpoints benefit from this presentation. Our study examined 24 genes encoding human leukocyte antigens (HLA). A significant reduction in immunological HLA gene expression was observed in Immunity_L, suggesting that tumor cells evade immune surveillance by presenting antigens in a compromised manner ( Figure 2B). GO enrichment analysis revealed that several genes involved in Immunity_H were associated with lymphocytemediated immunity, positive cell activation regulation, immunedependent cell death, positive regulation of leukocyte activation, positive regulation of lymphocyte activation, B cell-mediated immunity, immunoglobulin-mediated immune response, and immune response-activating cell surface receptor signaling pathway, suggesting that this subtype is closely related to the immunotherapy ( Figure 2C).
Correlation analysis of antigen-presenting cells
Further analysis of hub genes was performed with TIMER. Figure 2D demonstrates the internal correlation between different immune cells. Genes involved in Immunity_H were significantly upregulated in CD8+ T Cells and M1 macrophages. Genes involved in Immunity_L were significantly upregulated in activated dendritic cells and resting mast cells. These findings indicate that the genes involved in different immune subtypes have potential immunostimulatory effects. The genes promote the processing and presentation of immune cells by antigenpresenting cells to induce tumor response (Figures 3A, B).
Construction of the prognosis-related predictive models based on immune subtype analysis
We obtained clinical information from RCC patients to further explore key genes in the candidate gene set. Then, we performed univariate and multivariate Cox and LASSO regressions to identify the genes associated with RCC. Based on univariate Cox regression results (p-value < 0.01), 2,033 genes associated with prognosis were identified. Then, we screened 27 prognosis-related genes (p-value < 0.01) by LASSO regression analysis (Figures 4D, E). Finally, we obtained the best risk score (risk score = AL161669.1 × -0.250430734267225 + BCL3 × 0.0337407354902806 + AC138649.2 × 0.401250063642218 + DBH-AS1 × 0.204477476109481 + NARF × 0.178383885858721 + AC002456.1 × 0.407989595839647 + DONSON × 0.19222775376306 + RTKN × -0.0510580913817709 + AC068338.3 × -0.720995150501497+ ZFYVE19 × -0.19065271282695 + AC069200.1 × 0.363732217837645 + TM4SF19-AS1 × 0.59294311636509 + UPK3B × 0.233561624709354) value for subsequent analysis by multivariate Cox regression. A Kaplan-Meier curve was used to analyze risk scores and divide patients into high-and low-risk groups. There was a significant difference in OS between the high-and low-risk groups for RCC ( Figure 4F). inflammatory mediators, and cancer cells, have different properties and chemical characteristics. Diagnosis, survival outcomes, and drug sensitivity strongly influence the tumor microenvironment. There was a significant difference in memory B-cell counts, CD4+ T memory cell counts, and CD4+ Th2 cell counts between the two groups with different risk scores (Figures 5A-G). There was a comparison of immune checkpoint-related gene expression levels among different risk groups. Significant differences were seen in expression levels between immune checkpoint subtypes for several genes related to immune checkpoints. Based on these results, there were differences in immune regulatory pathways between different risk groups. Immune response dysregulation can lead to different prognoses in the two groups of patients ( Figure 6A). In RCC patients, the tumor microenvironment significantly affects survival outcomes ( Figure 6B).
Relationship between prognosis-related predictive models, immune cells, immune checkpoints, immunotherapy, and tumor microenvironment
Additionally, immunotherapy using CTLA4-negative and PD1positive patients showed promise in high-risk groups (Figures 6C-F). Therefore, surgery in early RCC in conjunction with chemotherapy has proven to be a promising treatment. Furthermore, the chemotherapy sensitivity of the tumor sample was predicted using GDSC database and "pRRophetic" R package. All prognosis-related predictive models were found to have significant associations with aicar, axitinib, bicalutamide, bortezomib, bosutinib, and cisplatin sensitivity, and cytarabine, docetaxel, imatinib, lapatinib, lenalidomide, and sunitinib ( Figures 7A-L).
Functional enrichment of immune gene coexpression modules
According to the differential expression analysis and prognostic prediction model, DBH-AS1 may be a key gene in RCCs. Consequently, we performed a GSEA and GSVA enrichment analysis on DBH-AS1. GSEA enrichment analysis indicated that DBH-AS1 was significantly enriched in various pathways. GSEA enrichment analysis showed that genes were enriched in pathways involved in complement activation, complement activation regulation, blood microparticle, RNA binding involved in posttranscriptional gene silencing, immunoglobulin complex, humoral immune response regulation, B-cell-mediated immunity, humoral immune response mediated by circulating immunoglobulin, and humoral immune response. Furthermore, the specific signaling pathways associated with the DBH-AS1 were investigated, and potential molecular mechanisms involved in the RCC pathogenesis and progression were investigated ( Figures 8A, B). According to GSVA, the differential pathways between DBH-AS1 expression groups were mainly signaling pathways such as adaptive immune response, apoptotic process, biological adhesion, carbohydrate metabolic process, cell cycle, cell population proliferation, cellular response to DNA damage stimulus, central nervous system development, and cytoskeleton organization ( Figure 8C).
DBH-AS1 evaluation in the renal carcinoma cells
According to studies using CCK8, the proliferative capacity of 786-O and Caki-1 cells overexpressing DBH-AS1 was higher than that of cells transfected with a vector that was empty ( Figures 9A, B). In addition, the transwell assay of cell invasion revealed that 786-O and Caki-1 cells overexpressing DBH-AS1 showed higher invasion capacity compared with empty vector (Figure 9C).
Discussion
RCC is among the 10 most common cancers worldwide, with men being more likely to be affected (16). It is the seventh most common cancer among men, and the 12th most common cancer among women, accounting for 2.6% of all cancers in the United States (17). RCC is, therefore, a serious health threat to humans and a heavy economic burden. In addition to surgery and radiation therapy (RT), ablation therapy, chemotherapy, and emerging immunotherapies can be used to treat RCC (18). Over the past decade, there have been many changes in RCC treatment landscape as new treatments have altered the efficacy of conventional treatments efficacy (19). The immunotherapy goal is to enhance anti-tumor immune responses through immune system activation. This therapy has revolutionized cancer treatment by enhancing anti-tumor immune responses. However, immunotherapy may not have a significant effect on RCC prognosis. Although RCC is a known immunogenic disease, it can evade the immune system by downregulating human leukocyte antigen class I (20). Due to Fas ligand expression, antigen presentation becomes ineffective because T cells undergo apoptosis, and immune suppressants are secreted. Currently, a clinical trial is investigating immunotherapy efficacy in RCC patients (21).
According to ssGSEA analysis, RCC cohort patients were divided into Immunity_H and Immunity_L. Our results showed that many immune-related genes were considered DEGs after performing the differential expression analysis between Immunity_H and Immunity_L groups. Furthermore, DEGs are associated with enrichment pathways. A GO enrichment analysis revealed that several immune-related pathways, including lymphocyte-mediated immunity, B-cell-mediated immunity, immunoglobulin-mediated immune response, and immune-response-activating cell surface receptor signaling pathway, were closely related to different immune subtypes. The HLA class I molecule is involved in recognizing, presenting, and lysing tumor cells by cytotoxic T lymphocytes (CTLs), and their defects may facilitate tumor immune escape (22). Studies indicated that RCC patients treated with TKIs whose HLA class I expression is downregulated have a lower response rate and a worse prognosis. HLA class I expression correlates with tumor CTL infiltration and function (23). According to our study, the Immunity_L group had significantly reduced expression of most immune HLA genes, suggesting that impaired antigen presentation on tumor cells may contribute to cancer development.
We explored the genes correlated with RCC prognosis in further analysis. Based on 13 immune-related genes, we built a (A) Significant differences identified in the expression levels between immune checkpoint subtypes and prognostic-related genes. (B) TMB score analysis between high-and low-risk groups. (C) Immunotherapy evaluation of CTLA4 negative and PD1 negative patients between low-and high-risk groups. (D) Immunotherapy evaluation of CTLA4 negative and PD1 positive patients between low-and high-risk groups. (E) Immunotherapy evaluation of CTLA4 positive and PD1 negative patients between low-and high-risk groups. (F) Immunotherapy evaluation of CTLA4 positive and PD1 positive patients between low-and high-risk groups. *** P < 0.001; ** P < 0.01; * P < 0.05. some patients, and the molecular characteristics of TME are closely related to radiotherapy and chemotherapy. We found that DBH-AS1 may play a key role in the immunotherapy of RCC patients based on differential expression analysis and prognostic prediction models. The previous study discovered that DBH-AS1 is involved in many tumors. According to Ye et al., DBH-AS1 regulates the growth of pancreatic cancer and is a viable target for predicting gemcitabine responses in patients with pancreatic cancer (24). DBH-AS1 promotes hepatocellular carcinoma development through miR-138/FAK/Src/ERK pathways in hepatocellular carcinoma as well (25). GSEA and GSVA analyses were performed in this study to investigate potential pathways involved in DBH-AS1. Moreover, they demonstrated the association between DBH-AS1 and several immune-related pathways, including immunoglobulin complex, humoral immune response regulation, B-cell-mediated immunity, humoral immune response mediated by circulating immunoglobulin, and humoral immune response.
The present study showed that immune subtypes were significantly associated with immunotherapy drug sensitivity, including aicar, axitinib, bicalutamide, bortezomib, bosutinib, cisplatin, cytarabine, docetaxel, imatinib, lapatinib, lenalidomide, and sunitinib. Due to the restricted therapeutic alternatives and clinical benefits of chemotherapy, seeking more effective treatment methods was imperative. As tumor-specific and non-specific antigens are expressed in cancerous cells, immunotherapy represents a promising new approach to treating malignancies. In recent decades, treatment options for RCC patients have expanded rapidly, and targeted immunotherapy is now one of the most effective treatments. Several emerging drugs are currently being tested in clinical trials to boost anti-tumor immune responses. As more treatment options are available, it is essential to develop biomarkers that can help stratify patients and determine the best options for them and the treatment sequence that will overcome resistance to the treatments.
Conclusion
The immune-related genes (DBH-AS1) identified in the present study are potential targets for immunotherapy development against RCC. This study provides a theoretical basis for developing RCC immunotherapy. Furthermore, the immune subtypes can be used to explore strategies for improving immunotherapy response in RCC patients.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.
|
v2
|
2022-11-03T17:56:44.501Z
|
2022-11-03T00:00:00.000Z
|
253260449
|
s2orc/train
|
Retroperitoneal alveolar rhabdomyosarcoma intruding into spinal canal: A case report and literature review
Background Rhabdomyosarcoma (RMS) is the most frequent soft sarcoma in children and adolescents. Alveolar rhabdomyosarcoma (ARMS) is a relatively rare subtype that is characterized by aggressive behavior and an unsatisfactory prognosis. An ARMS can arise anywhere but most commonly occurs at extremity sites with a very small fraction in the retroperitoneum. The utility of 2-Deoxy-2-[fluorine-18]-fluoro-D-glucose (18F-FDG) positron emission tomography combined with computed tomography (PET/CT) remains to be established in ARMS. Case Report A 3-year-old female child was accidentally found with a large left upper abdominal mass for a day. CT examination indicated a huge soft tissue mass in the left retroperitoneum extending superiorly to the level of the left hilus renalis and inferiorly to the left acetabulum in the pelvic cavity, with intrusion into the lumbar foramens. 18F-FDG PET/CT found a mass in the left retroperitoneum from the level of T12 to the left acetabulum, with the maximum standardized uptake value (SUVmax) of about 7.0, and a CT value of about 39 HU, invading the left L3-5 intervertebral foramina and protruding into the spinal canal, with unclear boundary with the spinal cord. Retroperitoneal tumor resection and the repair operation of vascular exploration were performed. An ARMS was confirmed by postoperative biopsy, immunohistochemical staining, and genetic detection with the rupture of the fork head in rhabdomyosarcoma (FKHR). The patient received chemotherapy and was in a good condition with no recurrence and obvious complications. Conclusion Retroperitoneal ARMS is rare and indicates a poor outcome with the potential to involve vital organs and intrude into the spinal canal. Accurate diagnosis and staging using PET/CT would contribute to better risk stratifications and appropriate treatment individually.
Background: Rhabdomyosarcoma (RMS) is the most frequent soft sarcoma in children and adolescents. Alveolar rhabdomyosarcoma (ARMS) is a relatively rare subtype that is characterized by aggressive behavior and an unsatisfactory prognosis. An ARMS can arise anywhere but most commonly occurs at extremity sites with a very small fraction in the retroperitoneum. The utility of -Deoxy--[fluorine-]-fluoro-D-glucose ( F-FDG) positron emission tomography combined with computed tomography (PET/CT) remains to be established in ARMS.
Case Report: A -year-old female child was accidentally found with a large left upper abdominal mass for a day. CT examination indicated a huge soft tissue mass in the left retroperitoneum extending superiorly to the level of the left hilus renalis and inferiorly to the left acetabulum in the pelvic cavity, with intrusion into the lumbar foramens.
F-FDG PET/CT found a mass in the left retroperitoneum from the level of T to the left acetabulum, with the maximum standardized uptake value (SUV max ) of about . , and a CT value of about HU, invading the left L -intervertebral foramina and protruding into the spinal canal, with unclear boundary with the spinal cord. Retroperitoneal tumor resection and the repair operation of vascular exploration were performed. An ARMS was confirmed by postoperative biopsy, immunohistochemical staining, and genetic detection with the rupture of the fork head in rhabdomyosarcoma (FKHR). The patient received chemotherapy and was in a good condition with no recurrence and obvious complications.
Conclusion: Retroperitoneal ARMS is rare and indicates a poor outcome with the potential to involve vital organs and intrude into the spinal canal. Accurate diagnosis and staging using PET/CT would contribute to better risk stratifications and appropriate treatment individually. KEYWORDS alveolar rhabdomyosarcoma, F-FDG, PET/CT, retroperitoneum, spinal canal intrusion, case report Introduction Rhabdomyosarcoma (RMS), which is derived from primary mesenchymal cells that differentiate into skeletal muscle, is a type of high-grade rare malignant tumor with an incidence of 4.6 cases per million (1,2). RMS is usually classified into four histologic subtypes: embryonal rhabdomyosarcoma (ERMS), alveolar rhabdomyosarcoma (ARMS), pleomorphic rhabdomyosarcoma, and sclerosing/spindle cell rhabdomyosarcoma (3). An ARMS is a relatively rare subcategory with a prevalence of about 20-25% of all RMS, compared with ERMS of 50-60% (4). It is more aggressive and has a more unsatisfactory prognosis than other subtypes (5). The ARMS most commonly occurs at extremity sites and sometimes in the head and/or neck or torso, while ERMS typically occurs in the head and neck (6). However, retroperitoneal ARMS is extremely rare and has poor outcomes accompanied by the large tumor volume and frequent involvement of vital organs, leading to difficult resection when it is diagnosed (7). Due to the specific anatomical location, retroperitoneal, especially paraspinal, ARMS shows the potential to infiltrate the vertebrae and protrude into the spinal canal (8,9).
2-Deoxy-2-[fluorine-18]-fluoro-D-glucose ( 18 F-FDG) positron emission tomography combined with computed tomography (PET/CT) is an imaging technology illustrating detailed metabolic and functional molecular information and the precise anatomical region of the lesion (10). It is used in evaluating different tumors and represents an extremely promising investigation method for the diagnosis, staging, and prognosis of rhabdomyosarcoma. Herein, we described a rare case with retroperitoneal ARMS invading the intervertebral foramen and intruding into the spinal canal, and we reviewed the available literature on retroperitoneal ARMS. Contrast-enhanced CT examination indicated a huge soft tissue mass, measuring 12.3 × 7.4 cm at the larger section, in the left retroperitoneum extending superiorly to the level of the left hilus renalis and inferiorly to the left acetabulum in the pelvic cavity, with intrusion into the lumbar foramens, compression and squeezing of the left kidney and left abdominal bowel, local compression and stenosis of the inferior vena cava, and the abdominal aorta was displaced to the right side by compression ( Figure 2). The tissue mass showed mottling calcification, cystic degeneration, and necrosis. The patient was injected with 0.1 mCi/kg of 18 F-FDG after 6 h of fasting, and PET/CT images were acquired 60 min later. The tissue mass with a high FDG uptake was found using 18 Figure 4C). The final diagnosis was confirmed as ARMS with TNM stage 3 and IRS stage III, and the level of risk was high. Then, the patient received chemotherapy comprising vincristine, dactinomycin, and cyclophosphamide (VAC therapy) with supportive treatments to alleviate adverse reactions to chemotherapy. A combination of vincristine and irinotecan (VI) was administered 1 month later, and VAC was repeated 2 months later. The patient has survived for 9 months since the initial diagnosis with no recurrence and obvious complications.
Discussion
Rhabdomyosarcoma is the most common extracranial solid tumor in the pediatric population after neuroblastoma and Wilms tumor (8). A total of 63% of RMS occur in children under 10 years old, and the incidence rate reaches the peak between 2 and 5 years old (11). ARMS is a relatively rare subtype with a poor prognosis, of which the estimated 10-year overall survival (OS) rate is 29.4% compared with 52.1% in the nonalveolar type (12). The retroperitoneum is not the typical primary site of RMS, with ∼8% of the incidence of all sites (13). Approximately, 20% of the patients present with distant metastatic disease at diagnosis, and lung, bone, and bone marrow are the commonly involved sites (6,14).
We searched the literature in the PubMed database from 1997 to 2022 using the keywords containing alveolar rhabdomyosarcoma and retroperitoneal. In total, 5 available case reports were found. We summarize the case reports in Table 1 (7,8,(15)(16)(17). Of these cases, male children seem to be the more vulnerable group, which is consistent with overall incidence . /fmed. . (male:female ratio of 1.51, 95% confidence interval (CI): 1.27-1.80) (18). Clinical symptoms are variable, depending on the size and location of the retroperitoneal mass and the presence or absence of distant metastases. The tumor squeezing certain organs can cause relevant performance, and the huge mass often invades vital organs such as the kidneys, aorta, inferior vena cava, and bilateral iliac vessels. However, patients can sometimes present with a mass and no notable symptom similar to the case we described here, causing a delay in diagnosis. It should be noticed that retroperitoneal ARMS appeared to have the potential to infiltrate the vertebrae and extend into the spinal canal. This could provide opportunities for tumors to invade the intervertebral foramen and spread to the spinal cord and even the brain as reported in the case study 2 (8). Thus, it argues for more attention on the careful and precise detection of the tumor intruding into the spinal canal and metastatic to the central nervous system. Alveolar rhabdomyosarcoma recurrently, of ∼80%, harbors chromosomal translocations including a t(2;13)(q35;q14) or a t(1;13)(p36;q14), which can generate fusion genes PAX3and PAX7-FOXO1 respectively. And proteins produced by these fusion genes can function as oncoproteins promoting the proliferation and apoptosis of tumor cells (6,19). The diagnosis of ARMS requires histology and molecular pathology studies of the tumor tissue (6). ARMS is typically composed of densely packed, small, and round cells aggregating in areas at the edges of fibrous septa forming structures such as pulmonary alveoli (1,20). Immunohistochemical markers include myogenic markers such as MyoD (myf3) or myogenin (myf4), myosin, myoglobin, muscle-specific actin, or desmin (21). Medical imaging provides noninvasive methods which are essential for the evaluation of patients with ARMS. The sonographic feature of ARMS is substantive hypoechoic or complex-echoic mass. CDFI shows rich and disorderly color blood flow signals within the mass. It often appears as an equal or a slightly low-density mass in plain CT, with unclear borders. The tumors usually grow rapidly, and necrosis, as well as cystic degeneration, can be seen in the lesions as a result of the insufficient blood supply. Enhanced CT scan shows heterogeneous enhancement and sometimes rim-like enhancement. Areas without enhancement are tissues with necrosis and cystic degeneration. Hemorrhage and calcification occur rarely, but mottling calcification was observed in this case. Moreover, CT can detect adjacent bone involvement but ARMS frequently destroy the bone. PET/CT reveals increased glucose metabolism of ARMS. As an advanced technology, PET/CT could provide more information about the lesions than conventional imaging detection methods. 18 F-FDG PET/CT imaging is useful for initial assessment, monitoring treatment response, and detection of recurrences with better accuracy for identifying primary sites, lymphatic involvement, and distant metastases (22, 23). Local lymph node metastasis has been considered a strong prognostic factor, calling for an emphasis on desirable detection modalities of lymphatic involvement (24). Compared with conventional imaging techniques, such as ultrasound, CT, and magnetic resonance imaging (MRI), PET/CT performs better in detecting lymph nodal metastasis with higher sensitivity and specificity (25). 18 F-FDG PET/CT can estimate the function and nature of nodes through the level of glucose metabolism in tissues and can help with accurate localization of the involved lymph nodes. In a prospective study by Völker et al. (25), the detection of involved lymph nodes using 18 F-FDG PET/CT reached a sensitivity of 93%, whereas conventional imaging modalities were only 36%. Ricard et al. (26) reported that 18 F-FDG PET/CT found 19 involved lymph nodes in 4 patients vs. 12 nodes by MRI and CT, and therefore, the results of PET/CT led to alteration of the lymph node staging and treatment strategies in some patients. Our case also observed similar advantages of PET/CT for discovering retroperitoneal lymphatic metastases, whereas negative in ultrasound and CT tests. The more accurate staging of regional lymph node involvement will benefit risk stratification and treatment decisions in patients with RMS. PET/CT also shows some potential superiorities in finding tumor invasion into the spinal canal. When evaluating the spinal canal involvement, all background tissues, including paraspinal musculature, vertebrae, spinal cord, nerve roots, and CSF, demonstrate relatively low metabolic activity using 18 F-FDG PET/CT, thus making it possible for differentiation between normal tissues and lesions (27,28). PET/CT allows for the identification of soft-tissue involvement such as neural foramen invasion and epidural extension of tumor in malignant involvement of the spine (29). However, PET/CT is inferior to MRI when used to detect spinal cord involvement. In recent years, the integrated PET and MR (PET/MR) imaging modality has been rapidly developed with the combined superiorities of quantification of radioactive tracer metabolism provided by PET and outstanding soft tissue contrast by MR (30). The value of PET/MR in clinical applications remains to be established, and we hope this innovative technology will provide more accurate diagnosis and ultimately improve patient prognosis. In addition, a study illustrated that metabolic parameters obtained from baseline PET/CT were potential to select patients sensitive to treatment (31). Features of patients including unfavorable sites of the primary tumor, older patient age at initial presentation, the alveolar subtype, and regional lymph node involvement are considered to be poor prognostic factors for RMS (5,32). Unfavorable sites include the prostate and bladder, cranial parameningeal sites, extremities, trunk, retroperitoneum, and other sites (13). Moreover, 18 F-FDG PET/CT may be an added prognostic predictor in RMS. High SUV max value is more prevalent among patients with less favorable features including unfavorable primary sites, alveolar pathology, and high-risk group (33). A study found that during diagnosis, patients with SUV max of <9 had an improved 3-year progression-free survival (62% of patients with SUV max of <9 vs. 39% of patients with SUV max of ≥9, p = 0.02) (34). In our case study, the SUV max of 7.0 might be associated with the patient's favorable prognosis.
Retroperitoneal ARMS should be differentiated from neuroblastoma. Neuroblastoma is the most common extracranial solid tumor in the pediatric population and almost 70% of the patients have abdominal neuroblastoma (35,36). Furthermore, neuroblastoma may arise from paraganglia and is likely to protrude into the spinal canal through the neural foramina. In addition to symptoms caused by the abdominal mass such as abdominal pain and fullness, patients usually present with elevated levels of vanillylmandelic acid (VMA) and homovanillic acid (HVA). The characteristic PET/CT findings of neuroblastoma in children include large size, mixed density with calcification, necrosis, and cystic degeneration, which specifically show peripheral hypermetabolic areas and central hypometabolic areas in the tumor, indicating central necrotic and cystic lesions. PET/CT is superior at revealing lymph nodes and distant organ metastases, which can provide an objective imaging basis for preoperative diagnosis and accurate staging of neuroblastoma in children, over conventional imaging. Calcification could be seen in almost 90% of cases, appearing as sandy, spotted, and mass shapes, which is a characteristic manifestation of neuroblastoma (37). 123 Overall, the cure rate of RMS could be increased with the improvements in risk stratifications and multimodal treatment including surgery, chemotherapy, and radiotherapy (40,41). As for retroperitoneal RMS, surgery, especially radical resection, is the principal choice, with a longer median OS than palliative surgery and conservative treatment (18 vs. 6 months) (9).
Conclusion
Retroperitoneal ARMS is relatively rare and characterized by its unfavorable outcome with the potential to involve vital organs and intrude into the spinal canal, and even spread to CNS. It should be noted that the retroperitoneal mass may be misdiagnosed as neuroblastoma and a biopsy is necessary for the final diagnosis. Accurate staging using PET/CT would contribute to better risk stratifications and appropriate treatment individually.
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
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 from the patients/participants legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements. The study was approved by the Institutional Review Board at the First Affiliated Hospital of Zhengzhou University and Peking University First Hospital.
|
v2
|
2022-11-03T18:00:59.107Z
|
2022-11-03T00:00:00.000Z
|
253259421
|
s2orc/train
|
Reduced expression of FRG1 facilitates breast cancer progression via GM-CSF/MEK-ERK axis by abating FRG1 mediated transcriptional repression of GM-CSF
Multiple molecular subtypes and distinct clinical outcomes in breast cancer, necessitate specific therapy. Moreover, despite the improvements in breast cancer therapy, it remains the fifth cause of cancer-related deaths, indicating the involvement of unknown genes. To identify novel contributors and molecular subtype independent therapeutic options, we report reduced expression of FRG1 in breast cancer patients, which regulates GM-CSF expression via direct binding to its promoter. Reduction in FRG1 expression enhanced EMT and increased cell proliferation, migration, and invasion, in breast cancer cell lines. Loss of FRG1 increased GM-CSF levels which activated MEK/ERK axis and prevented apoptosis by inhibiting p53 in an ERK-dependent manner. FRG1 depletion in the mouse model increased tumor volume, phospho-ERK, and EMT marker levels. The therapeutic potential of anti-GM-CSF therapy was evident by reduced tumor size, when tumors with decreased FRG1 were treated with anti-GM-CSF mAb. We found an inverse expression pattern of FRG1 and phospho-ERK levels in breast cancer patient tissues, corroborating the in vitro and mouse model-based findings. Our findings first time elucidate the role of FRG1 as a metastatic suppressor of breast cancer by regulating the GM-CSF/MEK-ERK axis.
INTRODUCTION
Breast cancer remains at the top of cancer-related deaths among women, with a mortality rate of 6.9% [1]. Epithelial to mesenchymal (EMT) transition, which comprises the detachment of cancer cells from their primary origin, intravasation, and formation of metastatic growth at distant sites, accounts for majority of the deaths associated with breast cancer [2]. Even after the significant progress in primary breast cancer therapy, metastasis and recurrence are the major cause of reduced survival [3]. Finding out the role and mechanism of additional genes which affect EMT can provide additional opportunities to improve the survival rate.
FSHD region gene 1 (FRG1), which is mainly known for being the candidate gene for facioscapulohumeral muscular dystrophy (FSHD), has recently shown its potential as a tumor suppressor gene. Our earlier study first time reported the correlation of FRG1 downregulation with oral cavity, colorectal and gastric carcinoma [4]. Depletion of FRG1 levels increased cancerous properties of prostate cancer cell lines via activation of the p38-MAPK [5]. Mechanistically, reports have suggested FRG1's involvement in pre-mRNA processing [6], F-actin-bundling [7], and angiogenesis [8]. The role of FRG1 in breast cancer is mostly unexplored. Expression profiling of the genes associated with triple-negative breast cancer (TNBC) using the MDA-MB-231 cell line showed FRG1 as one of the significantly downregulated genes which were connected with the migratory potential of breast cancer cells [9]. However, the exact biological function of FRG1 in breast cancer and its mechanism are entirely unknown.
So far, many cytokines such as IL6/8/10, TNFα, IFNγ, growth factors TGFβ, bFGF, and VEGF have been reported to involve with EMT in breast cancer [10]. Granulocyte-macrophage colonystimulation factor (GM-CSF), a potent hematopoietic growth factor mainly known for its immunomodulatory function in the tumor niche, has recently been reported to play a role in EMT in colon cancer [11]. Elevated expression of GM-CSF is clinically correlated with advanced histological grade, metastasis, and poor prognosis in patients with prostate cancer, breast cancer [12], and pancreatic ductal carcinoma [13]. Nevertheless, very little is known about the upstream regulation and downstream signaling mechanism coordinating GM-CSF-mediated metastatic colonization. Previously we found enhanced GM-CSF expression in FRG1depleted prostate cancer cells [5].
In the current study, we first report the FRG1-mediated regulation of GM-CSF in breast carcinoma. Prior to this, no information was available on the detailed mechanism of how GM-CSF promotes EMT in any cancer. Here we have shown that FRG1 binds to the GM-CSF promoter and inhibits its expression. The loss of FRG1 resulted in increased cell proliferation, migration, and invasion triggered by increased levels of GM-CSF and the activation of the MEK/ERK pathway, in both the cell lines of luminal (ER+; MCF7) and basal (TNBC; MDA-MB-231) origin. We have validated our in vitro findings in breast cancer patient tissues and mouse models. We have also shown the therapeutic potential of anti-GM-CSF antibody in the mouse model. Overall, here we report the role and molecular mechanism of a new gene, FRG1, in breast cancer, which has the potential to be explored as a therapeutic target irrespective of molecular subtypes.
RESULT FRG1 affects the tumorigenic properties of breast cancer cells
We prepared stable lines with FRG1 expression perturbation in estrogen receptor-positive (ER+) cells MCF7 which has moderate endogenous expression of FRG1 ( Supplementary Fig. S1A). The reduction of FRG1 expression in MCF7 increased the proliferation rate in MTS and colony formation assays compared to the control group (Fig. 1A, B). Correspondingly, increased FRG1 expression decreased the rate of proliferation in both assays (Fig. 1E, F). To explore the metastatic potential, we checked the effect of FRG1 expression on cell migration. The wound-healing assay showed enhanced cell migration due to reduced FRG1 expression (Fig. 1C). Matrigel invasion assay corroborated these findings (Fig. 1D). We observed the opposite effect on wound healing and invasion of MCF7 cells with ectopic expression of FRG1 (Fig. 1G, H). Besides, we generated FRG1 knockout MCF7 cells (FRG1_KO) and checked its effect on the tumorigenic properties ( Supplementary Fig. S1B, C). In accordance with our findings in the FRG1 knockdown group, we also observed increased cell proliferation ( Supplementary Fig. S1B) and migration ( Supplementary Fig. S1C) due to FRG1 knockout.
To determine if the effect of FRG1 is molecular subtype-specific, we ectopically expressed FRG1 in TNBC cell line MDA-MB-231, Fig. 1 Effect of FRG1 alteration on tumorigenic properties of breast cancer cells of different molecular subtypes. The levels of FRG1 were modulated in two breast cancer cell lines of different molecular subtypes and subjected to various cell-based assays; A, B Representative images of MTS assay (A) and colony formation assay (B) showing the difference in proliferative property of MCF7 cells due to FRG1 knockdown (FRG1_KD vs. Control_Sc). OD values were taken at 490 nm at 24 h. Bar diagrams show the difference between the two groups. C Representative images of scratch wound-healing assay in MCF7 cells with depleted FRG1 (FRG1_KD) and corresponding control (Control_Sc). The bar graph depicts the difference in the percentage of wound closure between the two groups. Scale bar, 100 μm. D Representative images of matrigel transwell invasion assay in MCF7 cells with reduced FRG1 level (FRG1_KD) and respective control (Control_Sc). The bar graph depicts the difference in the number of invaded cells between the two groups. Scale bar, 50 μm. E, F Representative images of MTS assay (E) and colony formation assay (F) in MCF7 cells with an elevated level of FRG1 (MCF7_Ex) and control (Control_Ev). Bar graph showing the difference between the two groups. G Representative images of scratch wound-healing assay in MCF7_Ex and Control_Ev. Bar graph showing the difference in wound closure between the two groups. Scale bar, 100 μm. H Representative images of matrigel transwell invasion assay in MCF7_Ex and Control_Ev. The difference in invaded cells between the two groups is shown in the bar graph. Scale bar, 50 μm. I, J Representative images of MTS assay (I) and colony formation assay (J) in MDA-MB-231 cells with elevated expression of FRG1 (FRG1_Ex) and their control (Control_Ev). Bar graphs indicate the difference between the two groups. K Representative images showing scratch wound-healing assay performed in MDA-MB-231_FRG1_Ex and Control_Ev. The bar diagram shows the difference in wound healing percentage between the two groups. Scale bar, 100 μm. L Representative images of transwell invasion assay, performed in MDA-MB-231_Ex and Control_Ev. The bar diagram shows the difference in invaded cells between the FRG1_Ex and Control_Ev. Scale bar, 50 μm. Experiments were performed in triplicate, two-tailed unpaired student's t-test was used to compare the two groups' differences. Results are presented as mean ± SD. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001. which has low endogenous FRG1 expression ( Supplementary Fig. S1A). In agreement with our observation in MCF7_FRG1_Ex cells, we observed a reduction in cell proliferation (Fig. 1I, J), migration (Fig. 1K), and invasion (Fig. 1L). Together these data suggest FRG1 expression can modulate the tumorigenic properties of breast cancer cell lines.
Reduction of FRG1 activates MEK/ERK signaling Intrigued by the effect of FRG1 expression level on the tumorigenic properties of breast cancer cells, we explored its effect on ERK and AKT, two frequently altered signaling pathways in cancers [14]. We observed that the reduced FRG1 level led to the activation of ERK and its upstream molecule MEK in MCF7 cells without the alteration in total-ERK or total-MEK levels ( Fig. 2A). A similar trend was observed in FRG1 knockout MCF7 cells, where knockout of FRG1 led to increased expression of phospho-ERK ( Supplementary Fig. S1D). On the other hand, ectopic expression of FRG1 decreased the activation of ERK and MEK in MCF7 (Fig. 2B) and MDA-MB-231 cells (Fig. 2C). As further corroboration of our findings, ectopic expression of FRG1 in MCF7_FRG1_KD cells nullified the change in phospho-ERK level (Fig. 2D).
Unexpectedly, we found that FRG1 depletion reduced AKT activation at the 308 position, and there was no effect at the 473 position in MCF7 cells ( Supplementary Fig. S2A, B). Increased FRG1 expression also affected the activation of AKT only at the 308 position, in the opposite manner ( Supplementary Fig. S2C, D). Previous studies have reported that activation in ERK signaling can suppress the AKT pathway [15]. We inhibited the ERK pathway in MCF7_FRG1_KD cells and found the reversal of the phospho-AKT 308 suppression ( Supplementary Fig. S2E), which confirmed that inhibition of AKT activation was ERK-mediated. As expected, no effect of ERK inhibition was found on the level of phospho-AKT 473 ( Supplementary Fig. S2E). This data suggests that reduced FRG1 expression might be involved in breast cancer progression by the activation of the ERK pathway.
Depletion of FRG1 suppresses apoptosis by ERK-mediated inhibition of p53
Loss of apoptotic control is a hallmark of cancer which permits cancer cells to survive longer and gives more time to accumulate the mutations [16]. Reduction in FRG1 level in MCF7 significantly decreased the downstream effector caspase 3/7 level (Fig. 3A). We made a coherent observation in a flow cytometry-based analysis of Annexin V/ propidium-iodide where FRG1 depletion reduced apoptosis in MCF7 cells (Fig. 3B). Mechanistically, we found that depletion of FRG1 reduced phospho-p53 level (Fig. 3C). No change in phospho-p38 level indicates reduced apoptosis in FRG1_KD cells may be independent of phospho-p38. (Fig. 3C). . Experiments were performed in triplicates. The intensity of the bands in each blot was measured by ImageJ software and normalized to GAPDH. Two-tailed unpaired student's t-test was used to compare the difference between the two groups. Results are presented as mean ± SD. ns, P > 0.05, *P ≤ 0.05; ***P ≤ 0.001.
Several studies have reported ERK-mediated inhibition of p53 [17,18]. To test this, we inhibited ERK and checked the levels of phospho-p53 in FRG1-depleted MCF7 cells. We observed restoration of phospho-p53 level in MCF7_FRG1_KD group treated with ERK inhibitor (Fig. 3D).
Overall, our findings suggest that a reduced level of FRG1 leads to an ERK-mediated decrease of phospho-p53, which reduces apoptosis.
FRG1 depletion enhances the expression of EMT markers by activating MEK/ERK pathway It is well established that EMT increases cell migration, leading to metastasis in cancer [19]. As reduced FRG1 level increased the migration in cell-based assays, we checked its effect on EMT markers snail, slug, and twist, which have long been reported to trigger EMT through ERK signaling [20][21][22]. As hypothesized, we detected significant upregulation of snail, slug, and twist in MCF7_FRG1_KD cells (Fig. 4A). To validate, we inactivated the ERK pathway in MCF7_FRG1_KD cells and found abrogation of the ERK-mediated upregulation of EMT marker snail (Fig. 4B). Scratch wound-healing assay also revealed a similar effect (Fig. 4C). As expected, ectopic expression of FRG1 in MDA-MB-231 cells reduced the expression of EMT markers (Fig. 4D).
To further substantiate our findings, we treated MDA-MB-231_FRG1_Ex cells with ERK activator Ceramide, which restored the reduced level of phospho-ERK caused by ectopic expression of FRG1 (Fig. 4E). The same experimental setup also rescued snail expression levels ( Fig. 4E) and cell migration (Fig. 4F). These findings also suggest that the effect of FRG1 on EMT is consistent in both the breast cancer cell lines. Our previous work [5] found that FRG1 perturbation changed the levels of inflammatory cytokines CXCL1, CXCL8, GM-CSF, PDGFα, and PDGFβ, which have been reported to affect EMT [11,23,24]. We found a significant increase in CXCL1, CXCL8, GM-CSF, PDGFα, and PDGFβ transcript levels due to FRG1 depletion ( Fig. 4G) in MCF7 cells, the opposite effect was observed with the ectopic expression of FRG1 in MDA-MB-231 cells (Fig. 4I). Further analysis of the effect of altered FRG1 levels on GM-CSF concentration in the conditioned media harvested from MCF7 and MDA-MB-231 cells, confirmed our findings at protein levels ( Fig. 4H, J). To ascertain the specificity of the observation, we inhibited the ERK pathway in MCF7_FRG1_KD cells, which counterbalanced the increase in CXCL1, CXCL8, PDGFα, and PDGFβ transcript levels caused by FRG1 depletion, suggesting that the effect was downstream of ERK (Supplementary Fig. S2F).
To elucidate if CXCL1 and CXCL8 were responsible for ERK activation and downstream change in EMT markers [25] we inhibited their receptor CXCR2 by using CXCR2 antagonist Cpd 19 [26]. We found no effect in phospho-ERK and snail levels ( Fig. 4K), which suggests ERK activation is not downstream of CXCR2-CXCL1/8. Interestingly, we found that ERK pathway inhibition in MCF7_FRG1_KD could not nullify the increase in GM-CSF level caused by FRG1 reduction (Supplementary Fig. S2F), which implies that the effect on GM-CSF is upstream of ERK. The effect of FRG1 abrogation in apoptosis was assessed by caspase 3/7 assay and flow cytometry. The expression of major signaling molecules that control apoptosis was examined by immunoblotting. A Graphical representation depicts a change in luminescence level corresponding to caspase 3/7 in MCF7 cells with reduced FRG1 (FRG1_KD) and its control (Control_Sc) (n = 3). B Cell apoptosis was analyzed by flow cytometry using 488 nm excitation and 647 nm emission filters in MCF7_FRG1_KD and Control_Sc groups. The bar diagram shows the level of apoptotic cells in the MCF7_FRG1_KD and Control_Sc groups. (n = 1) C Effect of FRG1 expression depletion on p38 and p53 levels was assessed by immunoblotting in MCF7 cells (FRG1_KD vs. Control_Sc) (n = 3). D MCF7_FRG1_KD cells were treated with 10 μM of ERK inhibitor FR180204 (FRG1_KD + FR180204) and its solvent (FRG1_KD + DMSO) for 2 h; rescue in the activated ERK, and p53 levels was measured by Western blot (n = 2). The band intensity of each blot was measured by ImageJ and normalized to internal control GAPDH. In A, C, D, values are shown as mean ± SD. Two-tailed unpaired student's t-test was used to compare the difference between the two groups. ns, P > 0.05, *P ≤ 0.05.
We propose that reduced FRG1 leads to ERK-mediated upregulation of cytokines except for GM-CSF, and FRG1 could act upstream of GM-CSF.
Effect of FRG1 on ERK is GM-CSF mediated
So far, the role of GM-CSF in breast cancer has not been fully understood. We checked the effect of GM-CSF levels in MCF7 cells and found that GM-CSF enhanced the tumorigenic properties of MCF7 cells by upregulating cell proliferation and migration ( Supplementary Fig. S3A, B). Also, ectopic administration of GM-CSF upregulated the expression of phospho-ERK and snail ( Supplementary Fig. S3C). In order to confirm that GM-CSF is downstream of FRG1 and upstream of ERK, we perturbed GM-CSF levels in the cells with altered expression of FRG1. GM-CSF inhibition resulted in the downregulation of phospho-ERK and snail in MCF7_FRG1_KD cells (Fig. 5A), which eventually reduced cell migration in scratch wound-healing assay (Fig. 5B). Correspondingly, the opposite effect was observed in MCF7_FRG1_Ex cells upon treatment with exogenous human recombinant GM-CSF (Fig. 5C, D). These results suggest that the effect of FRG1 on the ERK pathway might be GM-CSF mediated.
FRG1 binds to the GM-CSF promoter and regulates its expression
Based on the observations above, we hypothesized that FRG1 might be a direct transcriptional regulator of GM-CSF. Our previous work has shown "CTGGG" as a binding site for FRG1. We found six "CTGGG" sequences in the GM-CSF promoter region within 907 bp upstream of the transcription start site (Fig. 5E). Luciferase reporter assay using the promoter of GM-CSF (907 bp upstream of transcription start site in pGL4.23) revealed increased luciferase activity in HEK 293 T cells compared to the empty vector control (Fig. 5F). On the contrary, no change in luciferase activity was observed in HEK 293 T with FRG1 depletion, between the GM-CSF promoter and the control groups (Fig. 5F). When MDA-MB-231 cells with increased FRG1 expression levels were transfected with the same constructs, we found a substantial decrease in luciferase activity in cells containing GM-CSF promoter than control, while no change was observed in relative luciferase intensity in MDA-MB-231 (Control_Ev) cells with basal FRG1 expression between the GM-CSF promoter and the control groups (Fig. 5G). This observation strengthened that FRG1 more likely possessed an inhibitory effect on GM-CSF expression, which agrees with our previous observations. Additionally, to confirm the binding of FRG1 on the GM-CSF promoter, a ChIP assay was performed in HEK 293 T cells. As shown in Fig. 5H, enrichment of FRG1 protein on GM-CSF promoter fragment was found after immunoprecipitation with anti FRG1 antibody but not by the negative control IgG. This approves our hypothesis that FRG1 binds to the GM-CSF promoter. To further validate, a competitive EMSA on labeled oligos was carried out with an increased amount of unlabeled oligos. The result showed that oligos were sufficient to compete for the binding. Thereby drastic reduction in the intensity of the shift was observed. Furthermore, when the binding complex was subjected to FRG1 specific antibody, a shift was noted, indicating the binding of FRG1 to the oligos (Fig. 5I). These results confirm in vitro binding of FRG1 on the CTGGG site of GM-CSF promoter.
Therefore, our data strongly support the notion that FRG1 hinders EMT progression in breast cancer by inhibiting GM-CSFmediated ERK activation.
Reduced expression of FRG1 in patient samples is inversely correlated with phospho-ERK expression GEPIA-based analysis revealed lower expression of FRG1 transcripts in cancer patients (n = 1085) compared to normal (n = 291) (Fig. 6A) [27]. GEPIA also depicted that the patient group with a high level of FRG1 had a higher probability of disease-free survival (Fig. 6B). Kaplan-Meier plotter-based survival analysis also showed that breast cancer patients (containing wild-type p53) with a high level of FRG1 had a higher probability of recurrence-free survival ( Supplementary Fig. S4A) [28].
Estrogen signaling is crucial for breast carcinogenesis as it profoundly contributes to the proliferation of breast cancer cells. So, we checked whether the loss of FRG1 activated the ER signaling. To find out the correlation, we segregated breast cancer patient tissue samples according to their ER/PR/HER2 status and checked the expression of FRG1. We did not see any significant difference in FRG1 expression levels between ER + breast cancer tissues and TNBC patients (median FRG1 AS: 6 versus 5) (Fig. 6F). In MCF7 cells, we found upregulation in phospho-ER due to FRG1 depletion ( Supplementary Fig. S4B), and precisely the opposite in the case of ectopic expression of FRG1 in MCF7 cells (Supplementary Fig. S4C). To examine whether ER activation was Two-tailed unpaired student's t-test was used to compare the difference between the two groups. ns, P > 0.05, *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001.
concomitant to ERK activation, we treated FRG1_KD MCF7 cells with ERK inhibitor FR180204 and found a significant reduction in phospho-ER levels. This suggests an ERK-mediated upregulation of ER in FRG1-depleted MCF7 cells (Supplementary Fig. S4D).
Taken together, our findings imply an inverse relation between FRG1 and ERK activation in breast cancer patients. As a downstream effect, FRG1 can affect ER activation in ERKdependent manner.
Loss of FRG1 promotes tumor development and metastasis in vivo via regulating GM-CSF/ERK
To demonstrate the impact of FRG1 expression levels in vivo, we developed an orthotopic mice model by implanting 4T1 cells with depleted and elevated levels of FRG1, in the mammary fat pad of BALB/c mice. FRG1 knockdown significantly increased tumor volume and weight ( Fig. 7A and Supplementary Fig. S5A). Correspondingly, ectopic expression of FRG1 significantly reduced tumor volume and weight ( Fig. 7B and Supplementary Fig. S5B). Parallel to our cell line-based data, we found a reduced expression of phospho-ERK and snail in the FRG1 over-expression group (Fig. 7C). To assess its metastasis potential, we injected FRG1depleted 4T1 cells in the tail vein of BALB/c mice. We observed more metastatic nodules in the lungs of the FRG1_KD group than in control (Fig. 7D). An opposite trend was identified in the set with FRG1 over-expression (Fig. 7E). These results corroborate our in vitro observations in the mouse model.
Anti-GM-CSF treatment abrogates FRG1 depletion-based tumors growth in mice
To investigate the therapeutic potential of anti-GM-CSF therapy in breast cancer patients with reduced FRG1 expression, we did a mouse model-based study. We injected 4T1_FRG1_KD and Control_Sc cells subcutaneously into mice (n = 4). After the tumor reached a palpable size (7 days), one set of 4T1_FRG1_KD mice (n = 4) was intraperitoneally administrated with anti-GM-CSF neutralizing monoclonal antibody (mAb) (10 mg/kg body weight), and the other set of 4T1_FRG1_KD mice (n = 4) was injected with control IgG antibody (10 mg/kg body weight) alternative days till day 21. We found a significant reduction in tumor size in the set treated with GM-CSF mAb compared to IgG (Fig. 7F and Supplementary Fig. S5C). In support of our in vitro data, we also observed a reduction in phospho-ERK and snail in GM-CSF treated group (Fig. 7G). Henceforth our in vivo findings establish the role of FRG1-mediated regulation of GM-CSF. It also indicates that loss of FRG1, disrupts the suppression of GM-CSF, which might promote the proliferation and EMT by activating ERK and its downstream targets.
DISCUSSION
Metastatic dissemination in breast cancer accounts for 90% of cancer-related deaths [29]. Therefore, elucidating the molecular mechanism underlying the metastatic process has attracted considerable attention from researchers. Loss of function of tumor suppressor genes is often associated with increased metastasis and poor patient survival. In the case of breast cancer, different molecular subtypes represent discrete clinical outcomes. Hence, the search for novel targets that can act independently of molecular subtypes, will contribute to the development of more effective therapy.
Since the discovery of the FRG1 in 1996, most of the reports have highlighted its involvement in FSHD pathophysiology, muscle development, actin-bundling, and angiogenesis. Very few reports have shown an indirect association of FRG1 with cancer. Whole-exome sequencing identified some deleterious mutations in the FRG1 gene in calcifying fibrous tumor of the pleura [30]. Another whole-exome sequencing study, done in six follicular thyroid cell lines, reported a mutation in the FRG1 gene in cancer cell lines [31]. Our group showed FRG1-mediated activation of the p38-MAPK pathway in prostate cancer [5]. In the current study, our findings first-ever report the inverse association of FRG1 expression levels with breast cancer and unravel the underlying molecular mechanism using multiple model systems. Activation of ERK plays a pivotal role in cell proliferation, angiogenesis, and malignant transformation [32]. It supports tumor growth and cancer progression by upregulating various EMT-inducing factors [33]. TNBC, which is considered more aggressive and therapeutically challenging [34], is often associated with shorter patient survival with high expression of ERK [35]. The cross-talk between ERK and ERα signaling in luminal carcinoma may lead the cells toward chemoresistance [36]. In our case, we have found that reduced FRG1 expression led to the activation of ERK in both the breast cancer cell lines. We have further observed ERK-mediated elevated expression of the EMT markers snail, slug, and twist, which is consistent both in vitro and in vivo. Our findings may create enormous scope for research to develop novel therapeutics that can target upstream regulators of ERK and act irrespective of molecular subtypes of breast cancer.
The cross-talk between ERK and AKT molecules is an important determinant of the cell fate towards survival or apoptosis [18]. Surprisingly, although there was no change in AKT 473 activation, the depletion of FRG1 reduced the activation of AKT 308. Previous studies have shown ERK-mediated suppression of AKT activation [37], as was the case in our study, where the suppression of AKT 308 activation was rescued upon ERK inhibition in MCF7 cells. This observation supports the previous finding that MEK downregulation decreases AKT activation in EGFR and HER2-driven breast cancer [15]. Also, the activation of AKT was found to protect HeLa cells from apoptosis [38]. In this context, several reports suggest multiple mechanisms for MEK/ ERK-mediated downregulation of AKT activation [15,38,39]. ERK has been shown to negatively regulate Gab1-PI3K binding, reducing the downstream AKT signaling [39], [15]. Also, inhibition of ERK, reduces EGFR phosphorylation, resulting in augmented AKT phosphorylation [40].
In addition to AKT-P38-P53 mediated activation of apoptosis, several studies have reported ERK-mediated regulation of p53 and apoptosis [41]. Inactivation of ERK promoted translocation of apoptosis-inducing factor in the nucleus, thus, causing apoptosis [18]. Another report suggests ERK promotes increased MDM2 expression and thus promotes the degradation of p53 [17]. We found reduced expression of phospho-p53 in FRG1-depleted MCF7 cells, which was rescued by the inhibition of ERK. This is in parallel to the literature that supports the MEK/ERK pathway can downregulate the activation of p53 and promote cell survival [18]. We put forward our hypothesis that FRG1 depletion activates the ERK pathway, which attenuates AKT and p53 phosphorylation that results in ERK-dependent inhibition of the apoptotic pathway. Fig. 5 The effect of FRG1 on ERK activation and EMT markers is mediated by direct binding to the GM-CSF promoter and controlling its expression. A-D MCF7 cells with reduced FRG1 (FRG1_KD) and elevated FRG1 (FRG1_Ex) levels were treated with anti-GM-CSF antibody and human recombinant GM-CSF (hGM-CSF), respectively. The effect of these modulations was assessed by immunoblotting and wound-healing assay. A pERK, and snail expression was measured by Western blot in MCF7 cells with FRG1 knockdown treated with 2 mg/ml of anti-GM-CSF antibody (FRG1_KD + GM-CSF Ab) and control IgG (FRG1_KD + IgG) for an hour, along with untreated Control_Sc. (n = 3). B Representative images show cell migration in wound-healing assay in the same set (as A). The bar diagram depicts the difference in wound closure percentage among FRG1_KD + Control IgG, FRG1_KD + GM-CSF Ab, and untreated Control_Sc. Scale bar, 100 μm. The fifth lane shows the shift (arrow) resulting from competitive binding between the excess amount of mutated unbiotinylated probe and biotinylated probe with the nuclear protein extract. Result values show mean ± SD value. Two-tailed unpaired student's t-test was used to compare the difference between the two groups. ns, P > 0.05,*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001.
Previously it was reported that GM-CSF promotes breast cancer pathogenesis by recruiting CCL-18+ macrophages into the tumor microenvironment [12]. Increased GM-CSF level in breast cancer is correlated with increased metastasis and poor patient survival [42]. Higher expression of GM-CSF receptors on nonhematopoietic cells in multiple tumor types has suggested its potential protumorigenic factor, which is yet to be validated [43]. Expression of GM-CSF in skin carcinoma cells enhanced the metastatic growth and proliferation of cancer cells [44]. Also, in the head and neck [45], glioma [46], and osteosarcoma [47], autocrine stimulation of GM-CSF is reported to promote tumor growth. Although several reports have shown altered GM-CSF levels in multiple cancers, the underlying molecular mechanism in GM-CSF-mediated metastasis needs to be further addressed. The effect of the GM-CSF on the MAPK/ERK/ZEB1 pathway has been reported in colon cancer [11], but the detailed mechanism remains unexplored. The present study first time, discovered the in-depth role of GM-CSF in breast cancer and found that FRG1 acts as its repressor. When the expression of FRG1 is less in the cells, it leads to the expression of more GM-CSF, which in turn activates ERK-mediated EMT. Our findings have provided many missing links from FRG1 to GM-CSF to ERK to EMT. Besides, we have observed treatment with anti-GM-CSF mAb reduced tumor volume and EMT marker in the mouse model. Collectively, our study highlights the therapeutic potential of anti-GM-CSF therapy in cancer samples with low FRG1 expression.
To validate our findings, we performed a retrospective study in clinical patient samples and found reduced FRG1 expression in 71% of breast cancer patients' tissues. Reports suggest expression of ER in breast cancer is associated with favorable clinical outcomes [48]. We observed that in ER+ and TNBC samples, there was no difference in FRG1 expression levels. It emphasizes that the role of FRG1 in breast cancer is not subtypespecific. Yet, our in vitro findings suggest activation of ER in FRG1-depleted MCF7 cell line in an ERK-dependent manner showing cross-talk between FRG1-mediated regulation of ERK and ER signaling. Taken together the in vitro and in vivo data, the therapeutic potential of FRG1-mediated signaling can be explored in all molecular subtypes of breast cancer. Furthermore, survival analysis in GEPIA and Kaplan-Meier plotter designates the association of high FRG1 level with a higher recurrence-free survival rate in breast cancer patients, indicating a favorable role of FRG1 towards prognosis determination. Observation in large and stage-specific cohorts can confirm the authenticity of this observation.
In conclusion, reduced expression of FRG1 in breast cancer leads to transcriptional activation of GM-CSF, which promotes activation of ERK and EMT (Fig. 8). Identification of GM-CSF as an Fig. 6 Low FRG1 is associated with breast carcinogenesis and elevated pERK expression in breast cancer patients. A, B Analysis of RNA sequencing data from TCGA and GTEx databases using GEPIA webserver. A Box plot shows the levels of FRG1 transcripts in cancer (n = 1085) and normal (n = 291) samples applying the default parameters p value cutoff 0.01, log 2 FC value cutoff 1, log scale, and jitter size 0.4. B shows the percentage of disease-free survival (RFS) and 95% confidence interval in the high FRG1 and low FRG1 expression groups. C Representative IHC images of FRG1 expression in human breast cancer and adjacent normal tissues. The Scatter plot shows the comparison of median Allred scores of FRG1 expression between breast cancer (n = 194) and adjacent normal (n = 56) tissues. The difference between the two groups in FRG1 expression was calculated using the Mann-Whitney U-test. D Graphical distribution depicts the frequency of cancer vs. normal tissues according to the Allred score of FRG1. The bar diagram shows the percent distribution of individuals having high, moderate, and low levels of FRG1 in breast cancer patients (n = 194) and normal tissue samples (n = 56). The X-axis represents two groups; cancer and adjacent normal. The Y-axis denotes the percentage of individuals having high, moderate, and low levels of FRG1. E Representative images of FRG1 and pERK expression levels detected in breast cancer patients using IHC (n = 10). F X-axis of the scatter plot shows median Allred scores for the FRG1 level in ER-positive (n = 78) and TNBC (n = 40) breast cancer patients. Y-axis denotes the Allred score for the FRG1 expression level. ns, P > 0.05, ****P ≤ 0.0001.
activator of ERK, and the cross-talk between AKT and ERK, can help in the development of a more efficient therapeutic strategy.
MATERIALS AND METHODS Breast cancer patient sample collection
Our study included 194 breast cancer tissue samples with 56 adjacent normal tissues. From SRL diagnostics-Bhubaneswar, Apollo Hospitals-Bhubaneswar, and AHRCC-Cuttack, between 2014 to 2019, we collected tissues from 104 breast cancer patients, out of which 46 had the adjacent normal tissues. The patients who did not undergo any prior treatment before the surgery with known ER/PR/HER2 status, stage, and grade information as per the TNM system, were only included in the study. This study was approved by the Institutional Ethics Committee of NISER (NISER/ IEC/2021-02) and AHRCC. For prospective sample collection, written consent was taken; for retrospective tissue FFPE blocks, the ethics committee waived the requirement of written consent. Additionally, a tissue microarray containing 90 tumors and 10 normal tissue samples was purchased (Biomax, MD, USA, #BC081120c).
Cell culture, plasmids, generation of stable cell lines, and cellbased assays Detailed methodology is provided in the supplementary information.
RNA extraction and quantitative real-time PCR
According to the manufacturer's protocol, total RNA was extracted from the cells using an RNeasy mini kit (Qiagen, Hilden, Germany). Reverse transcription was performed with 1 µg of RNA using the verso cDNA synthesis kit (Thermo Scientific, Lithuania). Each qPCR reaction was carried out using 10 ng of cDNA, 2x SYBR Green PCR Master Mix (Thermo Fisher, CA, USA), and respective primers (Supplementary Table 1) in Applied Biosystem 7500 system (Thermo Fisher, CA, USA). All reactions were done in triplicate. GAPDH was used as an internal control. The relative expression of each transcript was calculated using the ΔΔCt method.
Enzyme-linked Immunosorbent assay (ELISA)
ELISA was performed to quantify the level of GM-CSF present in MCF7 and MDA-MB-231 cells with altered FRG1 expression. In total, 1 × 10 6 cells were plated in a 100 mm culture dish in their respective culture media. After 12 h, the media was replaced by the media containing 2% FBS. After 3 days of incubation, the media was collected and centrifuged at 4000 rpm (4°C) for 10 min to get rid of the cellular debris. The supernatant was aliquoted and stored at −80°C till further use. 100 μl of this supernatant was used to carry out the ELISA using Human GM-CSF Quantikine ELISA Kit (R&D Systems, MN, USA) according to the manufacturer's protocol. OD values were taken at 450 nm and 540 nm in the Varioscan multimode microplate reader (Thermo). During the final analysis, values obtained at 450 nm was subtracted from the values at 540 nm.
Western blot
Cell lysates were prepared in ice-cold RIPA buffer (Thermo Scientific, IL, USA), supplemented with protease-phosphatase inhibitor (Thermo Scientific, IL, USA). Protein was quantified by the BCA protein estimation kit (Thermo Scientific, IL, USA). About 30-40 µg of protein samples were resolved on 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a PVDF membrane (Millipore, Bangalore, India). The membrane was blocked using 5% BSA (MP Biomedicals, OH, USA) for an hour and incubated overnight with the primary antibodies (Supplementary Table 2). Blots were washed in 1X TBST buffer and detected with respective horseradish peroxidase (HRP) conjugated secondary antibodies (Abgenex, Bhubaneswar, India). The chemiluminescence signal was developed by SuperSignalTM West Femto maximum sensitivity substrate (Thermo Scientific, IL, USA) and detected in Chemidoc XRS + (Bio-Rad, CA, USA). Images were analyzed in ImageJ (NIH, MD, USA) software. All the experiments were performed in triplicate.
Pharmacological compounds for inhibition and activation assay
Details are given in the supplementary information.
Apoptosis assay
The cellular apoptosis was measured in FRG1-depleted MCF7 cells using Annexin FITC and PI staining kit (BD Pharmingen™, NJ, USA) following the manufacturer's protocol, in FACS caliber (BD Biosciences, CA, USA). Flow cytometric data were analyzed by CellQuest Pro software (BD Biosciences).
Electrophoretic mobility shift assay (EMSA)
Nuclear extract was prepared from HEK 293 T cells expressing FRG1 using the NE-PER kit (Thermo Scientific, IL, USA). Oligonucleotides were commercially procured (IDT, IA, USA). The binding of FRG1 on the CTGGG site of the GM-CSF promoter was investigated using 3 fmol of biotinylated double-stranded oligonucleotides. Competitive EMSA was carried out using 3 pmol of unlabeled oligonucleotides. For the supershift assay, 1.5 µg of FRG1 antibody (Abcam, Cambridge, UK) was used following standard reaction conditions per the manufacturer's protocol. Protein-DNA complexes were separated on 10% native polyacrylamide gels in 0.5X TBE, Fig. 7 Reduced expression of FRG1 promotes tumor growth and metastasis in BALB/c mice. In vivo experiments were carried out to validate the tumorigenesis and metastatic potential of FRG1. A-C BALB/c mice were injected with 4T1 cells with reduced (4T1_FRG1_KD) and elevated levels of FRG1 (4T1_FRG1_Ex), along with controls (Control_Sc and Control_Ev, respectively). Tumor volume and weight were measured. Protein extracts were isolated from the tumors and subjected to immunoblots to observe the expression of pERK and snail. A Representative images of tumors in mice bearing 4T1_FRG1_KD cells and Control_Sc cells. A graphical representation of the same depicts the change in tumor volume and weight between the two groups (n = 4). B Representative images of tumors in mice bearing 4T1_FRG1_Ex cells and Control_Ev cells. A graphical representation of the same depicts the change in tumor volume and weight between the two groups (n = 4). C Tumor harvested from 4T1_FRG1_Ex, and Control_Ev mice show expression of pERK and EMT marker snail. D, E 4T1 cells with altered FRG1 levels were injected in the tail vein of BALB/c mice. After that, nodules in the lungs were counted. Representative images of the lung nodules as observed in mice bearing 4T1_FRG1_Ex (D) and 4T1_FRG1_KD (E) cells with their respective controls. F, G BALB/c mice were injected with 4T1_FRG1_KD cells to develop tumors and the corresponding control (Control_Sc). After 7 days, 4T1_FRG1_KD group mice were treated with anti-GM-CSF antibody (n = 4) or control_IgG (n = 4), till day 21. F Images of tumors harvested from the three groups (FRG1_KD + Control IgG, FRG1_KD + GM-CSF mAb, and Control_Sc). The graph shows tumor volume measured on different days. G Protein extracts were isolated from the three sets and subjected to immunoblotting to observe pERK and snail expression. Bar graphs show the difference in expression levels of pERK and snail levels among the three groups (n = 3). Immunoblots are showing the expression of pERK, with the GAPDH as a loading control. Results are presented as mean ± SD. Two-tailed unpaired student's t-test was used to compare the difference between the two groups. ns, P > 0.05, *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001. In the presence of an optimal amount of FRG1 and lesser GM-CSF, activation of ERK gets reduced, which in turn downregulates snail, slug, and twist mediated EMT. On the other hand, a depleted level of pERK promotes the activation of AKT and p53 that facilitates apoptosis of the cells. Thus, the presence of FRG1 protects cells from GM-CSF-mediated tumorigenesis. Right panel shows the opposite effect that occurs due to a depleted level of FRG1. FRG1 downregulation leads to the upregulation of pERK. An increased amount of activated ERK inhibits the expression of pAKT and pP53 simultaneously, which in turn, inhibits cell apoptosis. Also, the upregulation of pERK increases the expression of EMT markers that leads cells toward EMT. Thus our model demonstrates FRG1-mediated suppression of breast carcinoma via GM-CSF/MEK/ERK signaling.
|
v2
|
2022-11-03T18:09:36.011Z
|
2022-11-03T00:00:00.000Z
|
253259372
|
s2orc/train
|
Complete response to nivolumab in Kirsten rat sarcoma virus oncogene KRAS-G12C mutant metastatic lung adenocarcinoma: a case report
Background The advent of immunotherapies has ushered in a new era in the treatment of non-small cell lung carcinoma. Although immunotherapies are associated with improved clinical outcomes, studies report a median overall survival of 11 months with progression-free survival of 2.5 months with the use of nivolumab for pretreated metastatic non-small cell lung cancer. Herein, we describe a case of advanced non-small cell lung carcinoma that has shown exceptional response to immunotherapy, with the patient being in complete response for the past 6 years since commencement of nivolumab. Case presentation We report the case of a 58-year-old female Caucasian, an ex-smoker with 40-pack-year history of smoking, who presented with cough and chest pain and was subsequently diagnosed with metastatic pulmonary adenocarcinoma. The tumor was positive for Kirsten rat sarcoma virus oncogene KRAS-G12C mutation and had high programmed death-1 ligand expression. She was commenced on first-line chemotherapy with carboplatin and gemcitabine with disease response, then continued on maintenance pemetrexed. She was then commenced on immunotherapy with nivolumab, with complete response for a total of 6 years. She does not report any adverse events. Currently, she shows no evidence of recurrence of non-small cell lung carcinoma. Conclusion The exceptional response to immunotherapy seen in this case may be explained by the presence of Kirsten rat sarcoma virus oncogene mutation, which is associated with enhanced clinical response to programmed death-1 ligand inhibitors. This report emphasizes the urgent need for further studies evaluating the role of Kirsten rat sarcoma virus oncogene mutation in determining the clinical efficacy of immunotherapies. This would enable us to make effective evidence-based clinical interventions in the treatment of non-small cell lung carcinoma.
Background
Non-small cell lung cancer (NSCLC) accounts for 85% of all lung cancer cases worldwide [1]. The emergence of immunotherapy has transformed the therapeutic landscape of NSCLC. Immune checkpoint inhibitors (ICIs) directed against programmed death-1 ligand (PD-L1), programmed death-1 (PD-1), and cytotoxic T lymphocyte-associated 4 (CTLA-4) proteins are associated with decreased mortality as well as longer progression-free survival in NSCLC patients [2,3]. Since their introduction in 2015, the role of PD-1 inhibitors has shown tremendous growth, from ancillary treatment to first-line therapies in NSCLC. Pembrolizumab, an antibody against PD-1 receptor, is currently indicated in both firstand second-line therapy in advanced cases of NSCLC Open Access *Correspondence: [email protected] without driver mutation, with greater benefit for those with high PD-L1 expression [4]. Nivolumab (anti-PD-1) and atezolizumab (anti-PD-L1) are two other immunemodulating agents recommended as second-line therapy in NSCLC, irrespective of PD-L1 levels [5,6]. Mutations in the Kirsten rat sarcoma viral (KRAS) oncogene are one of the most common genetic alterations seen in patients with NSCLC. The KRAS gene normally encodes a guanosine triphosphate binding protein that is involved in regulating the activity of the RAS gene, which is required for proper growth and development of cells [7]. Mutated KRAS results in constitutive activation of the RAS gene, leading to unregulated cell growth and development of carcinoma. Patients with KRAS mutations generally show decreased response to systemic chemotherapies and were previously associated with poor prognosis [7]. However, there is currently evidence showing increased survival and improved clinical response to ICIs in metastatic NSCLC patients harboring KRAS G12C mutation [8]. Sotorasib, which was recently approved for the treatment of NSCLC, is the first agent available against KRAS G12C mutation. This calls for further research into the development of novel agents against therapeutic targets in the KRAS gene. We report herein a case of advanced NSCLC with KRAS mutation and high PD-L1 expression with complete response to immunotherapy after chemotherapy.
Case presentation
A 58-year-old Caucasian female with significant history of smoking and no past medical history presented to the emergency department with chest pain and cough at T = 0 in January 2015. Computed tomography (CT) pulmonary angiogram revealed a large pericardial effusion measuring 40 mm in diameter along with moderate right pleural effusion (Fig. 1).
There was a spiculated anterior subpleural lesion measuring 28 mm extending to the anterolateral pleural surface of the right upper lobe (Fig. 2) and significant enlargement of the mediastinal with right paratracheal lymph node measuring 15 mm and subcarinal lymph node measuring 18 mm and left axillary lymph nodes measuring 12.5 mm. Video-assisted thoracoscopic surgery (VATS) pleurodesis was carried out.
The pericardial and pleural fluid cytology contained small numbers of malignant cells, presenting predominately singly in a bloodstained background. These cells have enlarged pleomorphic nuclei with granular to coarse chromatin, prominent nucleoli, and irregular nuclear contours. Immunohistochemical studies showed positive staining of tumor cells for thyroid transcription factor-1 (TTF-1), napsin A, and Cytokeratin 7 (CK7). Staining for Estrogen receptor (ER), Cytokeratin 20 (CK20), Caudal Type Homeobox 2 (CDX2), and calretinin were all negative. These findings are in support of adenocarcinoma of lung origin.
In 2015, the fluid sample sent was not sufficient to undergo further testing to determine PD-L1 expression, immunohistochemistry testing for c-ros oncogene 1 (ROS1) / anaplastic lymphoma kinase (ALK) mutation, or molecular testing for epidermal growth factor receptor (EGFR).
With the imaging and histology findings, she was diagnosed with stage IV (T3N2M1a) according to the TNM classification of the Union of International Cancer Control (UICC), 7th edition lung adenocarcinoma [9]. She had Eastern Cooperative Oncology Group (ECOG) performance status of 1 after VATS pleurodesis. The patient quit smoking at the time of diagnosis. She was then commenced on first-line chemotherapy, carboplatin (AUC 5), and gemcitabine (1000 mg/m 2 ) 3 weekly at T = 1 month in February 2015. She completed four cycles with no dose reduction. Follow-up CT imaging was done 3 months later (T = 4 months) demonstrated good clinical and radiological response. According to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 [10], complete response was achieved with disappearance of the pleural based right upper lobe lesion. There is complete response to the mediastinal, hilar, and axillary lymph nodes. She was clinically well with maintenance of her performance status.
Given the excellent response after first-line chemotherapy, she was then commenced on maintenance chemotherapy with pemetrexed at a dose of 500 mg/m 2 every 21 days. This is based on evidence that it prolongs overall survival and progression-free survival [11]. The patient completed eight cycles of pemetrexed without significant toxicities.
In December 2015, there was single lesion progression with a new right upper lobe lesion (18 × 15 mm 2 ) (Fig. 3). The patient declined biopsy of this lesion. Given disease progression as per RECIST version 1.1, she was commenced on nivolumab 240 mg every 2 weeks at around T = 12 months in February 2016 owing to emerging evidence of immunotherapy in this setting for previously treated metastatic non-small cell lung cancer.
After the commencement of immunotherapy, her follow-up scan done in May 2016 showed complete response of single lung lesion as per RECIST version 1.1. Further imaging subsequently every 3 months with CT of brain, chest, abdomen and pelvis continues to show no evidence of recurrence or new metastatic disease.
At T = 59 months in November 2019, she had an elective cholecystectomy for chronic cholecystitis secondary to multiple cholelithiases. Her nivolumab was not withheld during the surgery, and she did not develop any complications during the perioperative period.
At T = 63 months in March 2020, somatic gene mutation analysis was requested on her previous pericardial fluid and tissue sample done early in the diagnosis as the patient was not keen for rebiopsy. Reanalysis of the sample could be done using molecular testing. A mutation was detected in the KRAS gene p.(Gly12Cys): G12C (16%). No clinically relevant mutation was found in the tested regions of EGFR, BRAF, MET, RET, or ERBB2. It was negative for rearrangements in ALK and ROS1 genes. There was high level PD-L1 expression with a tumor proportion score (TPS) of 80%.
18-Fluorodeoxyglucose positron emission tomography (FDG-PET) scan at T = 83 months in November 2021 showed no suspicious fluorodeoxyglucose (FDG) uptake within lungs. The foci of intense FDG uptake surrounding right lung pleura is consistent with a previous talc pleurodesis (Fig. 4). The patient is currently continuing her immunotherapy and has no evidence of lung malignancy.
Discussion and conclusions
The advent of immunotherapies has brought about a monumental change in the prognosis as well as management of NSCLC. Therapies with ICIs are associated with improved patient survival, better tolerance, and reduced adverse effects. Although immunotherapies have tremendously enhanced patient outcomes in NSCLC, little is known about its predictive biomarkers and their efficacy in cancers with driver gene mutations. In this report, we describe a case of advanced NSCLC with KRAS mutation which has shown outstanding response to immunomodulators and in which the patient has been in complete response since she was started on nivolumab 6 years back.
Although ICIs are associated with improved clinical outcomes, studies assessing the efficacy of nivolumab in pretreated patients with advanced NSCLC reported a median overall survival of 11 months and progressionfree survival of 2.5 months as per the 5-year outcome of the Checkmate 017 and Checkmate 057 landmark trials [12]. While ICIs are associated with long-term clinical benefits, most patients develop resistance to these agents within 1-2 years [6]. However, we report a rare case in which the patient has shown excellent response to nivolumab with complete and durable response for the past 6 years, with no signs of resistance. This may be partly explained by the presence of KRAS mutation. KRAS mutation is seen in 25-30% of patients with NSCLC in the Western world, with the majority of them occurring in codon 12. The common KRAS mutations include G12C (43%), G12V (18%), and G12D (11%) [13]. KRAS mutation is also strongly associated with smoking history, and only 5% of KRAS mutations are seen in nonsmokers [7]. Moreover, G12C and G12V are more common among smokers, compared with G12D in nonsmokers [13]. KRAS mutation is associated with enhanced expression of PD-L1 ligand in NSCLC owing to stimulation of downstream pathways [14,15]. PD-L1 expression also depends on environmental factors. KRAS mutant lung cancers in smokers are reported to have greater PD-L1 expression compared with those in nonsmokers [16]. There is compelling evidence to suggest that KRAS mutation can result in longer progressionfree survival, greater clinical benefits, and improved survival with anti-PD-1 treatment compared with wild-type patients [15,17]. Liu et al. also reported that the clinical benefit of monotherapy with anti-PD-1 agents is comparable to that of combination therapy with docetaxel in KRAS mutated individuals [15]. In May 2021, the Food and Drug Administration (FDA) approved the use of sotorasib, the first-ever KRAS inhibitor to be used in targeted therapy. It was approved as a first-line agent in NSCLC patients with G12C mutation who have received at least one line of previous systemic therapy based on phase II data from the CodeBreaK100 study [18].
It is hypothesized that mitochondrial function can play a role in determining the clinical response of anti-PD-1 agents [19]. Factors such as smoking, advanced age, and obesity are associated with impaired mitochondrial function and greater expression of PD-1, resulting in improved clinical response [19]. This can explain the impressive response to immunotherapy for this patient, who was a heavy smoker.
Although immunotherapy has ushered in a new era in cancer therapy, it is not without its limitations. It has been proven that many patients derive significant clinical benefit from immunomodulators, but a large section of patients show no or dismal response to immunotherapies (primary resistance). Low tumor mutational burden (TMB) and expression of heterogeneous antigens leading to decreased tumor immunogenicity, alternate immune checkpoints, and release of immunosuppressive cells such as myeloid-derived suppressive cells and M2 macrophages into the tumor environment are some mechanisms that could contribute to immunotherapy resistance [19]. Host factors such as diet, microbiome, and autoimmunity also contribute to the development of resistance [19]. Aberrations in the interferon-gamma (IFN-γ) signaling pathway leading to defective antigen presentation are now understood as a mechanism of acquired resistance [20]. Even in large clinical trials, the overall response rate was between 47% and 63% [21]. Moreover, the development of acquired resistance after 1-2 years of treatment is also seen. ICIs are also associated with adverse effects, though to a much lesser extent compared withto conventional chemotherapies. Grade 3 or higher adverse effects are seen in 7-13% of total patients treated with PD-1 inhibitors [21]. Serious side effects associated with PD-1 inhibitors include myocarditis, pneumonitis, autoimmune hepatitis, encephalitis, and hypothyroidism [1,[21][22][23][24][25]. PD-L1 assay, though FDA approved, has reduced sensitivity and specificity compared with other diagnostic assays and has limited utility in predicting the clinical benefits of ICI [26]. Hyperprogression of There is no scan evidence of pericardial effusion. The foci of intense FDG uptake surrounding right lung pleura are consistent with a previous talc pleurodesis disease is another uncommon adverse effect of anti-PD-1 and PD-L1 therapies, which is characterized by rapid tumor growth or accelerated disease progression once the patient is started on ICI. A meta-analysis by Park et al. evaluated the incidence of hyperprogression among patients receiving ICIs and reported a pooled incidence of 13.4% [27].
Herein we present the case of a 58-year-old female with advanced NSCLC who has shown exceptional response to nivolumab therapy after chemotherapy. Given her good response, we plan to continue her treatment until she has disease progression or develops significant toxicities. This decision is supported by the conclusion from the Checkmate 153 trial that overall survival was significantly longer in the continuous treatment with nivolumab arm versus the 1-year fixedduration arm with retreatment on progression [28].
ICIs are no doubt a landmark development in oncology with the potential to revolutionize the management of lung cancer in the coming years. However, further studies assessing their efficacy should be carried out before reaching definitive conclusions regarding their clinical utility. Targeted therapies against KRAS must be given special emphasis because of the tremendous benefit shown in patients with NSCLC. Future research should also focus on the discovery of novel therapeutic targets, newer immunomodulators, and better biomarkers for predicting the clinical benefits of immunotherapies.
|
v2
|
2022-11-03T18:14:39.295Z
|
2022-11-03T00:00:00.000Z
|
253260830
|
s2orc/train
|
Orchestration of mesenchymal plasticity and immune evasiveness via rewiring of the metabolic program in pancreatic ductal adenocarcinoma
Pancreatic ductal adenocarcinoma (PDAC) is the most fatal cancer in humans, due to its difficulty of early detection and its high metastatic ability. The occurrence of epithelial to mesenchymal transition in preinvasive pancreatic lesions has been implicated in the early dissemination, drug resistance, and cancer stemness of PDAC. PDAC cells also have a reprogrammed metabolism, regulated by driver mutation-mediated pathways, a desmoplastic tumor microenvironment (TME), and interactions with stromal cells, including pancreatic stellate cells, fibroblasts, endothelial cells, and immune cells. Such metabolic reprogramming and its functional metabolites lead to enhanced mesenchymal plasticity, and creates an acidic and immunosuppressive TME, resulting in the augmentation of protumor immunity via cancer-associated inflammation. In this review, we summarize our recent understanding of how PDAC cells acquire and augment mesenchymal features via metabolic and immunological changes during tumor progression, and how mesenchymal malignancies induce metabolic network rewiring and facilitate an immune evasive TME. In addition, we also present our recent findings on the interesting relevance of the small G protein ADP-ribosylation factor 6-based signaling pathway driven by KRAS/TP53 mutations, inflammatory amplification signals mediated by the proinflammatory cytokine interleukin 6 and RNA-binding protein ARID5A on PDAC metabolic reprogramming and immune evasion, and finally discuss potential therapeutic strategies for the quasi-mesenchymal subtype of PDAC.
Introduction
Pancreatic ductal adenocarcinoma (PDAC) originates from epithelial cells of the exocrine pancreas, which is composed of secretory acinar cells and ductal cells (1). PDAC patients often have an unfavorable prognosis, and the 5-year overall survival rate has been reported to be only 11% in the United States (2). Only 20% of PDACs are confined to pancreatic tissue at diagnosis, approximately 30% have metastasized to regional lymph nodes, and more than 50% have disseminated to other tissues, primarily the liver and lungs (3).
Four major driver mutations have been identified in PDAC, including KRAS, TP53, CDKN2A, and SMAD4/DPC4 mutations (4)(5)(6). Constitutive active mutations of KRAS occur in more than 90% of patients, often demonstrate oncogenic activity, and have been shown to be involved in the initiating event of PDAC tumorigenesis (6)(7)(8)(9). In addition, oncogenic KRAS has been shown to promote tumor signaling through metabolic reprogramming (10) and stromal interactions (11) to facilitate tumor growth. Mutations in TP53 also often result in oncogenic activity, and are present in up to 70% of PDACs, typically occurring at late stages of PDAC carcinogenesis, and are frequently associated with invasive and metastatic phenotypes (6,12). Furthermore, TP53 mutations play an important role in inducing platelet-derived growth factor (PDGF) receptor B expression, which associated with reduced disease-free survival in PDAC patients (13).
Because of the lack of effective diagnostic biomarkers for PDAC and the absence of early symptoms, the diagnosis of PDAC is often made at advanced, terminal stages. Current treatment options include surgery, if possible, or chemotherapy (gemcitabine, FOLFIRINOX [fluorouracil, leucovorin, irinotecan, and oxaliplatin], etc.), and radiation therapy, all with limited efficiency and achieving only slightly prolonged survival (14, 15). Immune checkpoint-based immunotherapies have been incorporated, albeit to a limited extent, into treatment modalities for some other cancers, but clinical trials targeting checkpoint molecules, such as CTLA4, PD-1/PD-L1, or their other cognate ligands have been unsuccessful for the treatment of PDAC. So far, there have been no successful clinical trials against PDAC, even those targeting multiple immune checkpoints (16-18).
PDAC cells also demonstrate a poor nutritional status, high levels of oxidative stress, inflammatory stress, extracellular acidosis, hypoxia, and decreased angiogenesis (15,19,20). Consistently, these are strong selection pressures that enable only cells that have adapted their metabolism to these hostile conditions to survive and proliferate. Notably, accumulating lines of evidence suggest that these adaptations also make PDAC cells more invasive, metastatic, stem cell-like, and resistant to therapeutic treatments (21). Consistently, several genome-wide gene expression profiling and genomic sequencing approaches to elucidate the molecular landscape of PDAC have demonstrated that the so-called basal-like (also known as quasimesenchymal-like or squamous) subtype is associated with a less favorable prognosis than other subtypes (22-25). Importantly, PDAC metabolite profiling and transcriptional analysis confirmed that the quasi-mesenchymal-like subtype is associated with the glycolytic subtype (26-28). This reorganization of pancreatic cancer cell metabolism opens the way for new therapeutic opportunities (20). However, the substantial heterogeneity in gene expression and metabolic characteristics, the plasticity of pancreatic cancer cells, and the pathological changes associated with their linked physicochemical and biological changes in the tumor microenvironment (TME) make PDAC a challenging disease to cure (26,27,29).
In this review, we summarize recent studies on how gene expression changes via intrinsic genetic mutations and epigenetic alterations involved in the acquisition of mesenchymal traits in PDAC cells, particularly posttranscriptional dysregulation of expression, are linked to metabolic reorganization associated with immunosuppressive TME formation during the development and malignant progression of PDAC.
Recently, PDAC has been hypothesized to be associated with two morphologically distinct precursors, i.e., pancreatic intraepithelial neoplasia (PanIN) and intrapapillary mucinous neoplasia (IPMN). PanIN can progress to invasive carcinoma in a stepwise and linear manner, which is an established mechanism of PDAC progression (30). Multiple studies have reported the sequential accumulation of PDAC driver gene mutations in PanIN, with KRAS mutations being the earliest known genetic alterations, being present in more than 90% of all PanINs regardless of cancer grade (31). On the other hand, the inactivation of CDKN2A is rare in low-grade PanIN, but has been reported to occur in more than 70% of high-grade PanIN (32). Mutations in TP53 and SMAD4 occur during the late stages of PanIN progression, and are almost exclusively found in highgrade PanIN and invasive PDAC. In contrast, IPMN is driven by four driver gene mutations of pancreatic tumorigenesis similar to PanIN, including early mutations in KRAS and late mutations in CDKN2A, TP53, and SMAD4 (33). However, there are also two frequently altered driver genes specific to the IPMN pathway. Mutations in the oncogenic hotspot of GNAS are known to occur early in IPMN tumorigenesis (33-35). In addition, although inactivating mutations in Ring finger protein 43 (RNF43), which encodes a ubiquitin ligase involved in WNT signaling (often with loss of heterozygosity) are also common in IPMNs (36), the precise timing of the occurrence of RNF43 mutations in IPMN tumorigenesis has not yet been clarified to date.
In addition, we present our recent findings on the intriguing relevance of the small G protein ADP-ribosylation factor 6 (ARF6)-based signaling pathway driven by KRAS/TP53 mutations, as well as the inflammation amplifying signals mediated by the inflammatory cytokine interleukin 6 (IL-6) and the RNA-binding protein AT-rich interactive domain 5a (Arid5a) on PDAC metabolic reprogramming and immune evasion. We will present our recent findings on the relevance of these pathways, and finally discuss potential therapeutic strategies for the quasi-mesenchymal subtype of PDAC.
Plasticity of adult pancreatic tissues
The pancreas is an important organ responsible for metabolic control in the body, and is composed of two morphologically and functionally distinct components. The exocrine pancreas, accounting for more than 95% of total organ mass, is composed of acinar cells, which produce digestive enzymes, and ductal cells, which deliver these enzymes to the intestine. On the other hand, the endocrine islets of Langerhans consist of five different cell groups (a, b, d, PP, and e cells) that secrete various hormones, such as insulin and glucagon, and play crucial roles in the regulation of glucose metabolism. The exocrine and endocrine pancreas are associated with different diseases. Pancreatitis and pancreatic cancer, mostly PDAC, arise from the exocrine pancreas, whereas rare pancreatic neuroendocrine tumors arise from the endocrine islets, and diabetes is also a result of endocrine islet dysfunction (37). The mammalian pancreas has the capacity for regeneration after injury even in adults, with the acinar compartment having the highest plasticity in humans. Through epigenetic transcriptional regulation, acinar cells can dedifferentiate into an embryonic progenitor-like phenotype, and commit to either insulin + b-cells (38) or ductal cells (known as acinar to ductal metaplasia [ADM]) (39, 40). ADM transdifferentiation occurs in chronic pancreatitis via nuclear factor-kB (NF-kB) activation, and is associated with pancreatic intraepithelial neoplasia, which is a necessary step for the generation of neoplastic precursor lesions called PanINs (Pancreatic intraepithelial neoplasia) (41-43). Thus, it has been speculated that the acinar cells of the exocrine pancreas maintain plasticity to adapt to changes in the external environment, and that their dysregulation leads to pancreatitis and pancreatic cancer. progenitor, and aberrantly differentiated endocrine exocrine) (24). Each of these classifications has been able to predict the prognosis of patients with resected PDAC on multivariate analysis. Notably, in about half of PDAC tumors, increased expression levels of hypoxiaassociated genes were observed by RNA sequencing (RNAseq), and were substantially associated with basal-like subtypes, although there was no redundancy in the identified gene sets (44). Regarding morphology, PDACs are classified as having more or less than 40% glandular histogenesis, and are strongly associated with classical or basal subtypes, respectively (45). The squamous morphology found in more than 30% of invasive tumors has also been associated with basal-like tumors by several groups (16, 45). However, the mechanism by which PDAC diverges into various subtypes in the process of tumor evolution remains unclear.
Heterogeneity of PDAC
Recently, it has been reported that re-categorization of PDAC subtypes in a combined cohort of primary and metastatic tumors using single-cell RNAseq (scRNAseq) can lead to the extension of the two groups of basal-like and classical into five groups: "basal-like A", "basal-like B", "classic A", "classic B", and "hybrid" (46). These data sets, combined with cohort of patients with PDAC, enable the broad categorization of basal-like A and basal-like B into two disease subtypes, localized and metastatic disease, respectively. Thus, it is suggested that PDAC proceeds as a mixture of both expressed phenotypes, and that the behavior of the dominant phenotype and subtype is due to plasticity in both (46). The driver mutations for the classical and basal-like subtypes were shown to be biallelic loss of SMAD4 with GATA6 amplification, and biallelic loss of TP53 and/or CDKN2A with mutant KRAS allele amplification, respectively, but none of the features were completely exclusive (45, 46). Therefore, whereas scRNAseq analysis of precancerous lesions to determine whether these expression phenotypes are established in PanIN has not been performed to date, the early acquisition of asymmetric driver gene mutations is itself dynamic, presumably dictating PDAC behavior, suggesting that both clonality and plasticity of PDAC cells are responsible for the histological and biological heterogeneity.
Current diagnosis and treatment methods of pancreatic tumors
Symptoms of PDAC and its diagnosis Symptoms of PDAC are often vague and nonspecific, and hence it is sometimes referred to as the 'silent killer'; in fact, 30% to 35% of patients are diagnosed with locally advanced stages and 50% to 55% with metastatic stages of disease. Biomarkers for the early detection of PDAC have not yet been identified. The most common site of this tumor is the head of the pancreas, which causes biliary obstruction, resulting in dark urine, jaundice, appetite loss, fatigue, weight loss, and exocrine pancreatic insufficiency (47).
As early symptoms of PDAC are less frequent than those of any other cancer, and a method for its early diagnosis has not been established, multidisciplinary examinations are required to detect the pancreatic tumor. The pancreas is a digestive organ that also acts as an endocrine system, and hence has abundant blood vessels. This feature makes PDAC easy to metastasize and difficult to resect. There are four clinical stages in PDAC; 1) I-II resectable (5-year survival rate, 35%-45%), 2) II-III borderline resectable (10%-15%), 3) II-III locally advanced (10%-15%), and 4) metastatic (< 5%). Pancreas computed tomography (CT) angiography with chest and pelvis CT can be used for assessment of the vascular anatomy of the pancreas. The degree of contact between the tumor and local blood vessels is classified into three levels; uninvolved, abutted, or encased. The difference between abutment and encasement is the degree of circumferential tumor-vessel involvement; existence of the tumor more than 180 degrees around the vessel implies encasement. Magnetic resonance imaging and cholangiopancreatography are also helpful to assess the possibility of metastasis in indeterminate liver lesions, and are also useful for the identification of cancers that are poorly characterized on CT imaging (47).
Conventional treatments and ongoing notable clinical trials
Patients with nonresectable tumors are treated by chemotherapy according to their cancer stage and Eastern Cooperative Oncology Group (ECOG) performance status (48). Combinations of cytotoxic chemotherapies were developed in the previous decade and are still the basis of current treatments for metastatic pancreatic cancer ( Figure 1) (49). Two multidrug regimens are now offered; FOLFIRINOX, and gemcitabine combined with nanoparticle albumin-bound paclitaxel (nab-paclitaxel). Gemcitabine alone is offered to patients with ECOG performance status 2 (within the five ECOG criteria, the groups in which patients are capable of self-care but are unable to carry out any work activities; i.e., patients are up and about > 50% of their waking hours, Figure 1). To classify patients eligible for either FOLFIRINOX or gemcitabine plus nab-paclitaxel as first-line drugs, Knox et al. demonstrated that a low level of GATA6, which is a characteristic of basal-like tumors, is a useful biomarker for selecting gemcitabine plus nab-paclitaxel in first-line therapy (50). The PASS-01 study analyzing the usefulness of GATA6 as a surrogate marker is now ongoing (NCT04469556).
Recent scientific advances have made incremental progress for the treatment of specific subgroups of pancreatic tumors. The American Society of Clinical Oncology guidelines updated in 2020 state three recommendations for pancreatic cancer; [1] early testing of both germline and tumor cells for microsatellite instability/mismatch repair deficiency, BRCA mutations, and NTRK gene fusions, [2] larotrectinib or entrectinib after first-line therapy for patients with tumors harboring NTRK fusions, and [3] continued treatment, including chemotherapy or olaparib, for patients with a germline BRCA1 or BRCA2 mutation who have received first-line platinum-based chemotherapy (51).
Although oncogenic KRAS mutations are observed in almost 90% of PDACs, a lack of drug-accessible pockets in the KRAS protein has hindered the development of their inhibitors for many years. However, X-ray crystallography identified a cryptic pocket of KRAS G12C potentially useful for drug development (52, FIGURE 1 Conventional and updated guidelines for metastatic PDAC. Patients are classified by ECOG performance status for first-line chemotherapy. For second-line therapy, three options are recommended in the 2020 updated guidelines. The strength of recommendation is indicated by the number of stars. 53). A phase 1/2 clinical trial for the clinical-grade KRAS G12C inhibitor AMG-510 (sotorasib) is currently ongoing (NCT03600883). Other drugs targeting mutant KRAS proteins are also being developed (54).
Clinical trials of immune checkpoint inhibitors (ICIs) against PDAC were started with great expectations, but let researchers down because of their limited efficacy compared with their efficacy against other solid tumors, including melanoma and lung cancer (16, 17). These disappointing results were attributed to the unique characteristics of PDAC, which are explained in the following sections. Given these facts, ICI treatment in combination with other types of agents to increase treatment efficacy have been widely considered for the treatment of PDAC (54).
The concept of targeting cancer metabolism has existed for almost a century, since Otto Warburg's observation of aerobic glycolysis in cancer cells and Sidney Farber's paper describing anti-folate-induced remission of childhood acute lymphocytic leukemia (55,56). Their concepts were eclipsed for some time during which knowledge on oncogenes accumulated and molecular-targeting therapies showed substantial effects on patient survival. However, recent technological innovations leading to various omics analyses have clarified the connection between tumor-associated genes and metabolism (57). Mitochondria, which play various roles in cancer metabolism and malignancy, are typical targets of metabolic agents (58). The lipoate analogue CPI-613, which inhibits pyruvate dehydrogenase and a-ketoglutarate dehydrogenase and therefore disrupts mitochondrial function (59), is being evaluated in a phase III trial of metastatic PDAC (NCT03504423) (60). In this trial, both groups (with or without CPI-613) are treated with FOLFIRINOX, because it has been reported that CPI-613 enhances FOLFIRINOX cytotoxicity in some PDAC cell lines (14). Another treatment target of PDAC is autophagy, which is activated in PDAC (61). A clinical trial of combinatorial treatment of hydroxychloroquine, an inhibitor of lysosomal scavenging, a MEK inhibitor, and ICIs for PDAC patients is now ongoing (NCT04214418). As discussed here, the paradigm of targeting not only tumor cells but also the TME, including immune cells, could bring a bright future to PDAC therapy.
Acquisition of mesenchymal plasticity of PDAC cells, and clinical implications of EMT in PDAC TME and mesenchymal plasticity of PDAC A variety of stimuli, including mechanical stress, low pH, hypoxia, innate and adaptive immune responses, changes in the extracellular matrix (ECM), and treatment with antitumor drugs can activate epithelial-mesenchymal transition (EMT) in cancer cells (62). It has been shown in real clinical settings that EMT plays a role in pancreatic cancer cell dissemination to distant organs in the precancerous stage prior to and/or in parallel with primary tumor formation in PDAC (63). The fact that almost all patients who undergo complete surgical resection and are free of metastases at that time eventually die within 5 years is consistent with the early-seeding model (64-66), suggesting important roles for EMT in PDAC progression and its contribution to the poor outcome.
PDAC has been well documented to be a desmoplastic stroma consisting of a dense ECM infiltrated with heterogeneous cell populations, including immune cells, endothelial cells, and cancer-associated fibroblasts (CAFs) (67). The high density of the stroma limits oxygen supply to and diffusion in the TME, leading to the creation of a hypoxic environment. Desmoplasia is observed in the bulk of the ECM, and contains collagen, fibronectin, laminin, and hyaluronic acid. These ECM components are primarily produced by CAFs. CAFs are also involved in producing various cytokines, such as transforming growth factor b (TGF-b), IL-1, IL-6, and tumor necrosis factor, and facilitate EMT signaling pathways (68).
PDACs are characterized by hypovascular tumors in a hypoxic microenvironment, in which high interstitial fluid pressure occurs owing to desmoplasia (69). However, microvessel density (MVD) has been shown to vary considerably among PDAC tumors with its decline being associated with poor survival in inverse correlation with stromal surface area (70). The hypoxic microenvironment has broad effects on the biological behavior and malignant phenotype of PDAC, including pathological angiogenesis and metabolic reprogramming, synergistically contributing to PDAC development and therapeutic resistance. Hypoxia-inducible factors (HIFs) are essential for hypoxia-induced angiogenesis in PDAC through transcriptional activation of various angiogenic factors, such as vascular endothelial growth factor (VEGF). It has been shown that under hypoxic conditions, NF-kB activates the transcription of HIF-1a and its target gene VEGF-A, resulting in the increased secretion of VEGF, and enhanced angiogenesis in hypoxic pancreatic cancer cells (71). Phosphorylated signal transducer and activator of transcription 3 (STAT3) is also a hypoxia-responsive nuclear transcription factor that has been shown to act synergistically with HIF-1a to regulate angiogenesis under hypoxia in pancreatic cancer cells (72). Indeed, increased production of VEGF has been demonstrated in human PDAC cell lines and resected PDAC tumor tissues (73), showing that VEGF is produced under the control of activated HIF-1a and STAT3 under conditions of oxygen deprivation (74,75). VEGF produced by human PDAC cell lines has functional activity to promote endothelial cell growth in vitro, and in large tumors in immunocompromised mouse xenograft models (76). In addition, the anti-VEGF strategy was shown to markedly reduce the growth of human PDAC cell lines orthotopically implanted into mice with a decrease in tumor MVD (77,78). Despite these preclinical data suggesting that angiogenesis is important in PDAC, the use of anti-angiogenic agents has not been clinically successful for treating PDAC. Chronic treatment with VEGF antibodies was found to induce hypoxia and lead to increased collagen deposition, epithelial plasticity, and metastatic burden (79). These results may underly the lack of success of angiogenesis inhibitors in clinical trials of PDAC.
We previously showed that ARF6 is activated by VEGF in endothelial cells and is required for VEGF-induced tubular formation and migration. Furthermore, we have shown that ARF6 signaling is involved in choroidal neovascularization, which is a major cause of vision loss in patients with ageassociated macular degeneration. We also found that ARF6 signaling is involved in VE-cadherin recycling, and may be involved in the sprouting process of angiogenesis associated with VE-cadherin-based cell-cell junctions as well as cell migration/tubular network formation activities (80). In addition, we found that high expression of the Arf6 effector AMAP1 is associated with the fibrosis of pancreatic cancer (81).
Treatment strategies for PDAC targeting angiogenesis have been pointed out as a way to normalize the tumor vasculature, such as strategies that prune immature and inefficient blood vessels, eliminate unproductive vasculature, and enable the reliable delivery of intravenous cancer drugs (82,83). The inhibition of ARF6 signaling, which is important for pathological angiogenesis and fibrosis, may contribute to therapeutic strategies for PDAC.
Recent analyses have redefined the view that cellular senescence is the onset of the tissue remodeling that operates during normal embryonic development and tissue damage. To this end, senescent cells cease their own proliferation and recruit phagocytotic immune cells to promote tissue regeneration (84). On the other hand, it is well known that senescence is associated with cancer; in PDAC, senescence appears to produce tumor suppressive effects at the earliest stages. However, some lines of evidence indicate that senescent cells in the TME can produce a senescence-associated secretory phenotype (SASP), mediated by NF-kB and CCAAT/ enhancer-binding protein-b, including the secretion of proinflammatory cytokines (IL-6 and IL-8), chemokines (monocyte chemoattractant proteins [MCPs], macrophage inflammatory proteins [MIPs], TGFb, and granulocytemacrophage colony-stimulating factor [GM-CSF]), and proteases (84), and play protumorigenic roles during tumor progression (85). SASPs have been shown to induce cell plasticity by stimulating cancer cell proliferation, motility, and invasion, and by generating an inflammatory TME (86). Thus, in the PDAC microenvironment, SASP may be involved in promoting EMT.
Role of EMT in PDAC metastasis
An important aspect of the EMT program in cancer biology may be its involvement in not only facilitating cellular motility and invasiveness, but also in orchestrating the cancer stem cell state (CSCs) via epithelial-mesenchymal plasticity (87-89). Mechanistically, intrinsic oncogenic mutations, epigenetic gene expression conversion, and extrinsic inflammatory signals may enable highly epithelial and highly mesenchymal non-CSCs to reversibly transition to an intermediate quasi-mesenchymal state; in the case of epithelial cells, the transition is accompanied by EMT, whereas in the case of highly mesenchymal cells, it is induced by mesenchymal-epithelial transition. Presumably, similar responses might occur in normal epithelial tissue when stem cells are lost. Thus, in the invasion-metastatic cascade, the EMT program is thought to enable the seeding of cells from the primary tumor into the parenchymal layer of distant tissues, and subsequently confers stemness, giving the disseminated tumor cells the ability to form metastatic colonies (87-89).
Although it is clear that EMT is involved in tumor metastasis, the exact function of EMT in cancer is still being debated. Indeed, some studies on the effects of the EMTtranscription factors (TFs) SNAIL and TWIST in pancreatic cancer have questioned the role of EMT in metastasis. A study using PDAC model KPC (Pdx1-cre; LSL-Kras G12D ;Tp53 R172H/+ ) mice, in which TWIST and SNAIL were independently conditionally knocked out, resulting in Pdx1-cre; LSL-Kras G12D ;Tp53 R172H/+ ;Twist1 flox/flox and Pdx1-cre; LSL-Kras G12D ;P53 R172H/+ ;Snai1 flox/flox mice, respectively, found that although EMT was suppressed, the deficiency of SNAIL or TWIST did not affect tumor progression, regional invasion, or dissemination. Thus, it has been argued that EMT is not required for invasive and metastatic activities of cancers. On the other hand, mice bearing abrogation of EMT-transcription factor (EMT-TF) have been shown to be correlated with chemosensitivity to gemcitabine, indicating EMT induces chemotherapy resistance in pancreatic cancer (90). Similar results have been reported in breast cancer models (91). However, other groups have shown using the same KPC mouse PDAC model that ZEB1 conditional knockout mice (Pdx1-cre; LSL-Kras G12D ;Tp53 R172H/+ ;Zeb1 flox/flox ) have significantly reduced PanIN and PDAC formation, and invasion and metastasis, thus clearly demonstrating a crucial role for the EMT-TF ZEB1 in the PDAC progression (92). Taken together, these studies indicate a trend toward the differential functions of EMT-TF; SNAIL and TWIST do not appear to be necessary, whereas ZEB1 conversely appears to be an important factor that is not compensated by other EMT-TFs.
Metabolic characteristics of PDAC Glucose metabolism
Glucose is the principal carbon and energy source for the growth and maintenance of mammalian cells. Glucose catabolism occurs by two metabolic pathways; glycolysis and the tricarboxylic acid (TCA) cycle. These pathways not only fuel adenosine triphosphate (ATP) production, but also produce carbon intermediates that support macromolecular biosynthesis. One contribution of oncogenic KRAS mutations to the oncogenesis and progression of pancreatic cancer is oncogenic KRAS mutation-driven metabolic rewiring. Transcriptome and metabolomic analyses indicated that the activity of oncogenic KRAS mutations promoted the upregulation of key metabolic enzymes involved in glucose metabolism, including glycolysis, hexosamine biosynthesis leading to the synthesis of uridine diphosphate N-acetylglucosamine, which is a significant substrate for protein glycosylation, and the pentose phosphate pathway producing NADPH and ribose 5-phosphate, which are essential for nucleic acid synthesis (10). This analysis also indicated that oncogenic KRAS mutations enhance glucose consumption in PDAC through the increase in transcription of the glucose transporter 1 (GLUT1, also known as solute carrier family 2 member 1 [SLC2A1]) the enzymes hexokinase 1 and hexokinase 2 (HK1 and HK2), and lactate dehydrogenase A (LDHA) ( Figure 2). Thus, KRAS contributes to the unregulated growth of pancreatic cancer cells, and directly targeting metabolic pathways as a therapeutic target is a major challenge (93).
We previously showed that mutant KRAS, which is a major driver gene in PDAC cells, acts in a eukaryotic translation initiation factor 4A (eIF4A)-dependent manner to promote the translation of ARF6 mRNA, which is a member of the ARF family of GTPases with a quadruplex structure in the 5′untranslated region, and upregulates ARF6 protein expression (94). Recently, it was also reported that silencing of ARF6 inhibits the Warburg effect, which is associated with aerobic glycolytic processes, in KRAS-mutated PDAC cells (95). The oncogene c-Myc is a transcription factor that regulates aerobic glycolysis through the upregulation of many key glycolytic genes, such as GLUT1, HK2, and LDHA (96, 97), and is Metabolic characteristics of PDAC associated with the ARF6-based pathway. The tumor microenvironment (TME) in PDAC is characterized by low vascular density, resulting in severe hypoxia and low nutrient levels. PDAC is also characterized by a dense desmoplastic stroma. In mammals, glucose and glutamine are two of the most abundant nutrients that support cell survival and growth. Oncogenic KRAS mutations induce metabolic reprogramming by triggering the uptake of glucose, leading to increased glycolytic flux, carbon donation to the pentose phosphate pathway and hexosamine biosynthetic pathway, and lactate production driving acidic TME. Glutamine is also used as an energy substrate in the TCA cycle, and maintains the intracellular redox state of PDAC cells in an oncogenic KRAS-driven manner. Double mutations of KRAS/p53 cooperatively promote the expression and activation of the ARF6-AMAP1 pathway, and ARF6 is involved in maintaining the Warburg effect to meet the abnormal nutritional and energy demands of PDAC cells, as well as those required for autophogosome and macropinosome formation. Mutant p53 promotes ARF6 activation via the enhanced expression of mevalonate pathway enzymes, and also intracellular trafficking of ARF6 mediated by geranylgeranylation of RAB11b. TCA, tricarboxylic acid; HK-1/2, hexokinase-1/2; G6P, glucose 6-phosphate; LDH-A, lactate dehydrogenase-A; MCT, monocarboxylate transporter; CA, carbonic anhydrase; Ac-CoA, Acetyl-CoA; ACLY, ATP-citrate lyase; FASN, fatty acid synthase; GLS, glutaminase; GLUD, glutamine dehydrogenase; BCAA, branched-chain amino acid; BCAT, branched-chain amino acid transaminase 1; IDO, indoleamine 2,3-dioxygenase. associated with the transcriptional activation of ARF6. ARF6 has also been shown to be associated with the regulation of the expression of GLUT1, LDHA, and HK2, as well as c-Myc (95). Thus, it is possible that ARF6 is involved in the regulation of aerobic glycolysis via the regulation of c-Myc in PDACs. Interestingly, in several cancers, including PDAC, the upregulation of GLUT1 in cancer cells correlates with the low infiltration rate of cytotoxic CD8 + T cells (98-100). This suggests that tumor cells successfully compete for glucose, suppressing antitumor immunity while simultaneously maintaining high metabolic and proliferative rates (101,102). Importantly, this also indicates that antitumor immune cells are unable to obtain sufficient energy, thus impairing their function.
As solid tumors progress, large areas of the tumor often become deprived of oxygen, which interferes with ability of the immune system to combat the tumor (103). PDAC is characterized by a very hypoxic TME, and it has been noted that the high malignancy and poor curative efficacy of PDAC are mostly due to the hypoxic TME (103, 104). PDAC also shows increased accumulation of stromal tissue, i.e., desmoplasia, which may collapse blood vessels, and subsequently impede perfusion and promote maintenance of the hypoxic TME. Hypoxia and desmoplasia induce the expression of HIF-1a and its stabilization (105). HIF-1a is a key regulator of cellular responses to changes in oxygen concentration, and supports tumor cell adaptation to hypoxia in an oxygen-deprived TME.
Under hypoxic conditions (usually below 3% to 5% O 2 ), the HIF-1a subunit stabilizes and forms a dimer with the b-subunit aryl hydrocarbon receptor nuclear translocator (ARNT), which translocates to the nucleus to promote O 2 -regulated gene expression. HIF-1a is considered to play a crucial role in of metabolic reprogramming (106). Several studies have confirmed that HIF-1a meets the metabolic needs of pancreatic cancer cells by increasing the expression of glycolysis-associated enzymes and the production of lactate (107)(108)(109)(110). Indeed, the stabilization of HIF-1a has been reported to induce GLUT1 expression in a HIF-1a -dependent manner, increasing cellular glucose uptake and supporting aerobic glycolysis in cancer cells (111, 112). HIF-1a is also known to enhance the expression of LDHA (113, 114) and monocarboxylate transporter 4 (MCT4; encoded by SLC16A3) (115). LDHA reduces the dependence on oxygendependent mitochondrial oxidative phosphorylation (OXPHOS) by converting pyruvate to lactate, and the cell preferentially uses oxygen-independent glycolytic pathways to maintain sufficient ATP production to meet bioenergetic requirements, whereas MCT4 removes lactate from the cell by transporting it out of cells. Thus, HIF-1a drives the conversion from oxidative to glycolytic metabolism during hypoxia, which is not only beneficial for bioenergetic homeostasis, but may also promote tumor survival and growth.
Interestingly, it has been shown that the stabilization of HIF-1a by di-methyl-oxaloylglycine treatment markedly increases the level of ARF6 mRNA (116), and ARF6 activity is significantly promoted under hypoxia (117). As mentioned previously, ARF6 is also associated with the enhanced expression of genes involved in glycolytic metabolism in malignant pancreatic cancer with KRAS mutations, so hypoxia may potently promote glycolytic metabolism through the induction of HIF-1a and ARF6, thereby regulating the adaptive responses to a hypoxic environment.
In addition to glucose deprivation via tumor cells in the TME, higher rates of aerobic glycolysis in tumor cells may promote the production of lactic acid, which in turn increases the acidity of the TME. Excess lactate produced in tumor cells can also suppress CD8 + T and NK cell activation, and enhance the function of immunosuppressive cells, such as myeloid subsets, and M2polarized macrophages to an immunosuppressive phenotype (118,119). This makes it difficult for immune cells to survive, but tumor cells can often adapt, survive, and multiply despite these harsh conditions. Tumor cells can respond to extracellular acidic pH conditions and regulate cellular acid homeostasis by altering the expression of proteins associated with pH regulation, such as monocarboxylate transporters and carbonic anhydrase (CA) (120). In in vitro models of melanoma, exposure to lactic acidosis has been shown to induce the EMT phenotype (121). In pancreatic cancer cells, lactate enhances the expression of IL-8 and contributes to EMT and metastasis (122)(123)(124), and tumor cells can use lactate as an alternative energy fuel to promote their proliferation (125). Indeed, high levels of lactate in PDAC have been shown to correlate with poor patient prognosis (126). Therefore, it is strongly suggested that the acidic environment in tumor tissue is involved in the acquisition of mesenchymal traits, and the augmentation of an immunosuppressive PDAC TME.
Lipid metabolism
Lipids are major components of biological molecules, and play important roles in various processes. Lipids are composed of thousands of different molecules, including phospholipids, sphingolipids, fatty acids, cholesterol, cholesteryl esters, and triglycerides. Such lipids are implicated in a variety of cellular processes, and are important components of biological membranes (127)(128)(129)(130)(131)(132). Lipid uptake, accumulation, and lipogenesis are increased in various cancers, including pancreatic cancer, and provide energy for rapid tumor growth. In the early step of de novo lipid synthesis, ATP-citrate lyase (ACLY) catalyzes the conversion of citrate to acetyl-CoA, which is then converted to malonyl-CoA by acetyl-CoA carboxylase. The acyl groups of malonyl-CoA and acetyl-CoA bind to the acyl-carrier protein domain of fatty acid synthase (FASN) in an NADPH-dependent way to produce long-chain saturated fatty acid (133) (Figure 2).
Expression levels of lipogenic enzymes, including ACLY, are known to be often upregulated in PDAC (134, 135). Inhibition of ACLY activity suppresses PDAC cell growth in xenograft tumor models (136). Furthermore, PDAC patients, highly expressing FASN, have been shown to have shorter overall survival than those expressing low levels of FASN (137). Overexpression of the FASN gene may be correlated with resistance to radiotherapy and gemcitabine in pancreatic cancer patients (138), and inhibition of FASN results in high cytotoxicity of this drug. As higher lipogenic activity has been shown in PDAC cells compared with normal cells, genetic and pharmacological inhibition of FASN and other lipogenic enzymes appears to be a promising therapeutic strategy.
The mevalonate pathway (MVP) is essential for cellular lipid metabolism, including cholesterol biosynthesis and the posttranslational prenylation of proteins (139). The rate-limiting enzyme in this pathway, 3-hydroxyl-3-methylglutaryl-CoA (HMG-CoA) reductase, has been considered as a prominent target for MVP inhibition, and is increased in a KRAS-driven PDAC mouse model (125,140). Statins, which are reductase inhibitors, are used for the treatment of hypercholesterolemia (141). The anticancer effects of statins have also been analyzed in vitro in various cancer cell lines. Several studies have reported that simvastatin inhibits cancer cell proliferation by promoting apoptosis and reducing cell cycle progression via several kinds of signaling pathways, including mitochondrial apoptotic signaling pathways and the Rho signaling pathway involved in cell cycle arrest (142,143). In addition, lipophilic statins (lovastatin, simvastatin, etc.) have been shown to be potent vaccine adjuvants via modulation of post-translational protein prenylation. Mechanistically, statins inhibit geranylgeranylation of the small GTPase Rab5, such as in antigen-presenting cells, causing inhibition of endosome maturation, sustained antigen retention, reinforced antigen presentation, and activation of T cells (144). Therefore, the MVP pathway is a potential target for cancer immunotherapy.
We have previously shown characteristic features that predict responders of MVP-based cancer treatment. We found that the Arf-GTPase ARF6, and its downstream effector AMAP1 (also called ASAP1/DDEF1), are often overexpressed in various types of cancer, including PDAC, and closely associated with poor patient survival (145)(146)(147)(148)(149). Interestingly, we found that the MVP is crucial for ARF6 activation in breast cancer cells. In this process, the MVP is essential for geranylgeranylation of RAB11b, which promotes intracellular trafficking of ARF6 to the plasma membrane where it is activated by RTKs. Furthermore, consistent with reports that gain-of-function mutants of p53 activate the MVP, it is clear that mutant p53 is essential for ARF6 activation (148, 150). Our in vitro experiments showed that the presence of statins improved the sensitivities of breast cancer cells to various drugs. In contrast, inhibition of MVP is ineffective when cancer cells do not overexpress components of the ARF6-based pathway. We have also shown that statins inhibit not only ARF6 activity and invasive potential but also recycling of the immune checkpoint molecule PD-L1 to the plasma membrane in pancreatic cancer cells (94). The chemopreventive effects of statins have been shown in pancreatic cancer cell lines (151-153) and pancreatic cancer model mice (154). Thus, the MVP may be crucial for promoting cancer cell invasion, metastasis, drug resistance, and PD-L1 recycling through the overexpressed ARF6 pathway activated by RTKs.
Glutamine metabolism
Glutamine addiction is common in various cancers, including PDAC (155-160). Glutamine may be a mitochondrial substrate for synthesis of macromolecules in cancer cells by supplying carbon to fuel the TCA cycle, and is a major nitrogen donor for the production of nucleotides and nonessential amino acids (155). In mitochondria, glutamine has essential roles in the synthesis of energy in the form of ATP through the TCA cycle and the OXPHOS process. Mitochondrial metabolism has been demonstrated to be important for tumor growth in several types of cancer, including PDAC (161, 162). Glutamine is the most abundant nonessential amino acid in the blood and plays various roles in cell metabolism (158, 163). Glutamine is first catalyzed to glutamate by the enzyme glutaminase. Glutamate is then converted to a-ketoglutarate through a deamination reaction catalyzed by glutamate dehydrogenase in the mitochondria. Subsequently, a-ketoglutarate enters the TCA cycle to supply metabolic intermediates, such as citrate and malate, producing NADH and FADH2 to generate ATP. Malate is converted to pyruvate leading to NADPH production, and then pyruvate is in turn transformed to lactate. Glutamine can also produce s u b s t a n t i a l a m o u n t s o f t h e co f a c t o r N A D P H b y glutaminolysis, in which malate is converted to pyruvate by malic enzyme. Glutamine-derived a-ketoglutarate is reductively carboxylated by mitochondrial isocitrate dehydrogenase 2 (IDH2) to isocitrate, which can then be isomerized to citrate. Citrate produced in the mitochondrial matrix is transported to the cytoplasm and then converted to isocitrate by aconitase in a reversible reaction. Cytosolic isocitrate is metabolized to aketoglutarate through cytosolic isoform of IDH1, which can also produce NADPH, which may be used for lipid synthesis. PDAC cells maintain cellular redox homeostasis, which is necessary for cell growth, by metabolizing glutamine in response to NADPH (157).
Circulating glutamine can be taken up via transporters, such as alanine-serine-cysteine transporter 2 (ASCT2, also known as SLC1A5), and can be exported or imported via large neutral amino acid transporter 1 (LAT1, also known as SLC7A5), in exchange for branched-chain amino acids (BCAAs; leucine, isoleucine, and valine). BCAAs are broken down by branchedchain amino acid transaminase 1 (BCAT1) on the cytosolic side and BCAT2 on the mitochondrial side to produce branchedchain a-keto acid and glutamate (Figure 2). Early-stage pancreatic cancer driven by mutant KRAS has been shown to increase plasma BCAA levels (164). BCAT2, but not BCAT1, has been shown to be highly expressed in PanIN and PDAC ductal cells. Thus, it has been noted that the BCAA-BCAT2 axis driven by KRAS is important for PDAC development (165). In addition, some amino acid transporters (ASCT2 and LAT1) are overexpressed in pancreatic cancer (166), and associated with poor prognosis. PDAC cells are known to be highly dependent upon glutamine for tumor growth (157, 167). However, whereas the treatment of BPTES, a glutaminase inhibitor to target the glutamine metabolism, significantly inhibited PDAC proliferation, it did not affect PDAC cell death. Glutamine deprivation has been reported to activate macropinocytosis-associated autophagy and maintain proper intracellular glutamine levels by regulating glutamine metabolism. Furthermore, both glutamine deprivation and autophagy inhibition have been shown to robustly activate apoptotic cell death (168). Glutamine plays various roles in PDAC metabolic processes, suggesting that therapeutic strategies targeting the acquisition and utilization of this amino acid may be promising. However, glutamine deprivation was shown to promote the EMT signature in vitro and in vivo through an increase in the EMT master regulator Slug via ERK signaling and ATF4 activation (169). Thus, evaluating the effects of the simultaneous inhibition of distinct aspects of glutamine metabolism, such as the induction of autophagy and EMT on PDAC growth and metastasis may lead to new therapeutic approaches.
Recently, comprehensive analysis of metabolic enzymes by large-scale targeted proteomics demonstrated an enhanced metabolic system in malignant cancers to utilize glutaminederived nitrogen for DNA synthesis (a shift in glutamine nitrogen metabolism) (170). In malignant cancer cells, the expression level of the metabolic enzyme phosphoribosyl pyrophosphate amidotransferase (PPAT), which transfers the nitrogen from glutamine to nucleic acid precursors, was markedly increased, whereas the metabolic enzyme responsible for glutaminolysis, namely, glutaminase (GLS) was decreased, indicating a shift toward nucleotide biosynthesis. In addition, meta-analyses of human cancers have shown that PPAT is most strongly associated with malignancy among the metabolic enzymes, particularly prominent in neuroendocrine cancers, including small cell lung cancer (SCLC) (170). Interestingly, the hazard ratio for PPAT is high in pancreatic cancer, whereas GLS expression does not significantly correlate with cancer prognosis. In PDAC mouse models, GLS inhibition does not demonstrate any anti-tumor effects in vivo, indicating an adaptive metabolic network that sustains proliferation (171). In cancers in which glutamine supply from the circulation is limited, such as PDAC, glutamine synthesis mediated by glutamate ammonia ligase, an enzyme involved in de novo glutamine synthesis, and the associated nitrogen assimilation and transfer to nitrogen-containing macromolecules, such as nucleotides, has been shown to be important (172). Thus, shifts in glutamine nitrogen metabolism that promote nucleotide biosynthesis via the increased expression of PPAT while suppressing the GLS response, as demonstrated in SCLC, are important in cancer malignancy, and may be a potential therapeutic target for pancreatic cancer in a glutaminelimited environment.
Autophagy/micropinocytosis
PDACs also rely upon metabolic pathways, such as autophagy and macropinocytosis, to survive and maintain metabolic homeostasis in harsh environments, such as those with low nutrient levels, hypoxia, desmoplasia, and high interstitial pressure. Autophagy is an indispensable intracellular pathway that provides intracellular energy by degrading unnecessary organelles and macromolecules in response to stimuli, such as starvation and accumulation of unfolded proteins (173). The molecular mechanism of autophagy is strictly regulated by more than 30 autophagyrelated (ATG) proteins that are responsible for the dynamic autophagy pathways, and can be divided into the following series of steps: phagophore (isolation membrane) growth, closed double-membrane vesicle (autophagosome) formation, autophagosome-lysosome fusion, degradation within the lysosome, and recycling of the degradation products.
One of the characteristic features of PDAC is known to be increased autophagy. This is because owing to the tumor microenvironment of PDAC, in which the low vascular density results in severe hypoxia and limited nutrient utilization (61,174), PDAC cells must rewire their metabolism to sustain proliferation. Indeed, the inhibition of autophagy by the genetic or pharmacological inhibitor chloroquine (an inhibitor of lysosomal acidification) resulted in mitochondrial metabolic abnormalities leading to decreased OXPHOS, reduced proliferation in vitro, and inhibited tumor growth in vivo (61). Furthermore, the significance of autophagy in PDAC tumorigenesis was confirmed by crossing a conditional knockout mouse of the autophagy essential gene Atg5 with a PDAC mouse model (175,176). This autophagy inhibition in mouse studies may exert anti-tumor effects by cooperating with the TME (177). Indeed, the crosstalk between stromal cells and tumor cells in PDAC is important, indicating that autophagy is required for stromal cells to secrete alanine, which is then taken up by tumor cells to support their growth (178). In a study using a PDAC mouse model expressing a tetracycline-inducible dominant-negative ATG4B protein which can reversibly and acutely inhibit autophagy in fully formed tumors, the inhibition of autophagy was shown to suppress tumor growth via intrinsic as well as extrinsic factors in tumor cells (61). This study also showed that the effect of inhibiting autophagy in the tumor itself on tumor regression was partially mediated by macrophages, indicating that induction of the immune system via autophagy inhibition is also important for the anti-tumor effects. This may mean that there is autophagy-dependent metabolic crosstalk between tumor cells and the stroma, and hence autophagy is necessary to support the metabolism, tumorigenesis, and survival under harsh conditions of tumors.
PDAC does not respond well to ICIs, such as anti-PD1 and anti-CTL4A antibodies, and typically has a highly immunosuppressive TME that is characterized by marked infiltration of myeloid-derived suppressor cells (MDSCs) and lack of active cytotoxic CD8 + T cells (179-182). Resistance to ICI therapy is known to be associated with major histocompatibility complex class I (MHC-I), which is essential for endogenous antigen presentation by cancer cells (183-185). PDAC cells have been shown to have reduced expression of MHC-I molecules on the cell surface, and instead localize predominantly to autophagosomes and lysosomes (186, 187). Indeed, it has been demonstrated in human and mouse PDAC that MHC-I is degraded by an autophagy-dependent mechanism to induce immune evasion (188). In addition, autophagy inhibition increased the surface levels of MHC-I, leading to the promotion of antigen presentation, enhanced anti-tumor activity of T-cell responses, and suppression of tumor growth in orthotopically transplanted syngeneic mice. Systemic autophagy inhibition by chloroquine, as well as the tumorspecific inhibition of autophagy, in combination with ICIs, showed synergistic effects. These findings provide a molecular mechanism by which autophagy promotes immune evasion, and provide a rationale for further research toward the development of new therapies targeting the autophagy-lysosome system in PDAC.
When glucose is deprived in PDAC cells, large amounts of reactive oxygen species are produced to activate autophagy, and provide the nutrients necessary for growth (189). On the other hand, glutamine starvation increases the degree of macropinocytosis in PDAC cells, and hence glutamine is important for regulating the degree of macropinocytosis in PDAC cells (190). Macropinocytosis is a process involving membrane ruffles, which are used to internalize extracellular materials, such as soluble molecules, nutrients, and antigens. After the nonspecific uptake of extracellular fluids by endocytic processes, the formation of vesicular structures, named macropinosomes, which contain the internalized proteins fuse with lysosomes, resulting in proteolytic degradation. The free amino acids produced by this process support the metabolic requirements of tumor cells (191). Thus, macropinocytosis is a nonselective endocytotic program capable of taking up content from extracellular fluid in a nutrient recycling and scavenging pathway that has been recognized as a key mechanism supporting pancreatic cancer growth (192).
PDAC cells expressing oncogenic KRAS mutation exhibited high enhancements of basal macropinocytosis consuming extracellular proteins for rapid tumor proliferation, which is closely linked to autophagy (174,(193)(194)(195)(196)(197)(198). It has been shown that autophagy is required for the micropinocytosis-mediated degradation of extracellular proteins, and autophagy plays an important role in the breakdown of macromolecules internalized by macropinocytosis, to provide amino acids, particularly glutamine, in PDAC cells (168). The dynamic balance between glutamine metabolism and macropinocytosis-associated autophagy may ensure PDAC cell growth. Although these studies suggest that macropinocytosis is a potential therapeutic target for PDAC, understanding how macropinocytosis and autophagy cooperate is crucial for establishing treatments for PDAC.
ARF6 has been shown to regulate autophagy and colocalize with proteins mediating the initiation of autophagosome formation, i.e., the formation of pre-autophagosomal structures and phagophores (199,200). Mechanistically, activation of the lipid-modifying enzyme PIP5K by ARF6 may contribute to autophagy, as PIP2 produced by PIP5K affects membrane trafficking for phagosome formation, by regulating plasma membrane endocytosis. Interestingly, ARF6 has been shown to be required for macropinocytosis in HT180 cells, a human fibrosarcoma cell line (201). In PDAC expressing high levels of ARF6, ARF6 may be a potential target for autophagy and micropinocytosis, and combination therapy, such as ICIs, may lead to a new treatment for PDAC. We also demonstrated that combination therapy with the eIF4A inhibitor silvestrol, which inhibits ARF6 protein production, and anti-PD-1 antibodies improves the efficacy of anti-PD-1 therapy in PDAC (202). However, it remains unclear whether ARF6 inhibition actually affects therapeutic efficacy by inhibiting autophagy and macropinocytosis.
Other types of metabolism
Amino acid availability in the TME, particularly arginine and tryptophan, is an important determinant of antitumor immunity. Increased arginine levels play an important role in T-cell activation by inducing metabolic changes, including a shift from glycolysis to OXPHOS, and the promotion of memory T-cell differentiation (203). Indoleamine 2,3-dioxygenase (IDO), which catalyzes the conversion of tryptophan to kynurenine, is often overexpressed in PDAC (204). Tryptophan depletion and kynurenine production in TME promote the establishment of a suppressive immune environment, and attenuate anti-tumor Tcell responses (205).
Extracellular ATP levels may be rapidly and robustly increased by hypoxia (206,207). ATP, which has immunostimulatory properties on its own, may be ultimately converted to the nucleoside adenosine through stepwise process. Canonically, ATP is first catalyzed to AMP via the ectonucleotidase CD39. AMP is then dephosphorylated by CD73 and degraded into adenosine. Adenosine can then act on purinergic receptors, such as A1, A2a, A2b, and A3 (208), and regulates various aspects of physiology and pathophysiology (209,210). A2a receptors and A2b receptors are primarily responsible for the downstream signaling of immunosuppression associated with intracellular cAMP accumulation (211). In PDAC, high expression of CD73 was demonstrated to be associated with an immunosuppressive TME and poor survival, as well as decreased CD4 + , CD8 + , and CD21 + TILs (212). Therefore, CD73 may also play a significant role in regulation of the immunosuppressive microenvironment of PDACs and promote their tumor progression.
Immunosuppressive TME in PDACs
The emergence of cancer immunotherapy, particularly ICIs, has offered hope to many patients with tumors that are not curable by conventional therapies. However, PDAC is known to be less sensitive to ICIs than other solid tumors, such as melanoma and lung adenocarcinoma. On the other hand, in PDAC patients, neoantigen quality has been shown to be associated with overall survival, suggesting that PDAC is associated with acquired immunity (213). In particular, the preclinical success of ICI therapy in PDAC patients with microsatellite instability (MSI high) and mismatch repair defects, as well as the therapeutic potential of autologous T-cell-based therapy in PDAC patients, holds promise for adaptive immune-based treatment strategies for PDAC (214, 215). At present, there is an ongoing study testing the effects of ICIs in patients with MSI-high PDACs (NCT02628067), which may provide insights into the subset of patients who respond to immunotherapy and the underlying mechanisms related to efficacy and resistance for ICIs. Overall, clinical results have been disappointing, but in some cases, correlative immunophenotypic studies have demonstrated that these therapies elicit adaptive T-cell responses. This suggests that immunosurveillance is operating in PDAC, however, a rational approach to countering its highly heterogenous and plastic immune evasiveness is needed.
TME of PDACs
Pancreatic cancer is known to have an immunologically cold microenvironment. Overall, immunosuppressive TME in PDAC is often associated with the presence of a tumor-promoting immune cell population (216). Analysis of PDAC mouse models has shown that the expression of oncogenic KRAS itself leads to robust inflammation, and initiates the cycle of inflammation associated with carcinogenesis (11, 179, 217, 218). Furthermore, whereas the expression of KRAS mutant during embryogenesis is sufficient to promote the onset of PDAC proliferation, chronic inflammation is required for malignant transformation in adult PDAC mouse models, indicating that oncogenic mutations alone cannot induce PDAC malignancy (97, 219, 220). Therefore, the inflammatory environment and oncogenic mutations work in concert to promote tumor progression. Thus, inflammation caused by cytokines and chemokines released from PDAC cells that have acquired mesenchymal traits is often associated with the infiltration of innate immune cells that facilitate an immunologically tolerant environment rather than an antitumor immune response (221). A low level of T-cell infiltration correlates with mortality in PDAC (222). Biochemical (production of chemokines and other factors in TME) and physical (deposition of the ECM) barriers in the stroma surrounding the TME inhibit T-cell infiltration (Figure 3).
Fibrosis
Although there are multiple factors that cause ICI treatment resistance, one of the main possible contributors is a dense fibrous stroma (desmoplasia) occupying 80% to 90% of the tumor mass in PDAC (223,224). Desmoplasia is caused by the proliferation of a-smooth muscle actin-positive fibroblasts or activated pancreatic stellate cells, and work as a physical barrier against drug and immune cells. The trigger that causes these cells to proliferate is still unknown, but the communication among tumor cells and these cells have been identified. Two main components constitute desmoplasia: cells including fibroblasts and infiltrating immune cells, and noncellular proteins, such as collagen types I, III, and IV, fibronectin, and hyaluronan. A comprehensive review of the pancreatic cancer stroma has been published recently (15).
Heterogeneous fibroblasts
Fibroblasts exist in every solid organ, to maintain their morphology and function by depositing ECM proteins and secreting soluble factors (225). For instance, TGF-b secreted from fibroblasts is used by epithelial cells to cure skin injuries. Histological similarities, such as mesenchymal morphology, are maintained among fibroblasts in various organs, but their genomic landscapes differ depending on the organ in which they are located (226). Many studies have demonstrated that some fibroblasts contribute to tumor initiation, progression, and metastasis, and they are known as CAFs (227). Pancreatic cancer has a dense fibrotic architecture, and therefore, it will be useful to clarify the biology of CAFs in PDAC. Recent studies have demonstrated that the functional roles of CAFs in PDAC TME are more complicated than their expected simple tumorpromoting role (15).
Three-dimensional in vitro coculture of pancreatic stellate cells (PSCs) and KPC mouse-derived PDAC organoids induced two kinds of CAFs (228). Cocultured directly, PSCs turned into myofibroblastic CAFs (myCAFs) with highly upregulated a-SMA expression and myofibroblastic gene profiles (228). Although CAFs are thought to literally be 'associated' with tumors, myCAFs have anti-tumor activity (229), which requires further investigation for elucidation of the mechanism. On the other hand, indirect coculture transforms PSCs into inflammatory CAFs that display inflammatory cytokines, such as IL-6. It is thought that CAF-derived IL-6 contributes to immune evasion (228). scRNAseq by several investigators supported the existence of these two populations, and moreover, identified two other groups, namely, mesenchymal stem cell CAFs (mscCAFs) and antigen-presenting CAFs (apCAFs) (230, 231). MscCAFs are characterized by the expression of previously identified mesenchymal stem cell markers (CD44, CD49a, CD73, and CD90), and originate from bone marrow (232). They preferentially express GM-CSF, thus promoting macrophage polarization towards an immunosuppressive phenotype that results in the inhibition of CTL activity (179, 233,234). Coexpression of MHC class II, including CD74 and Immunosuppressive TME in PDAC. One of the features of PDAC is dense fibrosis, which limits immune-cell filtration, drug delivery, and oxygen supply. A variety of cells exist and compose the immunosuppressive milieu. Each type of fibroblast appears to play a key pro-tumor or antitumor role. The EMT of PDAC is among the factors enhancing anti-tumor immunity through not just cell intrinsic functions, such as PD-L1 recycling, but also crosstalk with other immune cells.
podoplanin, a pan-CAF marker, is a signature of apCAFs (230). It has recently been reported that apCAFs are derived from mesothelial cells and can induce the transformation of naïve CD4 + T cells into regulatory T cells by their direct ligation with a specific antigen (235). This group also demonstrated that targeting mesothelin expressed in the mesothelium may be an effective treatment owing to the inhibition of apCAF formation and regulatory T (Treg) cell induction (235).
T cells
T-cell exclusion in tumors is primarily mediated by CAFs that express fibroblast activation protein, and secrete the chemokine CXCL12 (236). Additionally, activation of integrinbinding protein and non-receptor tyrosine kinase focal adhesion kinase (FAK) is associated with increased collagen I deposition and immunosuppression (237,238). Highly phosphorylated FAK levels in pancreatic cancer patients were associated with decreased tumor-infiltrating CD8 + T cells and reduced survivability (239). In addition, the expression of FAK in patients with PDAC is associated with decreased tumor cellularity and survival (239).
Acquiring a terminally differentiated T-cell state can diversely impact disease outcome, either countering tumor proliferation through antigen-limited tumor-killing immune responses, or promoting cancer progression by actively inducing immunosuppression (216, 240, 241). In particular, CD8 + cytotoxic T lymphocytes and polarized CD4 + T cells known as T helper type 1 (Th1) cells exert protective effects against tumors in PDAC mouse models, and have been shown to be associated with prolonged survival of human PDAC patients (242). Conversely, CD8 + T-cell deficiency, low amounts of neoantigens, and CD4 + Th2 and Treg cells are associated with tumor-permissive anergy (242-245). Cytokines produced by Th2, particularly IL-4 and IL-13, can not only reduce antitumor immune responses, but also can directly accelerate tumor growth induced by KRAS transformed cells (246). PDAC tumors are also accompanied with abundant lymphocyte infiltrates that are typically associated with the gastrointestinal mucosa (247). Th17 cells comprise approximately 5% of the CD4 + T cells in PDACs. The role of Th17 cells in the TME is also contextdependent. In PDAC, IL-17 secretion from gd T cells and Th17 cells may enhance antitumor immune responses (248). However, early stages in PDAC carcinogenesis, IL-17 has a direct mitogenic effect on KRAS mutation-induced PanIN cells expressing IL-17R (218). Whereas the effects of distinct T-cell subsets depend on the underlying immune context of the tumor due to various physiological conditions and environments, and may be altered during the tumor progression of PDAC, the regulation of differentiation and function of T cells in PDAC TMEs play crucial roles in tumor immunity.
B cells and myeloid cells
To date, it has become clear that distinct cell populations derived from lymphocyte and myeloid cells can act in a pro-or anti-tumor manner, depending on the situation. B-cell subsets have become apparent as key immunomodulators in PDAC (249-252). Furthermore, suppressive myeloid cell programming is a major cause of tolerogenic T-cell programming in PDAC. Macrophages are thought to serve a major function in the induction of immunosuppression in PDAC; IL-10 + Arg1 + MHCII lo tumorassociated macrophages (TAMs) are predominant in the PDAC TME, and are effective in promoting Th2 cell differentiation, but ineffective in inducing CD8 + T-cell immunity (253)(254)(255)(256)(257). Similarly, immature MDSCs, collectively referred to as bone marrow-derived Gr1 hi CD11b + granulocyte lineage MDSCs, are characterized by a short half-life and strong suppressive effects in the TME (258). Although endogenous normal dendritic cells (DC) in the TME can produce anti-tumor T cells, the number of DCs in PDAC is low and probably insufficient to sustain robust adaptive immune responses. Furthermore, tumor-derived colony-stimulating factor 3 is found to inhibit the development process of DC in the bone marrow (259). Certain DC subsets have been understood as activators of immune evasiveness in PDAC. CD11b + CD103 − DCs with high expression of IL-23 and TGF-b are predominant in PDAC, drive the differentiation of FoxP3 − tumor-promoting type I + T cells, and promote metastatic spread (260,261). Moreover, it has been shown that Treg cells directly interact with tumor-associated DCs and suppress anti-tumor immunity by downregulation of costimulatory ligands expression that are important for activation of CD8 + Tcell (262).
Stimuli that recruit myeloid cells to the TME in PDAC are only partially understood. In mouse models, tumor-derived factors have been presented to accumulate MDSCs in the PDAC TME. In the same way, CCL2 produced by tumor cells and CSF1 produced by tumor-associated fibroblasts contribute to the generation of M2-like macrophages (263,264), and CXCL1 production by tumor cells has been linked to increased myeloid cell populations and decreased tumor infiltration of cytotoxic CD8 + T cells (265). In particular, focusing on the CSF1-CSF1R, CCL2-CCR2, and CXC chemokine-CXCR2 axes to target in the PDAC TME may contribute to pancreatic cancer progression.
Microbiota
The normal pancreas has long been believed to be a sterile, protected site from bacteria. However, recent studies have shown that the pancreas contains bacteria that invade through the Vater's vastus. Interestingly, it has been reported that in the inflammatory environment of PDAC, the bacterial content of pancreatic tumor tissue increases by approximately 1,000-fold compared with normal tissue (266)(267)(268). Furthermore, the bacterial species found in the tumorigenic pancreas are different from those in the gut, and low microbial diversity in the tumor results in a low survival rate of patients with PDAC, whereas high tumor microbiome diversity is associated with long-term survival (269). Mechanistically, the primary PDAC microbiome has a potent immunosuppressive effect on the inflammatory TME, driving the protumor inflammatory responses of PDACs via the activation of Toll-like receptors on bone marrow-monocyte cells, and inducing the expansion of MDSCs and anti-inflammatory M2-polarized macrophages. These innate immune cells with tolerogenic functions enhance the differentiation of immunosuppressive CD4 + T-cell populations and inhibit the expansion of cytotoxic CD8 + T cell populations (268). Consistently, microbial ablation in mice resulted in increased infiltration of Th1-polarized CD4 + and CD8 + T cells, decreased accumulation of MDSCs, and a TAM reprogrammed to a tumor-protective M1-like phenotype (268,270). Potentially, targeting the microbiome by oral antibiotics might reverse myeloid cell-mediated adaptive immunosuppression and promote the efficacy of ICI therapy in PDAC.
Novel mechanisms bridging mesenchymal malignancy and immune evasiveness via rewiring of the metabolic program of PDAC Immune evasion is an essential characteristic of cancer. Every day, the adult body produces mutant cells owing to genetic mutations via various intrinsic and extrinsic factors, and most mutant cells are detected and eliminated by the immune surveillance system. However, in rare cases in which mutant cells acquire traits that enable them to evade the immune surveillance system, the cells evade attack by immune cells and proliferate, eventually manifesting as cancer.
As mentioned above, the major immunosuppressive factors in the TME of PDAC include hypoxia, a low-nutrient environment, expression of immune checkpoint molecules, accumulation of immunosuppressive cell populations, such as Tregs and MDSCs, production of immunosuppressive cytokines, such as TGF-b and IL-10, immunosuppressive metabolic enzymes, such as Ido, arginase, and CD39/CD73, and metabolites, such as lactate and kynurenine. In addition, cancer-associated inflammation induced by IL-6, IL-1, IL-17, IL-22, and IL-23 is not only a driver of carcinogenesis, but is also associated with tumor progression by inducing EMT, whereby epithelial cells acquire malignant mesenchymal properties, such as detachment from other cells, invasion into adjacent tissues, and accelerated metastatic spread to other distant organs (21, 271-275). Thus, it is easy to speculate that the factors involved in the suppression of the immune environment of the TME are diverse and complex in their mechanisms of action, as they are produced not only by cancer cells, but also by the various stroma cells, including several kinds of immune cells and heterogenous fibroblasts in the TME.
On the other hand, international cancer-related consortiums, such as The Cancer Genome Atlas (TCGA) have promoted comprehensive genome-wide gene expression analyses of various cancers, but these efforts have not led to the development of effective diagnosis and treatment methods, particularly in the case of PDAC. There may be various reasons, such as the fact that the collected tissue sections are bulk preparations containing not only cancer cells but also stromal cells. Recently, the transcriptome and proteome have been compared worldwide, and it has been shown that there is a very poor correlation between the mRNA and protein levels of most genes (276)(277)(278)(279)(280). This strongly suggests that posttranscriptional mechanisms play an important role in the regulation of gene expression. Here, we present our studies from two different aspects on the molecular mechanisms linking the acquisition of mesenchymal plasticity and immune evasion in PDAC, with a focus on post-transcriptional mechanisms.
Functional roles of ARF6-AMAP1 axis as a mesenchymal executioner in PDAC ARFs, a family within the Ras superfamily of small GTPases, are evolutionally the most ancient of the small GTPases. The ARFs are conserved throughout eukaryotes, including in species that branched off early, such as Giardia lamblia, in which no members of the Ras family nor heterotrimeric G-proteins are found (281, 282). Giardia lamblia is an anaerobic eukaryote parasite of the gut, which is evolutionally inferred to be an amitochondrial-type eukaryote that developed before the creation of mitochondria (283). This implies that eukaryotic cell features, such as nuclei and flagella, predate mitochondrial endosymbiosis, suggesting that ARF family molecules have been deeply involved in the maintenance of life homeostasis under anaerobic conditions during the evolution of eukaryotes. The human ARF family consists of six isoforms, ARF1-6, which are classified into three classes based on sequence homology, as follows: class I (ARF1-3), class II (ARF4-5), and class III (ARF6) (284). Class I and class II Arfs primarily regulate vesicular transport between the Golgi and endoplasmic reticulum (284, 285). Although ARF6, the only class III member, has virtually identical effector-interacting domains as the other ARFs, it is the most divergent of the ARF proteins, and predominantly localizes to the plasma membrane and recycling endosomal compartments, and functions in intracellular events associated with membrane dynamics, including recycling of plasma membrane components (including both endocytosis and recycling-back to the plasma membrane), as well as in actincytoskeletal rearrangement at the cell periphery (286)(287)(288).
We identified the ARFGAP protein AMAPs as molecules that are induced during macrophage differentiation, bind to the integrin-associated protein, paxillin, and are involved in its intracellular dynamics. Furthermore, we found that AMAPs are ARF6-specific ARFGAP proteins that are commonly involved in enhancement of the cell motility of macrophages and epithelial cells (289-291). In addition, we identified a novel mechanism of action in which AMAP functions as an effector of activated ARF6 through steady-state binding to GTP-bound ARF6 via its ARFGAP domain in the presence of Mg 2+ (290,291). Consistently, Wittinghofer and colleagues demonstrated that Ca 2+ spikes stimulate the ARF6-specific GAP activity of AMAPs, but not other members of the ArfGAP family (292).
Subsequently, we identified GEP100 as a guanine-nucleotide exchange factor that activates Arf6 in the acquisition of invasive and metastatic traits of breast cancer cells upon activation of the epidermal growth factor (EGF)receptor pathway (145). We also identified the mechanism of action by which GEP100 activates Arf6 by binding directly to the phosphotyrosine moiety of the activated EGF receptor via the Pleckstrin-homology domain. Furthermore, we found that the simultaneous expression of Arf6 and GEP100 in MCF7 human epithelial-like breast cancer cells induced EGF-stimulation-dependent EMT-like changes. Subsequently, pathological analysis demonstrated that GEP100 expression is present in approximately 80% of invasive breast cancers (145). Our present study suggests that Arf6-based signaling pathways play an important role in the acquisition of invasive and metastatic traits via EMT induction in cancer cells. In this pathway, AMAP1 binds to different proteins, such as cortactin, and protein kinase D2 to promote cortical actin remodeling and integrin recycling (293, 294). AMAP1 also binds to EPB41L5 (148, 149), which shows increased expression during TGF-b-induced EMT (295). Furthermore, we demonstrated that the EMT-TF ZEB1 is involved in EPB41L5 gene expression, and that high expression levels of ZEB1 and EPB41L5 in cancer cells are associated with p53 mutations. This study demonstrated that the ARF6 pathway is a signaling pathway responsible for advanced cancer-specific mesenchymal traits associated with mutant p53 (296).
A series of our studies have identified that high protein levels of ARF6, AMAP1, and EPB41L5 were associated with invasiveness of several kinds of solid tumors, including breast cancer, clear cell renal cell carcinoma, lung adenocarcinoma, and PDAC and importantly that these expression levels were statistically correlates with poor prognosis (94, 145, 146, 149).
Notably, ARF6 and AMAP1 mRNAs are both rich in G/C content in their 5′-untranslated regions (UTRs) (74% and 88%, respectively) (297). Moreover, ARF6 mRNA contains a G-quadruplex structure at the 5′-UTR (94), indicating that efficient translation is dependent upon the RNA helicase eIF4A, a member of Cap-dependent translation initiation factors (298, 299). On the other hand, the 5′-UTR of AMAP1 mRNA contains a 5′-terminal oligopyrimidine-like sequence, indicating the mTOR complex 1 kinase-dependent translation control (300,301). We found that the eIF4A inhibitor silvestrol suppresses protein levels of ARF6 in KRAS mutant cells, but only moderately in KRAS intact cells (202), and the mTOR inhibitors rapamycin and Torin1 suppress AMAP1 expression in KRAS mutant cells (94). Mechanistically, oncogenic KRAS mutations are the major cause of the aberrant overexpression of ARF6 and AMAP1, in which KRAS signaling enhances eIF4A-dependent ARF6 mRNA translation and eIF4E-dependent AMAP1 mRNA translation. In addition, gain of function mutations of TP53 promoted the activation of ARF6 by PDGF via MVPmediated geranylgeranyl lipid modification of Rab11b in PDAC cells (94,148). Moreover, we revealed that the ARF6-AMAP1 pathway is closely associated with immune evasion in a KPC mouse model. Thus, the cooperation between eIF4A/4E-dependent mRNA translation and MVP has been identified as a link in which representative pancreatic driver mutations empower an ARF6-based pathway, activated by external ligands, to promote tumor cell motility, PD-L1 dynamics, and immune evasion. A recent clinical study by another group confirmed the importance of ARF6 in this context (302). We hence propose that targeting eIF4A, or eIF4E, as well as mutant KRAS, provides novel methods to improve the efficacy of anti-PD-1 therapy, in which ARF6 and AMAP1 overexpression may act as biomarkers to identify patients with drug-resistant disease in PDAC. Additionally, the ARF6-AMAP1 pathway was also found to be involved in acidosis and fibrosis of the TME, both of which are well known to be barriers against immune attack to cancer cells (81,303), indicating that the ARF6-AMAP1 pathway may also be a valuable target in modifying the TME from pro-tumor, which makes PDAC resistant to treatment, towards an anti-tumor state. Taken together, given the importa nce of the A RF 6-AMAP1 pathway in the pathophysiology of PDAC, its clinical application as a therapeutic target may broaden options for the treatment of PDAC (Figures 2, 3).
Arid5a acts as a dual regulator in malignant PDAC to generate an immunosuppressive TME Arid5a was identified as an RNA-binding protein that binds directly to the 3′-UTR of Il6 and stabilizes Il6 mRNA (304). Recent studies have demonstrated that Arid5a plays an important role in innate and adaptive immune responses (305). In macrophages and embryonic fibroblasts, stimulation by LPS, IL-1, and IL-6 induces Arid5a expression (304,305). Importantly, in untreated rheumatoid arthritis (RA) patients, expression of Arid5a in CD4 + T cells is increased, whereas treatment with the anti-IL-6 receptor antibody tocilizumab is associated with decreased Arid5a expression (306), indicating that the IL-6-ARID5a axis may be involved in RA pathogenesis. Consistently, Arid5a has been shown to be involved in several immune-associated pathologies. For example, Arid5a deficiency reduces IL-6 production under LPS-induced endotoxemia. Furthermore, in an experimental autoimmune encephalomyelitis (EAE) model, Arid5a deficiency significantly suppresses Th17 cell differentiation and lowers IL-6 serum levels, resulting in the reduced development of EAE (304). In addition, Arid5a regulates the stability of mRNAs for other genes involved in immune regulation, such as Stat3, Tbx21, Ox40, and Il17 (307-310). In addition, IL-6 increases its own mRNA stability by increasing Arid5a levels via a positive feedback loop (311). Consistently, Arid5a-deficient mice show impaired LPS-stimulated Il6 and Ifng expression, and are resistant to lethal endotoxic shock (304,308). Thus, Arid5a-mediated upregulation of these factors may be involved in the enhancement of Th1 and Th17 cell polarity and function in acute inflammatory responses and autoimmune diseases.
Several cytokines have been shown to be actively involved in metabolic reprogramming in physiological and pathological conditions (312). For example, during cancer cachexia, the overproduction of cytokines significantly increases energy expenditure and leads to weight loss (313). In particular, circulating levels of IL-6 have been shown to positively correlate with cachexia in cancer patients, and importantly, IL-6 levels were found to negatively associate with their survival (314)(315)(316)(317). Furthermore, treatment with the humanized anti-IL-6 receptor antibody tocilizumab increased body weight and serum levels of triglycerides and cholesterol in human cancer patients (318). Il6deficient mice have been shown to develop adult-onset obesity with impaired glucose and lipid metabolism (319). The overexpression of IL-6 in high-fat diet-induced obese mice reduced their body weight and improved their obesity-induced fatty liver and insulin resistance (320). Consistently, Arid5a −/− mice showed reduced IL-6 production; mice with long-term loss of Arid5a developed adultonset severe obesity. In contrast, mice with forced expression of Arid5a are highly resistant to high-fat diet-induced obesity (321). These results suggest that Arid5a is involved in IL-6-mediated metabolic regulation.
Recently, we showed that Arid5a mRNA and protein expression levels were significantly increased in mesenchymal tumor subtypes of PDAC and colorectal cancer (CRC), such as quasi-mesenchymal and consensus molecular subtype 4 subtypes, respectively. In addition, Arid5a expression was enhanced in in vitro EMT models, induced by IL-6 and TGF-b stimulation (322) (Figure 4). Furthermore, Arid5a enables mesenchymal tumor models of PDAC and CRC to facilitate immune evasiveness via promoting tumor infiltration of immunosuppressive granulocytic MDSCs (gMDSCs; also known as polymorphonuclear MDSCs (323)) and Tregs (324), and suppressing the recruitment and activation of anti-tumor effector T cells (322). Interestingly, Arid5a acts as a dual regulator leading to the formation of immunosuppressive TMEs in malignant tumors, triggering the metabolic reprograming and recruitment of suppressive immune cells. First, Arid5a induces the inhibitory effect of Ido1 on effector CD4 + /CD8 + T cells via the post-transcriptional stabilization of Ido1 mRNA by binding to its 3′-UTR, and a reduction in intratumoral tryptophan concentration (325,326). Additionally, Ido1 expression in tumor tissues promotes Treg differentiation/activation by generating kynurenine through tryptophan catabolism, and ultimately activating aryl hydrocarbon receptors (AhR) (327, 328), and AhR activation extensively mobilizes gMDSCs (329). Second, Arid5a upregulates chemokine Ccl2 expression in the TME via posttranscriptional stabilization of its mRNA, and then Ccl2 leads to enhancement of the infiltration of immunosuppressor cells, such as Tregs and gMDSCs (330)(331)(332)(333), to the TME (322).
Therefore, these findings provide insights into the molecular basis of the acquisition of mesenchymal plasticity and immune evasiveness by PDAC and CRC via augmentation of the RNAbinding protein Arid5a, and indicate that Arid5a is a promising target for tumor immunotherapy, in addition to inflammatory diseases ( Figure 4).
Conclusion and perspectives
In tumorigenesis, metabolic changes and chronic inflammation associated with genetic mutations in normal cells enable transformed cells to escape the homeostatic defense mechanisms of tissues, and to reprogram their intrinsic signaling mechanisms, as well as reprogram populations of stromal cells within the TME and the metabolic balance of the entire organism. In this process, tumor cell populations that adapt to the abnormal microenvironment form diverse, hierarchically organized colonies, and eventually acquire mesenchymal plasticity that promotes their dissemination, reduces the immune system's ability to counter tumor growth, and finally directly causes death of the organism. Elucidating the metabolic adaptations that tumors rely on to promote these changes and maintain growth in a metabolically unfavorable environment, as well as the molecular mechanisms that trigger the acquisition of mesenchymal plasticity and immune evasion capacity, will help towards developing new diagnostic and therapeutic approaches and dietary combinations for the treatment of PDAC.
Increased levels of IL-6 in the serum have been associated with poor overall survival prognosis in patients with high-grade pancreatic cancer (334), and the increased activity of IL-6/ STAT3-mediated signaling has been reported to be associated with poor prognosis in post-resection PDAC patients (335). IL-6 also activates STAT3 and induces the mesenchymal phenotype in human pancreatic cancer cells via the induction of SNAI1 (336). In chronic pancreatitis, IL-6/STAT3-mediated ADM transdifferentiation occurs and is associated with PanIN, which is a necessary step for the generation of tumorigenic precursor lesions (220, 337). For example, in the KRAS-induced PDAC mouse model (220, 337), pancreatic epithelial cells with constitutively active KRAS mutations (KRAS G12D ) have been reported to cause inflammatory activation by recruiting immune cells. In particular, myeloid cells have been reported to promote the production of IL-6 and soluble IL6R (sIL6R), activate STAT3 via IL-6 trans-signaling, and furthermore, the complex of IL-6 and sIL6R binds to gp130-expressing cells (220). Aberrant STAT3 activation owing to the homozygous loss of SOCS3 in the pancreas results in the accelerated progression of PanIN and the development of PDAC (220). It has also been shown that KRAS activation increases the levels of cytokines, such as IL-6 and IL-11, in epithelial cells, followed by STAT3 activation in an autocrine manner, and that STAT3-triggered matrix metallopeptidase 7 is required for tumor progression but not tumor development, and may be regulated by other STAT3 targets (337). As already mentioned, the TME of PDAC is severely hypoxic, and nutrient availability is limited by a low vascular density, so PDAC cells show increased autophagy that rewires their metabolism to enable survival in a harsh environment, and to maintain metabolic homeostasis. In a mouse model of PDAC caused by KRAS mutations, IL-6induced STAT3 activation was shown to be involved in the increase in autophagy. As a mechanism, receptors for advanced glycation products have been reported to promote the IL-6driven activation of STAT3 signaling in mitochondria, bridging B A FIGURE 4 Involvement of Arid5a in the acquisition of mesenchymal plasticity and immune evasiveness in PDAC (A) Arid5a expression is associated with acquisition of the mesenchymal phenotypes of PDAC and CRC. Especially, cells showing partial EMT and mesenchymal-like cell lines show much higher expression levels of ARID5A than epithelial-like cell lines. During TGF-b-or IL-6-induced EMT, Arid5a level is augmented in cells that have acquired mesenchymal phenotypes. (B) Arid5a acts as a dual regulator in malignant tumors, such as the mesenchymal subtypes of PDAC, to promote an immunosuppressive TME; Arid5a upregulates Ido1 expression via post-transcriptional stabilization of its mRNA and then enhances the suppressive effects of Ido1 on anti-tumor immune cells, such as CD8 + T cells and CD4 + Th1 cells via a reduction in intratumoral tryptophan concentration, and a promotion of the differentiation and activation of Treg cells. Additionally, Arid5a post-transcriptionally induces the expression of the chemokine Ccl2 in the TME, which recruits immunosuppressive cells, such as Treg cells and gMDSCs, to the TME. autophagy and the IL-6/STAT3 signaling pathway (338). Furthermore, IL-6 signaling has been implicated in the pathogenesis of cachexia in PDAC patients, by inducing a metabolic rewiring (339). Thus, it is clear that the activation of IL-6/STAT3 signaling is involved in the development of PDAC from the PanIN stage, continuing to malignant transformation.
Since its approval in 2009, tocilizumab has been shown to inhibit IL-6/STAT3 signaling in patients with autoimmune diseases, such as rheumatoid arthritis caused by the overexpression of IL-6, acute inflammatory diseases caused by chimeric antigen receptor T-cell therapy, and cytokine storms associated with SARS-CoV-2 infection. On the other hand, in clinical practice, few effective therapeutics have been developed as cancer treatments targeting IL-6/STAT3 signaling (340-343). As mentioned above, cancer is caused by a complex interplay of diverse cell populations, which leads to malignant transformation. Therefore, analysis of the expression and function of molecules associated with IL-6/STAT3 activation may enable the assessment of the local malignant potential and steady state of cancer, but may not be sufficient to predict the stage and detailed course of cancer. Furthermore, it has become clear that not only IL-6/STAT3 signaling, but also various groups of molecules are involved in cancer development. The mode of interaction between these molecules also requires further study.
In the future, it will be essential to introduce spatiotemporal gene expression analysis technology that analyzes multiple cell populations, improve detection technology to analyze the associations among aging, inflammation, and metabolism, and develop artificial intelligence technology to analyze cancer development and progression, and mathematical analysis technology to integrate these technologies. To this end, it is also indispensable to enhance the convergence of life science, physical science, engineering, and computational science to create the next generation of cancer diagnostics and therapeutics.
Author contributions
AH, HH, and ShH conceived and designed the manuscript. All authors wrote the manuscript, and approved the final manuscript.
Funding
This work was supported by Grants-in-aid from the Ministry of Education, Science, Sports and Culture of Japan to AH (grant no. 17K08614 and 22K06890) and ShH (grant no. 17K07151 and 22K07203), grants from the Suzuken Memorial Foundation to AH, and Kishimoto foundation to ShH.
|
v2
|
2022-11-03T18:15:43.081Z
|
2022-11-03T00:00:00.000Z
|
253260854
|
s2orc/train
|
Frailty affects prognosis in patients with colorectal cancer: A systematic review and meta-analysis
Background The prevalence of colorectal cancer has remained high. Most patients have already developed into the middle and advanced stage when they are diagnosed with colorectal cancer, and a small number of them are accompanied by metastasis. In recent years, frailty has been recognized as an important factor affecting the prognosis of colorectal cancer. The aim of this study was to assess the value of frailty on prognosis in patients with colorectal cancer after treatment. Method We systematically searched PubMed, Embase, Web Of Science databases up until March2022. A total of 18 studies were retrieved that met the inclusion criteria, including 9 prospective studies and 9 retrospective studies. Frailty screening tools, proportion of frail patients, and outcomes of colorectal cancer patients after treatment were recorded. Result 18 studies were included with a total of 352,535 participants. Regardless of differences in frailty screening and treatment approaches, outcomes for frailty patients were less favorable in all studies. Compared with the non-frail group, the frail group had higher mortality, more serious complications, more postoperative blood transfusions and delirium, and more support outside the home. Conclusion Although there is no uniform standard for frailty screening, assessing the frailty of colorectal cancer patients is of great significance for predicting prognosis of patients after treatment.
Introduction
Colorectal cancer has become the third most common cancer in the world and the second most deadly cancer in the world (1). It mainly occurs in the elderly, with the highest incidence around the age of 80 (2). Although the standard of therapy for rectal cancer remains surgery with or without neoadjuvant therapy (3), proportion of elderly patients undergoing surgery declines with age due to frailty (4). Frailty is a complex multifactorial syndrome, characterized by a clinically significant increase in vulnerability and worsened health outcomes (5). It affects morbidity and mortality in patients with various cancers (6,7). Frailty is not only seen in older patients, but younger adults can also fulfil the criteria for frailty (8). Young colorectal cancer patients should also be a group of concern. Cancer patients and those undergoing surgery are more likely to be infirm and have more adverse outcomes than those who are not infirm (9). As a result, oncology societies such as the International Society for Geriatric Oncology (SIOG) recommend frailty screening for older cancer patients (10). Although, some studies have been conducted on frailty and postoperative outcomes and prognosis and a number of frailty screening tools have been invented to assist clinicians in diagnosing (11)(12)(13), there is no standard assessment (14). What role does frailty play in the progression of colorectal cancer patients, and What changes have it brought to the prognosis of colorectal cancer patients? There are still many controversies in many studies. We decided to conduct further study on this.
Search strategy
The PubMed, Embase, Web Of Science databases were searched to identify all studies describing frailty and colorectal cancer. The search terms used were related to the following key words: "colon", "rectum", "tumor", "colorectal cancer", "frailty". The search string is included in detail in Appendix A. The search was completed on April 5, 2022. This study was conducted in accordance with established guidelines [PRISMA (15)].
Frailty screening tools
The range of frailty tools available to researchers and clinicians can be overwhelming. Due to the diversity of tools, we recommend choosing a frail tool for clinical or research application in cancer patients based on 1) the intent and feasibility of applying the tool to practice and 2) considering specific clinical or research needs, while also taking into account the limitations of available data. The commonly used screening tools are Canadian Study of Health and Aging-Clinical Frailty Scale (CSHA-CFS) (16), American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) Modified frailty indices (17,18), the Edmonton Frail Scale (19), Groningen Frailty Indicator (20), The Kihon Checklist (KCL) (21), Onco-geriatric G8 questionnaire and frailty phenotype (22) and etc.
Inclusion, exclusion and quality evaluation
We set inclusion and exclusion criteria for prospective and retrospective studies. The degree of frailty must be determined in a clinical setting, and patients were screened for frailty and divided into 2 groups. He specific inclusion and exclusion criteria are as follows.
Inclusion criteria:1. Study type are Case-control studies, cohort studies, cross-sectional studies or RCT;2. Screening of frail patients applies to internationally recognized frailty screening tools;3. Divide patients into frail and non-frail groups for study;4. Primary colorectal cancer without combining other tumors;5. Full text is available, and the data is complete;5. Published publicly, excluding meeting minutes and reviews.
Exclusion criteria:1. non-clinical research;2. Fail to identify or diagnose frailty;3. Patients were not divided into frail and non-frail groups for study;4. non-primary colorectal cancer or colorectal cancer combined with other types of tumors;5. The full text and complete data are not available.
Assignments
Two independent investigators (J.Y.P and L.J.Y) assessed the studies for eligibility, reached consensus by discussing which studies to include. When two investigators disagreed, a third investigator (C.M.H) was asked to decide on eligibility.
Extracted study data by (C.M.H), including the first author, publication year, study population, study type, sample size of frail patients, frailty assessment tools, and patient quality methods.
Quality assessment was done by two members of the research team (J.Y.P and L.J.Y). They assessed the quality of included studies using the Newcastle-Ottawa Cohort Study Scale. This scale was used to assess 8 questions in three domains. One point is awarded for each satisfactory answer, with a maximum of 9 points. When the score is greater than 5 points, it is considered to be eligible for inclusion. Each study was rated as low (6 points), moderate (7-8 points) or high quality (9 points). If the scores are inconsistent, they will be resolved through negotiation. The evaluation score is shown in Table 1.
Research indicators
The indicators we observed were the outcomes of colorectal cancer patients (frail and non-frail groups) after treatment. The primary outcome measure was mortality and complication rate, and the secondary outcome measures were delirium, postoperative blood transfusion, discharge destination other than home, readmission, and length of hospital stay.
Statistical analysis
We extracted data from all publications to calculate standard mean difference (SMD) and associated 95% confidence interval (CI) for continuous outcomes. P<0.05 was considered statistically significant. The presence of statistical heterogeneity of the results was assessed by using the I 2 measure, with I 2 >50% considered significant when P ≤ 0.10. If there was no We performed a meta-analysis of all studies, and Subgroup analyses were also performed for mortality and complication classification. STATA15.1 software was used for the standard meta−analysis and the sensitivity analysis.
Result
After initial screening of PubMed, Embase and Web Of Science databases, a total of 2373 studies were identified: 704 from Web Of Science, 701 from PubMed, and 968 from Embase. We performed title/abstract screening and full-text reading after adding constraints such as publication year, repetition, full-text reviews, and cross-references. Finally, 18 studies and 352,535 patients were included in this meta-analysis. Of the 18 studies, 9 were from Europe, 3 from North America, 5 from Asia, and 1 from Oceania. All included subjects were over 18 years old, and mainly consisted of the elderly over 65 years old. The ratio of males and females is relatively equal. The retrieval process is shown in Figure 1.
Among the 18 studies, retrospective studies and prospective studies each accounted for 9. Mortality was assessed in 12 studies. Complications was assessed in 12 studies. Delirium was assessed in 3 studies. Postoperative blood transfusion was assessed in 3 studies. Discharge destination not home (nursing facility or other) was assessed in 4 studies. Readmission was assessed in 4 studies and hospital stay was assessed in 9 studies. Among them, mortality (30-day, 90-day, 1-year, 2-year, 5-year mortality) and complications (according to Clavien-Dindo grade (23) 1-2 for minor, ≥3 for severe) were evaluated in subgroup analysis.
Each study assessed frailty differently, and the tools used to assess frailty are shown in Table 1.
These studies have an average score of 7.6 in the quality evaluation. The full score is 9 points. All studies fulfilled the inclusion criteria.
Frailty and mortality
A total of 10 studies involving 4,721 patients were included. Heterogeneity test was performed. The group of 30-day mortality: I-squared=0.0%, p=0.526, the heterogeneity was not significant, a fixed effects model was adopted, RR (95%CI) =6.02 (2.25, 16.15). The difference was statistically significant between the frail group and the non-frail group; The group of 90-day mortality: I-squared=51.6%, p=0.127, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =6.17 (1.24, 30.65). The difference was statistically significant between the frail Frontiers in Oncology frontiersin.org group and the non-frail group; The group of 1-year mortality: I-squared=85.5%, p=0, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =3.50 (1.43, 8.57). The difference was statistically significant between the frail group and the non-frail group; The group of 2-year mortality: I-squared=93.3%, p=0, the heterogeneity was significant, a random effects model was adopted, RR (95% CI) =3.15 (1.11, 8.89). The difference was statistically significant between the frail group and the non-frail group; The group of 5-year mortality: I-squared=92.0%, p=0, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =2.26 (1.21, 4.22), The difference was statistically significant between the frail group and the non-frail group. Thus, frailty was associated with increased mortality in patients with colorectal cancer after treatment (Figures 2-6).
Frailty and complications
A total of 7 studies involving 54,835 patients were included. Heterogeneity test was performed. The group of total complications: I-squared=22.3%, p=0.259, the heterogeneity was not significant, a fixed effects model was adopted, RR (95%CI) =1.59 (1.55, 1.64). The difference was statistically significant between the frail group and the non-frail group; The group of minor complications: I-squared=78.8%, p=0.001, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =1.28 (0.83, 1.99). There was no significant difference between the frail group and the non-frail group; The group of severe complications: I-squared=67.8%, p=0.008, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =2.26 (1.50, 3.39). The difference was statistically significant between the frail group and the non-frail group. We found that frailty did not appear to have a significant effect on minor complications after treatment but had a significant effect on severe complications (Figures 7-9).
Frailty and delirium
A total of 3 studies involving 500 patients were included. Heterogeneity test was performed. The group of delirium: I-squared=25.5%, p=0.261, the heterogeneity was not significant, a fixed effects model was adopted, RR (95%CI) =3.08 (1.32, 7.17). The difference was statistically significant between the frail group and the non-frail group. It can be discovered that frailty may be associated with high incidence of delirium ( Figure 10). 90-day mortality. 1-year mortality. 2-year mortality. 5-year mortality.
Frailty and postoperative blood transfusion
A total of 3 studies involving 295,724 patients were included. Heterogeneity test was performed. The group of postoperative blood transfusion: I-squared=0.0%, p=0.798, the heterogeneity was not significant, a fixed effects model was adopted, RR (95% CI) =1.87 (1.83, 1.91). The difference was statistically significant between the frail group and the non-frail group. We can find that frailty may be associated with high likelihood of postoperative blood transfusion (Figure 11).
Frailty and discharge destination not home
A total of 4 studies involving 295,983 patients were included. Heterogeneity test was performed. The group of discharge Total complications. destination not home: I-squared=75.7%, p=0.006, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =5.29 (2.56, 10.93). The difference was statistically significant between the frail group and the non-frail group. The frail group could be found to have a higher risk of readmission ( Figure 12).
Frailty and readmission
A total of 4 studies involving 295,843 patients were included. Heterogeneity test was performed. The group of readmissions: I-squared=63.4%, p=0.042, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =1.90 (1.02, FIGURE 9 Severe complications. 3.53). The difference was statistically significant between the frail group and the non-frail group. The frail group could be found to have a higher risk of readmission ( Figure 13).
Frailty and hospital stay
A total of 9 studies involving 54,920 patients were included. Heterogeneity test was performed. The group of hospital stay: I-squared=97.9%, p=0.000, the heterogeneity was significant, a random effects model was adopted, RR (95%CI) =1.40 (0.74, 2.06). The difference was statistically significant between the frail group and the non-frail group. The frail group could be found to have a higher risk of readmission ( Figure 14).
Subgroup analysis
Frailty and mortality at different follow-up times A meta-analysis of studies grouped according to time to death after treatment showed that patients classified as frail had FIGURE 11 Postoperative blood transfusion.
FIGURE 12
Discharge destination not home. higher mortality rates than non-frail patients, either 30 days after treatment or 5 years after treatment. Both short-and long-term survival declines in colorectal cancer patients were associated with frailty. Heterogeneity between studies was observed in subgroup analyses (Overall: I-squared=88.8%, p=0.000, RR (95%CI) =3.36 (2.22, 5.10)) ( Figure 15).
Frailty and complications of varying degrees after treatment
A meta-analysis of grouped studies with minor or major complications after treatment showed that the incidence of minor complications was not statistically different between the frail and non-frail groups, suggesting that the likelihood of FIGURE 13 Readmission. Hospital stay. (Figure 16).
Publication bias and sensitivity analysis
We performed publication bias and sensitivity analyses for each outcome. Eliminating each study one by one did not change the direction of the effect size of any results, verifying that the results were stable (Figures 17-19).
FIGURE 15
Mortality at different follow-up times. Complications of varying degrees after treatment.
Discussion
Our meta-analysis showed that frail colorectal cancer patients had poor short-term or long-term outcomes. Shortterm and long-term mortality and length of hospital stay were higher in frail patients than in non-frail patients. The odds of readmission, postoperative blood transfusion, insanity, and discharge destination other than home were also higher in frail patients than in non-frail patients. In terms of complications, there was no significant difference in the probability of minor complications between the two groups. In terms of serious complications, the frail group had a higher incidence than the non-frail group. We also found that frailty has an important impact on the prognosis of patients with colorectal cancer, regardless of frailty screening tools, reported from multiple studies (24-26). Funnel chart. In tests for heterogeneity, we found heterogeneity among some of the findings, especially in terms of hospital stay. Considering the interference of multiple factors such as study environment, study method, publication year, frailty screening methods, tumor location and stage, we speculate that the source of high heterogeneity may be very complex. Based on data from our research, we cannot complete the analysis of all sources of high heterogeneity. Although some studies included only colon cancer patients, some included emergency surgery in the analysis, and some had smaller sample sizes, these factors may not have had a significant effect on the overall results. It also did not show large errors in publication bias and sensitivity analysis. Overall, the studies had relatively reliable quality ratings.
It has been proved that the prevalence of frailty in older patients with colorectal cancer and an indication for surgery ranges from 25 to 46 percent, depending on the population studied and the tools used to measure it (27). Several frail screening tools have been shown to be useful in predicting surgical and chemotherapy outcomes (28), although not all validated tools have been studied. Studies have shown that the sensitivity, specificity, positive predictive value, and negative predictive value of predicting CGA depend on the tool used, the vulnerability in the sample, and the cutoff value chosen (29). Therefore, some limitations of existing fragile tools and existing fragile literature must be kept in mind when selecting fragile tools. The prevalence of frailty varies slightly from study to study depending on the frailty tool used; furthermore, the varying tools often do not identify exactly the same group of people (30,31). Studies have also found that these scales differ in their ability to predict prognostic outcomes because different subgroups are analyzed (32). We hypothesized that the magnitude of the risk of death in frail colorectal cancer patients may depend on the type of frailty assessment scale used, and we had to examine these findings separately because there are so many frailty screening tools available. Screening for frailty has a variety of additional tools, including cognitive impairment, disability, and comorbidities. Thus there is still some debate as to which frailty screening tool is the yardstick. Even social and economic factors of frailty (e.g., poverty, social isolation) are raised. But whether these additional tools have the same validity as existing frailty tools requires more validation.
The underlying mechanisms between frailty and poor prognosis in colorectal cancer have not been extensively studied. Nonetheless, several studies have reported elevated levels of C-reactive protein, interleukin-6, and tumor necrosis factor alpha in frail patients, suggesting that chronic inflammation may play a role (33, 34). Thus, overt chronic inflammation in frail patients may compromise their immune system and further reduce their functional reserve to adapt to stress (34,35). Therefore, they cannot tolerate the side effects of the treatment, resulting in incomplete treatment (36). Whether it's surgery or chemotherapy, clinicians worry about whether patients, especially frail patients, will be able to tolerate the trauma and side effects of treatment (37-39). They may be more willing to reduce the risk, and the benefit of the treatment is also reduced. Frail patients may also have other geriatric syndromes and poor postoperative outcomes, which can also negatively impact their long-term prognosis. These factors may explain the worse prognosis observed in frail colorectal cancer patients.
In addition, factors affecting the prognosis of patients with colorectal cancer are not limited to frailty. Several studies have found that sarcopenia is also common in cancer patients and predicts longer hospital stay LOS, worse postoperative complications, susceptibility to chemotherapy toxicity, decreased quality of life, and poor survival (40)(41)(42)(43). Studies suggest that inflammatory markers are related to sarcopenia and play a major role in the development of sarcopenia (44). When concentrations of inflammatory factors such as tumor necrosis factor and interleukin-6 are elevated, they activate multiple metabolic pathways, leading to reduced protein degradation and synthesis, and by disrupting insulin signaling, leading to insulin resistance, which further reduces muscle mass. Low- FIGURE 19 Begg & egger's test. grade systemic inflammation caused by tumors may lead to local muscle inflammation, which in turn leads to muscle degeneration (45). Muscles are the basis for maintaining normal physiological activities of the human body. Therefore, under the influence of inflammation and sarcopenia, the prognosis of colorectal cancer patients is not optimistic. Likewise, malnutrition affects outcomes in patients with colorectal cancer. Preoperative malnutrition in colorectal cancer patients is associated with many adverse postoperative outcomes and poorer prognosis. Malnourished patients have significant weight loss after surgery, are more likely to develop septic shock, and have increased requirements for postoperative blood transfusion, mechanical ventilation, and reoperation (46). Malnutrition may also lead to immunosuppression and, as a result, post-operative inflammation and infection problems are more frequent. In addition, micronutrient deficiencies may also lead to increased inflammation, lower serum albumin levels, and increased incidence of anastomotic leakage in patients (47). Patients with mild to severe malnutrition have significantly longer hospital stays and longer recovery of gastrointestinal function than well-nourished patients (48).
After patients are discharged from the hospital, they go to many different places. We consider that those patients who recover well will go back home to live with their families because they have retained some self-care ability. Those who have lost their independence mostly go to some nursing institutions or nursing homes, and they must live with the help of others. The latter were mostly those who were identified as frail.
We also found the results of postoperative blood transfusions and found that frail patients were more likely to require blood transfusions than non-frail patients. This may be due to the fact that most frail patients are already in a state of anemia, coupled with a weaker physiological reserve, which increases the difficulty of surgery and increases the risk of bleeding. After experiencing external stimuli such as surgery, it is more difficult for oneself to maintain a steady state, resulting in an external means-blood transfusion to help recovery.
Notably, not all adverse outcomes were associated with preoperative frailty (49). We found from the included studies that the stage of the tumor (TNM stage), the size and location of the tumor, the method of treatment (laparoscopy, laparotomy, radiotherapy, chemotherapy), whether there was intestinal obstruction before surgery, and whether there was a surgical stoma (temporary or permanent) may affect the prognosis of patients with colorectal cancer. Another study performed in oncological patients with different types of tumors and cancer stages found no relationship between preoperative frailty and postsurgical mortality (50), suggesting that, in the case of malignancies, factors other than frailty (tumor location and the presence of metastases) likely play a major role. This also confirms our conjecture.
Similarly, frail patients with advanced tumors and preoperative bowel obstruction or perforation tend to have worse outcomes. And those who are already frail can only receive palliative chemotherapy or local surgical resection. In this palliative treatment approach, frailty is often aggravated, leading to a vicious circle with a far worse prognosis than nonfrail patients. In addition, after evaluation by MNA nutritional score, Barthel index, and ASA grading standard, patients under different grades also have different prognostic performance. But no research has yet confirmed their link to frailty, and whether they should be part of frailty screening.
Screening frailty as an independent risk stratification tool in colorectal cancer patients has become imperative. Standard treatment for able-bodied patients with colorectal cancer, while for frail patients with colorectal cancer, the need for an individualized treatment plan must be considered (51). Before treating patients, clinicians will use various evaluation tools to screen out frail patients so that they can receive more care and formulate more suitable programs.
A comprehensive frailty assessment of colorectal cancer patients not only facilitates the early identification and comprehensive management of frailty syndromes, but also can optimize clinical care by obtaining physical and psychological information about the patient. Future studies should evaluate the prognostic value of frailty in the diagnosis and treatment of colorectal cancer patients. Due to the increasing number of elderly CRC patients, their frailty is very common (52). However, in the current study, the lack of a unified screening tool for frailty and the incompleteness of the test results make the data for evaluating the prognosis of patients with colorectal cancer for frailty still lacking (53,54). It is also worth noting that, the study found frailty may also interact with colorectal cancer, accelerating disease progression or worsening prognosis (55). Therefore, frailty may affect tumor biology, which may be an important line of thought for future research.
Limitations
We searched extensively for eligible studies, but it is still possible that we missed some relevant studies in other languages or databases. In the heterogeneity test, only the length of hospital stay had a relatively large publication bias, but overall there was no obvious publication bias, and the results were relatively reliable. Also, the number of research articles we included and the sample size of the studies were not large enough. The mechanism between frailty and prognostic changes in patients with colorectal cancer is unclear. Therefore, more clinical data and mechanistic studies are needed to supplement.
Conclusion
Frailty has a huge impact on the prognosis of colorectal cancer patients, especially in mortality and complications after treatment. To further explore how frailty alters the outcomes of colorectal cancer patients, and how to reduce the impact of this poor prognosis, more authoritative frailty assessment criteria and more clinical data are needed.
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.
Author contributions
MC and YH conceived and designed this study; MC completed the literature search and screening; JL and YJ included and excluded the literature and completed the quality assessment; ZG completed the data extraction; MC and ZG completed statistical analysis; MC completed the manuscript. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-03T18:35:04.849Z
|
2022-11-03T00:00:00.000Z
|
253259977
|
s2orc/train
|
The association between chiropractic integration in an Ontario community health centre and continued prescription opioid use for chronic non-cancer spinal pain: a sequential explanatory mixed methods study
Background: Emerging evidence suggests that access to chiropractic care may reduce the likelihood of initiating an opioid prescription for spinal pain; however, the impact of chiropractic care for patients already prescribed opioids is uncertain. We undertook a sequential explanatory mixed methods study to evaluate the association between initiating chiropractic care and continued opioid use among adult patients attending an Ontario community health centre (CHC) and receiving opioid therapy for chronic non-cancer spinal pain. Methods: We conducted a retrospective cohort study of 210 patient records between January 1, 2014 and December 31, 2020. We used generalized estimating equations, adjusted for patient demographics, co-morbidities, visit frequency, and calendar year, to evaluate the association between receipt versus non-receipt of chiropractic services and continued opioid use (e.g., unique opioid fills, number of refills, and dosages) up to one year following the index chiropractic visit. We also completed follow-up interviews with 14 patients and nine general practitioners from the CHC and integrated these data with our quantitative findings. Results: Over 12-month follow-up, there were lower rates of opioid fills (incidence rate ratio [IRR] = 0.66; 95% confidence interval [CI], 0.52–0.83) and refills (IRR = 0.27; 95% CI, 0.17–0.42) among chiropractic recipients (n = 49) versus non-recipients (n = 161). Although patients who did and did not receive chiropractic care began the study with the same dose of opioids, recipients were less likely to be prescribed higher-dose opioids (i.e., ≥ 50 mg morphine equivalents daily) compared to non-recipients at three months (odds ratio [OR] = 0.14; 95% CI, 0.04–0.47), six months (OR = 0.14; 95% CI, 0.05–0.40), nine months (OR = 0.19; 95% CI, 0.07–0.57), and 12 months (OR = 0.22; 95% CI, 0.08–0.62). Interviews suggested that patient self-efficacy, limited effectiveness of opioids for chronic pain, stigma regarding use of opioids, and access to chiropractic treatment were important influencing factors. Conclusion: We found that continued prescription opioid use among patients with chronic non-cancer spinal pain who received chiropractic care was lower than in patients who did not receive chiropractic care. Four themes emerged in our qualitative interviews to help provide a richer understanding of this association. Randomized controlled trials are needed to establish the effect of chiropractic care on opioid use for chronic spinal pain. Supplementary Information The online version contains supplementary material available at 10.1186/s12913-022-08632-9.
Background
Chronic non-cancer pain affecting the spine or other musculoskeletal tissues is a prevalent and global health problem associated with considerable socioeconomic burden. Worldwide, approximately one in five people live with chronic non-cancer pain [1][2][3][4], with seniors, women, military veterans, indigenous populations, rural inhabitants, those with lower formal education, and individuals reporting low socioeconomic status being most affected [5][6][7]. In Canada, the annual economic cost of chronic non-cancer pain due to medical expenditures and lost productivity was estimated between $38 and $40 billion in 2019, and this cost is expected to rise by more than 36% by the year 2030 [8]. The annual cost of chronic non-cancer pain in the United States (US) was previously estimated to be between $560 and $635 billion [9]. Opioids are commonly prescribed to patients to relieve chronic non-cancer pain, particularly in North America [10]; however, opioids provide only modest benefits [11] and are associated with important dosedependent harms, including overdose and death [12][13][14][15]. Accordingly, governments, policy makers, and insurers have been called upon to improve support for non-opioid approaches to managing chronic non-cancer pain, particularly in vulnerable and marginalized populations [16].
Emerging evidence suggests that early access to chiropractic treatment is associated with lower initiation of opioid prescribing among patients with non-cancer spinal pain [17][18][19][20][21]. A 2020 systematic review and metaanalysis of six cohort studies found that patients with acute or chronic non-cancer spinal pain who received chiropractic services early in their complaint were 64% less likely than non-chiropractic users to be prescribed opioids (pooled odds ratio [OR] = 0.36; 95% confidence interval [CI], 0.30 to 0.43) [17]. A subsequent observational study of 216,504 opioid-naive patients with new-onset low back pain who received initial treatment from chiropractors versus primary care physicians had 90% lower odds of short-term opioid use (adjusted OR = 0.10; 95% CI, 0.09 to 0.10) and 78% lower odds of long-term opioid use (adjusted OR = 0.22; 95% CI, 0.18 to 0.26) [18,19]. Similar findings have been reported by two other recent observational studies [20,21]; however, the association between receipt of chiropractic services and continued opioid use in patients with existing opioid prescriptions is uncertain [22][23][24]. Moreover, previously published studies on the topic of chiropractic care and opioid prescribing have lacked in-depth, contextual understanding because they have been exclusively quantitative in nature [17][18][19][20][21][22][23][24].
To help address these knowledge gaps, we conducted a mixed methods study to evaluate the association between initiating chiropractic care and continued opioid use among adult patients with chronic non-cancer spinal pain attending an Ontario community health centre (CHC) [25,26]. We hypothesized that younger age, male sex, health-related co-morbidities, depressive symptoms, poor health behaviours (e.g., smoking), a higher frequency of healthcare provider visits, and earlier years of our 7-year study timeframe would be positively associated with opioid use. We also hypothesized that chiropractic care would be inversely associated with opioid use [25].
Ethical considerations
Our study was approved by the Hamilton Integrated Research Ethics Board at McMaster University (project number 2021-10930). Approval to conduct this study was also obtained from the Chief Executive Officer at the Langs CHC [26]. All methods were carried out in accordance with the relevant guidelines and regulations and the Declaration of Helsinki.
Study design
We used a sequential explanatory mixed methods design [27]. In the quantitative phase, we obtained data via chart review [28] of electronic medical records (EMRs) of both recipients and non-recipients of chiropractic services with at least one prescribed opioid for the treatment of a chronic non-cancer spinal pain-related diagnosis at the Langs CHC [26]. In the qualitative phase, we conducted one-on-one interviews with patients and general practitioners (GPs) to explore perceptions of chiropractic integration on opioid prescribing. Complementarity [29] was our rationale for using a mixed methods approach, that is, the results from the qualitative phase of our study were used to help clarify and explain our quantitative findings. See Fig. 1 for a diagram outlining our study procedures. We followed the STROBE statement [30], the COREQ criteria [31], and the Good Reporting of A Mixed Methods Study (GRAMMS) guidelines [32,33] for our study (Additional file 1).
Setting
The Langs CHC is located in Cambridge, Ontario, Canada [25,26], a medium-sized urban municipality (population: ~130,000) located 82 km southwest of Toronto. This Centre provides healthcare services to communities and vulnerable populations with high unemployment rates, multiple co-morbidities, and musculoskeletal disorders that are commonly managed with prescription opioids [25,26]. Because chiropractic services are not publicly funded in Canada, these populations have traditionally faced barriers to accessing chiropractic care [23,[34][35][36][37][38][39][40]. However, since January 1, 2014 [34] a partially subsidized chiropractic spine pain program that operates on two half days per week has been offered to patients at the Centre. To be eligible to receive these services, patients have to be referred into the program by their GP. The Centre also employs a team of medical doctors, nurse practitioners, registered nurses, dieticians, social workers, community health workers, and a physiotherapist. For more complete details of the CHC's chiropractic spine pain program, our conceptual framework, and a list of diagnostic codes Fig. 1 Study diagram of an explanatory sequential design of a mixed methods study on the association of chiropractic integration with opioid use for chronic non-cancer spinal pain at the Langs Community Health Centre. The quantitative and qualitative data collection and analysis phases are listed at the top of each step of the diagram. The two points of interface (or mixing) of the quantitative and qualitative phases occur in the third and final steps. The term "QUANTITATIVE" is capitalized to indicate prioritization of the quantitative phase in the study. The study procedures and outputs for each phase are listed in point-form at each step used for defining our study sample, we refer readers to our study protocol [25].
Quantitative sampling Participants and data sources
We included records for all adult patients (aged ≥ 18 years) who received one or more prescriptions for opioids dispensed over a minimum period of three consecutive months, and who attended two or more appointments relating to a diagnosis of chronic spine (i.e., back or neck) pain at the Langs CHC between January 1, 2014, and December 31, 2020. The start date for quantitative sampling was January 1, 2014, which was the inaugural date of the Langs CHC's chiropractic spine pain program [34]. Patients receiving treatment for opioid use disorder (e.g., methadone, naloxone) prior to their index visit, as well as those with spinal neoplasms or other contraindications to chiropractic treatment (i.e., fractures, infections, inflammatory arthritis, or cauda equina syndrome), were excluded from our cohort. As we were interested in patients receiving long-term opioid therapy [13], we excluded individuals who had been prescribed opioids for < 90 days at their index visit, or who did not receive any opioid fills or refills after their index visit.
We linked EMR records of all patients in our study to medical drug claims data at the Institute for Clinical Evaluative Sciences (ICES) (https://www.ices.on.ca) with their Ontario health card number. ICES is an independent, non-profit research institute whose legal status under Ontario's health information privacy law allows for the collection and analysis of healthcare and demographic data, without consent, for health system evaluation and improvement. Patients whose health card number was incorrectly recorded in their EMR were excluded.
Quantitative data collection Variables
Opioid prescription data were obtained from the Narcotics Monitoring System database by an independent research scientist at ICES, including the number of prescribed opioid fills, the number of prescribed opioid refills (measured in 30-day equivalents), and the prescribed opioid dosage. These outcomes were measured for up to 12 months from the date of first opioid prescription following a patient's index visit for chronic non-cancer spinal pain. To maintain temporality, the index visit for patients who received chiropractic care was their first chiropractic visit. Other variables that were extracted from the EMR included socio-demographics (age and sex), general health (smoking status and body mass index), co-morbidities (depression, anxiety, fibromyalgia, diabetes, and cardiovascular disease), and the total number of healthcare (i.e., GP or chiropractic) visits.
Quantitative data analysis Statistical methods
Baseline characteristics were compared between the exposed (receipt of chiropractic care) and non-exposed groups using the chi-squared test for categorical variables (or Fisher's exact test if there was a cell frequency of < 5) and the Mann-Whitney U test for skewed continuous variables. We used generalized estimating equations (GEEs) to explore the association between exposure to chiropractic care and opioid prescribing [49,50]. To account for potential data clustering within-subjects or between medical or chiropractic practitioners, we used a robust variance estimator to compute the standard errors for our coefficient estimates. We also conducted sensitivity analyses with different working correlation structures, including independent, autoregressive, and unstructured matrices [49,50]. The specified link function in our GEE models was based on the data distribution (e.g., log-linear for data fitting a Poisson distribution, binomial for binary data).
We used GEEs with a Poisson distribution when the outcomes were counts (i.e., total number of unique opioid fills and subsequent refills over the entire course of follow-up, tabulated at the end of follow-up). We estimated incidence rate ratios (IRRs) for differences between the chiropractic and non-chiropractic groups using Poisson log-linear GEEs and reported the associated 95% CIs and p-values.
We used GEEs with a binomial distribution when the outcome was opioid dosage. We assessed opioid dosages at 90-day intervals, dichotomized into higher (≥ 50 mg) morphine equivalents daily (MED) or lower (< 50 mg) MED [11] and compared these between the chiropractic and non-chiropractic groups from baseline to 12-month follow-up. We originally planned to dichotomize opioid dose using a different threshold (≥ 90 mg versus < 90 mg MED) [25], but we modified our approach to reflect the central tendency of MED in our patient sample. We estimated between-group differences for dosage using a binary logistic GEE and reported these with ORs, 95% CIs, and p-values. To calculate the MED for each prescribed opioid, we multiplied the quantity × the milligrams per unit dispensed × drug-specific conversion factors (Additional file 2) [11,13]. Emary
Quantitative variables and study size
For each outcome of interest, we built univariable and multivariable models to estimate the crude and adjusted associations, respectively, between patients that did or did not receive chiropractic care (1 = received; 0 = did not receive) and opioid use. We grouped covariates into blocks (e.g., socio-demographic, health-related, depressive symptoms, health behaviours, and healthcare visits) and these were sequentially entered into our models, with time (i.e., calendar year) as an additional covariate and chiropractic/non-chiropractic care as the main exposure variable. To guard against over-fitting of our regression models [51], we set a minimum threshold of 10 events per category for each independent variable (i.e., minimum sample of 150 patient records) to ensure that each variable had sufficient discriminant power to detect an association with opioid use, if an association existed.
We assessed model fit using the quasi-likelihood under the independence model criterion (QIC) [50,52]. Correlation structures with the lowest QIC scores (closest to zero) were judged as the best model fit for the data. We also explored variance inflation factors (VIFs) to assess collinearity between independent variables. If multicollinearity was detected between two or more variables (i.e., VIFs ≥ 5) [53], we compared regression models, each separately containing one of the collinear variables, to one another and selected the model containing the variable that produced the lowest Akaike information criterion (AIC) value. The two-sided statistical significance level ( α ) for all quantitative analyses was 5%, and all data and comparative analyses were performed using SPSS v28.0.1.0 (IBM SPSS Statistics).
Qualitative sampling
For the qualitative phase of our study, we used stratified purposive sampling to select a sub-sample of chiropractic and non-chiropractic patients, whose charts we examined in the quantitative phase, to participate in one-on-one interviews [54]. This was the first stage of integration between our quantitative and qualitative study phases [55]. We also recruited a purposive sample of GPs from the Langs CHC. The lead author (PCE) conducted recruitment via telephone or e-mail using participant contact information provided by the Langs CHC administration. We offered gift cards ($10 for GPs, $30 for patients) as incentives for participation. We used maximum variation [54] in choosing participants, based on age, sex, and the number of years attending the CHC (for patients) or years in practice (for GPs), to encourage a range of sociodemographic characteristics and perspectives. We also collected patients' primary spine pain complaint and current opioid dose. We aimed to interview a minimum of 12-20 patients and 6-10 GPs [54], with interviews continuing until saturation; the point at which no new information was obtained from participants in the GP, chiropractic, and non-chiropractic groups [56]. We used fundamental qualitative description [56,57] as our methodological orientation to underpin the qualitative phase of our study.
Qualitative data collection
The lead author (PCE), a health research methodologist with expertise in mixed methods and qualitative research, conducted one-on-one (individual) semi-structured interviews with participants. Interviews were conducted either virtually (n = 3) using the Zoom videoconferencing application (Zoom Video Communications, Inc.) or in-person (n = 20), based on participant preference. We promoted confidentiality by conducting the interviews in a private office separate from the medical clinic at the Langs CHC. We obtained informed consent from participants before the start of each interview. Five members of our research team (PCE, ALB, MO, LM, JWB) developed the patient and GP interview guides (see Additional files 3 and 4, respectively) based on relevant literature [17][18][19][20][21][22][23][24]27] and our quantitative findings. Three of the five members (PCE, ALB, JWB) also have content expertise in the subject area of our study.
We audio recorded virtual interviews using Zoom's built-in recording feature and in-person interviews using MacIntosh recording software (Audio Recorder v1.3, FIPLAB Ltd.). The lead author (PCE) also took field notes after each interview to document other observations and emergent themes. To promote trustworthiness in our qualitative data, we employed member-checking [27] by sending the raw transcripts and a summary of our results to participants for feedback or correction. We also kept an audit trail of our qualitative data collection and analysis procedures [56]. A summary of our investigator reflexivity is provided in Additional file 5.
Qualitative data analysis
We transferred all interview audio recordings into the software program, MAXQDA (http://www.maxqda. com), and the lead author (PCE) transcribed the audio recordings verbatim. After participant identifiers were removed, another member of the research team (JD) reviewed a random sample of 15% of the transcripts for accuracy and found only a few minor typographical errors. All transcripts were then independently coded by two investigators (PCE, ALB) using an inductive content analytic approach [56]. The aim of this strategy was to descriptively summarize the information to ensure the 'best fit to the data' [57]. We used both open and axial coding in our data analysis: open coding to develop concepts from the data, and axial coding to relate these codes (or concepts) to one another followed by the identification of themes, sub-themes and representative quotes [27].
The two investigators undertaking coding of transcripts met three times throughout the analysis (i.e., after every seven to eight interviews) to compare themes and arrive at a final, agreed-upon set of themes through discussion. We organized these themes into tabular form and selected representative quotations for each theme/subtheme [27]. We created joint display tables as part of our data integration procedures (Fig. 1), and our qualitative and quantitative results were further combined using contiguous narrative and weaving approaches [27,55]. We then drew upon our qualitative and quantitative results jointly to come to a set of conclusions (i.e., 'metainferences') [27].
Quantitative findings
We identified a total of 1,166 patient records, and 210 met eligibility criteria for inclusion in our quantitative analysis (Fig. 2).
Patients with an index visit date in a more recent calendar year also had a lower rate of opioid refills (IRR = 0.82; 95% CI, 0.73 to 0.93) and were less likely to be receiving higher dose opioids at three months (OR = 0.73; 95% CI, 0. 57 [c-e]). Contrary to our predictions, anxiety and obesity were negatively associated with opioid dosage (Additional file 7 [c, d, f ]), while younger age was not associated with opioid use in our patient sample (Additional file 7). All VIFs were less than 1.4, suggesting no important multicollinearity among independent variables.
Qualitative and mixed methods findings
Twenty-three patients were recruited for interviews and 14 participated. Five patients scheduled interviews but cancelled (two chiropractic recipients, three non-recipients), two scheduled interviews but did not attend (one recipient, one non-recipient), one declined for health reasons and one was not interested. Of those who were interviewed, eight were chiropractic recipients and six were non-recipients. Among GPs, four of six medical doctors and five of six nurse practitioners completed interviews. Two medical doctors declined participation because of lack of time, and one nurse practitioner expressed interest but did not respond to further interview requests. In total, 23 interviews were completed (14 patients, nine GPs). The median durations of interviews were 25 min (range, 19 to 56) for patients and 38 min (range, 20 to 40) for GPs.
The majority (79%) of the 14 patients we interviewed were female, most (86%) were either receiving disability benefits or were unemployed, and the majority (71%) had previously received at least one opioid prescription for chronic non-cancer spinal pain. The median dosage for those currently receiving opioid medications was 19 mg MED (range, 14 to 90). Among patients and GPs, there was a large range of ages (33 to 82) and number of years attending the Langs CHC (patients: 2 to 43) or years in practice (GPs: 1 to 26), demonstrating variability among participants (Additional file 8). Opioid prescription fills over 12-month follow-up. An incidence rate ratio < 1 indicates a lower rate of opioid fills in the recipient group c Opioid prescription refills (of 30 days or equivalent) over 12-month follow-up. An incidence rate ratio < 1 indicates a lower rate of opioid refills in the recipient group Among all 23 participants, one non-chiropractic patient and four GPs made minor revisions to clarify statements from their interviews during member-checking. No other participants requested content changes or corrections to their transcripts or results. We determined that data saturation had been reached when only two new codes emerged from chiropractic recipient interviews 6, 7 and 8 (with no new codes from interviews 7 and 8); only one new code emerged from non-recipient interview 4 (with no new codes from interviews 5 and 6); and only one new code emerged from GP interviews 7, 8 and 9. At this point, participant recruitment was concluded.
Coding tree
We identified 37 codes across interviews which were categorized into four major themes: (1) patient selfefficacy, (2) accessibility of non-pharmacological services, (3) stigma regarding use of opioids, and (4) impact of treatment. Codes pertaining to patient self-efficacy were stratified into two sub-themes, 'active versus passive approaches' and 'resistance to taking medication. ' This latter sub-theme was common among chiropractic patients. For our second theme, we created the subthemes 'lack of access to non-pharmacological treatment options' and 'access to chiropractic services at the Langs CHC. ' Lack of access to non-pharmacological services (e.g., chiropractic, physiotherapy) was identified in nearly all (21 of 23) participant interviews and was reported as a common facilitator of opioid use. Our third theme captured codes related to the opioid crisis such as negative media coverage or lived experiences. Some patients also expressed a sense of judgement from others for using prescription opioids. The remaining codes related to patients' or GPs' perspectives on the impact of treatment for chronic non-cancer spinal pain, including subthemes of pain relief, functionality, recognition of the limited effectiveness of opioids for chronic pain, and anxiety and fear surrounding opioid withdrawal. Descriptions and frequency counts of each of our major themes, sub-themes, and representative participant quotes are provided in Additional file 9. Our main quantitative findings are presented with qualitative data as joint displays in Tables 2 and 3.
Discussion
Among patients receiving long-term opioid therapy for chronic non-cancer spinal pain, we found that initiating chiropractic care was associated with fewer fills and refills for prescription opioids and, when prescribed, reduced dosage of opioids. Based on our qualitative findings, use of opioids was influenced by patients' selfefficacy and concerns about opioid-related harms, recognition of the limited effect that opioids may have on chronic pain, increasing stigma regarding use of opioids, and access to non-pharmacological treatment options.
Our findings are supported by other uncontrolled observational studies [22][23][24]. A retrospective analysis of quality assurance data from a CHC in Manitoba, Canada [23] found that patients referred for chiropractic services had a 22% decrease in the number of opioid tablets used after attending an average of five chiropractic visits. Between baseline and discharge, the number of chiropractic patients prescribed opioids within this health care centre decreased 26% [23]. Findings of reduced opioid usage among patients receiving chiropractic services in US Veteran Administration [22] and CHC [24] clinic settings have also been recently reported.
The integration of quantitative and qualitative methods in our study generated several insights into our results. As highlighted in our interviews, patients who were referred for chiropractic services at the Langs CHC may have been more resistant to taking opioid medication than patients not referred for chiropractic services, a sentiment supported by some published evidence [58]. In addition, GPs indicated that access to chiropractic treatment gave them another non-opioid pain management option. Lack of access to non-pharmacological services (e.g., chiropractic, physiotherapy) was reported by several participants as a facilitator of opioid use, while chiropractic patients and GPs identified negative stigma associated with the use of opioids as a common barrier. We also found in our cohort that the proportion of chiropractic recipients who discontinued using opioids was nearly double that of non-recipients. These factors may help explain why chiropractic recipients obtained fewer opioid prescriptions and were less likely to be receiving higher opioid doses up to one year after presentation.
Similar to previous research [42,44], we found that a higher frequency of healthcare visits was positively associated with opioid use. Patients with lower self-efficacy or experiencing greater difficulty coping with their pain may have been more likely to visit their healthcare providers more often and obtain opioid prescriptions on a more frequent basis and at higher doses. Recent evidence suggests that active pain self-management programs that include exercise, goal setting, education, and counselling on opioid discontinuation, as well as interventions aimed at supporting prescribers' adherence to guidelines (e.g., chart audits, tracked performance metrics related to high-dose prescribing), can increase the likelihood of patients reducing their opioid dose or discontinuing opioid treatment [59]. However, as was frequently mentioned by both GPs and patients in our interviews (see Theme #2 in Additional file 8), accessibility of non-pharmacological treatment options remains a challenge, particularly for persons who are unemployed or from low income backgrounds [26, 34-40, 42-44, 59].
We found that patients with an index visit date in a more recent calendar year had fewer opioid prescription refills and were less likely to receive higher opioid doses at 3-and 6-month follow-up. Current guidelines [13,60] recommend optimization of non-opioid and non-pharmacologic treatments prior to opioid use, while limiting opioid doses (when first used with patients) to less than 50 mg MED, and offering a trial of voluntary tapering if doses are already ≥ 90 mg MED. Accordingly, several GPs indicated in their interviews that a concerted effort, in the form of internal chart audits and clinical team meetings, had been made in recent years to reduce opioid prescribing at the Langs CHC. When controlling for calendar year in our analyses, however, we found that the number of opioid fills, refills, and dosages were still considerably lower among chiropractic recipients.
Several observational studies have reported an association between use of chiropractic services and reduced opioid prescribing [17][18][19][20][21]61] or reduced opioid use [22][23][24]. Previous observational research [34][35][36][37][38][39] also suggests that integrating chiropractic services with physician management of spine-related pain is associated with improved patient outcomes and potential for cost savings (e.g., reductions in advanced imaging, GP visits, and specialist referrals). When accessed as a first-line treatment, chiropractic services may also help to delay, and in some cases prevent, opioid prescription [17][18][19][20][21]61]. In one of our interviews (see Theme #2, first sub-theme, Additional file 8), the following GP expressed that, "…having access to any kind of additional modalities in a timely and efficient manner … would probably reduce the need for opioids in the first place. " GP 9 Our findings add to a growing body of observational evidence that suggests integration of chiropractic services into primary care centres [23,24,[34][35][36][37][38][39] and interdisciplinary spine care pathways [62] would reduce barriers to access and potentially reduce use of opioids among patients with chronic non-cancer spinal pain. However, since the efficacy of non-pharmacological interventions including chiropractic care for reducing opioid use CHC community health centre, DC doctor of chiropractic, GP general practitioner, IRR incidence rate ratio, MED morphine equivalents daily, OR odds ratio a Prescription opioid refills were measured in 30-day equivalents remains uncertain [59], and observational research is susceptible to selection bias and confounding [63], welldesigned randomized controlled trials are needed to confirm these findings. Our qualitative findings suggest that lower opioid use is also related to factors such as self-efficacy and concern about opioid-related harms, access to non-pharmacological care, stigma, and knowledge of opioid effectiveness on chronic pain. Future research should investigate these factors further to inform their association with opioid use.
Strengths and limitations
Our study has several strengths. First, we used patient health card numbers to link EMR data with medical drug claims data from the Narcotics Monitoring System database at ICES to verify patient opioid prescriptions and dosages. Second, we specified the anticipated direction of association for each independent variable in our regression models a priori to provide greater confidence in our findings. Third, we used GEEs to account for hierarchical clustering and to control for differences in confounding factors between our exposed (receipt of chiropractic care) and unexposed groups. To account for policy changes in opioid prescribing, we controlled for calendar year in our analyses. This helped to more clearly delineate between a reduction in opioid use associated with access to chiropractic services versus confounding by policy change. Additional strengths included limited missing data (< 1%), direct data export from the EMR to avoid extraction errors [28], and validation of our qualitative data via member-checking. A final strength of our study is our qualitative findings, which provided a richer understanding of the barriers and facilitators to opioid use and how chiropractic services may have been used by patients and GPs to reduce reliance on opioid prescribing for chronic non-cancer spinal pain.
Our study also has several limitations. Due to the retrospective design in our quantitative phase, certain variables that may be associated with opioid use were Patients whose index visit date was in a more recent calendar year had a lower rate of refilling opioid prescriptions and were less likely to be receiving higher dose (≥ 50 mg MED) opioids at 3-and 6-month followup. GPs at Langs have made a concerted effort in recent years to reduce opioid prescribing.
DC doctor of chiropractic, GP general practitioner, IRR incidence rate ratio, MED morphine equivalents daily, OR odds ratio a Healthcare visits constitute GP and chiropractic visits b Prescription opioid refills were measured in 30-day equivalents unavailable. For example, due to the constraints of data recorded in the Langs EMR, we were unable to extract information on co-interventions that patients may have received outside of the CHC, as well as baseline severity/ chronicity of patients' spine-related pain, and additional potential confounders such as employment status or other mental health and pain conditions. However, Langs CHC patients are unlikely to access private healthcare services elsewhere due to socioeconomic disadvantages [23][24][25][26][34][35][36][37][38][39][40]. In addition, we used receipt of opioid prescriptions over three consecutive months, combined with multiple clinic visits for a non-cancer spinal pain diagnosis at the Langs CHC, as a proxy for chronic noncancer spinal pain. Another limitation is that despite restricting our EMR data extraction to patient encounters related to non-cancer spinal pain, and only including opioid medications prescribed on or between these visit dates, it remains possible that opioids may have been prescribed for other indications. However, this would have attenuated the association between chiropractic care and opioid use [64]. Furthermore, our primary outcome measures (i.e., opioid prescriptions and dosages) are surrogates for patient-important outcomes such as functional improvement or pain reduction. An inherent limitation with using a sequential mixed methods design (i.e., quantitative followed by qualitative) is that 11 months elapsed between our quantitative and qualitative study phases, subsequently limiting our qualitative data collection. For instance, some individuals whom we attempted to recruit from the larger cohort were no longer available for interviews (e.g., moved out of city, phone number no longer in service, or were deceased). A further limitation of the qualitative phase of our study is that we did not pilottest our interview guides. However, one week before the interviews, participants received an information letter containing examples from the interview questions. Lastly, chiropractors engaged to provide care at the Langs CHC were selected for their focus on evidence-based, timelimited management of musculoskeletal complaints [25,34]; practice variability among chiropractors in Canada [65] may reduce the generalizability of our findings in other settings.
Conclusion
We found that patients with chronic non-cancer spinal pain who received chiropractic care obtained fewer and lower dose opioid prescriptions than patients who did not receive chiropractic care. Follow-up interviews suggested this relationship was influenced by patient self-efficacy and concerns about opioid-related harms, limited effectiveness of opioids for chronic pain, stigma regarding use of opioids, and access to non-pharmacological treatment options. Although overall results are promising, large rigorously-conducted randomized controlled trials are needed to establish the role of chiropractic care in reducing opioid use for chronic spinal pain.
|
v2
|
2022-11-03T18:35:33.479Z
|
2022-11-03T00:00:00.000Z
|
253259817
|
s2orc/train
|
Synthesis, characterization, and radiosynthesis of fluorine-18-AVT-011 as a Pgp chemoresistance imaging marker
P-glycoprotein (Pgp) is the most studied ATP-binding cassette (ABC) efflux transporter and contributes to chemoresistance. A few tracers have been developed to detect the in-vivo status of chemoresistance using positron emission tomography (PET) imaging. In our study, we have synthesized labeled AVT-011 with fluorine-18 (18F) followed by in-vitro and in-vivo analysis. Tosylate AVT-011 precursor was synthesized and characterized by 1H-NMR and 13C-NMR. AVT-011 was labeled with 18F using the nucleophilic substitution method, and a standard set of quality control was performed. The specificity for Pgp was tested in U87MG cells with and without an inhibitor (tariquidar). The biodistribution and in-vivo stability were tested in the small animals (mice). The biodistribution data of [18F]-AVT-011 was extracted from the PET-CT imaging of breast cancer patients (n = 6). The precursor was synthesized with 36 ± 4% yield and 97 ± 2% purity. The labeling was more than 95% with a 42 ± 2% yield, as evaluated by Radio-HPLC. The cell-binding assay showed a specificity of the tracer for Pgp as the uptake increased by twice after blocking the Pgp receptors. The radiotracer showed a hepatorenal excretion pathway for clearance in an animal study. The uptake was higher in the liver, lungs, spleen, and heart at 15 min and decreased at 60 min. The patients' distribution showed similar uptake patterns as observed in the small animals. [18F]AVT-011 was characterized successfully with high radiochemical purity and yield. The in-vitro and in-vivo studies proved its specificity for Pgp and safe for patient use.
www.nature.com/scientificreports/ ATP-binding cassette (ABC) family proteins, ABCB1 and ABCG2, contribute to multidrug resistance. In many studies, ABCB1/ABCG2 mRNA expression has been inversely correlated with the response to chemotherapy [9][10][11] . Therefore, the focus was shifted to image Pgp expression using PET imaging. Vlaming et al. (2015) synthesized [ 18 F]-gefitinib and tested it in the cell lines overexpressing murine and human ABCB1 and Abcg2. The PET-CT imaging was done in the wild-type, Abcg2 −/− , Abcb1a/1b −/− , and Abcb1a/1b; Abcg2 −/− mice. The results showed 2.3-fold increased brain levels of [ 18 F]-gefitinib in Abcb1a/1b; Abcg2 −/− mice, compared to wild-type. Though the levels in the knockout animals were not different from the wild-type, showing that Abcb1a/1b and Abcg2 together limit access of [ 18 F]-gefitinib to the brain.
[ 11 C]-Verapamil (VPM) was developed as the first Pgp substrate radiotracer and is considered the most successful tracer for Pgp imaging. The applications of (R)-[ 11 C]-verapamil in assessing the Pgp functions in different conditions like aging, depression, and neurodegenerative diseases have been reported and reviewed extensively 12 . Maria et al. (2020) have used [ 11 C]-verapamil PET for evaluating Pgp function in drug resistance epileptogenic developmental lesions. They have conducted a dynamic [ 11 C]-VPM-PET imaging on twelve healthy controls and two epilepsy patients. The imaging data were used to calculate the influx rate, VPM-K 1, using single-compartment modeling with a VPM plasma input function. Statistical parametric mapping (SPM) analysis showed a significantly lower uptake of VPM corresponding to the area of the epileptogenic developmental lesion compared to healthy controls 13 . Savolainen et al. (2020) used [ 18 F]MC225 in mice and non-human primates, the dynamic PET images used for calculating the volume of distribution (V T ), kinetic rate constants (K 1 , K 2 ) in the baseline and post-inhibition scans. They found that K 1 is the accurate parameter to measure the Pgp function. Still, it contradicted the other study carried out by different groups using [ 11 C] metoclopramide that rely on K 2 (efflux) as a measure of Pgp function 14 . Garcia-varela et al., 2021 have made a head-to-head comparison between (R)-[ 11 C]-verapamil and [ 18 F]MC225 in non-human primates for measuring P-glycoprotein function. In baseline scans, [ 18 F]MC225 V T values were higher and k 2 values were lower than (R)-[ 11 C]-verapamil, whereas k 1 values were not significantly different. After blocking, V T values were the same for both tracers, whereas the k 1 and k 2 values were higher for (R)-[ 11 C]-verapamil than [ 18 F]MC225. Their study concludes that k 1 of (R)-[ 11 C]verapamil may not be an adequate parameter to measure the Pgp function. The in-vitro studies showed [ 18 F]MC225 to be more specific than (R)-[ 11 C]verapamil 15 . Therefore, there is more to explore and understand the dynamics of the chemoresistance using Pgp as a marker. Kannan et al., 2020 have developed and characterized [ 18 F]AVT-011 as a new radiotracer for imaging MDR in tumors. This study has shown the potential of [ 18 F]AVT-011 to measure ABCB1 function in tumors. The study's major limitation was low yield, which may be due to multistep radiolabeling. Therefore, the authors emphasized developing a simple labeling method that can provide high yield and is feasible to translate into humans 16 .
We have designed, synthesized, characterized, and radiolabeled the tosylate precursor (AVT-011) using a single-step chemical reaction in the present work. The radiochemical purity, quality control, stability, animal bio-distribution, and breast cancer patients (n = 6) have been studied and reported in this paper.
Material and methods
All the chemicals and reagents were procured from Sigma Aldrich (USA). 18 F radioisotope was produced with our in-house 16.5 MeV Cyclotron (PETtrace 860, GE Healthcare, USA) by the proton bombardment on the enriched O-18 water using [ 18 O(p,n) 18 F] reaction. The proton bombardment was done with the beam current ranging from 30 to 65 µA for 10-30 min depending upon the requirement of 18 F. The 18 F was delivered to the synthesizer module (Tracerlab FX2N, GE Healthcare, Chicago, USA) using Helium (UHP-6.0) as a carrier gas. All sep-pak cartridges like C18 plus, tC18 plus, Wax were procured from Waters-India, and C18ec was procured from Macherey-Nagel (Dueren, Germany). The pH was tested using pH paper (Fisher Scientific, New Hampshire, USA) and a pH meter (Mettler Toledo, Ohio, USA). The intermediate-I (4) was donated by Avaant Imaging (Lexington, MA, USA), and the precursor was synthesized by Essente healthcare (Bengaluru, Karnataka, India). The radiochemical purity was measured using high-performance liquid chromatography (HPLC) methods. HPLC system (Dionex, California, USA) was equipped with UV-Vis coupled and radioactivity detector. The quantitative analysis was done on C18 column (5 um 4.6 × 250, Shim-pack GW, Schimadzu) using mobile phase composition of acetonitrile (0.1% trifluoroacetic acid) and water (0.1% trifluoroacetic acid), starting with 5% acetonitrile (0-5 min), 5% to 100% acetonitrile (5-20 min), then 100% acetonitrile (20-25 min) and again at 5% acetonitrile (25-30 min). The residual solvents were measured using Gas chromatography (GC) (Scion 436 GC, Netherlands) with a flame ionization detector (FID). The column was operated initially at 40 °C for the first 3 min and then rose to 50 °C/min for up to 8 min, and the final temperature was set at 240 °C, and the column was BR-200 ms, 0.32 mm ID. The makeup gas consists of Helium (28 mL/min), zero air (300 mL/min), and hydrogen gas (30 mL/min) flow at the rate of 2 mL/min. The endotoxin test was carried out using Endosafe PTS cartridges on a NexGen Endosafe PTS machine from Charles River (Massachusetts, USA). A dilution factor of 100 was made for each preparation, 25 µl of the sample was added to each well, and a sterility test was performed using tryptic soya broth. PET-CT images were acquired on Philips Ingenuity TF 128 (Cleveland, OH, USA). It combines modular, LYSO-based PET components with a 128-channel CT component. The PET-CT imaging was done in breast cancer patients at Sri Shankara Cancer Hospital, Bengaluru, Karnataka, India. The biodistribution data of patients were used in our study. www.nature.com/scientificreports/ p-Toluene sulfonyl chloride (1) (570 mg, 3 mmol) was added portion wise to the mixture of ethylene glycol (2) (0.6 mL, 10 mmol) and pyridine (0.5 mL) at 0-5 °C and stirred for 3 h. TLC monitored the completion of the reaction. The reaction temperature was raised to room temperature and poured into ice water (50 mL) with constant stirring. A white solid was obtained, which was filtered and washed in cold water (50 mL). The solid was dried under air for 12 h to obtain the crude product. Then the crude product was purified by washing with ether to obtain 1,2-bis(tosyloxy)ethane (3) as a white solid (Yield = 72 ± 2%). HPLC and 1H-NMR characterized the product.
(c) Fluorine-19 labeling and cold synthesis of AVT-011 (6)
Tosylate precursor (5) (2 mg, 2.5 µmol) was mixed with sodium fluoride (NaF) (1.5 mg, 15 equiv.) and Kryptofix (3.6 mg, 4 equiv.) in the acetonitrile in a sealed vessel at 85 °C (in a pre-heated oil bath) for 20 min (Fig. 2). Afterwards, the acetonitrile was removed using nitrogen gas, and the residue was dissolved in ethyl acetate (1 mL). The suspension was loaded onto a silica gel cartridge (isosolute SPE column). The column was washed with ethyl acetate (4 mL) and followed by 10% ethanol in ethyl acetate (4 mL), and the final product was www.nature.com/scientificreports/ eluted with 40% ethanol in ethyl acetate (4 mL). The solvent was evaporated using nitrogen gas to yield the pure product (6) as a white solid (0.95 mg). The final product was characterised by LC-MS and 1 H-NMR.
(d) Radiolabeling and synthesis of [ 18 F]AVT-011 (7)
The reactions were carried out in the FX2N tracer lab module; around 500 ± 50 mCi of 18 F was added to the chemistry module from the cyclotron. The 18 F was trapped on a preconditioned anion exchange cartridge (QMA, ABX, Germany). It was eluted using a solution of kryptofix (7.5-15 mg of K222 and 1.5-3.0 mg of potassium carbonate) dissolved in 0.9 mL of acetonitrile and 100 µL of water. The precursor AVT-011 (05) (4-8 mg) was dissolved in 1.5 mL of solvent [dimethylformamide, dimethylsulfoxide, and acetonitrile (by heating at 80 °C)]. The reaction was carried out at 110, 130, and 160 °C for variable time points of 10, 20, and 30 min. The K222 eluent, precursor, and diluent (30% acetonitrile) were filled in vial-1, 3, and 5, respectively. Columns C18, tC18, and C18ec were used to purify the final product [ 18 F]AVT-011 (07). The columns were preconditioned with ethanol (4 mL) and water (10 mL). The SPE cartridge was fixed between V17 and V15 at C18 (2) port. Water (4 mL), 8% ethanol (3 mL) and ethanol (1.5 mL) were filled in vial-12, 13 and 14 respectively. Quality control. The radiochemical purity of the preparation was estimated by HPLC, and the solvent phase was used as described in the materials section. The retention time of the cold AVT-011 standard was measured at λ max 254 nm using a UV-Vis detector, whereas the [ 18 F]AVT-011 (07) was detected using the radioactive detector. The quality control parameters like physical appearance, pH, radiochemical purity, radionuclide purity, chemical purities, residual solvents, sterility, endotoxin test, and stability (in-vitro) were done as per the standard protocol described elsewhere 17 .
Cell binding assay. The cell binding was studied in the cancer cell line. U-87 MG cells were purchased from the National Centre for Cell Science (NCCS), Pune. The cells were cultured in Dulbecco's Modified Eagle Medium (DMEM)-high glucose containing 10% fetal bovine serum and 1% antibiotic in a T75 flask. The cells (suspended in media) were cultured in a T75 flask and placed in a humidified incubator at 37 °C under 5% CO 2 /95% air. The cells, when confluent, were detached using 0.25% trypsin-EDTA, washed and re-suspended in 1x phosphate buffered saline (PBS) at a concentration of approximately 5 × 10 6 cells/mL. Each test tube contained 1 mL of cell suspension into which 10 µL (~ 0.74 MBq) of [ 18 F]AVT-011 was added and incubated for 15 min, 30 min, 60 min, and 120 min at 37 °C in a water bath.
The blocking studies were done by incubating the same cell suspension (1 mL) with 500-fold excess of Pgp blocking agent Tariquidar solution for 30 min at 37 °C in a water bath. After 30 min, 10 µL (~ 0.74 MBq) of [ 18 F]AVT-011 was added and incubated for 15 min, 30 min, 60 min, and 120 min at 37 °C in a water bath. The incubation was terminated by adding 0.5 mL of normal cold saline. The mixture was centrifuged at 3000 rpm for 10 min, and the supernatant from each wash was collected into marked test tubes. The pellet was washed thrice with normal saline, and the supernatant was collected separately in the significant tubes. The radioactivity associated with cells and supernatant was counted in a gamma counter (Capnitec Inc), and the percentage of radioactivity bound to the cells was calculated. , and the mice were sacrificed at 15-, 30-, 60-and 120-min post-injection. Mice were then euthanized by carbon dioxide inhalation, and tissues were dissected, washed free of blood, dried, and weighed. Concomitant radioactivity was counted with an 18 F standard solution and injected dose per gram (% ID/g) was then calculated. For in-vivo stability, cardiac blood was collected (n = 2) in the heparinized vials after sacrificing the mice at 1 h. The blood was centrifuged at 3000×g for 10 min at room temperature to separate the plasma. An equal volume of separated plasma and chloroform/methanol (4:1, v/v) were mixed and centrifuged at 6000×g for 5 min to precipitate the proteins. The supernatant was subjected to the HPLC analysis.
Human PET-CT imaging. The breast cancer patients were enrolled in the study before starting their chemotherapy after obtaining approval from the Institutional Ethics Committee (Sri Shankara cancer hospital and research center, Ref no. SSCHRC/IEC10/55 dated 20.08.2020). The patients were enrolled in the study only after signing the informed consent form (format enclosed), and all experiments were performed in accordance with IEC guidelines for human experiments. The vital parameters, including pulse, blood pressure, oxygen saturation levels, and respiratory rate, were measured in all the patients before and after injecting radiopharmaceuticals.
Briefly, 370 ± 40 MBq of [ 18 F]AVT-011 was injected intravenously in all the enrolled patients (n = 6). The wholebody images were acquired after 45-60 min post-injection. Firstly, a low dose CT was acquired without contrast followed by PET with a bed position for 3 min. The fused images were processed using iterative reconstruction in a dedicated 128 PET/CT ingenuity machine. The images were analyzed for biodistribution of the tracer. The uptake in the various organs was measured and expressed in terms of standard uptake values (SUV). The background was measured over the lung fields where no lesion was identified. The SUV was measured over the organs like the liver, lungs, kidneys, intestine, brain, and lesion site. A standard 1 cm of the region of interest (ROI) was drawn on the ascending aorta to obtain SUV of the blood pool. The lymph node of the breast lesion side was considered for calculating SUV.
Results
Characterization of tosylate precursor and cold labeling. The compound 1,2-bis(tosyloxy)ethane 13 The appearance of the singlet at δ 3.88 in the proton NMR confirmed the presence of the -OCH 3 function. In contrast, the singlet at 2.66 ppm indicated the insertion of the tosylate bridge in precursor 5. Aromatic protons were observed at δ 6.56-9.36, whereas the N-ethylene functions were found between 2.28-2.82 ppm in the 1 H NMR. The appearance of the amide carbons at δ 163.50 and 167.46 ppm in the 13 C NMR confirmed the formation of the precursor 5.
The cold labeling was confirmed by a reaction between precursor (5) The feasibility of the precursor for labeling with 18 F was evaluated by the reaction of precursor 5 with cold fluorine ( 19 F) and is depicted in Fig. 2 (6). The LCMS data showed the substitution of the 19 F with the tosylate group of the precursor. The appearance of a [M + H] + peak at an m/z value of 679.25 in the mass spectrum confirmed the formation of the 19 F labelled target derivative 6. The yield of the target compound (6) was found to be 33 ± 4% and the product purity was greater than 99%. The chemical characterization data of the precursor and final compounds 3, 5, 6, and 7 are given in supplementary Figs. 1-7. Radiochemistry. The various concentration of kryptofix and precursor were tried to obtain the molar ratio of AVT/K222 , which can provide the maximum yield. The data (Table 1) showed that the molar ratio of (0.25:1) AVT (8 mg)/K222 (15 mg) was appropriate to give the maximum possible yield. The suitable solvent for dissolving the precursor was acetonitrile (dissolved by heating), and the proper radiolabeling reaction temperature was 110 °C for 10 min (Fig. 2). After adding 500 ± 50 mCi of 18 F, 210 ± 30 mCi of [ 18 F]AVT-011 was obtained (decay corrected). The incorporation of the 18 F into the precursor was 40 ± 4% as measured through HPLC. The final product was eluted with 1.5 mL of ethanol and further diluted with 15 mL of physiological saline and passed through a 0.22µ filter. After using various cartridges, C18ec was found to provide higher radiochemical purity www.nature.com/scientificreports/ with a yield of 42 ± 2% ( Table 1). The molar activity was 700 ± 60 GBq/μmol. Therefore, the final loading conditions (in FX2N tracerlab module) were 15 mg of K 222 (3 mg of K 2 CO 3 ) in acetonitrile/water was loaded in tube 1. 8 mg of the precursor was dissolved in 1 mL of acetonitrile in tube 3. 5 mL of acidic water was loaded in tube 5, 5 mL of water, 3 mL of 8%ethanol and 1.5 mL of ethanol was loaded into the tube 12, 13 and 14 respectively. The preconditioned QMA was used for trapping 18 F, and C18ec was used for final purification of the product from the reaction mixture. The product was collected in a collection vial containing 15 mL of saline and then transferred to the product vial by passing through a 0.22µ filter (cathivex).
Quality control. The final solution appeared clear after physical inspection and had a pH of 5.5 ± 0.6. The radiochemical purity of the final preparation was 97 ± 2% as evaluated by HPLC with radioactive peak retention time at 16.4 ± 1.2 min (n = 8) as compared to free 18 F at 4.0 ± 0.5 min (Fig. 3). The peak of [ 18 F]AVT-011 was identified by comparing it with the UV/Vis peak of standard AVT-011 solution at 254 nm under identical conditions. The retention time of the standard AVT-011 solution was 15.8 ± 1.3 min (Fig. 4). The radiotracer was found to be stable under in-vitro conditions. The radionuclide purity was more than 95%, and a half-life of 111 ± 4 min proved the radionuclide identity as 18 F with 511 keV peak as a significant peak. As evaluated by previously described methods, the kryptofix levels were below 50 µg/mL 17 . The concentration of residual solvent DMF was 80 ± 8 ppm (reference value-810 ppm). The ethanol content in the final preparation was 50,000 ± 10,000 ppm (reference value < 5000 ppm) due to the elution of the final product using ethanol. Still, the final preparation has less than 10% ethanol after dilution. The endotoxin levels were 1.0 endotoxin units (EU)/dose compared to the prescribed limit of 175 EU/dose. The preparation showed no turbidity in a soya broth over 14 days. These tests confirmed the sterility of the final dose and were suitable for patient use.
Cell binding assay. The cell binding assay showed uptake of 41.2 ± 4.8% in the U87 cells at 15 min, decreasing to 30.4 ± 3.1% at 120 min ( Table 2). The blocking study showed a significant (p < 0.005) increase in the radi- www.nature.com/scientificreports/ otracer uptake by 40% with time. It showed that blocking the Pgp efflux pump leads to trapping the radiotracer inside the cells, and retention increases with time. It showed its specificity towards the Pgp.
Animal biodistribution and stability. The tissue distribution showed the highest percentage of the injected dose was found in the liver (39.3 ± 12.2%), spleen (27.9 ± 8.9%), lungs (20.5 ± 1.9%), and kidneys (17.8 ± 4.4%) at 15 min (Fig. 5). With time, the activity was washed out from the liver, spleen, and lungs. The radioactivity increased in the kidneys and intestine with time. It showed a hepato-renal excretory route for the radiotracer. The normal brain showed minimal uptake of the radiotracer. The in-vivo stability was more than 92% at 1 h.
PET-CT imaging.
There was a rapid clearance of the radiotracer from the blood pool with predominant entero-hepatic clearance and additional renal clearance, as noted in the animal studies. The highest concentration of the tracer was seen in the liver, spleen, colon, and myocardium. The SUV (Table 3) showed the highest liver and gall bladder uptake. The higher uptake in the liver may be due to the metabolic breakdown of the tracer in the liver; the activity was cleared from the liver through the gall bladder and intestine (Fig. 6). The uptake in the lesion site was low, but it must be correlated with the Pgp expression of the lesion from the biopsy sample to draw any conclusion.
Discussion
Pgp plays a significant role in both physiological and pathophysiological conditions. It can efflux substrates out of the cell into the luminal space and ultimately out of the body. Due to the excretory role, it is majorly expressed on the cell membranes of the organs such as kidneys, liver, and intestine. The functional activity of Pgp is decisive under many conditions. The non-invasive imaging technique could provide insight into the functional status of Pgp behavior. Chemoresistance is a significant problem to the therapy, and timely detection of the chemoresistance may help the treating physician to change the course of the chemotherapy regimen. It can also help . This may be due to the two steps involved in the labeling; despite low yield, they carried out animal studies. It showed higher brain uptake in the knockout (Abcb1a/b-/-) mice compared to the control mice. The blocking studies showed an eight-fold increase in brain uptake, proving its specificity. The doxorubicin-resistant mice showed 32% lower uptake and increased by 40% with tariquidar administration. Their results showed the specificity of the [ 18 F]AVT-011 for the Pgp.
The main aim of our study was to overcome the limitations of the two-step method and develop a single-step synthesis method, and we have succeeded in our approach. We have conjugated the 1,2-bis(tosyloxy)ethane (3) with 6-O-desmethyl tariquidar to form a tosylate-based precursor (6). The tosylate group was substituted during the labeling by 18 F, and a one-step radio-synthesis was developed. The precursor was characterized using HPLC, LC-MS, 1 H-NMR, and 13 C-NMR. It showed that the precursor was synthesized with more than 99% purity with a yield of 36%. Various kryptofix (7.5, 10, and 15 mg) were used to elute 18 F from QMA cartridges, and maximum yield was obtained using 15 mg kryptofix and 3 mg of K 2 CO 3 . The molar ratio of 0.25/1 (AVT-011/K222) has given the maximum yield and radiochemical purity of 42 ± 2% with 97 ± 2%, respectively. Therefore, the final reaction condition was set to use 15 mg of kryptofix (3 mg of K 2 CO 3 ) with 8 mg of precursor. [ 18 F] was dried at 120 °C for 6 min, and the total elapsed time of the step was 20 min. The precursor was added to the dried [ 18 F], and the reaction was carried out at 110 °C for 10 min. After completion of the reaction, the reaction mixture was cooled to 50 °C and diluted with 7 mL of acidic water. The diluent mixture was loaded onto the preconditioned C18ec cartridge and washed with 7 mL of water. The final product was eluted with 1.5 mL of ethanol in the www.nature.com/scientificreports/ collection vial and diluted with 15 mL of physiological saline. It is transferred to the sterile vial through a 0.22µ cathivex filter in the dispenser. The total time of the synthesis was around 40 min. The in-vitro studies were carried out in the glioma cell line (U87 cells), which expressed Pgp receptors 18 . Therefore, we have used these cells for in-vitro validation. The cells were incubated with [ 18 F]AVT-011 for a pre-set time, and uptake decreased with time and remained the same for 1 and 2 h, which proves its saturation at 1 h. It showed that the tracer was effluxed from the cell with time and maximum efflux was maximum at 1 h. Blocking the Pgp receptors by incubating with an excess of the tariquidar solution showed a increased uptake of [ 18 F]AVT-011 in the cells. On blocking Pgp receptors, radiotracer gets trapped in the cells, which leads to an increase in the uptake. The uptake remained consistent at around 70% from 30-120 min. The in-vitro studies showed their affinity and specificity towards Pgp receptors. It proved the specificity of the [ 18 F]AVT-011.
The animal biodistribution studies showed higher uptake in the liver, spleen, lungs, and kidneys at 15 min, even after 30 min, the uptake was higher in the liver, intestine, and kidney. This showed that the [ 18 F]AVT-011 got metabolized in the liver and excreted through the intestine. The renal system was the other route of clearance. Hence, it showed the hepato-renal dual excretory pathway.
After obtaining institute ethics clearance, the tracer was tested on breast cancer patients (part of another research project). The biodistribution pattern in humans was found to be on similar lines with the animal distribution. The tracer was cleared through the hepatic-renal pathway. The uptake at the lesion site must be correlated to the Pgp status; hence it will be elaborated on in our other study. The human data showed the biodistribution of the [ 18 F]AVT-011 in breast cancer patients and ensured human safety as all parameters were normal before and after the scan. The biodistribution pattern observed in our study was in similar lines as reported by Kannan et al. The major limitation of [ 18 F]AVT-011 would be its uptake in the abdominal region. It cannot be an ideal tracer for imaging tumors located in the abdominal region, especially in the liver and gastrointestinal parts.
The affinity of [ 18 F]AVT-011 for Pgp receptors has been proved in the previous studies conducted by Kannan et al., 2020. We have worked on the limitations pointed out in their studies and developed a single-step radiolabeling method with a high yield. We have increased the yield to 42 ± 2% as compared to 1-2% reported by Kannan et al. It has increased its chances to explore its clinical application. We are working further on using this tracer for gliomas and increasing its specific activity using the HPLC method of radiosynthesis. Though, solid phase purification is a convenient method and gives a high yield. Therefore, we have used this approach and developed a convenient, single-step method yielding high radiochemical purity.
Conclusion
We have synthesized a modified precursor to developing a single-step radiosynthesis protocol for [ 18 F]AVT-011 with a high yield and radiochemical purity. The high yield of the [ 18 F]AVT-011 is advantageous for supplying this tracer at a distant hospital (without a cyclotron facility) and allows an exploration of its applications in various disorders.
|
v2
|
2022-11-04T06:18:03.902Z
|
2022-11-03T00:00:00.000Z
|
253267518
|
s2ag/train
|
Epigenomic effects of vitamin D in colorectal cancer.
Vitamin D regulates a plethora of physiological processes in the human body and has been proposed to exert several anticancer effects. Epigenetics plays an important role in regulating vitamin D actions. In this review, we highlight the recent advances in the understanding of different epigenetic factors such as lncRNAs, miRNAs, methylation and acetylation influenced by vitamin D and its downstream targets in colorectal cancer to find more potential therapeutic targets. We discuss how vitamin D exerts anticancer properties through interactions between the vitamin D receptor and genes (e.g., SLC30A10), the microenvironment, microbiota and other factors in colorectal cancer. Developing therapeutic approaches targeting the vitamin D signaling system will be aided by a better knowledge of the epigenetic impact of vitamin D.
|
v2
|
2022-11-04T06:18:03.944Z
|
2022-11-03T00:00:00.000Z
|
253267568
|
s2ag/train
|
Protective effects of Capsicum fruits and their constituents on damages in TNF-α-stimulated human dermal fibroblasts.
BACKGROUND
Antioxidant and anti-inflammatory effects of natural products on skin cells have been proved to be effective in improving skin damages. Capsicum species contain capsaicinoids which have antioxidant and anti-inflammatory properties, and various subspecies are cultivated. In this study, the effects of four Capsicum fruits and major constituents on oxidative stress and inflammatory reactions were measured human dermal fibroblasts (HDFs) to verify their effects on skin damage.
RESULTS
The inhibitory effects of nitric oxide (NO), reactive oxygen species (ROS), and prostaglandin E2 (PGE2 ) by cucumber hot pepper (CHP), red pepper (RDP), Shishito pepper (SSP), and Cheongyang pepper (CYP) were determined in HDFs. RDP and SSP inhibited the production of NO, ROS, and PGE2 in tumor necrosis factor-alpha (TNF-α)-stimulated HDFs. Additionally, SSP seeds restored TNF-α-induced increase in matrix metalloproteinase (MMP)-1 and decreased procollagen I α1 (COLIA1). In HPLC analysis of capsaicinoids capsaicin (CAP) and dihydrocapsaicin (DHC), CAP was detected higher than DHC in the peel and seeds of all 4 types of Capsicum fruits, and the total amount of capsaicinoids was the highest in SSP. CAP and DHC, which are were major constituents of Capsicum fruits, also inhibited NO, ROS, and PGE2 and restored MMP-1 and COLIA1.
CONCLUSION
RDP and SSP were shown significant protective effect on skin damages including oxidative stress, inflammatory reactions and reduction of collagens. Capsaicinoids CAP and DHC were proved as active constituents. This research may provide basic data for developing of Capsicum fruits as ingredients to improve skin damage such as inflammation and skin aging. This article is protected by copyright. All rights reserved.
|
v2
|
2022-11-04T06:18:04.057Z
|
2022-11-03T00:00:00.000Z
|
253265948
|
s2ag/train
|
Neuroprotective effect of canagliflozin against cisplatin-induced cerebral cortex injury is mediated by regulation of HO-1/PPAR-γ, SIRT1/FOXO-3, JNK/AP-1, TLR4/iNOS, and Ang II/Ang 1-7 signals.
OBJECTIVES
Canagliflozin (CAN), a sodium-glucose co-transporter 2 inhibitor, is an anti-hyperglycemic drug that has been approved to treat diabetes. This study evaluated the beneficial effects of CAN on cerebral cortex intoxication induced by cisplatin (CIS).
MATERIALS AND METHODS
Rats were allocated into 4 groups: normal control, CAN (10 mg/kg, P.O.) for 10 days, CIS (8 mg/kg, i.p.) as a single dose on the 5th day of the experiment, and CAN + CIS group.
RESULTS
In comparison with CIS control rats, CAN significantly mitigated CIS-induced cortical changes in rats' behavior in the open field and forced swimming assessment as well as histological structure. Biochemically, CAN administration efficiently decreased lipid peroxidation biomarkers MDA and boosted the antioxidant status via a remarkable increase in the cortical GSH content as well as enzymatic activities of antioxidant enzymes SOD, GST, CAT, and GPx mediated by up-regulation of HO-1, PPAR-γ, and SIRT1/FOXO-3 signals. Additionally, pretreatment with CAN significantly decreased cortical MPO, NO2-, TNF-α, and IL-6 levels. At the same time, it elevated the IL-10 level associated with the downregulation of JNK/AP-1, TLR4/iNOS/NO, and Ang II/Ang 1-7 signals.
CONCLUSION
Because of the potent antioxidant and anti-inflammatory properties of CAN, our findings showed that CAN could be a good candidate for the protection against CIS-induced cortical intoxication in the patient receiving CIS.
|
v2
|
2022-11-04T06:18:04.081Z
|
2022-11-03T00:00:00.000Z
|
253267378
|
s2ag/train
|
MRI Fat-Saturated T2-Weighted Radiomics Model for Identifying the Ki-67 Index of Soft Tissue Sarcomas.
BACKGROUND
Ki-67 expression has been shown to be an important risk factor associated with prognosis in patients with soft tissue sarcomas (STSs). Its assessment requires fine-needle biopsy and its accuracy can be influenced by tumor heterogeneity.
PURPOSE
To develop and test an MRI-based radiomics nomogram for identifying the Ki-67 status of STSs.
STUDY TYPE
Retrospective.
POPULATION
A total of 149 patients at two independent institutions (training cohort [high Ki-67/low ki-67]: 102 [52/50], external validation cohort [high Ki-67/low ki-67]: 47 [28/19]) with STSs.
FIELD STRENGTH/SEQUENCE
Fat-saturated T2-weighted imaging (FS-T2WI) with a fat-suppressed fast spin/turbo spin echo sequence at 1.5 T or 3 T.
ASSESSMENT
After radiomics feature extraction, logistic regression (LR), random forest (RF), support vector machine (SVM), and k-nearest neighbor (KNN) were used to construct radiomics models to distinguish between high and low Ki-67 status. Clinical-MRI characteristics included age, gender, location, size, margin, and MRI morphological features (size, margin, signal intensity, and peritumoral hyperintensity) were assessed. Univariate and multivariate logistic regression analysis were applied for screening significant risk factors. Radiomics nomogram was constructed by radiomics signature and risk factors.
STATISTICAL TESTS
Model performances (discrimination, calibration, and clinical usefulness) were validated in the validation cohort. The nomogram was assessed using the Harrell index of concordance (C-index), calibration curve analysis. The clinical utility of the model was assessed by decision curve analysis (DCA).
RESULTS
LR, RF, SVM, and KNN models represented AUCs of 0.789, 0.755, 0.726, and 0.701 in the validation cohort (P > 0.05). The nomogram had a C-index of 0.895 (95% CI: 0.837-0.953) in the training cohort and 0.852 (95% CI: 0.796-0.957) in the validation cohort and it demonstrated good calibration and clinical utility (P = 0.972 for the training cohort and P = 0.727 for the validation cohort).
DATA CONCLUSION
This MRI-based radiomics nomogram developed showed good performance in identifying Ki-67 expression status in STSs.
LEVEL OF EVIDENCE: 3
TECHNICAL EFFICACY STAGE
2.
|
v2
|
2022-11-04T06:18:04.835Z
|
2022-11-03T00:00:00.000Z
|
253266303
|
s2orc/train
|
International benchmarking of childhood cancer survival by stage at diagnosis: The BENCHISTA project protocol
Background Several studies have shown significant variation in overall survival rates from childhood cancer between countries, using population-based cancer registry (PBCR) data for all cancers combined and for many individual tumour types among children. Without accurate and comparable data on Tumour stage at diagnosis, it is difficult to define the reasons for these survival differences. This is because measurement systems designed for adult cancers do not apply to children’s cancers and cancer registries often hold limited information on paediatric tumour stage and the data sources used to define it. Aims The BENCHISTA project aims to test the application of the international consensus “Toronto Staging Guidelines” (TG) for paediatric tumours by European and non-European PBCRs for six common paediatric solid tumours so that reliable comparisons of stage at diagnosis and survival rates by stage can be made to understand any differences. A secondary aim is to test the data availability and completeness of collection of several ‘Toronto’ consensus non-stage prognostic factors, treatment types given, occurrence of relapse/progression and cause of death as a descriptive feasibility study. Methods PBCRs will use their permitted data access channels to apply the Toronto staging guidelines to all incident cases of six solid childhood cancers (medulloblastoma, osteosarcoma, Ewings sarcoma, rhabdomyosarcoma, neuroblastoma and Wilms tumour) diagnosed in a consecutive three-year period within 2014–2017 in their population. Each registry will provide a de-identified patient-level dataset including tumour stage at diagnosis, with only the contributing registry holding the information that would be needed to re-identify the patients. Where available to the registry, patient-level data on ‘Toronto’ non-stage prognostic factors, treatments given and clinical outcomes (relapse/progression/cause of death) will be included. More than 60 PBCRs have been involved in defining the patient-level dataset items and intend to participate by contributing their population-level data. Tumour-specific on-line training workshops with clinical experts are available to cancer registry staff to assist them in applying the Toronto staging guidelines in a consistent manner. There is also a project-specific help desk for discussion of difficult cases and promotion of the CanStaging online tools, developed through the International Association of Cancer Registries, to further ensure standardisation of data collection. Country-specific stage distribution and observed survival by stage at diagnosis will be calculated for each tumour type to compare survival between countries or large geographical regions. Discussion This study will be promote and enhance the collection of standardized staging data for childhood cancer by European and non-European population-based cancer registries. Therefore, this project can be seen as a feasibility project of widespread use of Toronto Staging at a population-level by cancer registries, specifying the data sources used and testing how well standardized the processes can be. Variation in tumour stage distribution could be due to real differences, to different diagnostic practices between countries and/or to variability in how cancer registries assign Toronto stage. This work also aims to strengthen working relationships between cancer registries, clinical services and cancer-specific clinical study groups, which is important for improving patient outcomes and stimulating research.
Introduction
Interventions to improve survival probabilities for children with cancer require a detailed understanding of reasons for treatment failure (e.g., tumour relapse or progression or toxicityrelated death). The extent of tumour spread at diagnosis (tumour stage) is one of the most important prognostic factors determining the chance of a patient with childhood cancer (CC) being 'cured' and is also a determinant of the intensity of treatment required. Disparities in tumour stage at diagnosis between countries could explain part of the survival differences seen in some international population based benchmarking studies [1][2][3][4]. Additional factors that may explain survival differences across countries are differences in diagnostic accuracy and in treatment [5].
PBCRs collect information on all new cancer cases that occur in a well-defined population, corresponding to a specific geographic area, producing unbiased population-level cancer indicators. So far, most PBCRs hold incomplete data on tumour stage for CC. This is because staging systems used for adult cancers are not easily applicable to CC and access to necessary clinical data sources to assign tumour stage is difficult. In 2014, an international working group developed consensus staging guidelines for paediatric cancers, known as the "Toronto" guidelines (TG) [6]. The feasibility of applying these guidelines by PBCRs was thereafter tested in Europe and Australia [7,8]. These studies found that PBCRs are capable of assigning Toronto stage to a high proportion (>95%) of registered cases and demonstrated the resources required to acquire tumour staging information from clinical registries, treating hospitals or routine health care data sources.
The broad aims of the International Benchmarking of Childhood Cancer Survival by Stage (BENCHISTA) Project are to improve understanding of the reasons for variation in childhood cancer survival between countries and to highlight areas to be targeted for improvement. The project is expected to reveal variation in stage at diagnosis and survival by stage between some countries. If found, this would suggest that improvement initiatives should include efforts to achieve earlier diagnosis and to reduce variation in treatments given, respectively.
Documentation of tumour stage at diagnosis using the Toronto guidelines is now a recommended variable for routine collection by PBCRs in many jurisdictions [9]. The use of these international standardised guidelines for childhood cancer staging is crucial to allow meaningful comparisons. Thus, this project aims to encourage and enable the greatest number of European and wider international population-based cancer registries to apply the TG for staging patients affected by the most common solid paediatric cancers. Compared with the feasibility study performed in Europe [8], more tumour types and the expansion of PBCRs participation will be investigated. Moreover, the project will benefit from and help to disseminate the recent inclusion of Toronto staging guidelines in an international electronic cancer staging tool-with free online access or download-intended for use by all PBCRs globally and which covers all the tumour types included in the BENCHISTA Project [10]. Lastly, the project also aims to further enhance working relationships between PBCRs, clinical services and tumour-specific clinical study groups which is important for sustainable clinical outcomes research using routine healthcare data. The collaboration with clinicians and use of standardized guidelines for staging childhood cancer are key components in current and future research.
Objectives/Hypothesis
There are two main research questions aiming to explain variations in overall CC survival across countries: 1. Are there any differences in stage at diagnosis between countries? 2. Do survival probabilities by tumour stage vary between countries? Question 1 will be assessed by PBCRs through assignment of tumour stage at diagnosis using a standardised and internationally comparable framework-the "Toronto" consensus staging guidelines [6,11]. Question 2 will be answered through PBCRs collecting detailed follow-up data of CC cases for a minimum of 3 years. This information will be used to calculate overall survival probabilities and compare survival between countries or geographic regions.
To answer these questions, participating PBCRs will provide both routinely collected and project-specific information. Furthermore, a feasibility study will assess how easily PBCRs can collect data on first line treatment, tumour biology, non-stage prognosticators (NSP) [11], relapse/recurrence or progression of disease and cause of death.
Time frame
The study is conducted from 1st January 2021 to 31st January 2024. Data from PBCRs started to flow to the data controller from March 2022 and is expected to be completed by the end of 2022. CCs are defined using the International Classification of Childhood Cancers, third edition (ICCC-3) and the International Classification of Diseases for Oncology, third edition (ICD-O-3) for tumour site [12,13]. Only malignant behaviour is selected. The protocol appendix includes detailed recommendations for the correct identification of cancer topography and morphology and Toronto Stage assignment for the six cancer types included in the project.
Selection of tumour types
The project will study stage distribution and survival for six solid CCs: medulloblastoma, osteosarcoma, Ewing sarcoma, rhabdomyosarcoma, neuroblastoma, and Wilms tumour.
These tumours have been selected based on at least one of the following considerations: • They have generally good prognosis (Wilms tumour, localised neuroblastoma) and are curable using 'standard of care' treatment regimens.
• For some of these tumour types, important differences in outcomes have already been demonstrated between certain populations [1].
• These cancers have shown little or no improvement in survival probabilities over a long period (1999-2007) [1].
Together, these tumour types represent a considerable percentage (about 50%) of all childhood solid tumours [1]. The expected numbers of cases for the three year inclusion period of this study by cancers are shown in Table 1 by ICCC-3 classification. Approximately 9,000 cases are expected to be included: 2,300 cases of Neuroblastoma and ganglioneuroblastoma; 1,698 cases of Nephroblastoma and other nonepithelial renal tumours; 1,521 cases of Medulloblastoma; 1,113 cases of Osteosarcoma; 949 cases of Ewing sarcoma and related sarcomas of bone; 1,170 cases of Rhabdomyosarcoma. These cases numbers are mostly estimated from data held by the EUROCARE-6 project (period of diagnosis 2005-2013) [14]. The PBCRs collect all the cancers recorded in the population resident in a specific jurisdiction or country, so this study will be population based and is not biased in the ways that might affect institutional or clinical trial series.
Inclusion criteria
The updated European Network of Cancer Registries (ENCR) recommendations [15] should be followed to record the date of incidence used by PBCRs to identify cases meeting the inclusion criteria: • All children under 15 years of age will be included. For the three cancer types common in adolescents (osteosarcoma, Ewing sarcoma and rhabdomyosarcoma), cases aged <20 years will be included, whenever this data are held by the registry.
• Information for three consecutive and complete years of incidence must be identified and submitted by each PBCR.
• Cases must be diagnosed in the period between 1.1.2014-31.12.2017 and have at least 3 years of follow-up for the definition of life status, according to each PBCR's practice.
The start date of the three year window of the selected period of incidence may be up to one year back and forth within this 4 year time frame, to maximize the PBCR's participation in the project, provided that all three of the above criteria are met. All cases that cannot be staged (due to missing information etc) must be included. The different data sources and methods used to reconstruct the stage will also be collected.
Identification of population-based cancer registries
All European PBCRs included in the EUROCARE-6 study (31 countries) have been invited to contribute. In addition, other non-European cancer registries from Australia, Ontario (Canada), Brazil, and Japan confirmed their availability to reconstruct Toronto stage and to participate in the BENCHISTA Project. Data from large institutions or clinical networks that possess high quality information may be acceptable providing full coverage of the incident CCs in the specific population can be demonstrated. In this case, additional external checks to verify the coverage will be performed comparing the number of submitted cases with incidence reported in literature. Registries commit to apply the Toronto guidelines for tumour staging to their existing data, using their online and usual data sources such as clinical records, pathological reports, and hospital discharge administrative files.
Almost all participating PBCRs (~65), have already been checked for quality indicators by the International Agency for Research on Cancer (IARC) [16], ENCR [17], European Cancer Registry Based Study on Survival and Care of Cancer Patients (EUROCARE project, [14], and/ or by the Automated Childhood Cancer Information System (ACCIS) [18], which improves the expected completeness and quality of the information collected for incidence and survival. Basically, all registries included in the EUROCARE 6 survival analysis for CC will be accepted. Cases ascertained only by death certificate (DCO), number of cases diagnosed by cytology or histology (microscopically verified) and those with unspecified morphology codes (NOS) will be considered as data quality indicators for the completeness and accuracy of population-level registration of the six diagnostic groups. The number of cases lost to follow-up and censored before the end date of follow-up will be used to assess the follow-up data quality.
Staging process
Participating PBCRs will assign tumour stage at diagnosis using the TG supported by the detailed guidance based on the Australian experience [7] and translated in French, Spanish, Bulgarian, Portuguese, Italian and Japanese to overcome any language barriers. The rules have also been incorporated into a free electronic cancer staging tool available online for overview and download at www.canstaging.org [10, 19] The TG include a two-tiered system approach to define stage [6,7] where Tier 2 staging system is more detailed and intended for use in high resource settings. All PBCRs will be asked to provide Tier 2 stage if they can access clinical details, otherwise Tier 1 will be acceptable only for assessing proportions of localised vs metastatic tumours at diagnosis.
Toronto stage is defined as extent of disease at the time of diagnosis and is based on evidence acquired before treatment with two exceptions: • Staging of localised (non-metastatic) Wilms tumour resected after pre-operative (neo-adjuvant) chemotherapy, where stage is based on surgical and pathological assessment of the nephrectomy specimen and indicated by the prefix 'y'.
• Staging of tumours in which investigations to exclude distant metastases occured within a short time after surgery to the primary tumour and before any systemic therapy has started.
For rhabdomyosarcoma, tumour stage should always be defined at diagnosis according to standard clinical TNM rules with nodal involvement assessed by imaging and/or lymph node biopsy, if performed prior to chemotherapy. Tier 2 Toronto staging includes tumour size (</� 5cm) and classification of the anatomical site as 'favourable' or 'unfavourable'.
For all diagnostic groups including Wilms tumour, the presence of distant metastases is assessed clinically (including imaging) or pathologically at the time of diagnosis and before neoadjuvant therapy. Metastases are defined at diagnosis.
Standardization (training)
To ensure registries assign TG stage in a standardized way, three online training sessions have been held. Clinical experts nominated by the relevant European tumour-specific clinical trial groups led on-line training courses, including exercises held in collaboration with the Belgium PBCR personnel. Attendees included cancer registration officers, clinicians, PBCR directors, and other professionals involved in the project. Recordings of the three training workshops can be accessed here: • Osteosarcoma and Ewing sarcoma: https://mediacentral.ucl.ac.uk/Play/71900 • Neuroblastoma and Wilms tumour: https://mediacentral.ucl.ac.uk/Play/71797 • Rhabdomyosarcoma and Medulloblastoma: https://mediacentral.ucl.ac.uk/Play/72207 The topics covered were: • General principles of 'Toronto staging'.
• Introduction, general aspects, diagnosis, therapy and non-stage prognosticators for each tumour.
• Staging exercises based on fictional cases.
• Brief explanation of variables requested to the PBCRs.
• Use of cancer staging tool and applied exercises.
For quality assurance purposes, and more specifically to analyse standardization procedures in the application of TG across registries, an exercise consisting of fictitious cases for staging has been generated and made available to participating PBCRs. Moreover, a survey to understand methods of data collection (availability of imaging, participation in training sessions, etc.) and to identify country-specific difficulties is also under current completion by participating PBCRs. Furthermore, a help desk promoting the interection between registries and clinicians to clarify grey cases was established and a Question & Answers document with all the queries posed by registrars is constantly updated and available on the project website [20].
Variables to be collected and structure of the case records to be submitted
For each tumour, PBCRs must complete a record template including compulsory and optional variables. The structure of the record is presented in Table 2.
Compulsory variables. Each record (case) includes demographic variables such as sex, year of birth, age at diagnosis in months, basis of diagnosis, plus information on examinations performed and data sources (clinical documents, administrative database, pathological database, etc) used by registrars for staging (see structure of the record, Table 2). If the tumour represents a second or third malignancy, the ICCC-3 classification of the previous tumour and the corresponding year of diagnosis should be reported. PBCRs should use all available sources of hospital data and enlist input from appropriate clinical staff where required to ensure consistent clinical interpretation of diagnostic investigations. Follow-up data (life status and time in days from diagnosis to death or last follow up) must be ensured up to at least three years from diagnosis.
Optional variables. This project will also assess availability of relevant information from registries regarding all six tumour types and non-stage prognosticators (NSP), primary treatment modalities used, relapse/recurrence/progression and cause of death. This information will be used for descriptive analyses of data availability, quality and completeness across participating countries.
Moreover, these optional variables may be useful as additional factors to explain any changes found in survival by stage. These analyses will focus on registries reporting a high percentage of completeness and on achieving data quality assurance.
Specific NSPs are requested for medulloblastoma, rhabdomyosarcoma, neuroblastoma and Wilms tumour, utilising the more recent Toronto guidelines on NSPs [11]. Considering that for most registries the main source available for stage reconstruction was, according to the European Joint Action on Rare Cancer pilot study, the clinical records of major hospital admissions, NSPs should be available [8]. Even if they are not the major objective of the study, NSPs are important to better understand survival differences, as they characterize the behaviour of the tumour and are crucial for clinical risk stratification for treatment.
Individual information on treatment modalities given to each patient (surgery, chemotherapy, radiotherapy) is required. If a registry is not able to identify first line therapy, it is recommended to include all treatments given in the first 12 months following date of diagnosis.
Knowledge of relapse/recurrence or progression of the disease is important for understanding the success of first line therapy and for estimation of event-free survival. PBCRs are therefore asked to provide data on occurrence of relapse/recurrence/progression for all cases within the 3 years of follow-up to understand the feasibility of collecting this data item. The distinction between relapse, recurrence or progression is not requested as this is not standardized.
Cause of death, categorised as due to tumour, toxicity, comorbidity or other cause, is an additional optional variable. This categorization requires a clinical review of the information reported to the PBCRs on causes of death, which may be multiple. Collection of this data item is important to understand the feasibility of future studies testing the hypothesis that differences in survival rates between countries may be partially ascribed to variation in deaths due to toxicity of treatment. ICDO-3-Topography 3 Only the numeric part of the ICD-O-3 topography code will be reported (the "C" and "." will not be included)
PLOS ONE
The BENCHISTA project protocol
Data quality
Quality data checks on the database will be performed at the IRCCS Foundation National Cancer Institute of Milan (INT) and problematic cases will be resolved with each PBCR. Indicators of quality and completeness of incidence data will be collected (DCO, microscopically verified cases, etc). Some additional indicators to define the accuracy of sub-typing definition specific for the six CC studied will be investigated. Furthermore, the number of cases received will be checked comparing them with the incidence reported in literature: the Automated Childhood Cancer Information System and the EUROCARE project papers on childhood cancers incidence and survival recently published [1,18].
Statistical analysis
The formal assessment of statistical power to detect differences in stage distribution and survival probabilities between countries is limited by the total number of incident cases of these rare tumour types per country. Therefore, analyses of stage distribution and survival probabilities for each tumour type per country will be descriptive, with 95% confidence intervals reported. As a population-based study, these are the largest numbers available and are not biased in the ways that may affect institutional or clinical trial series. The expected number of cases available in Europe by cancer type and by area, in three years, is approximately 9000 (Table 1). These numbers will be increased by the participation of some non-European jurisdictions (Australia, Brazil, Japan, Ontario (POGO, Canada) Endpoint 1: To formally assess differences in tumour stage at diagnosis, we will group European countries according to the geographical regions used in EUROCARE 5, to achieve the necessary groups' size [1]. Non-European jurisdictions will be considered individually. For expected case numbers by registry (based on the number of cases obtained from the EURO-CARE 5 study) there is approximately 60% power to detect a 10% difference in lower stages (localised, loco-regional) versus more advanced stage (metastatic) between two countries or regional groupings where there is a total of approximately 250 cases of each tumour type in each geographical comparator group (medulloblastoma, Wilms tumour). For group sizes with around 300 cases (neuroblastoma), the power would be 70%. The sarcomas, that collectively comprise about 35% of the cohort, will be combined for assessment of differences in stage distribution at diagnosis, according to the same country groupings. Endpoint 2: Survival differences between countries/regional groups, and how much of these differences are explained by variations in stage distribution will be studied by a multivariable Cox regression. Stage and other relevant prognostic variables (age at diagnosis, sex and/or primary site for at least some diagnostic groups), and confounders (including stage migration) will be included in the model. NSPs and recurrence/progression data will be considered, whenever they are available.
In this project it is not possible to calculate a sample size or a minimum detectable effect with the information available.
PLOS ONE
The BENCHISTA project protocol
Ethical consideration
Ethical approval for the project has been given by the Research Ethics Committee of University College London on 22nd of April 2021. Also, the Ethical Committee of the Fondazione IRCCS "Istituto Nazionale dei Tumori" (INT) approved the project during the e-session on 25th of May 2021. Individual patient consent is not required as data are collected under existing permissions for cancer registration in each jurisdiction [21][22][23]. The project has a named Patient/Parent involvement and engagement (PPIE) Lead and dedicated PPIE working group to ensure representation of the patient voice throughout. Such external communication and dissemination with key stakeholders will be crucial to raise awareness of the project and its rationale. A formal communication and dissemination plan is available on the project website [20].
Data management plan
Participating PBCRs will submit their maximally depersonalized dataset of cases to INT for analysis. The database of the project will be securely stored at INT for a maximum of 10 years, after which it will be destroyed unless ethical and regulatory approval is granted for further research.
The project's database is under the governance of the BENCHISTA project working group (PWG) and should be used according to the purposes for which the ethical approval has been granted. Only the INT-personnel involved in the project-will be able to access the BENCH-ISTA database. The BENCHISTA data remain the property of the contributing registries, whose consent is required before they can be used for purposes other than those originally envisaged in the BENCHISTA protocol. All members of the PWG that provide data must be informed of any analysis being proposed and carried out. Moreover, they all agreed on a common policy about communication, dissemination and publication.
Privacy
A Data Transfer Agreement (DTA) is a formal contract that documents which data are being transferred, the format and level of pseudonymisation and how the data can be used. The agreement protects the PBCRs providing the data, ensuring that data will not be misused, and prevents miscommunication as any questions about data use are formally agreed in advance. Each PBCR has one or two representatives in the PWG. In PWG meetings, held quarterly throughout the year, the whole materials and all the major and critical points (e.g., data-transfer and data-use issues) were discussed to obtain a collaborative understanding on the project and protocol.
Given the large number of PBCRs participating in the BENCHISTA Project (more than 60 registries) a single DTA was proposed to create a Consortium agreement. As stated in the Transparency Statement document of the project, the information will be collected from different cancer registries across the world in line with their national regulations and GDPR for data collection and protection of data for research use. The list of the groups involved in the project along with all the important links, details and documents (e.g., the publication policy and newsletters) are available in the BENCHISTA website [20].
Project governance
To ensure understanding and awareness of individual health care data usage in research, the BENCHISTA Project will establish a relationship with associations of parents and patients focused on paediatric cancers at a national and international level. Also, this collaboration will improve the communication of the final results of BENCHISTA to people who make decisions and to ultimately improve the public health organization at the national level for paediatric cancers.
Discussion
The BENCHISTA Project was born from the excellent synergy and collaboration emerged in the leading team and in the working group of the European Joint Action on Rare Cancer pilot project [8]. All the PBCRs participating in the pilot study enthusiastically accepted the proposal to expand the number of cases and the type of cancer to be staged with the Toronto Guidelines. Thanks to this important feedback, the researchers made an effort to include as many PBCRs as possible in order to stimulate the collection, address specific research hypotheses and focus on service/quality improvement. Rare cancers like the paediatric ones need of a high participation to provide clear results and, even if the stage at diagnosis is not always consistently collected or defined by PBCRs some of them are starting to apply the internationally recognised Toronto staging guidelines. A summary of these guidelines has been included in the TNM Classification of Malignant Tumours, 8th edition (2017) [24], and in a specific electronic cancer staging tool [14].
Differences in stage distribution could be due to real differences in burden of disease at diagnosis or to different diagnostic and staging criteria used by clinicians between countries or to different capabilities of PBCRs to access the necessary clinical information and/or interpret it to assign Toronto stage to each case.
To understand whether different diagnostic and staging approaches between countries have an impact on the risk of dying, information about the examinations performed to stage patients will be considered. Moreover, to adjust the different sources used by the registrars to assign the Toronto stage, the examinations results used by registrars to reconstruct stage will also be considered in the analysis.
From the PBCRs' perspective, their straightforward access to correct information, the presence of trained registrars involved in the data collection, the possibility to discuss difficult cases with clinicians and the availability of enough professionals working at the registries impact the integrity of collected information. All this information are captured trough an online survey and will help to investigate reasons for disparities across participating PBCRs and to interpret variations or lack of data availability in terms of stage at diagnosis.
The results of the project will be critically discussed considering several other factors related with the outcome collected at country level (e.g., social inequalities, presence of centre of reference or network of hospitals with teams involved in clinical international/national studies) [25], the creation of a questionnaire could be considered to address these factors.
To overcome problems of language barriers and comparability between cancer registries in assigning Toronto stage, the collaboration of clinicians involved in the training course and in the governance of the project was/is crucial. The help desk, the Question & Answers document, the training courses and the test with the fictitious cases are all available in the project's website aiming to provide guidance to all the PBCRs' personnel and to improve standardization and data management.
Regarding the transfer and collection of data, it is important to highlight there is noticeable heterogeneity in DTA requirements across countries and the processes required to set up a general DTA have been exhaustive and time-consuming. Administrative barriers should be reduced to optimize the exchange of data for research [21][22][23]. Thus, permanent data transfer agreements among research institutes, or at least agreed templates should be established to ensure efficient implementation of digital solutions for the exchange of health data among researchers.
The BENCHISTA Project is focused on obtaining the necessary information to apply the Toronto Staging guidelines on recently diagnosed cases obviously balancing between the need to have at least three years of follow-up and a good access to the required clinical data to retrospectively stage the patients. Hence, this project can be viewed as a feasibility project of widespread use of Toronto Staging at a population level by CRs, specifying data sources used and how well standardised the processes can be.
Participation in the project is intended to encourage PBCRs to routinely and consistently apply TG to all prospectively registered cases of childhood cancer to ensure the quality of the data, an essential feature for proper future assessments.
Additionally, the project will produce practical recommendations on strengthening collaboration between PBCRs and clinical/hospital registries so that staging of CC patients becomes more accurate, efficient and complete. This will ultimately allow future benchmarking research in survival analysis by stage to be undertaken in a more sustainable manner as prospective clinical observational studies using routine health care data. Stage at diagnosis is a variable collected for adult cancers by many national PBCRs. For childhood cancers, the ENCR-Joint Research Center data call for the European Cancer Information System now includes the TG, which are endorsed by the Union for International Cancer Control, International Agency for Research on Cancer, and the international associations of PBCRs (IACRand the ENCR).
PBCRs aim to improve the standards of registry in terms of health service and populationbased recording of cancer, but also to encourage the collection of new important clinical variables such as stage at diagnosis, recurrences, NSP and cause of death. Some of these parameters are an essential component of risk stratification aiming to guide the treatment protocol, any potential changes in it and improving short and long-term outcomes. The collection of these variables at PBCRs level is important to develop future studies and stimulate closer interactions with clinicians responsible of clinical/hospital registries and databases to better understand outcomes differences between countries.
|
v2
|
2022-11-05T06:17:04.908Z
|
2022-11-03T00:00:00.000Z
|
253302034
|
s2ag/train
|
Evaluating orelabrutinib as a novel treatment option for relapsed/refractory chronic lymphocytic leukemia in China
ABSTRACT Introduction The covalent Bruton’s tyrosine kinase (BTK) inhibitor ibrutinib has been approved in the USA for B cell malignancies for almost ten years and has improved the survival of patients with chronic lymphocytic leukemia (CLL)/small lymphocytic lymphoma (SLL). Orelabrutinib is a novel, highly selective covalent BTK inhibitor with proven efficacy and acceptable safety profile. In 2020, it was approved for the treatment of relapsed/refractory (R/R) CLL/SLL in China. Areas covered In this review, we summarized the current clinical trials exploring orelabrutinib monotherapy or orelabrutinib-based combination regimens in CLL/SLL, especially R/R CLL/SLL. Pharmacodynamics, pharmacokinetics, clinical efficacy and safety of orelabrutinib are also discussed. Expert opinion Orelabrutinib selectively inhibits BTK via covalent binding and exhibits linear pharmacokinetics. BTK is the only kinase targeted by orelabrutinib, and a few off-target toxicities of orelabrutinib have been reported. The phase I/II trial demonstrated the efficacy and safety of orelabrutinib in patients with R/R CLL/SLL; however, further clinical trials are needed to compare orelabrutinib with ibrutinib in patients with R/R CLL/SLL and to evaluate its efficacy and safety in patients with treatment-naive CLL/SLL.
|
v2
|
2022-11-05T06:17:05.862Z
|
2022-11-03T00:00:00.000Z
|
253303215
|
s2ag/train
|
Ganoderic acid A enhances tumor suppression function of oxaliplatin via inducing the cytotoxicity of T cells.
BACKGROUND
Various natural products have been demonstrated for their anti-tumor activities. As a natural triterpenoid, the effects of ganoderic acid A on oxaliplatin chemotherapy for cancer treatment remain unclear.
METHODS
A xenograft mouse model of colon cancer was constructed using the HT-29 cells. Ganoderic acid A was intravenously administered with or without oxaliplatin. The CCK-8 method was performed to assess cell viability. Flow cytometry was used to determine cell apoptosis and subtyping of T cells. Cytotoxicity of the T cells was assayed using a lymphocyte-tumor co-culture system in vitro.
RESULTS
Ganoderic acid A enhanced tumor suppression of oxaliplatin in the xenograft model, while single administration showed no obvious anti-tumor effect. Ganoderic acid A didn't affect cell proliferation and apoptosis of HT-29 cells treated by oxaliplatin in vitro. Additionally, ganoderic acid A co-administered with oxaliplatin didn't impact T cell subtyping in the xenograft model. Cytotoxicity of T cells in co-administered mice was remarkably enhanced compared with oxaliplatin-treated mice.
CONCLUSION
Our findings reveal that ganoderic acid A synergistically enhances tumor suppression of oxaliplatin possibly via increasing the cytotoxicity of T cells.
|
v2
|
2022-11-05T15:07:05.954Z
|
2022-11-03T00:00:00.000Z
|
253333316
|
s2orc/train
|
Tumor‐associated N1 and N2 neutrophils predict prognosis in patients with resected pancreatic ductal adenocarcinoma: A preliminary study
Dear Editor, Pancreatic ductal adenocarcinoma (PDAC) has been widely considered as one of the most lethal malignancies with a 5-year overall survival (OS) rate of only 11%.1 Lots of efforts have beendevoted into its early diagnosis, treatment strategies, and pathogenesis, etc. PDAC is characterized as excessive desmoplastic tumor with abundant immune cells infiltration, mainly including macrophages, T cells, immature myeloid cells, and neutrophils. Neutrophils are the most abundant white blood cells in blood and have been regarded as a homogeneous group in the past, which belongs to innate immune system and participates in defensing against pathogens. Recently, accumulating evidences determined that neutrophils had an important role in PDAC and accounted for a substantial proportion of tumor-infiltrating immune and inflammatory cells. Jiang et al.2 found that increased neutrophil infiltration was discovered as a central and prominent affected feature, which occurred in the liver, lung, and stomach at the PanIN stage. Importantly, serum leukotriene B4 (LTB4), derived from neutrophils, was validated for the early detection of PDAC. Considering its prognostic role, Ino et al.3 described relationships between prognosis and infiltrating immune cells in a cohort enrolled 212 PDAC patients who received radical surgical resection and found that tumor-infiltrating neutrophils were positively correlated with macrophages and regulatory T cells infiltrations and closely associated with shorter OS and disease-free survival (FS). This indicated that high tumor-infiltrating neutrophils always present boost immunosuppressive microenvironment and contribute to poor survival. Apart from tumor-infiltrating neutrophils, higher circulating neutrophils have also been found a negative association with outcomes. In our previous report, we noted that circulating neutrophil counts, whether 3 days within preoperation or 1 day after operation, were associated with shorter recurrence-FS (RFS) but
Tumor-associated N1 and N2 neutrophils predict prognosis in patients with resected pancreatic ductal adenocarcinoma: A preliminary study
Dear Editor, Pancreatic ductal adenocarcinoma (PDAC) has been widely considered as one of the most lethal malignancies with a 5-year overall survival (OS) rate of only 11%. 1 Lots of efforts have been devoted into its early diagnosis, treatment strategies, and pathogenesis, etc. PDAC is characterized as excessive desmoplastic tumor with abundant immune cells infiltration, mainly including macrophages, T cells, immature myeloid cells, and neutrophils. Neutrophils are the most abundant white blood cells in blood and have been regarded as a homogeneous group in the past, which belongs to innate immune system and participates in defensing against pathogens. Recently, accumulating evidences determined that neutrophils had an important role in PDAC and accounted for a substantial proportion of tumor-infiltrating immune and inflammatory cells. Jiang et al. 2 found that increased neutrophil infiltration was discovered as a central and prominent affected feature, which occurred in the liver, lung, and stomach at the PanIN stage. Importantly, serum leukotriene B4 (LTB4), derived from neutrophils, was validated for the early detection of PDAC. Considering its prognostic role, Ino et al. 3 described relationships between prognosis and infiltrating immune cells in a cohort enrolled 212 PDAC patients who received radical surgical resection and found that tumor-infiltrating neutrophils were positively correlated with macrophages and regulatory T cells infiltrations and closely associated with shorter OS and disease-free survival (FS). This indicated that high tumor-infiltrating neutrophils always present boost immunosuppressive microenvironment and contribute to poor survival. Apart from tumor-infiltrating neutrophils, higher circulating neutrophils have also been found a negative association with outcomes. In our previous report, we noted that circulating neutrophil counts, whether 3 days within preoperation or 1 day after operation, were associated with shorter recurrence-FS (RFS) but not with OS, which may hinted that neutrophils exerted a pro-tumor effect in PDAC. 4 Tumor-infiltrating neutrophils were often referred as tumor-associated neutrophils (TANs) in most studies. Wang et al. 5 identified a pro-tumor subcluster of neutrophils in PDAC and uncovered the pro-tumor mechanisms of TANs in PDAC microenvironment and revealed the association between high glycolytic activity and protumor functions in TANs. Similar to M1/M2 nomenclature of macrophages, recent studies also suggested that TANs could be divided into N1 and N2 categories based on their distinctive phenotypes: antitumorigenic N1 neutrophils and pro-tumorigenic N2 neutrophils. According to current in vivo/vitro studies, the N1/N2 neutrophil polarization may be depended on the special cytokine milieu, mainly including interferon-β (IFN-β) and TGF-β. However, little is known about the prognostic value of tumor-associated N1/N2 neutrophils in PDAC.
We preliminarily collected the PDAC tissues after radical surgery from January 2012 to December 2015 in our institute and stained for tumor-associated N1 and N2 neutrophils as refered. 6,7 The typical cell markers of tumorassociated N1 neutrophils were MPO + CD11b + CD206 − , and those of tumor-associated N2 neutrophils were MPO + CD11b + CD206 + by three-color immunofluorescence (IF) staining ( Figure 1A). Studies have reported that N1 neutrophils may possess powerful antitumor properties through antibody-dependent or direct cytotoxicity 8 , ROSmediated coupling 9 , etc. On the contrary, N2 neutrophils contributed to tumor angiogenesis by secreting vascular endothelial growth factor and matrix metallopeptidase 9, etc. and suppressing CTLs function by arginase. 10 Table S1.
The median number of tumor-associated N1 and N2 neutrophils were 10.7 (IQR 7.2-25.2) and 21.0 (IQR 13.7-29.0), respectively. The unpaired analysis indicated that the mean number of tumor-associated N2 neutrophils was 23.0 higher than that of tumor-associated N1 neutrophils (16.7, p = 0.002, Figure 1B). Additionally, the discrepancy between the tumor-associated N1 and N2 neutrophils were also observed in the paired analysis (p = 0.011, Figure 1C). According to the optimal cutoffs of tumor-associated N1 and N2 neutrophils, 22 and 18 cases were respectively classified into the high N1 and N2 neutrophil infiltration group. As shown in Table S1, the lower tumor-associated N1 neutrophil infiltration was significantly associated with easier lymph node metastasis (p = 0.012) and higher TNM stage (p = 0.024), but not with other factors. The higher tumor-associated N2 neutrophil infiltration was significantly correlated with distal location (p = 0.039) and easier lymph node metastasis (p = 0.039), similar to the previous study that more aggressive PDAC preferentially recruited more neutrophils.
Accumulative evidences showed that neutrophils had complex interaction with tumor microenvironment and played an important role in PDAC progression. The predictive values of neutrophil relevant biomarkers have been found in PDAC patients and regarded as potential therapeutic target. High density of total TANs infiltration always indicated poor outcome in PDAC, and current strategies of anti-TANs mainly focused on depleting TANs entirely or blocking chemokines that functioned as TANs recruitment. Series of preclinical studies used Ly6G or CXCR2 to block TANs and had encouraging outcomes in different mouse models. However, these therapeutic strategies are fraught with difficulties for further clinical trials because of the vital function in defense pathogens of neutrophils.
It is worth to consider that TANs have the opposite phenotype between N1 and N2. Our study successfully found that both tumor-associated N1 and N2 neutrophils displayed significant and opposite prognostic values within PDAC. The N1/N2 polarization may depend on different stimuli, including TGF-β induces N2 phenotype whereas IFN-β signaling polarizes neutrophils to N1 phenotype, and the causes of opposite effect in N1 and N2 neutrophil may include the secretions of different cytokines that orchestrate immune cells recruitment, different capability of arginase and proteases synthesis, and direct or indirect cytotoxicity, etc. The aforementioned results hinted that precisely eliminating N2 by specific markers or modified/polarized TANs into N1 would be a more efficacious strategy with less toxicity.
In summary, the exploration of neutrophil polarization and its correlation to clinical features in PDAC would significantly enhance our understanding of its pathophysiology and enable us to develop better treatment options.
C O N F L I C T O F I N T E R E S T
The authors declare no conflict of interest.
D ATA AVA I L A B I L I T Y S TAT E M E N T
Data included in this study are available upon request by contact with the corresponding author.
E T H I C S S TAT E M E N T
The study protocol was reviewed and approved by the Ethics Committee of Zhongshan Hospital, Fudan University.
|
v2
|
2022-11-05T15:10:51.405Z
|
2022-11-03T00:00:00.000Z
|
253310054
|
s2orc/train
|
SRPX2 Promotes Tumor Proliferation and Migration via the FAK Pathway in Papillary Thyroid Carcinoma
Thyroid cancer is the most common form of endocrine cancer around the world, and among which papillary thyroid carcinoma (PTC) is the most ubiquitous pathological sub-kind. Sushi repeat-containing protein X-linked 2 (SRPX2) was reported to be an independent prognostic factor and significantly overexpressed in advanced PTC patients. However, the biological functions of SRPX2 remain ambiguous in PTC. Here, we explored SRPX2 expression profiles and functions in PTC, finding that SRPX2 expression was remarkably upregulated in PTC tissues and cell lines. Further colony formation, CCK-8, as well as transwell assay, suggested that SRPX2 silencing remarkably dampened PTC growth and migration. Mouse xenograft models were established to find that SRPX2 silence remarkably suppressed PTC proliferation and migration in vivo. Following mechanism studies revealed that SRPX2 realized its functions in the PTC process partially through activating the Focal adhesion kinase (FAK) phosphorylation. In conclusion, this study investigated the functions and mechanisms of the SRPX2/FAK pathway in PTC progression. SRPX2 could act as a prospective biologic signature and therapeutic target molecule for PTC.
Introduction
Tyroid cancer is often encountered among the most commonly found cancers, and the incidence is located in 9th worldwide in 2020 [1]. Globally, thyroid cancer incidence is increasing yearly, especially in high-income regions, which becomes a growing threat to human health [2], and the increase was almost entirely due to papillary thyroid carcinoma (PTC), the most frequent sub-kind [3]. Te prognosis of PTC is generally good nowadays. However, cancer recurrence and metastasis may occur and result in a poor prognosis of advanced PTC. Terefore, it is important to assess the molecular parameters of PTC to better predict the clinical course of PTC and plan a treatment strategy.
Sushi repeat-containing protein X-linked 2 (SRPX2), a chondroitin sulfate proteoglycan, is highly expressed in a variety of cancers, including esophageal squamous carcinoma or gastric cancer [4,5]. It has been reported that SRPX2 promotes the progress of malignant in osteosarcoma by activating YAP1, which is closely associated with drug resistance, malignant phenotypes as well as expansion of cancer stem cells [6]. In advanced PTC, SRPX2 has been found signifcantly overexpressed and correlated with poorer disease-free survival. Moreover, SRPX2 has the potential to be an independent prognostic biomarker for advanced PTC [7]. However, the biological functions of SRPX2 remain ambiguous in PTC.
We explored SRPX2 expression profles and functions in PTC tissues and cells, fnding that SRPX2 expression was remarkably upregulated. Further experiments suggested that SRPX2 silencing remarkably dampened growth as well as migration in PTC. Following mechanism studies revealed that SRPX2 realized its functions in the PTC process partially through the activation of focal adhesion kinase (FAK) phosphorylation. Our work explored the functions and mechanisms of the SRPX2/FAK pathway in PTC progression. SRPX2 could act as a prospective biologic signature and therapeutic target molecule for PTC.
Tissue Specimens.
Tyroid tissues from cancerous (Tumor group) or neighboring noncancerous (normal group) were obtained in surgery from Te First People's Hospital of Chenzhou. All fresh tissues were immediately saved in TRIzol (Invitrogen, USA), followed by qRT-PCR assay. Te Medical Ethics Committee of Te First People's Hospital of Chenzhou approved this study, which was carried out in conformity with the Declaration of Helsinki. Before taking part in the research, all patients furnished their written informed permission forms.
Cell Culture.
Human PTC cells (KTC-1, BCPAP, K1, TPC-1, IHH-4, NPA87) and normal thyroid cells (Nthy-ori3-1) were purchased from the American Type Culture Collection (ATCC, USA). Cells were cultured as documented by the manufacturer at the condition of 37°C and 5% CO 2 . Short tandem repeat DNA profling was adopted to reauthenticate the cells prior to usage.
Colony Formation Assay.
After transfection with sh-SRPX2 or the control vector sh-NC, cells were transferred into plates (6-well plates, 10 3 each). 14 d later, methanol was used to fx the colonies, followed by crystal violet (0.1%) staining. Te colonies were counted under microscopy later.
Transwell Assay.
After transfection with sh-SRPX2 or the control vector sh-NC, cells were digested 24 hours after transfection and then seeded into upper chambers (10 5 cells/ well). Te upper cells were incubated without FBS, and 20% were in lower cells. 48 h later, 0.6% crystal violet was used to stain the upper chamber cells after cold methanol fxing; the stained cells were counted under microscopy later.
Mouse Xenograft Model. Ethical approval was obtained from the Institute Research Ethics Committee of Te First
People's Hospital of Chenzhou, and animal experiments were conducted following the institutional standard guidelines from Te First People's Hospital of Chenzhou. 2 × 10 6 cells/mL TPC-1-sh-SRPX2 or TPC-1-sh-NC cells were injected into C57BL/6 nude mice (dorsal fanks, fve mice/group). Ten tumors were extracted for weight and volume measurement 4 weeks later. 10 5 TPC-1-sh-SRPX2 or TPC-1-sh-NC cells were inoculated through the tail vein into nude mice (fve mice/group) to construct a lung metastasis model. After implantation for 8 weeks. Te lungs were obtained followed by a pathology assessment. And the number of macroscopically evident lung metastatic nodules was enumerated and verifed via microscopy of HE-stained sections.
Statistical
Analysis. GraphPad Prism 8.0 software was implemented for data analyses. T-tests, as well as one-way variance analysis, were employed in data analysis. All data is given in the form of mean ± SD. When P < 0.05, statistical signifcance is established.
SRPX2 is Elevated in PTC.
SRPX2 participates in the progress of multiple cancers. However, the functions and potential mechanisms of SRPX2 in PTC remain unknown. Studies revealed that SRPX2 is overexpressed in advanced PTC patients and correlated with poorer disease-free survival [7]. Here, qRT-PCR was adopted for the assessment of SRPX2 expression in PTC cell lines as well as tissues. As shown in Figure 1(a), in comparison to the nonmalignant thyroid cell line Nthy-ori3-1, SRPX2 was elevated in PTC cells, particularly in TPC-1 as well as KTC-1. We further verifed SRPX2 expression in 40 paired neighboring noncancerous thyroid tissues and PTC tissues. And the data illustrated that, compared to the surrounding nonmalignant thyroid tissues, SRPX2 was remarkably elevated in PTC tissues (Figure 1(b)).
SRPX2 Inhibition Diminished PTC Cell Proliferation as well as Invasion.
Since SRPX2 was upregulated in PTC, we designed si-RNAs to silence SRPX2 and explore its function. As shown in Figure 2(a), SRPX2 expression was remarkably reduced in cells transfected with si-SRPX2#1, which was built into shRNA. CCK-8 proliferation assays illustrated that SRPX2 silencing dampened the proliferation potential of PTC cells (Figure 2(b)). Besides, the colony-forming ability of PTC cells was evidently suppressed when SRPX2 was knocked down (Figures 2(c) and 2(d)). Moreover, transwell assays revealed that SRPX2 silencing expression diminished the invasion ability of PTC cells (Figures 2(e) and 2(f )).
SRPX2 Inhibition Suppressed PTC Proliferation and
Migration in Vivo. Next, we continued to investigate the in vivo functions of SRPX2. And results manifested that inhibition of SRPX2 expression signifcantly suppressed tumor growth (Figures 3(a) and 3(b)). Furthermore, the knockdown of SRPX2 expression signifcantly decreased PTC lung metastasis (Figures 3(c)-3(e)), indicating that SRPX2 silencing curbed PTC proliferation as well as migration in vivo.
SRPX2 Inhibition Led to Decreased Phosphorylation Levels of FAK.
Protein tyrosine kinase FAK is involved in multiple cancer processes. It has been reported that SRPX2 realizes its functions partly via the FAK-dependent pathway. Terefore, we continued to explore if SRPX2 regulated FAK activating process in PTC cells. Te results revealed that p-FAK was strikingly reduced when SRPX2 was silenced in PTC cells (Figures 4(a) and 4(b)). Next, to explore the regulation of SRPX2 in the activation of FAK in PTC in vivo, IHC staining was performed in the above mouse xenograft tumor tissues. p-FAK expression was noticeably decreased after the knockdown of SRPX2 (Figure 4(c)). Tese results indicated that SRPX2 might function partly in a FAK-dependent pathway in PTC progression.
Discussion
Tyroid cancer has been reported as the most pervasive endocrine cancer in the world, and the incidence keeps increasing for the past decades [8]. And PTC is a type of high-frequency pathological subkind [9]. Although improved systemic therapies, including radioiodine and targeted drugs, are available for PTC, advanced PTC prognosis is still poor [10]. Terefore, it is quite necessary to develop biomarkers and targets that can help better treat PTC. Anomalous SRPX2 exerts substantial functions in multiple cancers. SRPX2 is a component in an extracellular matrix which involved in tumor formation including colorectal cancer [11], gastrointestinal cancer [12], and prostate cancer [13]. Additionally, SRPX2 exerts essential roles in pan-cancers because of participating in stem cell diferentiation of human embryonic [14,15]. In PTC, SRPX2 has also been found signifcantly overexpressed and correlated with poorer disease-free survival of advanced PTC [7]. However, the biological functions of SRPX2 in PTC have not yet been reported.
Here, we confrmed SRPX2 expression in PTC cells as well as tissues, manifesting SRPX2 was boosted in PTC ( Figure 1). Subsequent functional experiments evinced that SRPX2 silencing dampened the proliferation as well as migration potential in PTC (Figures 2 and 3). All these fndings revealed the important role SRPX2 play in PTC. Hence, SRPX2 had the potential to be a treatment target and clinical signature for PTC. Next, we continued exploring the mechanism involving SRPX2 in PTC progression. FAK belongs to the protein tyrosine kinase families, participating in the movement, progress, self-renewal, and gene expression of cells [16]. Increasing evidence has shown that targeting FAK may be a promising therapeutic strategy for multiple cancers [17]. It has been reported that SRPX2 realizes its functions partly by FAK-dependent pathway. In pancreatic ductal adenocarcinoma, SRPX2 is boosted and contributes to malignant processes through the FAK-dependent pathway [18]. Overexpression of SRPX2 boosts tumor progress by FAK/ SRC/ERK pathway in lung cancer [19]. SRPX2 expedites hepatocellular carcinoma by targeting the FAK/AKT pathway and regulating the expression of MMP2/9 [20]. However, the relationship between SRPX2 and FAK in PTC has not yet been reported. Here, we revealed that SRPX2 silencing remarkably decreased p-FAK levels in PTC cells and mouse xenograft tumor tissues (Figure 4), indicating that SRPX2 might function partly in a FAK-dependent pathway to accelerate PTC progression.
Conclusions
SRPX2 expression was remarkably boosted in PTC. SRPX2 silencing suppressed PTC cell proliferation as well as migration partially via the reduction of FAK phosphorylation. Tough the detailed mechanism needs further clarifcation, this research disclosed the crucial biological functions of SRPX2 in PTC progress. SRPX2 could act as a prospective biologic signature as well as a therapeutic target molecule.
Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request. E-mail could be sent to the address [email protected].
|
v2
|
2022-11-06T14:31:03.972Z
|
2022-11-03T00:00:00.000Z
|
253351845
|
s2ag/train
|
Using evolutionary constraint to define novel candidate driver genes in medulloblastoma
Current knowledge of cancer genomics is biased against non-coding mutations. Here, we use whole genome sequencing data from pediatric brain tumors, combined with evolutionary constraint inferred from 240 mammals to identify genes enriched in non-coding constraint mutations (NCCMs). We compare medulloblastoma (MB, malignant) to pilocytic astrocytoma (PA, benign) and find drastically different NCCM frequencies between the two. In PA, a high NCCM frequency only affects the BRAF locus, while in MB, >500 genes have high levels of NCCMs. Intriguingly, many genes are associated with different age of onset, such as HOXB1 in young patients and NUAK1 in adult patients. Our analysis points to different molecular pathways in different patient groups. These novel candidate driver genes may assist patient stratification in MB and may be useful for treatment options. One-Sentence Summary Non-coding constraint mutations implicate novel candidate genes to stratify medulloblastoma by age and subgroups.
|
v2
|
2022-11-06T16:21:30.113Z
|
2022-11-03T00:00:00.000Z
|
253354205
|
s2ag/train
|
Review of commercially available biomarkers in the diagnosis of prostate cancer
Introduction
Diagnosing prostate cancer is a complex process. Although PSA testing remains the basic laboratory study, new biomarkers and test are evolving quickly.
Aim
The aim of this review was to summarize available tests and markers for diagnosing prostate cancer.
Materials and methods
Literature search was conducted using PubMed and Cohrane databases.
Results and conclusions
Detailed description of ExoDx, PCA3, SelectMDx, Mi-prostate Score, SChLAP1, PSA, PHI, 4K Score tests was presented. Available test ease qualification for a prostate biopsy or observation. Patients should be qualified individually in deciding on a specific test to be performed. Urologists should be aware of each test mechanism and limitations.
|
v2
|
2022-12-21T16:08:14.285Z
|
2022-11-03T00:00:00.000Z
|
254909111
|
s2ag/train
|
Revisiting the effectiveness of a forgotten technique: the value of blue dye on sentinel lymph node biopsies of gynecological cancer
Sentinel lymph node biopsy with blue dye has been used extensively in the past for the evaluation of patients with gynecological cancer. While the technique has been negatively criticized as inaccurate in detecting an appropriate number of lymph nodes bilaterally, novel studies seem to suggest an acceptable detection rate that is directly attributed to the number of cases performed as well as the type of dye that is used. Taking into consideration the significant impact of comprehensive pelvic lymphadenectomy on the quality of life of these patients, it seems reasonable to suggest its use in the absence of appropriate economic resources, such as in low income countries. In the present study we summarize the available evidence concerning the diagnostic performance of the various techniques that are used in sentinel lymph node resection in gynecologic oncology, focusing on the potential benefits of blue dye in specific settings.
|
v2
|
2022-12-21T16:22:30.828Z
|
2022-11-03T00:00:00.000Z
|
254921867
|
s2ag/train
|
Laparoscopic surgery for clear cell endometrial carcinoma, is it a realistic approach?
Clear cell endometrial carcinoma is an aggressive form of endometrial cancer that has a poor prognosis for survival. Over the past ten years, laparoscopic surgery has replaced open surgery as the preferred method of treating early-stage endometrial cancer. However, its effect on the prognosis of patients with clear cell carcinoma is still unclear. In the current systematic review, we examined the information that was published in the following databases: Medline, Scopus, Clinicaltrials.gov, EMBASE, Cochrane Central Register of Controlled Trials, CENTRAL, and Google Scholar. Three studies were found to be significantly diverse; as a result, meta-analysis was not possible. Current research reveals that compared to traditional laparotomy, minimally invasive surgery does not significantly influence progression-free and overall survival rates; nevertheless, given the relative dearth of data, more research is required before the technique’s safety can be determined.
|
v2
|
2022-11-04T15:55:20.183Z
|
2022-11-04T00:00:00.000Z
|
253269106
|
s2orc/train
|
A comparative analysis of telomere length maintenance circuits in fission and budding yeast
The natural ends of the linear eukaryotic chromosomes are protected by telomeres, which also play an important role in aging and cancer development. Telomere length varies between species, but it is strictly controlled in all organisms. The process of Telomere Length Maintenance (TLM) involves many pathways, protein complexes and interactions that were first discovered in budding and fission yeast model organisms (Saccharomyces cerevisiae, Schizosaccharomyces pombe). In particular, large-scale systematic genetic screens in budding yeast uncovered a network of ≈ 500 genes that, when mutated, cause telomeres to lengthen or to shorten. In contrast, the TLM network in fission yeast remains largely unknown and systematic data is still lacking. In this work we try to close this gap and develop a unified interpretable machine learning framework for TLM gene discovery and phenotype prediction in both species. We demonstrate the utility of our framework in pinpointing the pathways by which TLM homeostasis is maintained and predicting novel TLM genes in fission yeast. The results of this study could be used for better understanding of telomere biology and serve as a step towards the adaptation of computational methods based on telomeric data for human prognosis.
Introduction
In most eukaryotes, the chromosomal ends are protected by telomeres, composed of short G-rich repeats and a special set of proteins (Blackburn, 1991). Telomeres play a pivotal role in chromosomal duplication, stability, and dynamics (Zakian, 1995).
Telomeres shrink with replicative age due to the inability of DNA polymerases to synthesize lagging-strand DNA after the removal of RNA primers at the extreme ends of the chromosome (Hayflick, 1965;Harley et al., 1994). This condition, referred to as the "end replication problem", is solved by the ribonucleoprotein telomerase, which uses its RNA subunit to reverse transcribe telomeric DNA (Greider and Blackburn, 1989).
OPEN ACCESS EDITED BY
Telomerase is expressed in stem cells, but is barely detected in somatic cells. Continuously growing microorganisms, such as yeasts, constitutively express telomerase and are excellent models to investigate the mechanisms that regulate telomere biology and have uncovered a complex network of factors required to maintain telomere length homeostasis (Wellinger and Zakian, 2012;Harari and Kupiec, 2014;Kupiec, 2014). In budding yeast (S. cerevisiae), telomerase recruitment and activity is mediated by several factors that are required for the elongation of the shortest telomeres in some of the cell cycles (Teixeira et al., 2004).
Large-scale systematic genetic screens in S. cerevisiae, which scored collections of gene-knockouts and hypomorphic alleles, discovered a network of genes that participate in controlling telomere length (Askree et al., 2004;Gatbonton et al., 2006;Ungar et al., 2009;Puddu et al., 2019). These Telomere Length Maintenance (TLM) gene products have a variety of biochemical roles, some of which were not previously identified to be connected with the regulation of telomere size. Whereas mutations in some of these genes lead to shorter telomeres, others cause telomeres to elongate. Thus, each and every one of the ≈500 genes identified controls in a positive or negative way the length of telomeres.
So far, to the extent that we know, no attention has been paid to computationally modeling the TLM network in S. pombe. Thus, the aim of this study is to close this gap and to create a machine learning framework for examining telomere maintenance in fission and budding yeast. Our proposed framework is first validated in S. cerevisiae for detecting telomeric length phenotype. Next, we test it on curated S. pombe TLM data, and pinpoint the most important pathways and protein complexes that make up TLM homeostasis. We follow by investigating the TLM phenotype within the yeast orthologs and suggest the most likely S. pombe genes that are currently unknown to be members of the TLM network in this organism. Last, we perform gene ontology (GO) enrichment analysis for these candidates and reveal that they are significantly enriched for biological processes known to be highly linked with telomere maintenance functions.
Telomere length data
The data for the S. cerevisiae TLM genes along with their corresponding telomere length category was obtained from Van Leeuwen et al. (2016) and Puddu et al. (2019). For the binary classification of telomere length, the categories were reduced to the following phenotypes: 'short' and 'long'. YIR016W was observed as both normal and 'very long', hence we assigned it as 'long'. The data also included the Decreased Abundance by mRNA Perturbation (DAmP) collection and we treated the two telomere length labels of 'DAmP Short' and 'DAmP Long' as 'short' and 'long', respectively. The S. pombe TLM genes were curated from the Fission Yeast Phenotype Ontology (FYPO) v2011-01-18 (Harris et al., 2013). The following FYPO terms were labeled as having 'short' ('FYPO: 0002239′, 'FYPO:0006511′, 'FYPO:0003106′, 'FYPO:0003107′) and 'long' ('FYPO:0002019′) telomere length phenotypes. Finally, we merged these genes with a list of genes that were found to regulate the homeostasis of telomeres (Liu et al., 2010).
Overall, we obtained 483 and 224 unique TLM genes with corresponding binary telomere length phenotypes for S. cerevisiae and S. pombe, respectively (Supplementary Table S1).
Genetic interaction data
For each S. cerevisiae TLM gene, raw Genetic Interaction (GI) scores were taken from the pairwise interaction format of the TheCellMap.org repository (Usaj et al., 2017). They were produced by systematic Synthetic Genetic Array (SGA) experiments and scored by comparing the fitness of the double mutant to the corresponding single mutants (Costanzo et al., 2016). We only considered the authors' lenient threshold, i.e., GIs with p − value < 0.05. In the case of multiple measurements per interaction, the GI score with the lowest p − value was used. The outcome was a score matrix of the TLM genes and 5850 genes sharing at least one GI with one of the TLM genes.
The S. pombe GI data was downloaded from BioGRID v4.4.207 (Oughtred et al., 2021). Unlike the S. cerevisiae data, the interactions only contain a verbal description and not a numerical score, which we mapped into a score of 1 when there was an interaction and 0 otherwise. We used all the data that are marked 'genetic' in their 'Experimental System Type' field.
Pathway data
To construct the pathway features, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database release 99.0 (Ogata et al., 1999) was parsed via the BioServices v1.10.1 (Cokelaer et al., 2013) API. Only non-global and nonoverview maps were considered (i.e., utilizing pathways that only contain genes and not ones that contain other pathways).
Protein complex data
For S. cerevisiae, the CYC2008 catalog was used. It contains 408 manually curated heteromeric protein complexes that were confirmed by small-scale experiments from the literature (Pu et al., 2009).
Frontiers in Genetics frontiersin.org 02 S. pombe complex information was downloaded from PomBase (Harris et al., 2022). It is based on the GO database, for terms that are classified under "macromolecular complex" (GO:0032991) and have fission yeast genes annotated, such that the most specific complex is retained.
Orthology data
A manually curated ortholog list of fission to budding yeast was retrieved from PomBase (Wood et al., 2012). For cases where there was more than one ortholog per gene, the gene with the maximum score from the Smith-Waterman alignment algorithm (Smith and Waterman, 1981) was selected. This was achieved using the Biopython v1.79 (Cock et al., 2009) pairwise2.align.localds function with the same parameters as in the web BLAST NCBI interface (http://blast.ncbi.nlm.nih.gov), i.e., BLOSUM62 scoring matrix, a gap cost of 11 and an extension cost of 1. The final set contained 3953 orthologous pairs (Supplementary Table S2).
GO data
In order to process the GO consortium database (Ashburner et al., 2000;Consortium, 2021) the Python package GOATOOLS v1.2.3 (Klopfenstein et al., 2018) was used. For the feature engineering, GO terms from Biological Process (BP) and Cellular Component (CC) categories were filtered to include only genes for which we have prior data (i.e., genes that appear in the S. cerevisiae GI dataset). Broad terms that contain more than 30 genes were excluded from further analysis, as well as terms with less than 3 genes.
GO enrichment analysis
For the GO enrichment analysis, we used the PANTHER web API (http://pantherdb.org/services/openAPISpec.jsp) with annotation files from March 2022 (GO Ontology database DOI:10.5281/zenodo.6399963 released on 2022-03-22). We employed the 'Enrichment (Overrepresentation)' test which computes a p − value using Fisher's exact test, and the False Discovery Rate (FDR) method was used for multiple hypothesis correction. We limit the enrichment testing to only include biological process terms. In order to avoid broad terms, we restricted the analysis to terms that contain at most 250 genes. The test cutoff was set to an FDR q − value < 0.05.
Feature generation
We designed four sets of features that span a variety of molecular functions including genetic interactions, pathway maps, biological processes, and protein complexes. We refer to a feature based on the main dataset it was derived from, namely, KEGG, GO BP, CYC2008, and GO CC. A summary of all the feature sets we evaluated for telomere length classification is presented in Table 1.
In order to extract the KEGG and GO BP features, we considered for each gene its proportion from the group that has a non-zero GI score with it. For KEGG and GO BP, respectively, this group consists of the genes that make up a pathway and GO term direct gene members, more details are provided in the Supplementary Material. The CYC2008 and GO CC features indicate for each protein complex a gene's membership in it. When combining a pair of feature sets, it was done by merging on the intersection of genes that are in both sets.
For classifying TLM genes and predicting S. pombe candidates, we defined another feature, namely, propagation to anchor genes. These are genes that act as an endpoint for TLM-related processes as described for S. cerevisiae (Shachar et al., 2008) and that we adapted for S. pombe (Supplementary Table S3). Producing this feature employs a random walk with a restart propagation process as described in (Cowen et al., 2017). Specifically, the following steps were taken: 1) We used the whole genome with Protein-Protein Interaction (PPI) binary scores from BioGRID v4.4.207 as the adjacency matrix (interactions labeled 'physical' in their 'Experimental System Type' field). We normalized this matrix by W = AD −1 , where A is the adjacency matrix and D is the diagonal degree matrix.
2) The starting vector p 0 was set to 1 |G| for each anchor gene and zero for all other genes, where |G| is the number of anchor genes.
3) The resulting feature vector was produced by a computation until convergence of the vector p k = 0.2p 0 + 0.8Wp k−1
Machine learning models
To evaluate classification performance, we used five standard machine learning models from the Python package SciKit-Learn (version 1.0.2), that have been proposed for obtaining good prediction accuracy in the bioinformatics domain (Olson et al., 2018). We retained the recommended hyperparameters that were set prior to all experimentations as summarized in Table 2.
This list of models was tweaked to include the Extreme Gradient Boosting Classifier (XGB) from the XGBoost package (v0.9) instead of the Gradient Boosting Classifier and Frontiers in Genetics frontiersin.org 03 the Logistic Regression was replaced by the Logistic Regression Cross Validation (LRCV) model. In addition, the Linear Support Vector Classifier (LSVC) model was added to have more than one linear model assessed, resulting in six models overall.
Evaluation setting
Our dataset is imbalanced among the telomere length classes. This leads to models that are overly conservative when predicting the minority classes. To address this issue we performed 5 repeated stratified 10-fold Cross-Validation (CV) experiments. This procedure is followed for each held-out test in a 10-fold CV and the whole process is repeated 5 times, producing different splits and held-out test sets in each repetition, but maintaining the percentage of samples for each class. For all the setups, the models were assessed in each run for the relevant evaluation metrics on the held-out test dataset and the median score (unless stated otherwise) across all experiments is reported.
The evaluation metrics included Matthew's Correlation Coefficient (MCC) and Area Under the receiver-operating characteristic Curve (AUC) as they are more robust to imbalanced label distribution (He and Garcia, 2009;Chicco and Jurman, 2020). The MCC is calculated as follows: where TP represents the True Positive; TN, the True Negative; FP, the False Positive; FN, the False Negative. It is the Pearson correlation coefficient between the predicted and true labels. The Receiver-Operating Characteristic curve depicts the true positive rate as a function of the false positive rate. The AUC is a metric for assessing a classifier's overall performance. The better the classifier is, the closer to one the AUC is.
In order to keep the number of features not greater than the number of samples (Ressom et al., 2008), feature sets that exceeded this threshold were reduced to match the size of the samples in all of the experiments. To this end, a Bernoulli Naive Bayes Classifier was applied during the training phase to determine the importance of each feature based on the observed log probability of features given a class. The top ranking features, up to the number of samples, were selected to be used for testing via SciKit-Learn's SelectFromModel.
Prediction of telomere length changes following gene knockout
We compiled a comprehensive collection of feature sets for telomere length prediction following gene knockout. First, we focused on the task of predicting shorter-than-normal vs. longerthan-normal length. We assessed each one of the feature sets individually after applying standardization scaling using the StandardScaler functionality from SciKit-Learn.
A heatmap of the metrics' median results across all test runs for each Machine Learning (ML) model and feature sets is presented ( Figure 1A). The results are ordered by the overall median score of a model, across all features. The Random Forest Classifier (RF) achieved the greatest performance across all features in the MCC metric with an overall median score of 0.3 and LRCV had the highest overall median AUC score of 0.69.
A comparison of the type of features used reveals that protein complex-based features (CYC2008 and GO CC) significantly outperform other features across all classification metrics and ML models in this context ( Figure 1B). Comparing the two in each metric shows that CYC2008 has a significantly higher MCC score (0.49 vs. 0.39, p < 0.0001), but achieves similar AUC scores (0.75 vs. 0.74). Therefore, ML models using a single set of protein complex-based features, and in particular, CYC2008, will lead to a better classification of telomere length samples than pathway features.
Next, we moved to examine a mixture of feature sets. We limited the search to pairwise feature space combinations, selecting the top features in the same manner as described above. The results are ordered by the median score per feature combination, across all models (Figure 2A). In both AUC and MCC measures, the highest-performing features across all models are made of the combination of protein complex features (CYC2008) and pathway features (KEGG). The highest overall performing model and feature set were the LRCV with CYC2008 and KEGG features, with a median AUC score of 0.834 and median MCC score of 0.513 across all experiments ( Figure 2B). In addition, the top two performing models in both metrics used the CYC2008 and KEGG feature combination, both exceeding the median AUC > 0.8 and MCC Frontiers in Genetics frontiersin.org 06 > 0.5 scores, attesting to the utility of these features. Therefore, our subsequent analysis utilized only CYC2008 and KEGG feature sets with the LRCV model pipeline.
Finally, we performed a meta-analysis of the predictions. Instead of the binary categorization to 'short' and 'long' phenotypes, we checked to see if the classifier's estimated probabilities for those predictions were higher in more extreme phenotypes. To this end, we looked at the original phenotypes, namely, 'very short', 'short', 'slightly short', 'slightly long', 'long', and 'very long'. They were originally (Askree et al., 2004) deemed as such by comparing them to a baseline of wild-type telomere length, measuring in bulk Southern blot. For example, strains of 385-420 telomeric nucleotides were considered 'long' when compared to the 350±35 nucleotides of the wild-type, whereas those with longer telomeres were designated 'very long'. We compared the estimated probabilities that were produced by the model for each such subtype ( Figure 2C). When focusing on the larger fraction (about 67%) of phenotypes that were reduced to 'short', we observed that severe phenotypes were assigned with higher confidence (median probability of 'very short' 0.88 vs. 'short' 0.84, p < 0.0005, and 'short' vs. 'slightly short' 0.77, p < 0.0005). This was not the case for the 'long' phenotypic length (no significant differences were found between 'slightly long', 'long', and 'very long'). One plausible explanation is the relatively small number of samples that were available during training and testing (for example, less than 1% is labeled 'very long').
Application to Schizosaccharomyces pombe
After establishing our classifier's performance, we wished to generalize it to data from S. pombe, a fungal species whose ancestors separated from S. cerevisiae ≈400 million years ago (thus these species are different from each other as either is from animals). To this end, we first mapped cerevisiae features to pombe features. KEGG pathways contain a consistent naming convention, allowing for these features to be mapped seamlessly. For the complex-based features, we retained the ones in CYC2008 that have a unique GO identifier and intersected that set with the S. pombe corresponding set (with the same GO ID). Overall, we could reproduce 127 features for 198 samples ('long' -158, 'short' -40). As before, we trained a 'short'/'long' classifier and evaluated its performance in crossvalidation. The median AUC score for classifying short and long telomere length was 0.81, mean AUC result of 0.79 with a standard deviation of 0.16 ( Figure 3A) and a median MCC of 0.49.
Next, we turned to analyze the feature importance that led to these results ( Figure 3B). To accomplish this, we took the fifteen coefficients with the highest mean absolute value that the LRCV model learned across all runs. Using the held-out set in each run, allowed us to detect which characteristics contribute the most to the examined model's generalization capability by applying the SciKit-Learn's permutation_importance with n_repeats = 5. The difference between the baseline scoring metric that is used by Frontiers in Genetics frontiersin.org 07 LRCV (the default accuracy measure, in our case) and the one from permuting the feature column is defined as the permutation importance. To summarize, after running all the experiments, we were left with the top-15 highest mean absolute coefficients that participated in all the runs and their respective mean permutation importance. Having our data scaled in the preprocessing step, allowed us to then look at the odds ratio for each feature in conjunction with its permutation importance score with regards to rest. This way, we could assess the highest influencing features in our system as-is, disregarding interacting terms. Among the features with a relative high permutation score, we find the telomerase, Mre11-Rad50-Xrs2 (MRX) and the HIstone Regulator (HIR) complexes and DNA damage response pathways, aligning with the findings of the S. cerevisiae TLM mechanisms (Askree et al., 2004;Rubinstein et al., 2014). Despite the difference between the data sets of the two organisms where in cerevisiae the majority group is the 'short' one, while in pombe the opposite is true, the performance generalized well to the pombe setting.
Prediction of S. pombe TLM genes from orthologs
Orthologs are the result of speciation events and are likely to be functionally related. In our context, previous research has demonstrated that gene dispensability is conserved for the majority of ortholog genes in budding and fission yeast (Kim et al., 2010). Based on this result, we set to assess if the same holds for TLM genes. According to data we have, out of 3953 orthologs, only 51 genes (1.29%) are TLM genes in both species. However, when focusing on the subset of TLM genes in either species, 9.94% of the genes (51/513) are TLM genes in both yeasts. A closer look into the telomere length phenotype within this subset of shared TLM orthologs demonstrates more robust conservation than the one that has been detailed so far. 60.78% (31/51) of shared TLM genes preserve the phenotype (p < 0.1). Overall, 29.4% (15/51) and 31.37% (16/51) of these orthologous pairs retain the TLM 'short' and 'long' phenotypes, respectively.
We postulate that there could be more S. pombe TLM genes within the orthologous pairs that are currently unknown to have this role. In order to predict TLM candidates, our prediction system was evaluated against the task of classifying between TLM and non-TLM genes. To this end, the LRCV model remained the same, apart from the additional setting of its class_weight parameter to 'balanced'. The KEGG and CYC2008 feature sets were built in the same manner, but this time with respect to the entire ortholog gene set. In total, we explored five methods for TLM prediction (as also summarized in Table 3 The performance of the five predictors is summarized in Table 4. Method (5) performed best; out of the other four methods, method (3)applying our framework as-is to this new taskdominated the rest in terms of AUC.
Next, we aimed to predict new TLM candidates in S. pombe using method (5). To this end, we executed the model in a leave-oneout setting so that we could use the entire dataset for training and make a prediction with respect to each gene in turn. The resulting top-30 predictions that are not known to be TLM genes, along with their S. cerevisiae orthologs, were subjected to GO enrichment analysis (Supplementary Tables S4-S6). The top-10 is presented in their ranked order (Table 5). Homing in on some of the genes shows that they are associated with the Target of Rapamycin (TOR) (Ungar et al., 2011;Rallis et al., 2017;Lie et al., 2018) a known participant in the regulation of subtelomeric and telomeric regions (Schonbrun et al., 2009;Cohen et al., 2018). Further inspection reveals that 10% of the S. pombe predicted genes are orthologous to known TLM genes in S. cerevisiae. This is reassuring as it is consistent with the prior experimental knowledge that was discussed above. Overall, our candidate TLM genes were significantly enriched (p < 0.001) for core DNA maintenance processes (DNA damage response, DNA repair, and DNA replication), DNA assembly or remodeling functions (chromatin organization, chromosome segregation), and mitotic and meiotic cell cycles (regulation of cell cycle process, regulation of mitotic cell cycle and meiotic cell cycle). Taken together, these enriched terms suggest that there is an association between our candidate genes and telomere maintenance homeostasis.
Conclusion
This study set out to create a general machine learning pipeline for telomere length maintenance analysis in fission and budding yeast. We have identified a set of features along with a simple linear model that can predict the telomere length phenotype under various settings. The framework can also pinpoint explanatory variables leading to its output while utilizing a broad range of data sources, including genetic interaction data, that is being used for the first time in this context, to the best of our knowledge.
The generalizability of these results is subject to certain limitations. For instance, the datasets used to derive the features are incomplete and so is our interpretation of the predictions. Furthermore, the small and imbalanced data, in some of the tasks we investigated, makes it hard to learn the underlying structure of the data.
Although this study focuses on yeast datasets, the suggested system may well have a bearing on human data, such as the UK Biobank (Bycroft et al., 2018) as telomere length is a promising biomarker for age-associated diseases and cancer. Considerably more work will need to be done in order to have high quality data of the S. pombe TLM network, and our predictions are a promising starting point for this investigation.
Data availability statement
The genetic interaction data used for S. cerevisiae contains no version control and was downloaded from https://thecellmap. protein kinase Gsk3 (*) indicates a known budding yeast TLM gene. Product description data was taken from PomBase (Harris et al., 2022).
Frontiers in Genetics frontiersin.org 09 org/costanzo2016/ on March 2022. For the rest of this work, publicly accessible datasets were examined, and the publication makes note of the relevant versions. All of the data and code used for this study are available at: https://github.com/Iftahp/ yeastTLM.
|
v2
|
2022-11-04T16:06:22.708Z
|
2022-11-04T00:00:00.000Z
|
253269063
|
s2ag/train
|
PD0166285 sensitizes esophageal squamous cell carcinoma to radiotherapy by dual inhibition of WEE1 and PKMYT1
Background Esophageal squamous cell carcinoma (ESCC) is an aggressive tumor with a 5-year survival rate of only 20%. More than 80% of ESCC patients possess TP53 mutation, which abolishes the G1/S checkpoint and accelerates the cell cycle. Thus, WEE1 and PKMYT1, regulators of G2/M phase in cell cycle, play essential roles in TP53-mutated cancer cells. PD0166285(PD) is a pyridopyrimidine compound that can inhibit WEE1 and PKMYT1 simultaneously, however, the effects of PD on ESCC, either as monotherapy or in combination therapy with radiotherapy, remain unclear. Methods To measure the anti-tumor efficacy of PD in ESCC cells, cell viability, cell cycle and cell apoptosis assays were examined in KYSE150 and TE1 cells with PD treatment. The combination therapy of PD and irradiation was also performed in ESCC cells to find whether PD can sensitize ESCC cells to irradiation. Vivo assays were also performed to investigate the efficacy of PD. Results We found that the IC50 values of PD among ESCC cells ranged from 234 to 694 nM, PD can regulate cell cycle and induce cell apoptosis in ESCC cells in a dose-dependent manner. When combined with irradiation, PD sensitized ESCC cells to irradiation by abolishing G2/M phase arrest, inducing a high ratio of mitosis catastrophe, eventually leading to cell death. We also demonstrated that PD can attenuate DNA damage repair by inhibiting Rad51, further research also found the interaction of WEE1 and Rad51. In vivo assays, PD inhibited the tumor growth in mice, combination therapy showed better therapeutic efficacy. Conclusion PD0166285 can exert antitumor effect by inhibiting the function of WEE1 and PKMYT1 in ESCC cells, and also sensitize ESCC cells to irradiation not only by abolishing G2/M arrest but also attenuating DNA repair directly. We believe PD0166285 can be a potent treatment option for ESCC in the future.
|
v2
|
2022-11-04T16:07:16.846Z
|
2022-11-04T00:00:00.000Z
|
253269190
|
s2orc/train
|
The benefits of propofol on cancer treatment: Decipher its modulation code to immunocytes
Anesthetics are essential for cancer surgery, but accumulated research have proven that some anesthetics promote the occurrence of certain cancers, leading to adverse effects in the lives of patients. Although anesthetic technology is mature, there is no golden drug selection standard for surgical cancer treatment. To afford the responsibility of human health, a more specific regimen for cancer resection is indeed necessary. Immunosuppression in oncologic surgery has an adverse influence on the outcomes of patients. The choice of anesthetic strategies influences perioperative immunity. Among anesthetics, propofol has shown positive effects on immunity. Apart from that, propofol’s anticancer effect has been generally reported, which makes it more significant in oncologic surgery. However, the immunoregulative function of propofol is not reorganized well. Herein, we have summarized the impact of propofol on different immunocytes, proposed its potential mechanism for the positive effect on cancer immunity, and offered a conceivable hypothesis on its regulation to postoperative inflammation. We conclude that the priority of propofol is high in oncologic surgery and propofol may be a promising immunomodulatory drug for tumor therapy.
Introduction
Tumor is a significant disease jeopardizing human health. According to the cancer statistics in 2022 (Siegel et al., 2022), in the United States, it has been estimated that 609,360 people will die from cancer and approximately 1,918,030 new cancer cases will be diagnosed, leading to a huge burden to families and society. For most solid cancers, surgical resection is the preferred treatment method. However, due to inevitable operative tissue lesion and the use of anesthetics, surgery is commonly accompanied by the decay of the immune system (Ogawa et al., 2000;Kim, 2017;Kim, 2018). Also, there are probably many adverse effects to bear after surgery due to the release of numerous cancer cells or This article was submitted to Pharmacology of Anti-Cancer Drugs, a section of the journal Frontiers in Pharmacology the suppression of the activity of anticancer lymphocytes. All of these possibly raise the recurrence risk of tumors or cause poor prognosis for patients. Research has indicated that there are intimate connections between anesthetics and cancer occurrence, development, recurrence, and surgical prognosis (Jiao et al., 2018;Moradkhani and Karimi, 2018). The selection of anesthetic strategies can be directly related to the recovery of the immune system. Propofol, first used as an anesthetic, has shown its new pharmacological functions in cancer treatment, promising to be an effective drug in the therapy of cancer.
The underlying mechanisms of propofol's anticancer effect have been expounded by some review articles (Jiang et al., 2018;Gao et al., 2020;Xu et al., 2020). However, despite its direct suppression of tumor cells, propofol also alters the immune system (Sanders et al., 2011;Kim, 2018), which is vital to the prognosis of cancer patients, but less elucidated. In this review, we analyzed its immunomodulation function in cancer treatment based on immunocytes-related experimental research and its preponderance in cancer resection based on the interpretation of the clinical research, revealing that propofol-mediated COX inhibition in macrophages contributed to its antitumor immunity and propofol-profited immunity recovery after surgery. We also propose a model of sterile inflammation after cancer resection, demonstrating the beneficial effect of propofol on wound healing. Also, we conclude that propofol is approximated to be a mild immunomodulatory drug that will favor tumor therapy.
Propofol: An anesthetic
Surgical resection is the primal strategy for tumor treatment, and the development of anesthetics provides the wings for surgery to attain its present achievement. Propofol is one of the most applied anesthetics in clinical practice. Early in 1987, propofol was first approved in surgery as an anesthetic inducer in the United Kingdom. Shortly afterward, because of its excellent sedative and anesthetic effect, propofol rapidly became a generally used intravenous anesthetic agent and was applied in cancer resection surgery.
Brief history of anesthetics
Anesthesia is the cornerstone of modern surgery. By inducing loss of consciousness, anesthetics can help patients relieve their physical and mental sufferings. The earliest anesthetics came from the extract of natural products, and the analgesic effect of this kind of crude extract was inadequate, and it was easy to cause drug poisoning.
In 1846, Dr Morton carried out the first public representation of ether for surgery, opening the history of modern anesthesiology. At the same time, it also began the dominance of ether as an anesthetic for about 100 years. Moreover, other volatiles were proposed to have analgesic effects, like nitrous oxide and chloroform.
With the consistent discovery of natural anesthetics, synthetic anesthetics have also been studied. In 1905, the local anesthetic procaine came out, which solved the shortcomings of cocaine as a local anesthetic. After that, lidocaine was also successfully invented. In the synthesis of general anesthetics, sodium thiopental, used in clinical trials in 1934, has been undoubtedly one of the most dominant general anesthetics in the 20th century. Also, halothane, which rapidly replaced ether and chloroform (Gyorfi and Kim, 2020) in 1956, could not shake its dominant position. As an inhalational anesthetic, halothane was replaced by isoflurane and sevoflurane in the 1990s.
The side effects of thiopental were gradually exposed after it being used for more than 50 years. Therefore, it was necessary to seek more safe and effective intravenous anesthetics, and propofol discovered by John Baird Glen emerged as the times require and has become a standard inducer for surgical anesthesia (Walsh, 2018). Propofol has a unique anesthetic balance and has minimal impact on respiration and heart rate. In the mouse study, It can be taken repeatedly in mice, and there is no cumulative "hangover" effect, indicating that the metabolism is fast. The pharmacokinetic and metabolic studies in humans showed that >99% of propofol was metabolized in the liver. The liver P450 enzyme can oxidize propofol to generate water-soluble, excretable metabolites which can be excreted by the kidneys. The rest is oxidized by liver P450 enzyme to generate water-soluble, excretable metabolites.
Mechanism of propofol's anesthetic effect
The molecular mechanism of anesthesia is not completely clarified. The Meyer-Overton rule was the most prevalent mechanism to explain anesthesia in the 20th century (Bovill, 2000). Meyer and Overton proposed that the potency of an anesthetic was proportional to its lipid solubility, and it affected nerve function by disturbing the lipid bilayer (Perouansky, 2015). However, with the revelation of the protein complexes, the classical anesthesia theory gradually gave way to the ion channel theory. According to this, general anesthetics regulated the ion flow of nerve cells by affecting the switch of specific ion channels on the cell membrane, such as the transmitter-gated ion channels (TGICs), to achieve anesthesia (Weir et al., 2017). The TGIC can be roughly divided into two types. The inhibitory type, represented by GABAAR, is composed of five transmembrane subunits. After binding with the receptor, the ligand can cause channel opening, anion influx, and neuronal hyperpolarization. The excitatory receptor, like Nmethyl-D-aspartate (NMDA), exhibited the opposite effect. Therefore, general anesthetics can enhance the inhibitory receptor or weaken the excitatory one to induce the anesthetic effect.
Propofol is a GABAAR agonist
GABA (γ-aminobutyric acid) is the most important inhibitory neurotransmitter in the central nervous system. There are two types of GABA receptors (GABARs) (Brohan and Goudra, 2017): GABAAR and GABABR. GABABR is a slow response receptor of GABA and can inhibit the release of presynaptic neurotransmitters (Manz et al., 2019) and promote the production of inhibitory postsynaptic potential on the postsynaptic membrane by activating downstream pathways. GABAAR is a kind of ligand-gated ion channel, which responds quickly to GABA and is the main target of propofol-induced sedative effect (Brohan and Goudra, 2017). GABAAR is a chloride ion channel formed by 2 α subunits, 2 β subunits, and 1 γ subunit. Propofol binds to its extracellular part to activate the channel directly, increasing the permeability of the cell membrane to chlorine, causing chlorine influx, leading to the hyperpolarization of the cell membrane, and finally inhibiting the excitability of neurons.
Other mechanisms for propofol to induce anesthesia
In addition to the activation of GABAAR, studies have shown that propofol can affect the opening of voltage-gated ion channels. Ratnakumari and Hemmings (1997) pointed out that propofol could induce the influx of sodium and release of glutamate by regulating the sodium channel. Protein kinase C was indicated to be involved in anesthesia of the CNS. The study by Hemmings et al. (1995)showed that propofol might induce anesthetic effect by regulating PKC phosphorylation.
Propofol: A direct cancer cell-inhibiting reagent
It has been proven that propofol displays an intimate association with the biological behavior of cancer. In some cancers, such as gallbladder cancer and breast cancer, propofol processed cancer-promoting effects, but most studies have shown its cancer-inhibiting function from not only affecting epigenetic pathways, such as those involving miRNA, lncRNA, and histone acetylation, but also modulating signaling pathways, such as the hypoxia, NF-κB, MAPK, SLUG, and Nrf2 pathways. The cancer types that could be inhibited by propofol include colon cancer (Miao et al., 2010), breast cancer (Li et al., 2012), cervical cancer (Zhang et al., 2015), glioma , non-small-cell lung cancer (Xing et al., 2018), cholangiocarcinoma , Leydig cell cancer (Kang et al., 2019), colorectal cancer (Chen et al., 2018), thyroid cancer (Chen et al., 2019a), leukemia stem cell (Chen et al., 2020), gastric cancer (Yang et al., 2017), oral squamous cell carcinoma , endometrial cancer (Du et al., 2018), cardia cancer (Su et al., 2020), and so on. It can not only inhibit cancer angiogenesis, invasion, and metastasis but also reduce cancer proliferation and induce cancer cell death such as apoptosis, which has been reported by numerous articles. So, in this passage, we discuss propofol's anticancer effect from a new direction, its immunity modulation effect, which has not been well organized by published articles.
Effect of propofol on immunocytes
Since many years, it has been proven that propofol has immunomodulation function which can regulate the resistance of the human body to cancer. Many review articles have mentioned its underlying mechanism incidentally. For instance, the effect of volatile anesthetics (sevoflurane, isoflurane, nitrous oxide, and halothane), intravenous anesthetics (propofol, ketamine, thiopental, and midazolam), and perioperative auxiliary drugs (morphine, fentanyl, sufentanil, remifentanil, alfentanil, and lidocaine) on immune function has been reviewed by Kim (2018). But due to the limited length of the article, the effect of propofol on immunity has not been well organized, and many valuable articles have been omitted. Herein, we focus on propofol's immunomodulation function and concentrate on its effect on immunocytes, uncovering its code in cancer treatment.
Native immunocytes are crucial for the process of innate immune response to prohibit infection, clean up tumor cells and damaged tissue, and realize homeostasis in a nonspecific manner. An adaptive immune response can powerfully sweep away unique targets. The immunomodulation of anesthetics consists of both a direct path and an indirect path. In addition to the effect on the neuroendocrine system, they can also regulate the function of immunocytes directly. As a generally used anesthetic, propofol exhibits diverse influences on multiple immune cells ( Figure 1 and Table 1). In this section, the detailed depiction is as follows.
Monocytes and macrophages
Leukocytes can be mainly divided into three kinds: monocytes/macrophages, granulocytes, and lymphocytes. The monocytes in the circulation come from the bone marrow and are collected into specific tissues with the effect of chemokines through the vascular endothelial cell, transforming into macrophages. Although monocytes exhibit phagocytosis and immune regulation function, macrophages are the more efficient formation. Macrophages have two kinds of polarization forms: classically activated macrophages (M1 type) and alternatively activated macrophages (M2 type).
The M1-type macrophages, which are induced by LPS, IFN-γ, or TNF-α, mainly secrete pro-inflammatory factors with the consequence of tissue damage, while the M2-type macrophages activated by IL-4, IL-10, IL-13, or TGF-β produce anti-inflammatory factors, participating in wound healing.
Monocytes
It has been reported that prostaglandin E2(PGE2) could inhibit immune function by suppressing the secretion of IL-12 or IFN-γ. While, propofol could repress, both in vitro and in vivo Inada et al., 2009a), the secretion of PGE2 from the monocytes by the inhibition of COX-2 activity, which might retard the immunosuppression effect in the postoperative phase. Besides, by directly suppressing the secretion of IL-10 ( Kambara et al., 2009), an immunosuppressive cytokine, and increasing TNF and IL-1 (Rossano et al., 1992), which could increase the capability of normal T cells and activate multiple immunocytes, propofol could also promote immune function.
A study that was carried out by Krumholz et al. (1999) tested the effect of many intravenous anesthetics on the chemotaxis of human monocytes. It concluded that unlike ketamine, midazolam, and droperidol, propofol did not impair monocyte chemotaxis, which was another beneficial factor for immunity.
Propofol participates in monocytes' regulation of postoperative coagulation reaction. Thrombomodulin (TM) and thromboxane A2 (TXA2) are two pro-coagulants. Also, Lin et al. (2011) have reported that propofol could raise the level of TM by the inactivation of NADPH oxidase and tristetraprolin (TTP) and the activation of HuR. However, another research that investigated the effect of propofol on thromboxane B2 (TXB2) production led to some inconsistencies. They found that TXB2 production (reflecting the content of TXA2) from the monocytes was decreased by suppressing the activity of cyclooxygenase in the presence of propofol (Inada et al., 2009a).
Macrophages
Although the viability of macrophages was not affected, an appropriate concentration of propofol could regulate their function.
High-dose propofol treatment damages the immune system, possibly leading to propofol infusion syndrome (PRIS), a life-threatening complication. In 2019, to discover the molecular mechanisms, a study by Sun et al. (2019) found that propofol overdose could induce macrophage pyroptosis, a kind of programmed cell death (PCD) via activating the NLRP3/ASC/caspase-1 pathway. However, it has been reported that the therapeutic concentrations of propofol did not cause a decrease in macrophage viability (Chen et al., 2003). Interestingly, in 2020, a new study reported that by downregulating the expression of TLR-4 and inhibiting caspase-1 activation, LPS-and ATP-induced pyroptosis could be reversed by the administration of propofol (Ji et al., 2020). Besides cell viability, Chen et al. (2003) found that propofol also suppressed the phagocytosis of macrophages, which was due to the inhibition of mitochondrial membrane potential and adenosine triphosphate synthesis. And another research demonstrated that propofol repressed the phagocytosis of macrophages by the activation of GABAA receptors and inhibition of p130cas phosphorylation (Shiratsuchi et al., 2009). However, research in 2019 has reported that distinct from isoflurane and sevoflurane, propofol did not abate macrophage phagocytosis mediated by opsonization-special phagocytosis enhanced by the antibody or complement bound to the target (Zha et al., 2019). As for cytokines, propofol could inhibit INF-γ mRNA synthesis in macrophages. Also, there were some differences in polarized macrophages. Probably by regulating the GABA(A) receptor and Nrf2-mediated signal transduction, propofol prevents the release of IL-6 and IL-1β in M1 macrophages to attenuate tissue damage-related inflammatory responses, but did not affect the function of M2 macrophages (Kochiyama et al., 2019). Moreover, it has been confirmed that propofol could consistently reduce the oxidative ability of macrophages (Chen et al., 2003). Also, chemotaxis was also suppressed.
Microglia
Microglia are special macrophages located in the brain and spine, participating in neuroinflammatory processes. Similarly, under a dose that did not affect the viability of microglial, propofol could still suppress its function (Liu et al., 2017).
It has been reported that the pretreatment of propofol could attenuate extracellular pressure-stimulated phagocytosis in human HMG030 cells (Yu et al., 2011). Also, under the treatment of propofol, the secretion of pro-inflammation cytokines was retarded (Yu et al., 2011). Through the inhibition of the NF-κB and p38 MAPK pathways, propofol alleviated the intermittent hypoxia-induced secretion of TNF-α and IL-6 (Liu et al., 2017). Also, the production of IL-1β induced by LPS could be suppressed by propofol via inhibiting extracellular signal-regulated kinase 1/2 (ERK 1/2) phosphorylation (Tanaka et al., 2013). Besides, the reduction
Granulocytes
Due to the diversity of the nucleus, granulocytes are also named polymorphonuclear leukocytes (PMN), composed of eosinophils, neutrophils, basophils, and mast cells. Given the abundance, PMN refers in particular to neutrophils under some circumstances.
Neutrophils
Although neutrophils are vital guardians against bacterial infection, they also exhibit tissue injury effects by producing a large number of toxic factors, and this harmful inflammation reaction could be suppressed by propofol.
It has been reported that chemotaxis, phagocytosis, and reactive oxygen species (ROS) production of neutrophils all could be blocked by propofol. Mikawa et al. (1998) examined the effect of propofol at clinically relevant concentrations and found that propofol could inhibit the function of neutrophils in a dose-dependent manner, which was supposed to contribute to the decreasing effect on [Ca2+]i in neutrophils. Another in vitro experiment came to the same conclusion and proved that propofol significantly decreased O 2− and H 2 O 2 formation and released myeloperoxidase (MPO) (Mühling et al., 2002). Particularly, Yang et al. (2013) have shown that by the blockade of formyl peptide receptor 1 (FPR1), propofol inhibited the generation of superoxide generation and elastase in neutrophils, attenuating neutrophil-mediated inflammatory damage.
Lymphocytes
Lymphocytes can be generally divided into nature killer (NK) cells, T helper (Th) cells, cytotoxic T cells (CTLs), γδ T cells, and B cells. Some of these have been reported to be regulated by propofol.
NK cells
The NK cells are a member of the innate immune system, which can recognize and destroy cancer or virus-infected cells in a nonspecific way. Dendritic cell (DC)-based vaccine injection is a special method to enhance anticancer immunity. Through experiments conducted in mice, Inada et al. (2009b) proved that propofol-differentiated DCs could significantly improve the activity of NK cells. Also, research in 2018 reported that by influencing the expression of activating or inhibitory receptors, the cytotoxicity of NK cells in postoperative patients with esophageal squamous cell carcinoma could be upregulated by propofol (Zhou et al., 2018). PGE2 could suppress the production of IFN-γ from NK cells via the EP4 receptor. The cytokines secreted by the NK cells, like IFN-γ, could be promoted in the presence of propofol, which might be relevant to the suppression of macrophage PGE2 production. The upregulated IFN-γ can increase the activity of macrophages, leading to increased production of IL-12 and IL-18 and resulting in further activation of NK cells (Inada et al., 2010). Also, the killing effect of NK cells on tumors would be enhanced in this way. This is consistent with the research that reported propofol as a powerful inducer of interferon-γ (Rossano et al., 1992).
T cells
It is believed that anesthetic agents have an adverse effect on human immunity. To examine the effects of intravenous (i.v.) agents on T lymphocytes, thiopentone, methohexitone, etomidate, and propofol were administrated. Devlin et al. (1994) confirmed that among them, propofol was the only one that showed no statistically significant depression of T-cell proliferation.
According to the difference in TCR, the thymus cell can be divided into αβ T cells and γδ T cells, and αβ T cells can be subdivided into CD4+T cells and CD8+T cells, which can differentiate into T helper (Th) cells and cytotoxic T cells (CTLs). Research by Kushida et al. (2007) reported that in vitro activity of CTLs against EL4 (mouse lymphoma cells) was obviously greater after propofol treatment, which is probably the reason why after propofol administration, the antitumor immunity of CTLs was obviously improved.
Th1 and Th2 cells are two kinds of Th cells coming from nondifferentiated Th0 cells, inducing cell-mediated immunity and humoral immunity, respectively. Salo et al. (1997) launched the research to compare the Th1/Th2 balance modulation function between thiopentone and propofol. The result illustrated that distinct from thiopentone, propofol did not suppress Th cell function, but shifted Th1/Th2 balance to induce Th1-like responses.
In general, by regulating the balance between Th1 and Th2 cells, propofol promotes the Th1-like responses, which may contribute to the activity of NK cells and TCLs, and the function of B cells may be not affected.
Dendritic cells
The antigen-presenting cell (APC), like a macrophage, B cell, or dendritic cell (DC), can present antigens to T cells, activating its function. Both the macrophage and the B cell can only present antigens to mature T cells, like activated or memory T cells. However, the DC is the most powerful APC, and the only cell that can activate naive T cells (Th0), which bridge innate and adaptive immunity. By processing antigens during innate immune responses to present them to Th0, DCs can initiate adaptive immunity. Beyond that, DCs can produce cytokines, leukotrienes, and prostanoids to regulate immune response as well. Research has Frontiers in Pharmacology frontiersin.org reported that propofol suppresses prostaglandin E2 production from DCs by inhibiting the activity of COX enzyme (Inada et al., 2011), but the impact is limited to DCs, with no influence on IL-12/IL-10 production and T-cell proliferation. Leukotrienes can include cysteinyl leukotrienes (CysLTs) and leukotriene B4 (LTB4 ). CysLTs can mediate DC migration and maturation, and LTB4 can also modulate multiple immune processes, such as enhancing phagocytic and antimicrobial activities of the neutrophils and macrophages and stimulating the secretion of immunoglobulins by lymphocytes. Inada et al. (2013) found that by the direct inhibition of the 5-LO enzyme, propofol could also suppress the production of CysLTs and LTB4. Also, another article (Sethi et al., 2012) pointed out that suppression of leukotriene production with propofol could decrease metastasis after tumor surgery.
5 Propofol: A potential immunoregulative remedy in cancer treatment
Mechanism of anticancer effect of immunocytes
The human immune system has three basic functions: immune defense (for eradicating foreign pathogens), immune surveillance (for eliminating abnormal inner components), and immune homeostasis (for preventing normal inner constituents from attacking). Among them, immune surveillance is the key point to protecting organisms from tumor invasion.
Adaptive immunity is crucial in recognizing and cleaning up the body's tumor cells. Generally, cellular immunity is the protagonist in immune surveillance, with the assistance of humoral immunity (Figure 2). The antigen is a special substance to stimulate the body to initiate adaptive immunity. The production of adaptive immunity like antibodies and the CTL can capture its special antigen powerfully, the tension of which is known as antigenicity. Immunogenicity is the other important characteristic of antigen, which can reflect the potency to stimulate adaptive immunity. Different from antigenicity, immunogenicity is not exhibited by all antigens, and even some of them do not have the ability to initiate adaptive immunity. For those tumor antigens possessing weak or no immunogenicity, native immunity seems to undertake more obligations toward tumor prevention. These anticancer immunocytes involve NK cells, macrophages, etc.
Designating to eradicate tumor cells directly, the CTL is the essential executor of acquired immunity for cancer resistance, which can be a self-propagating process expatiated by Chen and Mellman (2013). They named this the cancer-immunity cycle, systematically illustrating the generation of cancer immunity
FIGURE 2
Anticancer cycle of immunocytes. BC, TC, and NK are the fundamental immunocytes in tumor immunity. The antigen released from tumor cells will be acquired by DCs. Also, DCs can present the antigen to Th0. Once activated, Th0 proliferates and differentiates into Th2 or Th1, harmonizing the activation of PC and CTL. Then, the antibodies or CTL will traffic to the TME and recognize target tumor cells and produce an effect. NK can recognize and kill cancer cells in non-special ways. BC, B lymphocytes; TC, T lymphocyte; NK, natural killer cells; ab, antibody; CTL, cytotoxic lymphocyte; PC, plasma cell; TME, tumor microenvironment; DC, dendritic cell; Th0, non-differentiated type-0 cell; Th1, type 1 T helper cell; Th2, type 2 T helper cell.
Frontiers in Pharmacology frontiersin.org dominated by T-cell responses (Pio et al., 2019). The closed loop consists of seven stepwise events that include the release of cancer cell antigens, cancer antigen presentation, activation of T cells, trafficking of T cells to tumors, infiltration of T cells, recognition of cancer cells by the CTL, and killing of cancer cells (Chen and Mellman, 2017). With the implementation of this cycle, immunestimulatory factors are accumulated, amplifying the process of acquiring immunity response. DCs coming from the bone marrow enter into the peripheral blood, distributed throughout the whole body except the brain by the blood flow (Gardner and Ruffell, 2016). Also, the DCs settling in multiple organs and nonlymphoid tissue are immature. They can highly express membrane receptors like the Fc receptor and mannose receptor, with the assistance of which DCs act out the powerful ability of antigen uptake. After the uptake, immature DCs are transformed into mature DCs, gradually, and at the same time, their migration from the peripheral tissue into secondary lymphatic organs along the lymphatics occurs.
The membrane receptors or antigen uptake, like the Fc receptor, is downregulated in mature DCs, ending in the shutting down of uptake capacity. However, their ability of antigen presentation is sharpened due to the high expression of MHC II molecules and the abundance of co-stimulatory molecules like CD80, CD86, and CD40. The existence of a costimulatory signal is indispensable for the activation of DC4+ naive T cells, and the DC is the only professional APC that can trigger the activation of Th0 cells.
Once the combination of tumor antigen and membrane receptors is accomplished, endocytosis of the DCs forms the endosomes, trafficking antigen protein into the lysosomes. The antigens break up into short peptides containing 10-30 amino acid residues under the efficiency of acid hydrolases.
MHC II formed in the rough endoplasmic reticulum is transferred into the Golgi complex and carried by vesicles to endosomes. After that, MHC II captures the short peptides decomposed by hydrolases, forming an antigen-MHC II complex. This complex will be transferred into the cytomembrane for antigen presentation.
With the aid of surface adhesion molecules, Th0 cells can make a temporary touch with DCs. When the TCR of Th0 recognizes the specific antigen-MHC II complex on the surface of the DC, the stable combination is achieved. Then, the CD4 of Th0 binds with the MHC II of the DC, making it more stable. Meanwhile, the relative protein of Th0 can connect with various co-stimulatory molecules, promoting the activation of Th0.
In the local microenvironment, naïve CD4 + Th0 cells are regulated by different cytokines to make separate differentiation. Cytokines like IL-12 and IFN-γ will hasten Th0 to realize Th1 polarization, but IL-4 can prompt it into Th2.
The cytokines produced by Th1, like IL-2 and IFN-γ, can promote cellular immunity response. The progenitor of the CTL in the secondary immune organ that captures the antigen-MHC I complex by TCR, proliferates, and differentiates into the CTL under the effect of cytokines secreted by Th1.
Guided by the chemokines, the CTL departs the lymphatic tissue for the location of the tumor. Similarly, the CTL contacts tumor cells by the adhesion molecule and recognizes them by the combination of TCR with the antigen-MHC I complex. After that, the cancer cells are executed by the CTL through the secretion of perforin/granzyme or the Fas/Fasl pathway.
The antigen-antibody immune complex can be bound to the surface of the DCs with the aid of CD21, forming iccosomes. Naïve B cells can recognize iccosomes directly depending on APC processing. Swallowed and processed by the B cells, the iccosomes combine with the MHC I and are presented to the Th cells, promoting Th2 polarization. Meanwhile, the activated Th cells express CD40L, providing the second signal for B-cell activation.
B cells shift into the plasma cells (PCs) under the influence of cytokines like IL-4, IL-5, and IL-6 secreted by the Th2 cells. The PCs are terminal B cells, most of which immigrate into the bone marrow and produce antibodies for a long time.
Although tumor antigens can induce specific antibodies and kill tumor cells by ADCC, the spontaneously produced antibody is not the key factor for tumor suppression. Some antibodies can even interfere with the immune response and promote cancer proliferation (Stambrook et al., 2017;Tokunaga et al., 2019).
The NK cell is one of the main tumor executors for cancer that lacks immunogenicity due to its nonspecific effect (Myers and Miller, 2021). The NK cells come from the bone marrow and are mainly interspersed among the peripheral blood and spleen. They can put tumor cells to death directly without the sensitization of antigens. Similar to the pattern of the cancerimmunity cycle, Bald et al. (2020) proposed the NK cell-cancer cycle to illuminate the anticancer process of NK.
The tumor microenvironment (TME) is closely related to cancer development (Wu and Dai, 2017;Arneth, 2019), consisting chiefly of different types of cells and the extracellular matrix (ECM). First, endothelial cells emerge. Tumor vessels originate from the endothelial progenitor cells, providing nutrient substances for tumor survival. The fibroblast is another participant of the TME. It has been reported to promote hematogenous metastasis of primary carcinoma. Besides, diversified immune cells, such as granulocytes, lymphocytes, and macrophages, settle in the TME. For instance, the macrophages can aggravate metastasis of tumors. The ECM is an intricate net structure by macromolecules, whose concentration can affect the intensity of tumor mobility.
The first step of the NK cell-cancer cycle is the recruitment of NK cells into the TME. Chemokines like CCL5 secreted from the immunocytes in the TME are crucial for the recruitment. Then, the NK cell membrane receptors, like the killer immunoglobulinlike receptors (KIRs) and natural cytotoxicity receptors (NCRs), recognize the tumor cells, leading to the activation of NK cells.
After this, death receptor signaling or cytotoxic granules expelled by the NK cells can sentence tumor cells to death. The NK cells can also orchestrate adaptive immune responses. The cross-talk between the NK cells and DCs can bring about profound adaptive immunity to cancer.
Outlook of propofol in cancer suppression by immunomodulation
The influence of propofol on tumor immunology is relatively legible. Despite propofol possessing the ability to depress cancer proliferation directly, stimulating immunity of the human body is a more positive way to confront potential residual cancer cells.
The frame of tumor immune is simplified in Figure 3A. First, we discussed the effect of propofol on cellular immunity and humoral immunity. IFN-γ and TNFβ produced by the human Th1 subset can induce cell-mediated immunity. IL-4, IL-5, and IL-9 produced by the Th2 subset can promote humoral immunity. Rossano et al. found that propofol could promote unstimulated lymphocytes to produce both IL-4 and IFN-γ, while other anesthetics like ketamine or thiopentone could only upregulate the production of IL-4 but not IFN-γ. This might indicate that in the presence of propofol, both cellular immunity and humoral immunity were upregulated. Further research carried out by Salo et al. indicates that propofol could alter the Th1/Th2 balance judging by the production of IFNγ and IL-4. The increased IFN-γ/IL-4 ratio probably indicates that propofol could improve cell-mediated immunity by shifting the balance toward Th1-like responses. Given the IFNγ/IL-4 ratio and the complicated effect of antibodies on immunity, humoral immunity does not play an important role in propofol-induced cancer immunity.
The chief factor for anticancer immunity of propofol is focused on its influence on Th1 cells. It has been reported that the activity of CTL against EL4 cancer cells was significantly improved after propofol administration, which might be related to its stimulative effect on Th1 cells that could promote the activation of CTL.
As for NK cells, propofol could upregulate the cytotoxicity of NK cells against esophageal squamous cell carcinoma. This phenomenon may have a significant bearing on the DCs. The research by Inada et al. proved that the activity of NK cells could be upregulated by propofol-differentiated DCs. Besides, it has been reported that propofol could regulate the migration and maturation of DCs by affecting the level of leukotrienes. Also, the upregulated IFN-γ by propofol could also enhance cytotoxicity of the NK cells. All of these indicate that cell-mediated cytotoxic innate immunity participates in propofol-induced cancer immunity.
As shown in Figure 3A, the increased production of IL-12 is the reason for the upregulation of anticancer immunity by activating NK cells and CTLs. Cytokines like IL-12 can hasten Th0 to realize Th1 polarization, activating cellular immunity. Also, IL-12 secreted by the DCs, macrophages, or other cells can raise the activation of NK cells. Figure 3B explains the mechanism of IL-12 accumulation, and the decrease of prostaglandin E2 (PGE2) secretion due to the COX2 inhibition is pivotal. PGE2 is important in modulating immunity, by suppressing the production of IFN-γ secreted from the NK or Th cells. Consequently, innate immunity of NK cells and adaptive immunity activated by Th1 cells could be downregulated by PGE2. Because of its ability to produce a vast amount of PGE2, the macrophage is one of the most powerful cells in the regulation of immunity. Inada et al. found that propofol could suppress the activity of COX, especially COX-2, due to the similarity of its chemical structure to a COX inhibitor, γ-tocopherol, suppressing the production of PGE2. The suppression of PGE2 production could raise the production of IL-12, which promotes the activity of CTL by promoting Th1 polarization. E-prostanoid receptors include EP1, EP2, EP3, and EP4. By binding with EP4, PGE2 can directly inhibit IFN-γ production by NK cells (Walker and Rotondo, 2004). It has been reported that the suppression of PGE2 by propofol led to the upregulation of IFN-γ secreted by NK cells. In turn, increased IFN-γ enhanced IL-12 and IL-8 secretion by macrophages, leading to further activation of NK cells in the macrophages: NK cell co-culture. This indicates the underlying mechanism of activation of the NK cells by propofol.
Overall, the promotion of the anticancer effect of CTL and NK cells is probably due to the inhibition of COX-2 in macrophages by propofol, the descending secretion of PGE2. The decrease in PGE2 likely contributes to the upregulation of IFN-γ and IL-12, finally promoting the activities of NK cells and CTLs.
FIGURE 4
Simplified mechanism of immunosuppression induced by surgery damage. Tissue injury caused by surgery stimulates the afferent neuron (AE). The cerebrum accepts the stimulation, processes the information, and sends them to a special function unit. The locus coeruleus-noradrenergic system (LC/NE system) rapidly transmits the instruction to the immune organs and adrenal medulla through the efferent neuron (EN) and sympathetic nervous system (SNS). Affected by the hypothalamus, the anterior pituitary secretes the adrenocorticotropic hormone (ACTH) to promote the release of cortisol. Both of these contribute to the inhibition of the immune system.
Frontiers in Pharmacology
frontiersin.org
Immunomodulation function of propofol in cancer resection
Surgery is the primary selection for most cancer treatments, and in this section, we discuss the function of propofol in cancer resection based on immunity.
Surgery and immunity suppression
The immune system is the guardian of the organism. Its normal operation protects the human body from external threats and updates the internal components to maintain their functional conditions. The disorder of immunity, which
FIGURE 5
Frame of inflammation and the key modulating point of propofol to the monocyte-macrophage system. Tissue damage stimulates the occurrence of inflammation by the DAMP release. By inhibiting M1 polarization of macrophages, propofol may prevent the local tissue from excessive inflammation.
FIGURE 6
Modulation of propofol to neutrophils. The local damaged tissue can activate neutrophils by the binding of FPR1. Propofol can weaken this activation, preventing the excessive emission of proteolytic enzymes or toxic ROS, which can limit local tissue damage.
Frontiers in Pharmacology frontiersin.org 12 includes both enhancement and reduction, brings troubles. Immunity can be affected by many factors, like the emotional state, nutritional status, and drug administration. Herein, we briefly demonstrate the variations of immunity in the patients undergoing surgery.
Surgical lesion and immunosuppression
Immunosuppression is a common biological response induced by surgery stress and the stress system, which includes the hypothalamic-pituitary-adrenal (HPA) axis and sympathetic nervous system (SNS), participates in the process (Charmandari et al., 2005). Without the consideration of tranquilizers and anesthetics, tissue damage under surgical operation can cause local inflammation and pain (Manou-Stathopoulou et al., 2019), both of which stimulate the sensors of the afferent nerves. After neurotransmission, the brain receives nerve impulses and commands the whole body through the HPA axis and SNS (showed in Figure 4).
Influenced by the corticotrophin-releasing hormone (CRH) coming from the paraventricular nucleus of the hypothalamus, the anterior pituitary secretes the adrenocorticotropic hormone (ACTH), prompting the adrenal cortex to release glucocorticoids (e.g., cortisol). As a lipophilic hormone, glucocorticoids can easily pass through the cell membrane and bind to the corresponding receptors, the glucocorticoid receptors (GRs), in the cytoplasm. After being activated by the glucocorticoids, the GR can translocate into the nucleus, exhibiting its genomic effects in the form of a homodimer or monomer to inhibit inflammation (Vandewalle et al., 2018).
Distinct from the parasympathetic nervous system, the SNS is crucial in the regulation of immune responses (Janig, 2014). Both central immune organs, which include the bone marrow and thymus, and peripheral immune organs, like the lymph nodes, spleen, and mucosal-associated lymphoid tissue (MALT) are innervated by the SNS, and the immune cells with adrenoceptors may be dominated by the noradrenaline released from the sympathetic postganglionic neurons, causing adrenergic stress-induced immunosuppressive effects. The locus coeruleus-noradrenergic system (LC/NE system) (Aston-Jones and Waterhouse, 2016; Benarroch, 2018), comprising the locus coeruleus and other noradrenergic cell groups in the brainstem, is another vital functional unit during stress response. In receiving stimulation from the brain, it can not only activate the HPA axis to increase the secretion of cortisol but also modulate the SNS. The spinal cord efferent fibers from the neurons in the LC/NE system end up in preganglionic sympathetic neurons, igniting the activity of the SNS. As mentioned previously, via acetylcholine, fibers from the preganglionic sympathetic neurons activate the sympathetic postganglionic neurons, which in turn innervate the primary and secondary lymphoid organs with norepinephrine. Besides, the preganglionic sympathetic neurons can also directly activate the adrenal medulla through acetylcholine, which is also termed the sympathetic-adrenal medulla system, to increase the secretion of catecholamines such as norepinephrine and epinephrine. Given the fact that adrenergic receptors extensively exist on the surface of the immune cells, the circulating catecholamines may also play a role in immunomodulation (Sharif et al., 2018).
Generally, despite the powerful immunosuppressive functions of the glucocorticoids, catecholamines also participate in the regulation. By the combination of rapid sympathetic nerve modulation to the immune organs and the gradual raise in circulating glucocorticoids and catecholamines secreted from the adrenal, the immunosuppression of surgery trauma is achieved.
Perioperative period and immune suppression
The perioperative timeframe is a period of time embracing surgery, starting with the patients' receival of operation notification and ending with their achieving the related treatment; the patient's body generally returning to the basic level. Psychological stress is the basal challenge in the preoperative phase. Studies have demonstrated that 60%-80% of patients have anxiety due to the awareness of their surgery (Gursoy et al., 2016). Also, it has been confirmed that anxiety from the incoming uncharted surgery arouse stress responses through the HPA axis. After being sent into an operating room, patients begin the intraoperative experience until their transfer to the postanesthesia care unit (PACU) is completed. During the intraoperative period, the administration of anesthetics, analgesic agents, and muscle relaxants is essential. In addition to the traumatic damage to the body, anesthetic and analgesic agents can also produce immunosuppression (Longhini et al., 2020). The local anesthetic administration may but mitigate the HPA axis and SNS response by blocking the afferent neural transmission to alleviate the inhibition (Xu et al., 2016). The postoperative phase is a period of body recovery, which includes immune recovery. The intraoperative phase-induced immunosuppressive effect will last for several days, and the selection of different anesthetics can influence the suppressive duration.
Propofol's immunoregulation function in cancer resection compared with volatile anesthetics
The combined intravenous-inhalation anesthesia is the most commonly used general anesthetic pattern in the neoplasm resection. Therefore, we focus on the immunomodulation of intravenous (i.v.) and volatile anesthetics here. The i.v. anesthetics usually exhibit immunosuppressive characteristics. Ketamine, midazolam, and droperidol have been proven to suppress the chemotaxis of monocytes, attenuating the immune function of postoperative patients (Krumholz et al., Frontiers in Pharmacology frontiersin.org 1999), and thiopentone, methohexitone, and etomidate have been reported to reduce the function of T cells (Devlin et al., 1994), which may easily promote tumor dissemination and recurrence. Compared to these, propofol seems to have but little side impact on immunity. When it comes to volatile anesthetics, are there still advantages for propofol in immunity after cancer resection?
Breast cancer
There are some complications in the effect of propofol and sevoflurane on immune function in patients undergoing breast cancer surgery. Some clinical researchers have reported that there are not many differences in lymphocyte function or cytokine secretion after propofol or sevoflurane treatment (Lim et al., 2018;Oh et al., 2018) (Deegan et al., 2010). Lim et al. had detected no differences in the NK cell count, CTL count, or cytokines between blood samples of the studied groups. In 2018, similarly, Oh et al. had reported that there were no intergroup differences in type 1 and type 17 T helper cells, NK cells, cytotoxic T cells, cytokines, or the neutrophil-to-lymphocyte ratio. The serum concentrations of 11 cytokines [interleukin 1β (IL-1β), IL-2, IL-4, IL-5, IL-6, IL-8, IL-10, IL-12p70, IL-13, interferon γ, and tumor necrosis factor α) and three MMPs (MMP-1, MMP-3, and MMP-9) were measured by research, and only a minority of cytokines were different between the two groups, which means these two drugs exhibit a similar function in regulating perioperative cancer immunity. Nevertheless, when compared to sevoflurane, propofol had a better impact on immunity due to its preservation for NK cell cytotoxicity (NKCC) (Cho et al., 2017). A total of 50 patients undergoing breast cancer resection were analyzed by Cho et al. to demonstrate the influence of these on the function of NK cells, and the results revealed that when compared with sevoflurane, propofol significantly increased NKCC (%). As for desflurane, while both these had a favorable impact on immunity, desflurane seemed to be the better of the two (Woo et al., 2015). Research in 2015 indicated that in terms of preservation of IL-2/IL-4 and CD4 + /CD8+T cell ratio, both propofol and desflurane anesthesia induced a favorable immune response in the perioperative period. But when it came to leukocytes and NK cells, desflurane exhibited fewer adverse immune responses than did propofol.
Colorectal carcinoma
Research has revealed that propofol had less influence on lymphocytes than sevoflurane (Chen et al., 2015). This research showed that the percentage of CD3⁺, CD4⁺, and CD19⁺ subtypes increased immediately after surgery, while the percentage of NK cells significantly decreased. But the proportion of the lymphocyte subtype recovered to the preoperational baseline sooner in the propofol group than it did in the sevoflurane group. CD45RO+ cells are the functional form of T cells, while CD45RA+ cells are inactive T cells. In another survey, the levels of CD45RA+ and CD45RO+ of both groups significantly decreased after surgery, while the CD45RA+ of all recovered at the postoperative 72 h, whereas the level of CD45RO+ recovered less in the sevoflurane group. According to the level of CD45RA+/CD45RO+, the postoperative immunosuppression in the treatment group of sevoflurane lasted a longer time than it did in the propofol group (Yu et al., 2019).
Other cancers
In tongue cancer, the percentages of CD3(+) cells, CD3(+) CD4(+) cells, NK cells, and the CD4(+)/CD8(+) ratios after an operation were significantly lower in the sevoflurane groups (Zhang et al., 2014), while propofol exhibited a better effect on the recovery of immunity. The activation and differentiation of T helper cells are essential in perioperative antitumor immunity. The percentage of CD4(+)CD28(+) and ratio of interferon-gamma:interleukin-4 were evaluated in a research. Compared to isoflurane in non-small-cell lung cancer, propofol showed a more significant effect on the activation of peripheral Th cells (Ren et al., 2010). In addition, Liu et al. (2016) found that the counts of CD3 + cells, CD4 + cells, NK cells, and the CD4+/CD8+ ratios were significantly lower in the sevoflurane group, which meant that the function of the lymphocyte subsets in cervical cancer patients undergoing radical hysterectomy seemed to be better in the presence of propofol than it is with sevoflurane. However, in kidney cancer surgery, a study analyzing the amount of NK cells, total T lymphocytes, regulatory T cells, T helper cells, cytotoxic T lymphocytes, and their subpopulations concluded that there were no significant differences between propofol and sevoflurane in many kinds of lymphocytes (Efremov et al., 2020).
Generally speaking, in terms of the clinical results in many kinds of different cancer resections (showed in Table 2), it is probable that propofol possesses more benefits in postoperative immunoregulation than do sevoflurane and isoflurane, but less than desflurane. Despite that, there are still some divergences and more research is needed.
Propofol promotes wound healing by modulating inflammation in surgical cancer treatment
Inflammation is an ineluctable process after surgical treatment, and in spite of the immune reaction, immunocytes also participate in the process of inflammation.
Model of inflammation after cancer resection
Inflammation is a defense reaction of living creatures to resist harming factors. Its function is to clean up damaging cells and Frontiers in Pharmacology frontiersin.org recover tissue injury, which include the three basic processes-alteration, exudation, and proliferation. It is an extremely delicate and intricate biological process with hundreds of thousands of uncovering mysteries. Herein, we depict the basic mode of the inflammation process and put forward the key regulating points of propofol in it. The inflammatory response is coordinated by a large range of mediators that form complex regulatory networks (Kuprash and Nedospasov, 2016). To dissect these complex networks, it is helpful to place these signals into functional categories and distinguish between inducers and mediators of inflammation. Inducers are signals that initiate the inflammatory response. They activate specialized sensors, which then elicit the production of specific sets of mediators. The mediators, in turn, alter the functional states of tissues and organs, where the effectors settle, in a way that allows them to adapt to the conditions indicated by the particular inducer of inflammation. Thus, a generic inflammatory 'pathway' consists of inducers, sensors, mediators, and effectors, which determines the type of inflammatory response (Medzhitov, 2008;Medzhitov, 2010). The conserved microbial products, such as lipopolysaccharide, are referred to as pathogen-associated molecular patterns (PAMPs), and they activate pattern recognition receptors (PRRs). The PRR signaling pathways have been well characterized as the initiators of cascades that eventually lead to the migration of leukocytes to the site of infection. In spite of this, inflammation in the absence of pathogens and their products is referred to as sterile inflammation, which is induced by sterile injury like cancer resection. The immunostimulatory molecular patterns in sterile inflammation differ from microbial patterns and are canonically associated with damage (Zindel and Kubes, 2020); thus, they are called damage-associated molecular patterns (DAMPs). DAMPs are released during tissue damage and initiate an inflammatory response. Sterile inflammation and subsequent tissue repair depend on a well-orchestrated migration sequence of leukocytes to and from the site of injury.
Sensors, such as toll-like receptors (TLRs), are expressed on specialized sentinel cells, such as tissue-resident macrophages, DCs, and mast cells. They induce the production of mediators, which include cytokines, chemokines, bioactive amines, eicosanoids, and products of proteolytic cascades such as bradykinin. Mast cells are present in most tissues characteristically surrounding blood vessels and nerves and are especially prominent near the boundaries between the outside world and the internal milieu, such as the skin, mucosa of the lungs, and digestive tract, as well as the mouth, conjunctiva, and nose. Herein, we assign the conception of the sentinel cell to mastocyte and exhibit the process of sterile inflammation. The mast cell is the initiator of inflammation and plays a key role in the inflammatory process (da Silva et al., 2014). Mast cells can be stimulated to degranulate by allergens through cross-linking with immunoglobulin E receptors (e.g., FcεRI), physical injury through pattern recognition receptors for damage-associated molecular patterns (DAMPs), microbial pathogens through pattern recognition receptors for pathogenassociated molecular patterns (PAMPs), and various compounds through their associated G-protein-coupled receptors (e.g., morphine through opioid receptors) or ligand-gated ion channels (Redegeld et al., 2018). When activated, a mast cell can either selectively release (piecemeal degranulation) or rapidly release (anaphylactic degranulation) "mediators," or compounds that induce inflammation, from storage granules into the local microenvironment (Mukai et al., 2018). In the phase of alternation, the stimulators from injuring tissue activate mast cells. Once activated, the multiple inflammatory mediators (like histamine, leukotriene, prostaglandin D2, IL-1, IL-4, IL-8, TNF, and so on) are released. Some of them (like histamine, leukotriene, IL-1, and TNF) induce the contraction of vascular endothelial cells, enlarging the endothelial cell gap, and finally increasing vascular permeability. Exudation of white cells is another feature of inflammation. Cytokines like TNF and IL-1 can increase the expression of integrin ligands in the endothelial cell, which can promote the adhesion of white cells. Under the action of chemotactic agents (like LTB4 and IL-8), white cells are attracted to the injured place through the blood vessel. The neutrophils are the most abundant leukocytes in the blood and rapidly infiltrate the target tissues. Monocytes are recruited second in the phase of exudation due to the release of the monocyte chemotactic agent from neutrophils (Kolaczkowska and Kubes, 2013). The leukocytes, mainly referring to neutrophils, recognize and degrade damaging tissue and hazard factors through phagocytosis to maintain homeostasis, and then the regeneration of the peripheral tissue realizes the healing process.
In spite of the fact that neutrophils exhibit positive aspects in tissue recovery, during phagocytosis, the leakage of lysosomal content into the extracellular stroma exacerbates tissue damage (Amulic et al., 2012). This is also a common phenomenon causing signs of swelling, pain, redness, heat, and loss of function, which makes the proper regulation of neutrophil function significant.
Propofol possesses beneficial influence on sterile inflammation in tissue recovery
By regulating inflammation-associated immune cells, propofol plays a role in wound healing (showed in Figure 5). Although mastocytes initiate inflammation, recent research have reported that propofol could inhibit this function. Neutrophils are still the direct participants and the modality of neutrophils decides the destiny of inflammation, a favorable or detrimental one. Propofol has a direct impact on neutrophils and downregulates its function even at clinically relevant Frontiers in Pharmacology frontiersin.org concentrations. Although neutrophils play a crucial role in the host defense mechanism as a component of nonspecific cellmediated immunity, they are also thought to play an important role in the pathogenesis of auto tissue injury, leading to multiple organ dysfunction (Liew and Kubes, 2019). Overproduction of reactive oxygen species (ROS) by neutrophils that accumulate in the organs in response to chemotactic factors in the surrounding milieu contributes to this process. Thus, the immune function of neutrophils is a double-edged sword, and how to properly regulate its capability and prevent its overactivation is crucial. The suppression of neutrophil functions by anesthetics may be favorable to attenuate organ dysfunction mediated by the mechanism of auto tissue injury. Propofol inhibits the chemotaxis, a movement toward certain chemicals, of neutrophils, which probably attenuates its aggregation in damaged tissue. Also, phagocytosis of neutrophils is also inhibited by propofol; similar to chemotaxis, the underlying mechanism of propofols' suppression of phagocytosis is unidentified. The decreasing effect on the intracellular calcium ion by propofol might result in depression of chemotaxis, phagocytosis, and ROS production. Delightfully, further research on propofols' inhibition of the production of ROS and released proteolysis enzymes might imply its positive effect on inflammation (showed in Figure 6). The formyl peptide receptor (FPR) belongs to a class of G protein-coupled receptors and the class of receptors possessing seven hydrophobic transmembrane domains located in the cellular membrane of neutrophils. Human neutrophils express two members of this family: FPR1 and FPR2. Through the selective and competitive binding affinity to FPR1, propofol achieves its influence on neutrophils. FPR1 is activated by N-formyl peptides, which can be derived from mitochondrial proteins. By binding to FPR1, N-formyl peptides released from damaged tissue can activate neutrophils and induce severe inflammatory responses. Once overwhelmingly activated, neutrophils are transformed into destructive cells, releasing toxic ROS and proteolytic enzymes that destroy the surrounding tissue. It has been proven that propofol can reduce the overproduction of ROS by inhibiting the Ca 2+ and MAPK signaling pathways. After binding with ligands, FPR1 can activate phospholipase C (PLC), which catalyzes the conversion of phosphoinositol 4,5-biphosphate to inositol 1,4,5triphosphate (IP3) to cause the rapid release of Ca 2+ . Besides, the ERK 1/2 MAPK cascades are also activated by FPR1, increasing the production of ROS. Both of these can just be suppressed by propofol. But how propofol represses the elastase, a major serine protease secreted by stimulated human neutrophils, released from the granules into their surroundings is still elusive. Also, the roles of PI3K/AKT signaling and p38 MAPK cascades inhibited by propofol via FPR1 still need further consideration.
Neutrophils are the executors of inflammation in their early phase, directly impacting the fate of inflammation, favorable or destructive. Also, the macrophage, recruited by the neutrophil to the damaging place, is probably the switch deciding the fate. During inflammatory processes, monocyte-derived (M0) macrophages undergo polarization to classically (M1) and alternatively (M2) activate macrophages, depending on the local tissue environment. Two cytokines, namely, macrophage colony-stimulating factor (M-CSF) and granulocyte M-CSF (GM-CSF), are important for priming monocyte to macrophage differentiation. M-CSF and GM-CSF stimulate monocytes to give rise to phenotypically different subsets of macrophages. M-CSF has been shown to stimulate monocyte differentiation to an anti-inflammatory, immunosuppressive macrophage phenotype (M2), while GM-CSF has been shown to stimulate a pro-inflammatory macrophage phenotype (M1). M1 macrophages are the dominating phenotype observed in the early stages of inflammation and are activated by mediators like interferon-γ (IFN-γ), tumor necrosis factor (TNF), and damageassociated molecular patterns (DAMPs). These mediator molecules create a pro-inflammatory response that in return produces pro-inflammatory cytokines. M1 macrophages recruited during the early phases of inflammation promote the production of interleukin-6, IL-1β, and tumor necrosis factor α (TNF-α), exacerbating inflammation and contributing to tissue destruction. By contrast, M2 macrophages are characterized by their involvement in immune regulation and homeostatic functions associated with wound healing. Antiinflammatory cytokines such as interleukin-4 (IL-4) and interleukin-13 (IL-13) stimulate M2 macrophage polarization and the M2 "repair" designation (also referred to as alternatively activated macrophages) broadly refers to macrophages that function in constructive processes like wound healing and tissue repair and those that turn off damaging immune system activation by producing anti-inflammatory cytokines like IL-10. So the perioperative management to control macrophage differentiation is the point, and propofol is exactly involved in this regulation. Inhibition of IL-6 production may suppress systemic inflammation induced by surgical trauma, thereby reducing postoperative complications. It has been reported that propofol suppresses IL-6 and IL-1β expressions during human M1 macrophage polarization, suggesting that propofol plays a protective role in the development and progression of inflammation. Also, the activity of M2 macrophage is not affected by propofol.
Conclusion
Postoperative immunosuppression is a common phenomenon and has a disadvantageous impact on patient prognosis of surgical cancer resection. Propofol, a generally used anesthetic in oncologic surgery, has been proven to have beneficial effects on immune recovery. Despite the fact that propofol cannot reverse surgery-induced immune suppression related to the activation of HPA and SNS, it can directly regulate Frontiers in Pharmacology frontiersin.org the function of immunocytes. Although the inhibition of innate immune cells, like macrophages and neutrophils, has been generally reported, propofol can improve anticancer immunity by activating the function of lymphocytes like NK cells and TCLs. The enhancement of the cytotoxic effect of these immunocytes may be beneficial to oncologic surgery. The repression of the immune system induced by cancer surgery can facilitate the dissemination of different kinds of cancer cells (Chen et al., 2019b), which leaves a hidden trouble for cancer patients. The elevation of cell-mediated immunity by propofol is likely to make sense in this condition. In addition, multiple articles have reported the anticancer effect of propofol. The comparison with other anesthetics concludes that the priority of propofol is high in cancer surgery for its relatively less adverse impact on immunity. Moreover, propofol is hopefully to be a favorable immunoregulatory agent for cancer treatment and a potential drug for wound healing by the regulation of sterile inflammation. However, the limitation of propofol in clinical application still exists. The commonly used anesthetics clinically include volatile anesthetics, intravenous anesthetics, and perioperative auxiliary drugs. Each type of drug has unique effects on immunity. Although we recommend propofol in cancer resection, combination administration is common in clinical practice. What kind of combinations of anesthetics and adjuvant drugs is the best for propofol in oncological surgery? Also, are there other intravenous anesthetics that possess better effects after an improved combination? Then, the existing articles mainly focus on breast and colorectal cancer, but the research on other tumors is relatively insufficient. To clarify the effect of propofol on immunity, studies on other major cancer types are needed. Finally, in spite of neutrophils, studies of propofol on immunocytes rarely reach the molecular level. More meticulous basic studies are still essential for more precise demonstration.
Author contributions
Study concepts: CW and LW. Study design: CW and LG. Manuscript preparation: LG and XP. Manuscript editing: LG, LW, and XP. Manuscript review: XP, CW, LW, and LG.
|
v2
|
2022-11-04T16:07:20.288Z
|
2022-11-04T00:00:00.000Z
|
253269116
|
s2orc/train
|
Cardiac arrhythmias associated with immune checkpoint inhibitors: A comprehensive disproportionality analysis of the FDA adverse event reporting system
Introduction: With the widespread application of Immune checkpoint inhibitors (ICIs), it is important to explore the association between ICIs and cardiac arrhythmias and to characterize the clinical features of ICI-associated cardiac arrhythmias in real-world studies. Objective: The purpose of this study was to characterize the main features of ICI-related cardiac arrhythmias. Methods: From January 2017 to June 2021, data in the Food and Drug Administration Adverse Event Reporting System (FAERS) database were retrieved to conduct the disproportionality analysis. For the ICI-related cardiac arrhythmia detection, signals were detected by reporting odds ratio (ROR) and information component (IC), calculated using two-by-two contingency tables The clinical characteristics of patients reported with ICI-related cardiac arrhythmias were compared between fatal and non-fatal groups, and the time to onset (TTO) following different ICI regimens was further investigated. Multivariate logistic regression was used to evaluate the association between concurrent cardiotoxicities and ICI-associated arrhythmias. Results: We identified a total of 1957 ICI–associated cardiac arrhythmias reports which appeared to influence more men (64.44%) than women (30.76%), with a median age of 68 [interquartile range (IQR) 60–75] years. Cardiac arrhythmias were reported most often in patients with lung, pleura, thymus and heart cancers (38.02% of 1957 patients). Compared with the full database, ICIs were detected with pharmacovigilance of cardiac arrhythmias (ROR025 = 1.16, IC025 = 0.19). Anti-PD-1 and anti-PD-L1 monotherapies were found to be related to higher reporting of arrhythmias, corresponding to ROR025 = 1.03, IC025 = 0.06 and ROR025 = 1.27, IC025 = 0.29, respectively, with the exception of anti-CTLA-4 monotherapies (ROR025 = 0.57, IC025 = −1.21). The spectrum of arrhythmias induced by ICIs differed among therapeutic regimens. There was no significant difference in the onset time between monotherapy and combination regimen. Moreover, reports of ICI-associated arrhythmias were associated with other concurrent cardiotoxicity, including cardiac failure [ROR 2.61 (2.20–3.09)], coronary artery disorders [ROR 2.28 (1.83–2.85)], myocardial disorders [ROR 5.25 (4.44–6.22)], pericardial disorders [ROR 2.76 (2.09–3.64)] and cardiac valve disorders [ROR 3.21 (1.34–7.68)]. Conclusion: ICI monotherapy and combination therapy can lead to cardiac arrhythmias that can result in serious outcomes and tend to occur early. Our findings underscore the importance of early recognition and management of ICI-related cardiac arrhythmias.
Whereas, immune-related adverse events (irAEs) can affect multiple organ systems (Zhai et al., 2019;Hu et al., 2020;Mikami et al., 2021;Bomze et al., 2022), including the cardiovascular system (Salem et al., 2018;Ma et al., 2021). Due to its rarity, primary evidence regarding ICIs-associated cardiac arrhythmias is derived from case reports (Katsume et al., 2018;Bukamur et al., 2019;Prevel et al., 2020;Alhumaid et al., 2021;Savarapu et al., 2021) and clinical trials (Joseph et al., 2021), which have not systematically focused on ICI-induced arrhythmias. Cardiac arrhythmias associated with ICIs have been reported to occur in the setting of myocarditis (Katsume et al., 2018), which implies that ICI-related arrhythmias may be associated with concurrent cardiotoxicity. Besides, the overviewed relationship between arrhythmias and ICIs, the spectrum of potential signals, the factors related to fatality, as well as the clinical information of ICIassociated arrhythmias remain unknown.
In this pharmacovigilance study, we investigated the FDA's Adverse Event Reporting System (FAERS) to identify the association between arrhythmias and different ICI regimens, detect a comprehensive spectrum of 17 potential signals, and present comprehensive information (patient characterizations, prognosis outcomes, the onset time and the association between concurrent cardiotoxicities and ICI-associated arrhythmias).
Data source
We conducted a retrospective pharmacovigilance study based on data from January 2017 to June 2021 in the FAERS database. The FAERS database is a spontaneous reporting system (SRS), which collects adverse events (AEs) reports by health professionals, consumers, pharmaceutical manufacturers, patients, and other non-healthcare workers. OpenVigil FDA, a pharmacovigilance tool, was adapted to extract FAERS data using the openFDA API for accessing the FDA drug-event database with the additional openFDA duplicate detection functionality.
Frontiers in Pharmacology frontiersin.org
Statistical analysis
We used descriptive statistics to present the clinical characteristics of the ICI-associated arrhythmias. A comparison of categorical variables was made between fatal and non-fatal group using the chi-squared test. We used t test and nonparametric test to analyze the normally distributed and not normally distributed continuous variables respectively, and p < 0.05 was considered significant. Multivariate logistic regression was used to examine concurrent cardiotoxicities related to ICIrelated arrhythmias. The reporting odds ratio (ROR) with 95% confidence intervals (CIs) and Bayesian confidence propagation neural networks of information components (IC) were two specific indices to calculate disproportionality in pharmacovigilance (Noren et al., 2013;Zhai et al., 2019), which could detect potential signals in our investigation. The calculation formulas for ROR and IC are as follows: N expected : the number of case reports expected for the target drug AEs. N observed : the observed number of case reports for the target drug AEs. N drug : the total number of case reports for the target drug, regardless of adverse reactions. N event : the total number of case reports for the target AEs, regardless of drug. N total : the total number of case reports in the database. For IC, a significant signal was considered when the lower limit of the IC 95% confidence interval (IC 025 ) value was greater than zero (Bate et al., 1998;Noren et al., 2013). For ROR, a significant signal was considered when the lower end of the 95% credibility interval (ROR 025 ) exceeded 1, with at least 3 cases (Rothman et al., 2004). One of the two algorithms meeting the criteria should be considered as a positive signal of arrhythmia. All the data analysis was performed by SPSS 24.0 (SPSS Inc, Chicago, IL, United States).
As shown in Table 1, no significant differences were found in patient age, reporter and reporting year for fatal vs non-fatal reports. Use of different ICI regimens were similar in fatal vs. non-fatal ICI-related arrhythmia reports. There was a significant difference between fatal and non-fatal reports in tumor indications (p = 0.002), with the highest percentage of reported deaths (45.38%, 59/130) in digestive system patients. Notably, patient gender was statistically different between the two groups (p = 0.009), and the proportion of fatal reports in male patients was higher than that in female patients (69.85 vs. 25.93%). Concurrent cardiotoxicity was also different in fatal vs. non-fatal ICI-related arrhythmias reports (p = 0.001). Moreover, there was a significant difference in the reporting region between the two groups (p < 0.001), with the highest percentage of fatality occurring in South America (42.59%, 23/54).
Signal values related to different immunotherapy regimens
In general, ICIs were significantly associated with the reporting frequency of arrhythmias [ROR 1.20 (1.16-1.24), IC025 0.19] (
The signal spectrum of cardiac arrhythmias differs in immune therapies
The cardiac arrhythmia signal spectrum of different ICI strategies was shown in Table 3, where the IC 025 was regarded as an indicator. As shown in Table 3, ipilimumab plus nivolumab presented a broadest spectrum of cardiac arrhythmias AEs with 8 PTs detected as signals, ranging from cardiac arrest (IC 025 = 0.01) to sudden death (IC 025 = 1.90). For nivolumab, a total of 7 PTs as signals were observed, with signal values ranging from IC 025 = 0.19 (atrial fibrillation) to IC 025 = 1.34 (sudden death). There were 6 PTs both statistically associated with pembrolizumab and atezolizumab receiving. However, the drug with the least PTs were ipilimumab and cemiplimab, with no signal detected, followed by tremelimumab, with only one signal (tachycardia, IC 025 = 0.48) detected. Interestingly, both drugs (ipilimumab and tremelimumab) were all anti-CTLA-4 drugs, with no or only one reported AEs. Cemiplimab, as one of the three anti-PD-1 drugs, presented no signal due to the rare application.
Atrioventricular block complete, atrial fibrillation and sudden death were three overlapping PTs. Among these, sudden death was the most frequent PT, also detected as the second strongest signal (IC 025 = 1.90). Both atrioventricular block complete and atrial fibrillation were found significantly associated with nivolumab, pembrolizumab, atezolizumab, avelumab, and ipilimumab plus nivolumab, with atrioventricular block complete detected as the strongest signal in pembrolizumab (IC 025 = 2.00).
Associations between concurrent cardiotoxicity and ICI-associated arrhythmias
The associations between concurrent cardiotoxicity reports and specific arrhythmias reports under each HLT was diverse, with myocardial disorders having significantly elevated reporting of four specific arrhythmias in HLT level but cardiac valve disorders only increasing the risk of supraventricular arrhythmias.
Discussion
To the best of our knowledge, this is the first pharmacovigilance study on cardiac arrhythmias reports associated with ICIs based on the FAERS database. Our research presented a comprehensive description on cardiac arrhythmias associated to different ICI regimens, resulting in certain systematical and accurate conclusions. Importantly, our study detected a significant signal between cardiac arrhythmias and ICI therapy. Notably, our study revealed that immune-mediated arrhythmias were disproportionately more frequent reported in concurrent cardiotoxicity, which was concordant to what is observed in prior studies (Johnson et al., 2016;Salem et al., 2018;Herrmann, 2020;Nso et al., 2020;Baik et al., 2021;Stein-Merlob et al., 2021). The ICI-associated arrhythmias reports indicated a complicated clinical course, which prompted an evaluation for the presence of other cardiotoxicities. Similarly, all patients presenting with symptoms concerning for ICI-associated cardiotoxicity should have a 12-lead ECG to assess for arrhythmias.
In our study, concurrent cardiotoxicity increasing the reporting risk of ICI-related arrhythmias included cardiac failure [ROR 2.61 (2.20-3.09)], coronary artery disorders [ROR 2.28 (1.83-2.85)], myocardial disorders [ROR 5.25 Frontiers in Pharmacology frontiersin.org (Escudier et al., 2017;Mir et al., 2018;Stein-Merlob et al., 2021), presented consistent association with concurrent myocardial disorders and different correlation with other four kinds of concurrent cardiotoxicity. Previous studies suggested that reports of supraventricular arrhythmias following ICI therapy were associated with other concurrent irAEs (Salem et al., 2018) or T-lymphocyte-mediated inflammation in the sinoatrial and atrioventricular nodes (Johnson et al., 2016;Nso et al., 2020). Rate and rhythm disorders reports (including arrhythmia, bradycardia and tachycardia) could be seen in the setting of high degrees of conduction block. In patients receiving ICI therapy, ventricular arrhythmias and conduction block might be a result of the T-lymphocyte-mediated inflammatory infiltration into the myocardium (Johnson et al., 2016;Herrmann, 2020;Baik et al., 2021;Stein-Merlob et al., 2021). Whereas, precise mechanisms underlying of ICI-associated arrhythmias remain to be elucidated. It is unclear whether the increased reporting of arrhythmias following ICI therapy was due to concurrent cardiotoxicities versus due to ICI treatment itself. This study showed that ICI-associated arrhythmias were over-reported for anti-PD-1/PD-L1 vs anti-CTLA-4 monotherapy (Zhou et al., 2019;Ma et al., 2021). The nonsusceptibility of anti-CTLA-4 to pericarditis and myocarditis was due to the difference of disease-specific effects (Salem et al., 2018;Ma et al., 2021) and mechanism (Grabie et al., 2019), respectively. Due to the correlation between ICI related arrhythmias and the both two concurrent cardiotoxicities, the low reporting risk of anti-CTLA-4 in myocarditis and pericarditis may lead to low reporting risk of arrhythmias. Owing to a lack of studies on immunotherapy-induced arrhythmias, the rationale for no signal for anti-CTLA-4 drugs need to be further elucidated and explored.
Our study showed that most reports of ICI-associated arrhythmias occurred early after ICI initiation, with no significant difference between different ICI regimens. The median time to onset of arrhythmias reports associated with ICIs was 32 (IQR 10-109) days, and most reports (33.57%) appeared within the first 30 days after the initiation of ICI, which suggested the importance of cardiac monitoring during the higher-risk time window of 30 days. The median onset time reported with dual ICI therapies were 32.5 (IQR 12-96.25) days, and no earlier onset of arrhythmias was reported with ICI combination therapies than with ICI monotherapies. This finding was inconsistent with those of previously published case reports of ICI-associated cardiotoxicity, which reported that cardiotoxicity occurred earlier when two ICIs were combined (Zhou et al., 2019).
This study involves certain limitations that should be recognized. Firstly, resulting from the signal mining of FAERS database, our study may be associated with inevitable underreporting and selective reporting. Firstly, as a spontaneous reporting system (SRS), there are some Frontiers in Pharmacology frontiersin.org limitations inherent to FAERS database, including missing data, partial clinical features of AEs, and reporting bias (e.g., inevitable underreporting, selective reporting and the potential for the data to be misunderstood). Secondly, it is difficult to control for confounding factors such as history of arrhythmias or concomitant medications, both of which might influence the risk of cardiac arrhythmias. Lastly, due to lack of the number of patients exposed to ICIs without AEs, FAERS data can neither be used to calculate the incidence of an adverse reaction nor quantify adverse reaction signals based on the total number of AEs.
Conclusion
This study comprehensively evaluated the relationship between ICIs and cardiac arrhythmias based on the FAERS database, as well as exploring the associations between concurrent cardiotoxicity and ICI-related arrhythmias, which can assist medication monitoring, clinical practice, and future investigations. Further studies are needed to address the mechanisms underlying ICI-related arrhythmias and to validate the results in our study.
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
|
2022-11-04T16:09:38.563Z
|
2022-11-04T00:00:00.000Z
|
253269105
|
s2ag/train
|
Core functions and forms of Bright IDEAS: A multi-methods evaluation of the adoption of an evidence-based psychosocial training program through iterative adaptation
Background Despite efforts to widely disseminate interventions designed to increase access to quality supportive care to pediatric cancer patients and their families, many of these interventions fail to meet expectations once deployed in real-life clinical settings. This study identifies the functions and forms of Bright IDEAS: Problem-Solving Skills Training, an evidence based psychosocial intervention for caregivers of children recently diagnosed with cancer, to identify pragmatic program adaptations in its real-world clinical implementation. We compare intervention adoption before and after adaptations to the Bright IDEAS training program as part of a national training program designed to disseminate the intervention. Methods 209 pediatric psychosocial oncology practitioners representing 134 unique institutions were trained during 10 in-person 8-hour workshops (2015–2019). Functions and forms of Bright IDEAS were identified, and adaptations made to the training agenda and curriculum based on practitioner feedback following implementation in local institutions. Mixed method evaluation included longitudinal surveys at 6- and 12-months post training; and qualitative interviews among a subgroup of practitioners (N = 47) to understand and compare perspectives on intervention adoption and barriers to implementation before and after adaptations to the Bright IDEAS training program. The RE-AIM framework was used to guide dissemination evaluation. Results A total of four adaptations were tailored to the identified forms of the intervention: case studies; pre-training reading materials; training videos; and letters of institutional support from primary supervisor. Pre- and post-training adaptations to the Bright IDEAS training program were mapped to RE-AIM constructs. Quantitative findings demonstrate that adaptations appeared to improve adoption and usage overall. Conclusion This study provides insight into how contextual factors influence psychosocial practitioners' capacity to adopt, implement, and maintain Bright IDEAS in the clinical setting. This study demonstrates the use of real-time stakeholder feedback to guide intervention translation from research to practice settings.
|
v2
|
2022-11-05T06:17:05.922Z
|
2022-11-04T00:00:00.000Z
|
253302678
|
s2ag/train
|
The telomerase connection of the brain and its implications for neurodegenerative diseases.
Telomerase, consisting of the protein subunit TERT and RNA component TERC, is best known for maintaining and extending human telomeres, the ends of linear chromosomes, in tissues, where it is active, such as stem cells, germline cells, lymphocytes and endothelial cells. This function is considered as canonical. However, various non-canonical functions for the protein part TERT have been discovered. There are multiple such roles which can interfere with several signalling pathways, cancer development and many other processes. One of these non-canonical functions includes shuttling of the TERT protein out of the nucleus upon increased oxidative stress into the cytoplasm and organelles such as mitochondria. Mitochondrial TERT is able to protect cells from oxidative stress, DNA damage and apoptosis although the exact mechanisms are incompletely understood. Recently, a protective role for TERT was described in brain neurons. Here TERT is able to counteract effects of toxic neurodegenerative proteins via changes in gene expression, activation of neurotrophic factors as well as activation of protein degrading pathways such as autophagy. Protein degradation processes are prominently involved in degrading toxic proteins in the brain like amyloid-β, pathological tau and α-synuclein that are responsible for various neurodegenerative diseases. These new findings can have implications for the development of novel treatment strategies for neurodegenerative diseases. The current review summarises our knowledge on the role of the telomerase protein TERT in brain function, in particular, under the aspect of age-related neurodegenerative diseases. It also describes various strategies to increase TERT levels in the brain.
|
v2
|
2022-11-05T06:17:05.965Z
|
2022-11-04T00:00:00.000Z
|
253303149
|
s2ag/train
|
The role of the 47-kDa membrane lipoprotein of Treponema pallidum in promoting maturation of peripheral blood monocyte-derived dendritic cells without enhancing C-C chemokine receptor type 7-mediated dendritic cell migration.
BACKGROUND
The 47-kDa membrane lipoprotein (Tp47) is the most representative membrane protein of Treponema pallidum (T. pallidum). Dendritic cells (DCs) are the most potent professional antigen-presenting cells (APCs) that connect innate and acquired immunity. The regulatory role of Tp47 on DCs remains unclear.
OBJECTIVES
To evaluate the effects of Tp47 on DC maturation and migration, and research the changes of the main chemokine C-C chemokine receptor type 7 (CCR7) involved in DC migration.
MATERIAL AND METHODS
A transwell assay was applied to assess the migration of DCs. Cytokines (interleukin (IL)-6, IL-10, IL-12, and tumor necrosis factor alpha (TNF-α)) in the supernatants were measured using enzyme-linked immunosorbent assay (ELISA), and the expression of cell surface markers (CD80, CD86, CD40, and human leukocyte antigen (HLA)-DR) and CCR7 was assessed using flow cytometry. The expression of CCR7 in DCs was analyzed using quantitative real-time polymerase chain reaction (qRT-PCR).
RESULTS
The Tp47 promoted DC phenotypic maturation, such as increased CD40, CD80, CD86, and HLA-DR expression, as well as DC functional maturation, thus stimulating DCs to secrete inflammatory cytokines, including IL-6, IL-10, IL-12, and TNF-α. At the same time, Tp47 did not enhance DC migration and did not increase the expression of CCR7.
CONCLUSIONS
The Tp47 promoted the maturation of DCs while not enhancing CCR7-mediated DC migration ability. This may be one of the mechanisms by which T. pallidum escapes host immune clearance.
|
v2
|
2022-11-05T06:17:06.036Z
|
2022-11-04T00:00:00.000Z
|
253302694
|
s2ag/train
|
The SLC22A2 gene is a determinant of hematological toxicity of oxaliplatin in patients with colorectal cancer.
OBJECTIVE
To investigate the association between polymorphisms in the SLC22A2 gene and the hematological toxicity of oxaliplatin in colorectal cancer (CRC) patients receiving chemotherapy.
MATERIALS AND METHODS
A total of 81 patients with colon or rectal cancer were included in the study. The single nucleotide polymorphisms (SNPs) rs3127573, rs316019, and rs1869641 of the SLC22A2 gene were selected for genotyping using the polymerase chain reaction (PCR) and sequence analysis. Oxaliplatin-associated hematological toxicities were evaluated using the Common Toxicity Criteria for Adverse Events (CTCAE, Version 5.0).
RESULTS
The rs1869641 genotype was significantly associated with the occurrence of thrombocytopenia (p = 0.047), whereas the rs316019 genotype was significantly associated with severity of leucopenia and neutropenia (p = 0.004 and 0.001, respectively). The rs3127573 genotype was not associated with hematological toxicities arising during chemotherapy with oxaliplatin.
CONCLUSION
It is shown here, for the first time, that the rs316019 gene variant of the SLC22A2 gene may be associated with the hematological toxicity of oxaliplatin. Patients with genotype CA/AA of rs316019 are more likely to develop serious hematological adverse effects.
|
v2
|
2022-11-05T06:17:06.110Z
|
2022-11-04T00:00:00.000Z
|
253302195
|
s2ag/train
|
Human amniotic membrane as a multifunctional biomaterial: recent advances and applications.
The developing fetus is wrapped by a human amniotic membrane or amnion. Amnion is a promising human tissue allograft in clinical application because of its chemical composition, collagen-based, and mechanical properties of the extracellular matrix. In addition, amnion contains cells and growth factors; therefore, meets the essential parameters of tissue engineering. No donor morbidity, easy processing and storage, fewer ethical issue, anti-inflammatory, antioxidant, antibacterial, and non-immunogenic properties are other advantages of amnion usage. For these reasons, amnion can resolve some bottlenecks in the regenerative medicine issues such as tissue engineering and cell therapy. Over the last decades, biomedical applications of amnion have evolved from a simple sheet for skin or cornea repair to high-technology applications such as amnion nanocomposite, powder, or hydrogel for the regeneration of cartilage, muscle, tendon, and heart. Furthermore, amnion has anticancer as well as drug/cell delivery capacity. This review highlights various ancient and new applications of amnion in research and clinical applications, from regenerative medicine to cancer therapy, focusing on articles published during the last decade that also revealed information regarding amnion-based products. Challenges and future perspectives of the amnion in regenerative medicine are also discussed.
|
v2
|
2022-11-05T06:17:06.309Z
|
2022-11-04T00:00:00.000Z
|
253303077
|
s2ag/train
|
A 1,2,3-Triazole Derivative of Quinazoline Exhibits Antitumor Activity by Tethering RNF168 to SQSTM1/P62.
Quinazoline and its derivatives have drawn much attention in the development of potential antitumor agents. Here, we synthesized a series of 1,2,3-triazole derivatives of quinazoline at the C6 position and evaluated for their cytotoxic activity in various human cancer cell lines. We found that compound 5a was the most cytotoxic to HCT-116 cells (IC50, 0.36 μM). Target profiling found that 5a directly binds to both the autophagy-associated protein SQSTM1/P62 and the E3 ligase RNF168, promoting their interaction. Consistently, 5a treatment induces a decrease in RNF168-mediated H2A ubiquitination and compromises homologous recombination-mediated DNA repair, thus increasing the sensitivity of HCT-116 to X-ray radiation. Moreover, 5a suppressed xenografted tumor growth in mice in a dose-dependent manner. Taken together, the 1,2,3-triazole derivative of quinazoline 5a may serve as a novel compound for tumor therapy based on its role in promoting a P62/RNF168 interaction.
|
v2
|
2022-11-05T06:17:06.541Z
|
2022-11-04T00:00:00.000Z
|
253302708
|
s2orc/train
|
CT vs. bioluminescence: A comparison of imaging techniques for orthotopic prostate tumors in mice
Prostate cancer is one of the most diagnosed cancers in men in the United States. In mouse models, orthotopic tumors are favored for their biological relevance and simulation of growth in a microenvironment akin to that found in humans. However, to monitor the disease course, animal models require consistent and noninvasive surveillance. In vivo bioluminescent imaging has become a mainstay imaging modality due to its flexibility and ease of use. However, with some orthotopic prostate tumor models, bioluminescence fails to describe disease progression due to optical scattering and signal attenuation. CT scanning, in addition to its utility in human cancer diagnosis and surveillance, can be applied to mouse models with improved results. However, CT imaging has poor definition when imaging soft tissues and is not routinely used in prostate cancer models. Using an orthotopic prostate cancer model, our results demonstrate that, when compared to bioluminescent imaging, CT imaging correlates more closely to orthotopic prostate tumor growth in mice. Based on the data from this study, we conclude that CT imaging can be used as an alternative to the more commonly used bioluminescent imaging for measuring orthotopic prostate cancer growth over time.
Introduction
As the second leading cause of death in the United States, approximately 39.5% of people will be diagnosed with cancer at some point in their lifetime, highlighting the importance of developing novel diagnostics and therapeutics for this widespread disease [1]. Although research traditionally assays human cancer cells in vitro, mouse models allow for a more comprehensive study of pathogenesis inside a living subject [2]. Subcutaneous tumor models can be used to model tumor growth inside a living being, but orthotopically implanted tumors allow interaction with the relevant tumor microenvironment, as well as modeling natural metastasis [3].
One specific area of cancer research that benefits from orthotopic mouse models is prostate cancer. 1 in 8 men will be diagnosed with prostate cancer in their lifetime, and although many cases are indolent and prognosis is generally favorable, clinically aggressive prostate cancers are more likely to invade and metastasize regardless of treatment, resulting in worse outcomes [4]. Patients with distant disease have only a 31% five year survival rate, which is much lower when compared to the near-100% five year survival rate found in earlier detected, local disease [5]. Characterizing prostate cancer progression from primary to metastatic is vital to improving outcomes. Mouse models and orthotopic implantation of human cancer cells are especially important for studying prostate cancer, as it allows focused inquiry into the primary growth cascade that occurs in aggressive phenotypes [6]. Prostate cancer is also affected by its microenvironment; thus, studying the tumor within the confines of its natural anatomical position better mimics the human disease [7]. Unfortunately, monitoring orthotopic prostate tumor growth is difficult without sacrificing the animal. Imaging allows for a series of tumor measurements over time, providing valuable data on size and disease progression. Optical imaging modalities such as whole animal bioluminescent imaging are simple and widely used due to their specificity and versatility, as cells can either be transfected with genes for bioluminescent enzymes or tagged with a fluorescent probe [8]. One of the most widely used methods to monitor orthotopically implanted tumors uses cells transfected with luciferase, an enzyme that emits light when it interacts with its substrate, luciferin. The substrate is injected, allowing it to be imaged noninvasively, a technique which takes images and provides pixel intensities for a specific region of interest to estimate tumor growth [9]. However, any optical imaging modalities are subject to light scattering and light absorption. As such, optical imaging can be hindered by tissue depth and is generally limited to small animal rodent models [10].
The bioluminescent technique was developed, and its accuracy was reported, using subcutaneous tumor implants [11,12]. While bioluminescence has shown correlation to actual tumor volume in orthotopic models, it is more variable [12,13]. Accurately measuring tumors becomes challenging when characterizing orthotopic growth, particularly when tumors start to grow beyond normal anatomical scope, as they can grow to large sizes without impeding animal movement or viability. Further, larger tumor volumes can result in necrotic centers, alterations in blood supply causing hypoxia, and aberrant metabolism, all of which can lead to diminished luciferin concentrations and enzyme activity, decreasing fluorescent signals and falsely diminishing fluorescent intensity [14]. These problems have the likelihood to increase variability, which diminishes the utility of this technique to evaluate primary cancer growth and metastasis.
RM-1 cells, a mouse prostate carcinoma cell line, have mutated ras and myc oncogenes, resulting in androgen insensitivity and aggressivity [15]. RM-1 cells produce fast-growing, aggressive tumors when implanted in mice with a functioning immune system, providing a useful model for studying aggressive prostate cancer. LNCaP cells are an androgen sensitive human prostate cancer cell line well-known as a model for early prostate cancer progression [16]. Traditionally, long-term imaging of both hind flank and orthotopic prostate tumors is done using bioluminescent imaging due to its ease and convenience [17]. However, we were unable to transfect RM-1 cells with luciferase, which made bioluminescent imaging impossible. In addition, bioluminescent imaging was found to be inaccurate in measuring luciferase expressing LNCaP tumors.
Therefore, to monitor long-term tumor growth, we opted to use an alternate imaging method. Computed tomography (CT) imaging is used in humans to locate and characterize tumors, track invasion and metastasis, and plan therapeutic intervention and treatment [18]. In prostate cancer, CT is used for evaluating metastases, mostly by detecting enlarged lymph nodes, and for planning external beam radiation therapy to delineate potential healthy tissue involvement in patients' treatment [19]. For small animals, CT imaging provides a viable modality to assess a primary tumor's growth over time but is generally not employed in prostate cancer. CT imaging usually has poor contrast with soft tissues and is not used as a standard method to measure prostate tumor growth. We have found that we can use anatomical landmarks to reliably measure the prostate and prostate tumor. Thus, we hypothesize that CT tumor measurements correlate more closely to tumor size as compared to bioluminescent imaging in small animals with orthotopically implanted prostate tumors and that CT imaging can be used as an alternative method for measuring prostate tumor growth over time.
Iodine contrast dye enhances CT imaging and identification of prostate
Reading and interpreting CT scans requires knowledge of a subject's normal anatomy, so we validated the ability to accurately identify and image the margins of a normal prostate using iodine contrast with CT imaging in a deceased mouse (Fig 1). This provided a frame of reference to locate the prostate's position to other peritoneal structures such as the bladder and rectum that have pronounced contrast in CT imaging. The bladder is often filled with urine and the rectum/colon is filled with air providing distinct contrast as opposed to a solid tissue like the prostate. The organs and space surrounding the prostate are different densities so this region is uniquely arranged to allow for identification of these organs on CT despite being soft tissues.
Anatomical geography is a reliable method to locate prostate on CT
Once the prostate was located using a contrast agent, this method was applied in a live, breathing mouse after an operator was trained to find the prostate without the use of contrast. Although breathing causes imaging artifacts, we found that, especially when the mouse has a full bladder, the prostate was easily identified and was measurable within a few tenths of a millimeter (Fig 2). In Fig 2A, other anatomical landmarks are highlighted along with the prostate. The bladder is shown in yellow, prostate in red, seminal vesicles in green and the colon in blue in all three planes. In Fig 2B and 2C, the anatomical structural landmarks are unlabeled, and the prostate is denoted with the red line to highlight prostate dimensions in the different planes. We found the optimal conditions for CT imaging to be early morning due to mice eating and drinking during the dark cycle. In 6-8-week-old C57Bl/6J mice, normal prostate dimensions range between 4.6-5.3 mm in either dimension. For NU/J mice at 6-8 weeks of age, normal prostate dimensions range from 3.7-4.5 mm in either dimension.
CT can accurately measure tumor growth over time
RM-1 cells, although ideal for studying prostate cancer growth and metastasis in immune competent mice, were not able to be stably transfected to express luciferase. Therefore, an imaging modality other than bioluminescence was required to construct a tumor growth curve. Starting at 21 days post implantation, animals were imaged using CT regularly until most had entered an exponential growth phase (Fig 3A), at which point the tumors were removed. The final CT, at 46 days post-implantation measurement was compared to gross measurement of the excised tumor using calipers (Fig 3B). For 8 of 9 animals, the measurements used to calculate the volume of the tumors differed by less than 1mm (average 0.15mm), resulting in estimated volumes that differed by less than 100mm 3 . In one animal, the tumor was overestimated in two dimensions by more than one millimeter, which resulted in an overestimated volume of 255 mm 3 . When comparing the tumor volumes for all 9 mice using CT vs excised ex vivo measurements, there was no significant difference in average calculated tumor volume using these two methods ( Fig 3C). These data indicate that using CT imaging measurements were no different than external measurements made on excised tumors using calipers.
CT provides more accurate tumor growth measurements as compared to bioluminescent imaging in a prostate orthotopic model
To monitor the growth of tumors using optical imaging techniques, LNCaP cells were transduced to express the enzyme luciferase, which can be supplemented with the substrate d-luciferin to emit light (LNCaP-Luc). We then orthotopically implanted LNCaP-Luc prostate cancer cells into NU/J mice and tracked their growth using either bioluminescence imaging with IVIS ( Fig 4A) or CT imaging ( Fig 4B). Expected exponential growth is evident in almost all mice when tumors are measured by CT imaging (Fig 4B). In contrast, using bioluminescent imaging to estimate tumor growth over time resulted in erratic tumor growth curves (Fig 4A). A few representative tumor bearing mice from this cohort have been plotted to emphasize the difference of measurements between the two methods in Fig 4C. To provide a direct visual comparison of bioluminescence imaging, CT, and ex vivo measurements, a representative comparison of each imaging modality is shown in Fig 5A-5C. Orthotopic LNCaP tumor measurements at 8 weeks post-implantation obtained from CT correspond more closely to the ex vivo measurements as compared to the bioluminescence method.
To validate each measurement over the course of the experiment, cages of animals were randomly selected for termination to remove the tumors at 8-, 10-, and 12-weeks post- implantation. Tumor measurements were verified externally by caliper. Prior to sacrifice the tumor size was measured via CT or bioluminescence imaging. We found no statistical correlation between CT measurements and bioluminescence imaging measurements (Fig 6A). CT imaging measurements were significantly correlated to caliper measurements taken upon removal of the tumor (Fig 6B). Bioluminescence imaging measurements did not correlate significantly with caliper measurements taken after tumor removal ( Fig 6C). Thus, CT imaging of orthotopic Luc-LNCaP cells was a more accurate method to measure tumor growth over time as compared to bioluminescent imaging.
Discussion
CT imaging more accurately correlates to ex vivo tumor size when measuring orthotopically implanted prostate tumors when compared with bioluminescent imaging. In fact, bioluminescent imaging measurements were highly variable in our LNCaP model. Bioluminescence has proven to be a reliable method for many tumor models, including orthotopic models, but there are certain models where bioluminescence does not accurately measure a growing tumor [13]. Previous studies have shown a bioluminescent plateau, where signal detects a mass but underestimates total burden as size increases [20,21]. There are limitations to correlating total flux data to a growth curve, especially in internal organs, primarily due to optical scattering and interference. The farther bioluminescent rays travel through tissue, the more likely the signal will be lost. Since correlating bioluminescence data to size was first validated using flank tumors with minimal scatter and signal interference, it follows that the variability would diminish reliability and measurement accuracy. Indeed, this variable correlation has been found in some orthotopic tumor studies [13]. Additionally, signal detection requires either bioluminescent transfection or fluorescent tagging. In the case of luciferase transfection, the signal intensity depends on a variety of factors. A large tumor not only increases scattering, but adjusts to its microenvironment, resulting in alterations in blood supply, pH, and oxidative substrates, all key components that regulate oxidoreductase reactions, like luciferase [14]. Larger tumors may contain necrotic cores, aberrant blood supply, and adjustments in surrounding stroma that do not allow luciferin access, falsely diminishing signal and underestimating size [22,23]. Finally, even if bioluminescent transfection or fluorescent tagging is possible, it may not be desirable. Primary human cancer and metastases are neither transfectable nor taggable, and a diagnostic animal model meant to mimic human physiology benefits from using clinically relevant, translatable imaging methods.
CT allows for a three-dimensional picture to be composed, providing volumetric data, without probes or tags. While bioluminescence data can detect a mass with variable correlation to its true size, a CT scan can be used to accurately measure the tumor volume. CT's ability to detect metastasis, especially when combined with contrast or positron emission tomography, is more precise than bioluminescence by providing anatomical feedback to the extent of disease. Although our present study did not investigate metastases, machines such as micro-CTs, which provide high resolution CT images, have demonstrated competency when imaging lung and liver metastases in other cancer models, suggesting its utility when examining metastatic disease as well [24,25]. Alternatively, MRI gives good soft tissue contrast, but can be costly and time consuming to perform [26]. 3D ultrasound is another alternative method to measure prostate cancer growth over time but takes substantial amount of training to adequately perform these measurements [27].
3D volume measurements would likely have provided a higher correlation with excised tumor measurements. Although we only used 2-dimensional approximation in CT, by scanning the length and width of the tumor, we were able to measure based on the largest section and found it sufficient to provide an R 2 value of 0.9373 when compared to ex vivo tumor measurements. If quick, longitudinal imaging is the goal, such as it is with bioluminescent imaging, 2D approximations will adequately provide this information. The relative speed, ease of use, and the versatility of detecting bioluminescent and fluorescent signals all make bioluminescence an attractive option for quantifying tumor growth. IVIS can image five mice at once, which is fast and efficient when compared to CT's ability to image one animal at a time. However, as bioluminescent signal varies, one animal with particularly intense signal can drown out animals with less intense signal, requiring the operator to reposition animals. As CT imaging is also relatively quick (five minutes per mouse), the logistical advantage of IVIS over CT diminishes. A major concern for using CT over bioluminescence is the need for ionizing radiation, which, in mice, could potentially cause off-target effects [28]. Using low doses of radiation to gather CT images could cause long-term off-target effects in mice. However, the dose required for sub-millimeter resolution imaging is less than 1 centigray, which is far below the recommended dose of 76-80 gray for definitive prostate radiotherapy in humans or 5-7 gray whole body irradiation to result in mouse lethality [29,30]. These tumor bearing animals only live a month or two after irradiation, which will not provide them the time to potentially develop any of these side effects. Our data also supports that tumors are growing exponentially despite regular CT imaging, suggesting that repeated exposure to CT radiation is not sufficient to interfere with tumor growth studies.
Despite its comparable speed and superior accuracy, CT's biggest drawback when compared to IVIS is that it requires prior anatomical knowledge, especially without the use of labeled probes. In humans, reading CT scans requires years of medical training, and mouse anatomy is similarly challenging to learn and understand. However, as we have demonstrated, with practice and the use of exploratory techniques such as performing an initial experiment using contrast agents or browsing existing anatomical scans, it is possible for researchers and technicians to reliably image and identify structures. Further, with focused scans and consistent models, such as those used in prostate cancer, the necessary scope of anatomical knowledge narrows. As CT is versatile and applies to many dimensions of cancer research, it is not only feasible to learn to read CT scans, but also advantageous, as it allows researchers to gather higher quality data. To our knowledge this is one of the first studies that demonstrate the ability to use CT to accurately measure orthotopic prostate growth over time.
To conclude, we compared tumor growth measurements using bioluminescence and CT to quantify orthotopic prostate tumors. Our in vivo studies demonstrated that bioluminescence growth measurements are highly variable and do not reflect true tumor growth in the LNCaP orthotopically implanted model. Conversely, CT scans are highly accurate and precise at estimating true tumor size and dimensionality over time in two orthotopic prostate cancer models. CT scanning is also preferable when bioluminescent transfection or fluorescent tagging is not possible, like in the case of the RM-1 cell line. As CT scanning has become commonplace in detecting cancer clinically, adapting experimental models to reflect clinical approaches are important, especially when the alternatives maybe less precise.
Cell culture
RM-1 cells were obtained from Dr. Leah Cook (Department of Pathology and Microbiology, UNMC). The RM-1 cells were maintained in RPMI-1640 media supplemented with 10% FBS and 1% penicillin/streptomycin in a humidified incubator at 37˚C, 95% air and 5% CO2. LNCaP cells were purchased from ATCC (clone FGC cat. CRL-1740) and transduced with lentivirus expressing firefly luciferase (Genecopeia). LNCaP-Luc cells were maintained in RPMI-1640 media supplemented with 5% FBS and 1% penicillin/streptomycin in a humidified incubator at 37˚C, 95% air and 5% CO2. Cells were tested regularly for mycoplasma contamination.
Ethical approval of the study
This study was approved and performed under the institutional animal care and use committee at UNMC (20-019-03-FC). This study was carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Mice were sacrificed using CO2 exposure followed by cervical dislocation. Isoflurane was used for anesthesia for surgery and imaging of mice. Animals were treated with buprenorphine as an analgesic after surgery. Animals were euthanized immediately upon displaying signs of reduced ambulatory function or distress.
Experimental animals
6 to 8-week-old male C57BL/6J mice or athymic NU/J (Jackson Laboratories, Bar Harbor, ME) were used in this study. The mice were housed at the University of Nebraska Medical Center (UNMC), exposed to a 12 h light/dark cycle, and fed and watered ad libitum. For the contrast study, one mouse was used. For the RM1 CT experiment, nine mice were used. For the LnCAP CT vs. bioluminescence experiment, twenty mice were used.
Orthotopic implantation of tumor cells
Mice were anesthetized by continuous flow of 2-3% isoflurane with 1% oxygen using a mouse anesthesia machine. The lower abdomen was cleaned with alcohol and iodine and a one cm midline incision was made. To expose the prostate gland, the bladder was gently retracted. RM-1 cells (in C57Bl/6J) or LNCaP-Luc cells (in NU/J) were injected into the dorsal prostatic lobe in a total volume of 50 μL containing 2.0 × 10 6 cells in 50% Matrigel using a 30-gauge needle. The peritoneal muscle tissue was closed using absorbable catgut sutures (cat. 563B, Surgical Specialties, Tijuana, Mexico) and the skin was closed with wound clips (cat. 1111C15, Thomas Scientific, Swedesboro, NJ, USA). Buprenorphine (0.1 mg/kg, Reckitt Benckiser Healthcare, UK Ltd., Hull, UK) was administrated by intraperitoneal injection just after the surgery followed by three doses at 6, 24, and 48 h post-surgery as an analgesic. Sterile surgical procedures were maintained for the entire process. Wound clips were removed after ten days, and animals were monitored daily for re-opening of the incision site, infection, or distress.
CT imaging
Three weeks (RM-1) or six weeks (LNCaP-Luc) post implantation, and weekly thereafter, animals were CT imaged to monitor tumor growth. Animals were anesthetized by continuous flow of 2.7% isoflurane with oxygen. Cone beam CT images were acquired with the small animal radiation research platform, SARRP (XSTRAHL, Suwanee, GA). The animals are placed on a bed that rotates 360˚between a stationary x-ray source (providing 60kVp, 0.8mA continuously) and a flat panel amorphous silicon detector. Cone beam CT images were acquired at a 0.6 × 0.6 × 0.6 mm 3 voxel resolution.
For training purposes, a preliminary contrast-enhanced imaging study was performed on a single euthanized mouse to clearly identify the prostate. For contrast imaging, a euthanized mouse's prostate was exposed with a three cm midline incision. The peritoneum was disturbed but internal organs were left unmanipulated to preserve their anatomical location. Fifty microliters of Isovue-370 (Bracco Diagnostics, Singen, Germany), an organically bound iodine contrast agent, was injected directly into the prostate lobes. The peritoneum and muscle layers were gently sutured to restore any displaced organs. A CT was acquired immediately to so that minimal leakage of iodine would occur to distort the visualization of the prostate.
The prostate is connected directly to the bladder and seminal vesicles and lays on top of the colon when a mouse is in the supine position. These geographical landmarks (bladder, the seminal vesicles and colon) provide enough contrast to identify the prostate on all three planes of a CT scan. A CT operator was trained on both images with contrast agent and without contrast agent to reliably measure the prostate before tumor implantation. Once the tumor begins growing the prostate tumor is more easily identified and measured. Prostates of healthy animals were dissected and measured, and the volumes were calculated and compared to CT images to validate the accuracy of the CT measurements. The operator was trained on ten mice.
Tumor volume calculation
For both CT and ex vivo measurements with calipers, tumors were measured along two perpendicular axes usually in the transverse and sagittal (dorsal to ventral) planes. In CT imaging, the coronal plane was traversed to ensure the largest slice of the tumor was measured. The longest measurement was assigned to length (l) and the shorter measurement assigned to width (w). The volume was then estimated with the equation V = w 2� l/2 and is expressed in units of mm 3 .
Bioluminescent imaging
D-Luciferin potassium salt (100 mg/kg, PerkinElmer, cat. 122799, Waltham, MA, USA) was dissolved in PBS and then sterile filtered. LNCaP-Luc tumor bearing mice were injected intraperitoneally 15 minutes prior to imaging based on previous imaging studies (data not shown). Mice were anesthetized using continuous flow of 2.5% isoflurane with oxygen and placed in the Xenogen IVIS Spectrum bioluminescence imaging system (PerkinElmer, MA, USA). Luminescence was acquired for 1 second with the maximum field of view (FOV) of 24.5 cm, a bin size of 8 and lens aperture set to f1 (fully open for maximal photon collection). Images were analyzed using Living Image 4.5.1 software (Caliper Life Sciences, Hopkinton, MA, USA). Regions of interest (ROI) were determined visually to include the entire area demonstrating light output, and total flux was reported as a measurement of photons/sec.
Statistical analysis
To compare CT to ex vivo caliper measurements, a student's t-test was performed. We performed linear regression analysis and used Pearson correlation coefficients to evaluate the relationship between tumor growth measured by CT, bioluminescence imaging, and ex vivo measurements pairwise. Statistical significance was defined as p <0.05. All analysis was performed with Graph Pad Prism.
|
v2
|
2022-11-05T13:40:52.474Z
|
2022-11-04T00:00:00.000Z
|
253304569
|
s2orc/train
|
Combinatorial targeting of menin and the histone methyltransferase DOT1L as a novel therapeutic strategy for treatment of chemotherapy-resistant ovarian cancer
Background Ovarian cancer (OC) is characterized by a low response rate and high frequency of resistance development to currently available treatments. The therapeutic potential of histone methyltransferase DOT1L inhibitor in OC cells has been demonstrated, but optimal efficacy and safety of this targeted therapy approach still require improvement. We set forth to evaluate if this problem can be overcome by combinatorial targeting of this epigenetic modifier and menin, one of its functional partners in chromatin. Methods siRNA-mediated gene knock-down and pharmacological inhibition of menin, a key component of the MLL/SET1 complex and a fitness gene in OC cells, coupled to cell proliferation assays on a panel of high grade serous OC cell lines, including chemotherapy-sensitive and -resistant clones, were applied in order to evaluate how depletion or blockade of this enzyme influences growth and viability of OC cells. RNA sequencing was applied to identify menin target genes and pathways, and the effects of combined inhibition of menin and DOT1L on growth and transcriptome of these OC models were evaluated. Results Silencing and pharmacological inhibition of menin exert antiproliferative effects in all OC cells tested and, in PEO1 and PEO4 cells, a profound impact on transcriptome via down-regulation of cell cycle regulatory pathways, aryl hydrocarbon receptor, MYC and KRAS signalling. We demonstrated association of menin and DOT1L in OC cells and identified a subset of genes co-regulated by the two factors. Interestingly, co-treatment with DOT1L and menin pharmacological inhibitors exerts an additive effect on growth inhibition on chemotherapy-sensitive and -refractory OC cells mediated by transcriptome changes controlled by menin and DOT1L activities. Conclusion These results indicate that menin functionally cooperates with DOT1L in OC cells modulating transcription of genes involved in key cellular functions including, among others, cell proliferation and survival, that are strongly affected by combined inhibition of these two epigenetic regulators, suggesting that this may represent a novel therapeutic strategy for chemotherapy-resistant OCs. Trial registration NA; The manuscript does not contain clinical trials. Supplementary Information The online version contains supplementary material available at 10.1186/s12935-022-02740-6.
Introduction
Ovarian cancer (OC) is the leading cause of death from gynaecological malignancies, accounting for 313,959 new cases in 2020 and more than 200,000 victims [1]. Due to the lack of specific symptoms, OC is frequently diagnosed at advanced stages and is characterized by low response rate and the high frequency of resistance to current treatments [2]. These tumours are highly heterogeneous and classified into several subtypes, each characterized by distinct gene expression, epigenetic and mutational patterns, and, consequently, critically differing one from another. Among OCs, high-grade serous ovarian cancer (HGSOC) represents the most common, aggressive and lethal form of epithelial ovarian cancer [3], pointing out to an urgent need to identify novel therapeutic targets and treatment approaches for this tumour subtype.
Our group recently demonstrated that the histone methyltransferase DOT1L represents an effective therapeutic target in ERα-expressing OC cells where pharmacological blockade of this epigenetic enzyme induces inhibition of estrogen signalling [4]. Other studies highlighted ERα-independent DOT1L role in OC mediated by transcriptional regulation of cell cycle [5] and multi-drug resistance genes [6] and suggested DOT1L as a valuable prognostic biomarker in OC. Evaluation of therapeutic potential of DOT1L pharmacological inhibition has proved to be effective in treatment of multiple cancer types including ovarian, breast, prostate and other solid tumours (reviewed in [7]). However, DOT1L inhibitors use in clinics is still hampered by low tolerance and severe side effects caused by drug administration indicating to the need of therapeutic efficacy improvement for these epigenetic drugs [7]. Therefore, new therapies are being sought to allow the doses of DOT1L-targeting drugs to be reduced, including application of drug combination treatments.
Most of the proteins in a cell do not act alone but through physical interaction with multiple co-factors, form multiprotein complexes that govern cellular processes. Thus, targeting components of protein assemblies containing factors known to exert mitogenic activity in cancer cells may represent a plausible strategy for novel drug target discovery for treatment of this malignancy [8]. Moreover, simultaneous inhibition of activity of proteins belonging to the same molecular complex have potential to enhance targeted therapy effectiveness and safety using the sub-optimal drug concentrations [9].
Scaffold protein menin, encoded by multiple endocrine neoplasia 1 gene (MEN1) and known to be involved in histone modification and epigenetic gene regulation, has been shown to physically bind DOT1L in leukaemia and breast cancer cells [10], [11]. Moreover, menin interaction with MLL is involved in development of acute leukaemias with translocations of the MLL gene [12], [13], as well as of solid tumours, including castration-resistant prostate cancer [14] and hepatocellular carcinoma [15]. Importantly, menin-MLL inhibitors are currently being clinically tested for relapsed or refractory acute myeloid leukaemias treatment and therefore may soon be introduced in clinical practice for treatment of this disease, reducing time required for its approval for treatment of other cancer types.
In the present study we characterized the effects of menin silencing and its pharmacological inhibition on the proliferation of a panel of HGSOC cells and extensively investigated transcriptional changes induced by its depletion and blockade in chemotherapy-sensitive and resistant relapsed OC models. We showed here that menin targeting may be effective in reinforcement of DOT1L inhibition effect since their cooperative blockade in OC cells has additive antiproliferative effect.
Ovarian tissue and cell lines data analysis
Data for 264 ovarian cancers and 180 normal ovarian tissue samples were collected from web resources ICGC [16] and GTEx, respectively, and were used for comparative analysis of MEN1 mRNA expression level in cancer and normal tissues (derived from normal regions of the left and or/ right ovary of women that got traumatic injury, cerebrovascular, heart, liver, renal, respiratory or neurological diseases). CRISPR-Cas9 gene knock-down data were analysed using DepMap portal (https://score. depmap.sanger.ac.uk).
Transient small interfering RNA transfection
Reverse transfection of MEN1-specific siRNAs (IDs: s8682, s8683, and s8684), targeting different gene regions, and scrambled negative control (ID: s813), all purchased from Ambion (Thermo Fisher Scientific, Austin, TX, USA), was performed into all cell lines using Lipofectamine RNAiMAX (Invitrogen, Thermo Fisher Scientific), according to the manufacturer's instructions. First, 0,25 µl of 2 pM siRNA were diluted in 4,75 µl of Opti-MEM reduced serum medium (Gibco, Thermo Fisher Scientific), supplemented with 100 U/ml penicillin, 100 mg/mL streptomycin, and 250 ng/mL Amphotericin-B. Then, a mix of 0.3 µl of Lipofectamine RNAiMAX and 4,7 µl of Opti-MEM medium was added to diluted siRNA (1:1 v/v ratio), vortexed and incubated for 30 min at room temperature. At the end of incubation, 40 µl of Opti-MEM medium was added to the mix, and 50 µl of the final master mix was aliquoted in a well of 96-well TPP tissue culture plate (Merck). Finally, 15*10 3 cells resuspended in a 50 µl of Opti-MEM medium were plated on the top of siRNA-lipid complex. Cells were incubated for 4 h at 37 °C in the presence of 5% CO 2 . After that, 100 µl of RPMI (in case of PEO1, PEO4, PEO14 and OVCAR-3 cells) or DMEM (in case of Caov-3 cells), medium supplemented with 20 or 40% FBS were added to each well reaching a final volume of 200 µl. Each treatment was performed in sextuplicate. After 96 h, cells were harvested for protein, RNA extraction or used for measurement of cell viability by MTT assay.
Compounds
All cell lines were incubated with increasing concentrations of the menin-MLL interaction inhibitors MI-136 (S7815) or MI-503 (S7817), both purchased from Selleckchem (Houston, TX, USA) or vehicle (DMSO, Merck) as control. PEO1 and PEO4 cell lines were exposed to a combination of MI-136 and DOT1L inhibitor EPZ5676 (S7062, Selleckchem), at the indicated concentrations or vehicle (DMSO) as control.
Cell proliferation analysis
96 h post-transfection, cell proliferation was evaluated by MTT (3-(4,5-Dimethylthiazol-2-yr)-2,5-Diphenyltetrazolium Bromide)-based colorimetric assay (Invitrogen, Carlsbad, CA), according to the manufacturer instructions. Absorbance was measured by the VICTOR Multilabel Plate Reader (PerkinElmer, Milan, Italy) at 570 nm wavelength, subtracting background values read at 620 nm wavelengths for each sample. In order to assess the effects of the compounds on cell growth, 3*10 3 cells per well were seeded in sextuplicate in a 96-well TPP tissue culture plate. Cell proliferation was determined as described above after cells exposure to drugs, drugs combination or vehicle for 3, 6, 9, and 12 days.
Total protein extraction
For total protein extraction, cells were harvested, washed twice with ice-cold PBS-EDTA (0.5 mM EDTA), and lysed for 15 min on ice in high salt buffer (Tris-HCl pH 7.5 50 mM, NaCl 180 mM, NP40 0.15%, Glycerol 10%, MgCl2 1.5 mM, NaMo4 1mM, and NaF 0.5 mM). At the end of incubation, samples were centrifuged at 13.000 rpm for 30 min at 4 °C, and the supernatant containing total protein extract was diluted with two volumes of low salt buffer (Tris-HCl pH 7.5 50 mM, NP40 0.15%, Glycerol 10%). Protein concentration was measured using a Bradford protein assay.
Total RNA extraction
Total RNA was extracted from all cell lines using TRIzol reagent (Life Technologies, Carlsbad, CA), according to the manufacturer's instructions. In case of siRNA-transfected cells, total RNA samples both for RT-qPCR and RNA sequencing were extracted 96 h post transfection using 50 µl of TRIzol reagent per well of 96-well plate. All treatments were performed in sextuplicate, pooled together and further processed as unique sample following TRIzol extraction protocol until phase separation step after which the supernatant was collected and further purified using RNA Clean & Concentrator − 5 kit (Zymo Research, Irvine, CA, USA) according to the manufacturer's instructions. Two independent biological replicates were prepared for both MEN1-targeting siRNA and scramble control. RNA extraction from PEO1 and PEO4 cells treated with 3.2 µM EPZ5676, 0.8 µM MI-136, their combination, or vehicle was performed after nine-day cultivation of OC cells in the presence of the indicated compounds. Three independent biological replicates were prepared for all treatments and for validation experiments. Before use, RNA purity was assessed by NanoDrop™ 2000/2000c spectrophotometer (Thermo Fisher Scientific), whereas its concentration and integrity were measured using Qubit RNA assay kit and fluorimeter (Life Technologies) and Agilent 4200 Tapestation System (Agilent, Santa Clara, CA, USA), respectively.
RT-qPCR
One µg of total RNA was retro-transcribed using Sensi-FAST cDNA Synthesis Kit (Meridian Bioscience, Cincinnati, OH, United States). qPCR was carried out in triplicate, using 50 ng of cDNA, SensiFAST SYBR Lo-ROX qPCR mix (Meridian Bioscience) and LightCycler 480 II instrument (Roche, Basel, Switzerland). GAPDH was used as a reference gene for the normalization of target gene expression.
Drug combination analysis
Combenefit software [21] was used for drug combination analysis. The effect of increasing sub-optimal concentrations of MI-136 and EPZ5676 after 9 and 12 days of treatment was calculated by data procession with classical Loewe synergy model.
RNA sequencing and data analysis
For each sample, 1 µg of total RNA were used for rRNA depletion by using MGIEasy rRNA Depletion Kit (MGI, Shenzhen, Guangdong, China). For validation experiment. Then, the purified RNA was used as an input for indexed libraries preparation using MGIeasy RNA Directional Library Prep Kit V2.0 (MGI) according to manufacturer's instructions. Equimolar pools were prepared, circularized using MGIEasy Circularization Module (MGI) and subjected to DNA nanoballs generation followed by sequencing (paired-end, 2 × 100 cycles) on the DNBSEQ-G50RS (MGI). For validation experiments, indexed libraries were prepared starting from 1 µg total RNA according to Illumina Stranded Total RNA preparation kit (Illumina Inc., San Diego, CA, USA) and sequenced at a concentration of 1,7 pM on the NextSeq 500 platform (Illumina Inc.). In both cases FastQC tool (http://www.bioinformatics.babraham.ac.uk/projects/ fastqc) was used for quality control analysis on generated raw sequencing files (.fastq) followed by adapter trimming using Cutadapt v.4.0 [22]. STAR tool (v.2.7.5a) [23] with the standard parameters was used for alignment of reads on human genome (assembly hg38) considering present in GenCode Release 36 (GRCh38.p12) genes. FeatureCounts [24] and DESeq2 [25] were utilized for expressed genes quantification and differentially expressed genes quantification, respectively. Genes characterized by |Fold-Change|≥Q1 (first quartile) and padj ≤ 0.05 were considered as differentially expressed.
Functional analyses and pathway analyses
Investigation of modulated signaling pathways was performed using Ingenuity Pathway Analysis (IPA) and Gene Set Enrichment Analysis (GSEA) [26] tools using the lists of differentially expressed genes. GOplot [27] was used for the Circos plot generation.
Statistical analyses
Statistical analyses were performed using R (version 4.0.2). Error bars represent means ± SD of independent replicates. Comparisons between two groups were performed by Student's t-test. Values of p ≤ 0.05 were considered as statistically significant.
Evaluation of menin expression in OCs
Lately, a pro-proliferative role of menin in multiple cancer types including leukaemia, hepatocellular carcinoma, breast, endometrial and prostate cancers has been demonstrated [13], [28], [29], [30], [31]. Moreover, elevated expression of gene encoding for menin protein (MEN1) has been found in endometrioid and breast cancers, [30], [11], whereas in hepatocellular and prostate cancer MEN1 overexpression has been correlated to disease progression [28], [31]. In order to investigate menin role in OC development, we analysed RNA-seq data from ICGC [16] and GTEx expression profiling datasets and found out that MEN1 expression level is increased in ovarian tumours in comparison to normal tissues (Fig. 1A). We next explored whether menin expression is essential for proliferation and survival of OC cells by interrogating the data of genome-wide CRISPR-Cas9 loss-of-function genetic screen [32], first to place MEN1 in the context of fitness and non-fitness genes in OC (Fig. 1B), then revealing that menin expression is required for optimal survival for most of analysed OC cell lines (Fig. 1C). Moreover, further analysis demonstrated that cell models isolated from metastatic OC display a statistically significant elevated dependence from menin expression compared to the ones derived from primary tumours (Fig. 1D). Altogether, these results indicate a possible involvement of menin in OC development and progression.
We next adopted, as experimental model, a panel of five cell lines isolated from HGSOC tumours, representing the most frequent, aggressive, and lethal form of OC [33]. Among selected cell lines PEO14 and Caov-3 were isolated from patient before treatment, PEO1 -after the first disease recurrence, whereas OVCAR-3 and PEO4 cells are derived from relapsed tumours established after the acquisition of chemo-resistance. These cell models are characterized by typical HGSOC genetic background since all of them carry pathogenic mutations in TP53, a hallmark genetic aberration found in approximately 96% of such tumours [34], [35]. Moreover, PEO1 cell line bears pathogenic BRCA2 mutation and therefore represents an OC tumour type defective for double stranded DNA repair, known to be associated with hereditary breast and ovarian cancer [36]. Investigation of MEN1 mRNA and menin protein expression in selected cell lines revealed heterogeneous expression of this molecule with the highest expression level of both of them in PEO1 and PEO4 cells, that we adopted as principal models within the manuscript (Fig. S1A). Interestingly, according to the data of genome-wide CRISPR-Cas9 loss-of-function genetic screen [32], MEN1 resulted to be a key fitness gene in these cells, indicating that menin represents a plausible therapeutic target in this biological context (Fig. 1C, D).
Menin silencing induces proliferation reduction and transcriptome deregulation in OC cells
To evaluate menin effect on tumorigenic potential of OC cells, siRNA-mediated MEN1 knock-down coupled to MTT assay was applied to selected experimental models. Exponentially growing OC cells were transiently transfected with three siRNAs targeting different regions of menin-encoding transcript together with scramble siRNA as negative control. Knock down effect on MEN1 mRNA, menin protein expression and cell proliferation were evaluated by qPCR, western blot and MTT assay, [32] in all OC cell lines. Box plots showing MEN1 gene essentiality score according to [32] in all OC cell lines (C) and in cell lines derived from different tumor types (D) (** p ≤ 0.005) respectively, 96 hours after transfection ( Fig. 2A-C, Additional file 1: Fig. S1B-D). Out of three tested siRNAs, two successfully inhibited mRNA and protein expression in all tested cell lines ( Fig. 2A, B, Additional file 1: Fig. S1B, C). Menin depletion inhibited proliferation of all cell lines (Fig. 2C, Additional file 1: Fig. S1D), except for the primary tumour-derived Caov-3 cells, confirming results obtained by CRISPR-Cas9 screening according to which loss of MEN1 expression does not exert significant impact on Caov-3 cell proliferation (Fig. 1C, D).
Since menin is involved in gene expression control [37], we then focused our attention on analysis of transcriptome changes induced by MEN1 knock-down in PEO1 and PEO4 cells representing relapsed and chemotherapyresistant OC tumours and chosen while displaying the highest expression of menin protein. To this aim, from the three MEN1-targeting siRNAs, siRNA #3 was chosen, as far as its transfection induced maximum decrease of menin protein expression in all tested cell lines. As shown in Fig. 2D, menin depletion induced profound Table S1A and S1B) were found to be deregulated in PEO1 and PEO4 cells respectively (|fold-change| ≥ first quartile (Q1), padj ≤ 0.05). Comparison of transcriptome changes between the two cell lines revealed significant similarity since more than a half of genes, differentially expressed in PEO4 cells, were concordantly deregulated in PEO1 (Fig. 2D). Analysis of affected processes by IPA revealed concordant activation of synaptogenesis, integrin and sumoylation pathways and inhibition of aryl hydrocarbon receptor (AhR) and cell cycle-related signalling in both OC cell lines (Fig. 2E). These results were further confirmed and extended by interpretation of gene expression data with GSEA method that highlighted downregulation of MYC and DNA repair-related signalling pathways (Fig. 2F). The results obtained from these two analyses, which are based on different database searches and complement to each other, highlighted common features (Additional file 3: Table S1A and S1B) allowing us to suggest that MEN1 silencing induces inhibition of cell proliferation by suppression of genes involved in cell cycle regulation. To validate the results obtained here, we employed additional transcriptome sequencing experiments, starting from biological replicate of both cell lines (PEO1 and PEO4) and using a different sequencing technology and platform. As reported in Fig. S2A the correlation between analysed samples resulted very high and differentially expressed genes involved in aryl hydrocarbon receptor (AhR), integrin and cell cycle-related signalling pathways confirmed (Additional file 1: Fig. S2B-D).
Menin pharmacological inhibition negatively affects OC cell proliferation through deregulation of gene expression
Recent investigations demonstrated that pharmacological blockade of menin activity represents an effective strategy for treatment of acute leukaemia and a subset of solid tumours (reviewed in [38]). Importantly, a number of compounds, inhibiting menin activity via disruption of menin interaction with one of its functional partners MLL [39] have been synthesized and some of them are currently being tested in clinical trials [38]. Evaluation of menin blockade effects on the proliferation of the two selected models PEO1 and PEO4 (Fig. 3A) and of the additional OC cell lines (Additional file 1: Fig. S3) revealed that MI-136 treatment induces a profound dosedependent reduction of cell proliferation. This result was confirmed also for another menin pharmacological inhibitor -MI-503 (Additional file 1: Fig. S4), a compound developed on the basis of MI-136 and characterized by modified molecular scaffold [13].
We next explored transcriptome changes induced by menin pharmacological blockade in PEO1 and PEO4 cells, previously used for analysis of MEN1 silencing effect on gene expression. As shown in Figs. 3B 9 dayslong menin inhibition with MI-136 induced a profound effect on PEO4 transcriptome with 6331 (|fold-change| ≥ Q1, padj < 0.05, Additional file 2: Table S1D) deregulated genes and lower impact on PEO1 mRNA profile (1463, |fold-change| ≥ Q1, padj < 0.05, Additional file 2: Table S1C). 186 and 353 genes were concordantly downand up-regulated respectively in both cell lines (Fig. 3B). Moreover, comparison of menin inhibition-induced changes with the ones observed upon MEN1 knockdown revealed that 16 down-and 66 up-regulated genes, summarized in Fig. 3C, respond in the same way both to menin blockade and silencing in PEO1 and PEO4 cells in statistically significant manner in at least three comparisons indicating that expression of these genes may be regulated by menin-MLL complex. GSEA analysis performed on this subset of differentially expressed transcripts revealed concordant upregulation of KRAS signalling in both cell lines (Fig. 3D), whereas IPA analysis highlighted downregulation of AhR signalling and tendency for upregulation of integrin signalling pathways (Fig. 3E). Interestingly, these pathways were found to be concordantly deregulated also upon MEN1 silencing (Fig. 2E), suggesting that menin-MLL interaction mediates AhR and integrin signalling pathways activity.
Combinatorial pharmacological inhibition of menin and DOT1L exerts synergetic effect on proliferation and transcriptome changes in OC cells
Recently, evidences of complementary activities of menin and DOT1L inhibitors in NPM1-mutant and MLLrearranged leukaemia have been demonstrated [40], [10]. Moreover, the same result was obtained for estrogen receptor-positive breast cancer cells where synergic effect of pharmacological blockade of these two proteins on the proliferation has been shown and it was highlighted that menin represents DOT1L and ERα co-factor of in this cancer type [11]. Considering the profound effect of DOT1L inhibition on proliferation of PEO1 and PEO4 cells that has been evaluated in our previously published study where we demonstrated that treatment of OC cells with DOT1L inhibitors EPZ004777, EPZ5676 and SGC0946 induced a concentration-dependent inhibition of cell growth in PEO1 and PEO4 cells [4], we investigated whether the interaction between the two proteins occurs also in this cancer type and to estimate the effect of their simultaneous inhibition on proliferation of OC cells. Menin immunoprecipitation performed on nuclear extracts from PEO4 cells, having the higher expression of menin compared to the other selected cell lines, confirmed the presence of DOT1L among coimmunoprecipitated proteins (Additional file 1: Fig. S5). To further confirm the association of these two proteins at chromatin level, we compared the changes induced by MI-136 and EPZ5676 in OC cells (Additional file 2: Tables S1E, S1F) and found out that out of 186 concordantly down-and 353 up-regulated upon MI-136 treatment genes, common for two OC cell lines, 47 and 151 were concordantly down-and up-regulated, respectively also upon EPZ5676 treatment in both cell lines. Similar result was obtained also upon comparison of transcriptional changes induced by EPZ5676 treatment with the effect of MEN1 silencing that revealed 34 downand 127 up-regulated transcripts, common for both of OC tumours that expresses both proteins. Next, we tested the effect of a combination of MI-136 with DOT1L inhibitor EPZ5676 that currently undergoes clinical trials (reviewed in [7]). We demonstrated that administration of these two drugs exert an additive effect on the growth of PEO1 cells and on proliferation of chemotherapy-resistant PEO4 cells while using suboptimal concentrations of both drugs (Fig. 4B), with a week synergy observable on the growth of the PEO4 cells upon MI-136 treatment (Fig. 4B, right panel).
To further investigate the molecular mechanisms underlying the interplay of MI-136 and EPZ5676 in OC cells, we analysed transcriptome changes induced by their administration for 9 days, a time point at which the additive effect of the two drugs was already evident (Additional file 1: Fig. S6). We found that also in this case differential expression profiles of the two cell lines (in total 3446 and 3389 deregulated transcripts (Additional file 2: Tables S1G, S1H) with |fold-change| ≥ Q1, padj ≤ 0.05 for PEO1 and PEO4, respectively) share significant portion of concordantly deregulated mRNAs that comprised 813 down-and 1332 up-regulated transcripts (Fig. 4C). Further comparison of transcriptional changes induced by combination of MI-136 and EPZ5676 with expression changes induced by these two drugs alone allowed us to determine a subgroup of genes (90 up-and 53 down-regulated) characterized by enhanced deregulation in the presence of drugs combination respect to the single drug-induced expression change (Fig. 4D). Evaluation of the effect on signalling pathway activity induced by the two drugs by IPA and GSEA revealed that AhR, integrin signalling, KRAS, MYC and cell cycle-related pathways, previously found to be affected upon menin silencing or blockade, were affected also in these case (Fig. 4E, F). Moreover, such pathways as IL-8, ovarian cancer and estrogen receptor signalling pathways, previously described to be deregulated upon inhibition of DOT1L alone [4] or in combination with menin pharmacological blockade in other biological contexts [11] were also influenced. Altogether, these results indicate that combinatorial treatment of OC cells with menin and DOT1L inhibitors induces an additive antiproliferative effect on the cell growth, driven by pronounced deregulation of gene expression and signalling pathways.
Discussion
OC is the most lethal gynaecologic malignancy since more than 75% of affected women are diagnosed at an advanced stage of the disease and less than one-half of patients survive for more than five years after diagnosis. It embraces a heterogenous group of malignancies with different characteristics such as aetiology, molecular biology and others. HGSOC represents the deadliest OC tumour subtype since most patients develop disseminated disease already by the time of diagnosis [41]. Despite standard therapy, that includes cytoreductive surgery followed by platinum-paclitaxel chemotherapy initially gives good response, relapse occurs within two years in around 70-80% of HGSOC patients and eventually almost all recurrent tumours develop chemoresistance, leaving limited therapeutic options for further treatment [42]. Thus, the discovery of novel therapeutic targets including single molecules or signalling pathways, involved in drug resistance and metastasis has become the main focus of the on-going research, representing a promising approach for treatment of these deadly tumours. Among them, epigenetic modulators represent the most promising class of druggable targets due to their potential to effectively reverse transcriptional and epigenetic abnormalities induced by their aberrant activity [43].
In this study we report that targeting MEN1 geneencoded protein menin, known as an important DOT1L cofactor in MLL-rearranged leukaemia and antiestrogen therapy-resistant breast cancer cells [10], [7] may represent an effective therapeutic approach against OC. Starting from the observation of increased MEN1 mRNA expression in OC cells and its relevant essentiality for survival of metastasis-derived cell lines, we demonstrated that siRNA-mediated depletion of MEN1 expression exerts an antiproliferative effect on the growth of OC cells, that may be explained by inhibition of cell cycle-related signalling and DNA repair obtained by RNA sequencing. Importantly, MEN1 depletion induced inhibition of activity of MYC proto-oncogene, known to be amplified in around 64% of OC tumours [44] and activated in over half of human cancers including OC [45]. Similar results have previously been observed in fibrosarcoma and liver cancer cells where menin was characterized as a critical cofactor for MYC-mediated transcription, that promotes growth of tumours with deregulated MYC expression [46]. Moreover, in androgen receptor-dependent prostate cancer cells menin is involved in MYC-mediated activation of androgen receptor transcription [47], indicating a cooperative action of the two proteins. Activity of another transcription factor -AhR, was also found to be suppressed upon MEN1 silencing, whose role in controlling proliferation, migration, and tumour cell invasion has not been extensively determined yet, however there are some indications for its tumour-promoting role in OC since AhR nuclear localization has been associated with worth outcome for OC patients [48] and evidences of AhR receptor role in promotion of cell growth, stemness and metastatic potential of OC [49].
We next demonstrated that pharmacological inhibition of menin-MLL interaction exerts dose-dependent antiproliferative effect on the growth of OC and that also in this case inhibition of AhR signalling may be in charge of the observed effect, indicating the possibility of cooperative regulation of this signalling pathways by menin-MLL complex. Interestingly, we found that MI-136 treatment induced downregulation of KRAS signalling, whose constitutive activation is typically associated with low-grade ovarian carcinoma, endometrioid and mucinous ovarian tumours since these OC subtypes usually bear KRAS activating mutations [50]. However, recent study demonstrated that targeting RAS signalling in HGSOC cells with ADT-006, a small molecular inhibitor of RAS-effector interactions, displayed high sensitivity to this compound in vitro [51], indicating an antitumoral effect of KRAS signalling inhibition also in these cells and implying that antiproliferative effect of menin-MLL interaction blockade may be mediated by this signal transduction pathway. Finally, the investigation of the relationship between menin and DOT1L confirmed their functional interplay in OC since nuclear interaction of the two proteins was observed and comparison of transcription profile changes induced by menin or DOT1L pharmacological blockade revealed a subset of commonly deregulated genes. Supporting this result, we demonstrated that simultaneous administration of DOT1L and menin small molecule inhibitors have an additive antiproliferative effect on chemotherapy-sensitive and -refractory relapsed OC cells, and that combinatorial treatment of OC cells with sub-optimal doses of the two drugs induced profound effect on OC transcriptome that included cell cycle, MYC, AhR and KRAS signalling found to be deregulated upon MEN1 silencing or menin pharmacological blockade.
Conclusion
The results obtained in this study suggest that menin functionally cooperates with DOT1L in OC cells modulating transcriptome changes involved in key cellular functions including cell proliferation. Combinatorial blockade of DOT1L and menin activity show an additive effect on OC cell growth, thus, we assume that the inhibition of these two epigenetic regulators may represent a worth exploring approach to improve the therapy, response and survival of OC patients. However, it has to be noted that our results provide a starting point in the investigation of DOT1L and menin interplay in OC and further research are needed to determine which of the here discovered functional processes underly the observed antiproliferative effect. Moreover, the mutational status of the cell lines used in this study need to be considered and further validations of DOT1L and menin inhibitors on patient-derived OC models and xenografts are necessary to confirm the beneficial effect of the drug combination in clinically relevant cancer models.
|
v2
|
2022-11-06T06:17:53.465Z
|
2022-11-04T00:00:00.000Z
|
253351492
|
s2ag/train
|
Multispecific CAR T Cells Deprive Lymphomas of Escape via Antigen Loss.
Chimeric antigen receptor (CAR) modified T cell therapy has transformed the management of relapsed/refractory B cell malignancies. Despite high overall response rates, relapse post CAR T treatment remains a clinical challenge. Loss of target antigen, specifically CD19, is one well-defined mechanism of disease relapse. The mechanism of CD19 loss and which patients are at higher risk of CD19 loss remain poorly understood. To overcome CD19 loss, CARs targeting multiple antigens are being tested in clinical trials. CD19/20 and CD19/22 bispecific CARs demonstrate cytotoxicity against CD19-negative cells in preclinical studies. These CARs have also shown efficacy, safety, and a relatively low rate of CD19-negative relapse in phase I trials. These small studies suggest that multispecific CAR T cells can deprive lymphomas of escape via antigen loss. However, the selection of an ideal target, the right CAR construct, and whether these multispecific CARs can induce long-term remissions are still under investigation. Expected final online publication date for the Annual Review of Medicine, Volume 74 is January 2023. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.
|
v2
|
2022-11-06T06:17:53.830Z
|
2022-11-04T00:00:00.000Z
|
253360339
|
s2orc/train
|
Admission Rates, Time Trends, Risk Factors, and Outcomes of Ischemic and Hemorrhagic Stroke From German Nationwide Data
Background and Objectives In the past decade, there have been major improvements in the control of risk factors, acute stroke therapies, and rehabilitation after the availability of high-quality evidence and guidelines on best practices in the acute phase. In this changing landscape, we aimed to investigate the stroke admission rates, time trends, risk factors, and outcomes during the period of 2014–2019 using German nationwide data. Methods We obtained data of all acute stroke hospitalizations by the Federal Statistical Office. All hospitalized cases of adults (age 18 years or older) with acute stroke from the years 2014–2019 were analyzed regarding time trends, risk factors, treatments, morbidity, and in-hospital mortality according to stroke subtype (all-cause/ischemic/hemorrhagic). Results Between 2014 and 2019, overall stroke hospitalizations in adults (median age = 76 years, [IQR: 65–83 years]) initially increased from 306,425 in 2014 to peak at 318,849 in 2017 before falling to again to 312,692 in 2019, whereas percentage stroke hospitalizations that resulted in death remained stable during this period at 8.5% in 2014 and 8.6% in 2019. In a multivariate model of 1,882,930 cases, the strongest predictors of in-hospital stroke mortality were hemorrhagic subtype (adjusted OR [aOR] = 3.06, 95% CI 3.02–3.10; p < 0.001), cancer (aOR = 2.11, 2.06–2.16; p < 0.001), congestive heart failure (aOR = 1.70, 1.67–1.73; p < 0.001), and lower extremity arterial disease (aOR = 1.76, 1.67–1.84; p < 0.001). Discussion Despite recent advances in acute stroke care over the past decade, the percentage of stroke hospitalizations resulting in death remained unchanged. Further research is needed to determine how best to optimize stroke care pathways for multimorbid patients.
Stroke is the second leading cause of death and a major cause of disability worldwide. 1 The global burden of stroke has increased dramatically over the past 30 years in part due to population growth and aging, but also as a result of rising risk factor prevalence including obesity, sedentary lifestyle, hypertension, high alcohol consumption, and chronic kidney disease (CKD). 2 Lower-and middle-income countries continue to account for a large share of this burden.
There have been major developments in stroke care over the past decade with advances in prevention, acute therapies, and neurorehabilitation. 3 Several landmark trials for both ischemic and hemorrhagic stroke were published during this period that led to the widespread implementation of evidence-based interventions. These include mechanical thrombectomy, 4,5 an extended time window for thrombolysis using MRI-based or perfusion-based imaging, 6,7 and intensive blood pressure lowering for intracerebral hemorrhage. 8 However, it remains unclear whether this progress has translated into improved admission rates, risk factor profiles, and survival for both hemorrhagic and ischemic stroke in the general population. Some studies have observed decline in stroke admissions and mortality, [9][10][11][12] but up-to-date, countryspecific admissions and survival data by stroke-subtype are scarce, and the most recent Global Burden of Disease study highlights the absence of original, good-quality stroke epidemiologic studies for most countries. 1 Regular monitoring of these trends by stroke subtype is also necessary to monitor the disease burden at a population level. 13 In Germany, all hospitals are required by federal law to transfer data on all in-patient hospitalizations to the Institute for the Hospital Remuneration System (Institut fur das Entgeltsystem im Krankenhaus, InEK; Siegburg, Germany; gdrg.de) since 2002. The Federal Statistical Office has made a large part of this data set available for scientific purposes. Based on these nationwide data from the years 2014 to 2019, we analyzed recent temporal trends in admission rates, risk factors, and mortality after ischemic and hemorrhagic stroke.
Methods
A diagnosis and procedure-related remuneration system (German Diagnosis Related Groups, G-DRG system) was introduced for all in-hospitalization services in Germany in 2003, enabling accurate and comprehensive acquisition of defined cases of illness. 14 Consistent documentation and billing are achieved by the use of detailed mandatory coding guidelines Owing to federal law, all hospitals are required to transfer patient data on diagnoses, comorbidities, medical services, or procedures and procedure-related complications and to health insurance to receive reimbursement. Afterward, these data were collected by the Institute for the Hospital Remuneration System (InEK) for calculation of the current DRG-system and by the Federal Statistical Office.
Data Source
The Research Data Centers of the Federal Statistical Office and the Statistical Offices of the Laender (Statistisches Bundesamt, DESTATIS; destatis.de) provided data for the years 2014-2019 for analysis with respect to risk factors, in-hospital outcomes, and time trends related to acute stroke. The database contains all in-patient treated patients on a case base per year, except for treatments in psychiatric or psychosomatic units. 15 We excluded medical care provided by officebased specialists with special admitting. Because of data privacy protection, all subgroups less than 6 cases were excluded from the analysis. There was only remote access to the anonymized original data. A statistical analysis program written in SAS (SAS 9.2: SAS Institute Inc., Cary, NC, USA) was executed by the Research Data Center.
Standard Protocol Approvals, Registrations, and Patient Consents
The data presented here were examined as part of the Gen-derVasc research project. This project was approved by the Ethics Committee of the Landesaerztekammer Westfalen-Lippe and the Medical Faculty of the University of Muenster (No 2019-21-f-S).
Diagnoses and Procedure Codes
Diagnoses, patient characteristics, comorbidities, and stroke complications are uniformly coded according to the International Statistical Classification of Diseases and Related Health Problems, 10th revision, German modification (ICD-10-GM) (eTable 1 in the supplement, links.lww.com/WNL/ C325). Similarly, endovascular and surgical procedures are uniformly coded according to the German Operation and Procedure Classification (OPS) (see eTable 2 in the supplement, links.lww.com/WNL/C325). Coding guidelines and annual adaptation by the German Institute for Medical Documentation and Information (Cologne, Germany) ensure uniform documentation. All hospitalized adult patients with an acute stroke (age 18 years or older, diagnosis code: ICD I60-64) as their principal diagnosis were included in the analysis.
A stroke was defined as a syndrome of rapidly developing clinical symptoms and/or signs of focal, and at times global (applied to patients in deep coma and to those with subarachnoid hemorrhage), loss of brain function, with symptoms lasting more than 24 hours or leading to death, with no apparent cause other than that of vascular origin. 16 Stroke subtype classification (ischemic or hemorrhagic) corresponded with the ICD-10-GM codes (see eTable 1 in the supplement, links.lww.com/WNL/C325).
Statistical Analysis
Frequencies are given as case number per 100,000 population based on the German population of the respective year. Proportions of hospitalizations are case numbers per 100,000 of total in-hospital cases. Mortality displays the percentage of in-hospital deaths within a designated subgroup. The analysis comprises all in-patient treated acute stroke cases in Germany and does not represent a subgroup or sample. The Mantel-Haenszel χ 2 test was performed to evaluate time trends of categorical variables. Trends over time for continuous variables were analyzed using Spearman correlation analysis. Logistic regression analysis was used to assess the relationship between various demographic factors, vascular risk factors, treatments, and in-hospital stroke mortality, with the final model adjusting for the following covariates: age, sex, hypertension, dyslipidemia, diabetes mellitus, obesity, dementia, smoking, previous stroke, atrial fibrillation/flutter, lower extremity arterial disease (LEAD), cardiovascular disease, congestive heart failure, chronic kidney disease, cancer, carotid interventions, decompressive craniectomy, extracranial/ intracranial hematoma evacuations, mechanical thrombectomy, and intravenous (IV) thrombolysis treatment. All analyses were fully explorative, and all p-values were interpreted accordingly, that is, all p-values are two-sided, and pvalues <0.05 were considered statistically noticeable. Statistical analysis was performed using SAS software (SAS 9.3: SAS Institute Inc., Cary, NC, USA), IBM SPSS statistics software (IBM SPSS Statistics for Windows, version 28.0. Armonk, NY: IBM Corp) and the web-based statistics software VassarStats (vassarstats.net; R. Lowry).
Data Availability
The data used in this study cannot be made available in the manuscript, the supplemental files, or in a public repository because of German data protection laws ("Bundesdatenschutzgesetz", BDSG). Generally, access to data of statutory health insurance funds for research purposes is possible only under the conditions defined in German Social Law (SGB V § 287). Requests for data access can be sent as a formal proposal specifying the recipient and purpose of the data transfer to the appropriate data protection agency. Access to the data used in this study can only be provided to external parties under the conditions of the cooperation contract of this research project and after written approval by the sickness fund.
Results
Overall, between 2014 and 2019, there were a total of 1,882,930 hospitalizations coded as acute stroke in Germany, corresponding to an average 1.62% of all in-patient admissions. The frequency of hospitalized stroke cases remained relatively constant from 377.38 per 100,000 population in 2014 to 375.98 in 2019. Table 1 shows the absolute ischemic and hemorrhagic stroke admission rates for this period according to age group and year. The absolute number of ischemic stroke admissions peaked at 323.28 per 100,000 population in 2016 and then fell steadily to 317.75 in 2019. The ICH admission rate also had its nadir in 2019 at 58.23 per 100,000 population. During this time period, there were decreases in the stroke admission rates in the 40-49 and 70-79 year age groups while rates increased for those aged 60-69 and 80-89 years (Table 1 and Figure 1).
Summary statistics for baseline demographic and clinical characteristics according to stroke subtype during this time period are shown in Table 2. Patients presenting with ischemic stroke were slightly older than those with a hemorrhagic stroke (median age 76 vs 74 years). In total, 905,667 (48.1%) of all patients were female with comparable rates between the subtypes. Hypertension was the most prevalent comorbidity, affecting 74.4% of all acute stroke patients. Comorbidities were generally more prevalent in ischemic stroke patients apart from cancer, which was more common in ICH (3.5 vs 2.9%). Time trends of demographic characteristics and vascular risk factors remained stable throughout the study period for both ischemic and hemorrhagic stroke subtypes (see eTables 3 and 4 in the supplement, links.lww.com/WNL/ C325).
The therapies received and the complications and outcomes experienced by those hospitalized with acute stroke from 2014 to 2019 are summarized in eTable 5, links.lww.com/ WNL/C325 in the supplement and detailed according to year and subtypes in Tables 3 and 4. During this period, 27,097 (1.7%) and 16,479 (1%) of ischemic stroke patients were treated with carotid endarterectomy and stenting, respectively. The rate of carotid stenting increased steadily from 0.7% in 2014 to 1.3% in 2019 (p < 0.001) ( Table 3). Craniectomy rates remained stable throughout this period. Mechanical thrombectomy rates progressively increased from 1.9% in 2014 to 6.1% in 2019 (p < 0.001). Similarly, IV thrombolysis rates rose from 11.2% in 2014 to 14.1% in 2019 (p < 0.001). In line with these changes, there was also a trend toward greater bleeding and hemorrhagic transformation events (1.3%-1.6% and 2.8%-5.1% from 2014 to 2019, both p < 0.001). However, there was no change in the requirement for blood transfusions. There was a small rise in the incidence of acute kidney injury (AKI) (1.4%-2.7%; p < 0.001). Associated health care costs also rose during this period to peak at €7,634.42 in 2019 (p < 0.001).
Among patients hospitalized with acute hemorrhagic stroke, the rate of neurosurgical interventions remained stable over the 6-year period at approximately 1.9% and 10% for craniectomy and evacuation procedures, respectively (Table 4 and eTable 5 in the supplement, links.lww.com/WNL/ C325). 6.6% of patients had other bleeding events or required a blood transfusion. The rate of AKI or need for renal replacement therapy increased from 2.5% in 2014 to 3.7% in 2019 (p < 0.001), higher than that of those with ischemic stroke (2.5%; p < 0.001). 22% of patients with ICH required mechanical ventilation in comparison with 5% of those with ischemic stroke (p < 0.001). The mean length of hospital stay was also greater for patients with ICH at 15.4 days in comparison with 12 days for ischemic stroke patients (p < 0.001). The mean charges per case was also nearly double that of their ischemic stroke counterparts at €11,397.59.
Discussion
We present nationwide data regarding the rates of hospitalized stroke in Germany from 2014 to 2019 along with subtype-specific demographics, clinical characteristics, complications, and outcomes. Overall, there were stable rates of stroke hospitalization during this time period with a small decline observed in recent years. Use of IV thrombolysis and mechanical thrombectomy progressively increased during this period along a parallel increase in the rate of hemorrhagic transformation. However, despite greater therapeutic intervention, in-hospital stroke mortality rates did not improve over this period, with hemorrhagic subtype and several comorbidities including CHF, LEAD, and cancer all predictive of a greater risk of death.
These national data show small decreases in rates of hospitalized stroke for both ischemic and hemorrhagic subtypes in Germany in 2018 and 2019. Our results are in keeping with reductions in stroke hospitalization and severity that have been observed in other developed countries including Israel, 17 Norway, 18 and the United Kingdom. 19 These changes may reflect improved control of cardiovascular risk factors, the availability of high-equality evidence and guidelines on best practices in primary and secondary stroke prevention, 20,21 and more proactive implementation of stroke prevention clinics.
Better hypertension control and improved antihypertensive use have also been reported in Germany during this time period. 22 Reassuringly, in contrast to some previous population-based studies, we did not find an increasing stroke admission rate in younger age groups. [23][24][25] This discrepancy may relate to the more recent time period of this study because those studies that previously reported increases in "young stroke" were generally comparing the 1990s with the early 2000 period, after which this apparent increase appeared to plateau. 24,26 Although one may have expected greater reductions in hospitalization rates by the end of the period, improvements in diagnosis over time could have resulted in more accurate diagnoses in later years, thus affecting rates.
Few studies have looked at recent temporal trends in patient characteristics among acute stroke patients. 27 Overall, the number, median age, and risk factor profile among stroke patients remained stable between 2014 and 2019 in Germany. Ischemic stroke patients were more multimorbid with a greater prevalence of hypertension, atrial fibrillation, coronary heart disease, CKD, diabetes mellitus, dyslipidemia, and dementia compared with hemorrhagic stroke patients. A premorbid diagnosis of cancer was more prevalent in the latter group. These risk factor differences between the major stroke subtypes appear to be consistent over time and with prior literature. 28 Multimorbidity burden in stroke has important implications because it has been associated with an increasing risk of allcause mortality, with mortality risk more than doubled for those with at least 5 morbidities compared with those with none. 29 For therapies and complications, there was nearly a three-fold increase in mechanical thrombectomy rates along with a smaller increase in IVT numbers during this time period. However, this was accompanied by increased hemorrhagic transformation. These changes align with the growing evidence during this period that thrombolytic therapy and mechanical thrombectomy among eligible patients were associated with less death and dependency despite a relative increase in symptomatic ICH in those treated with IVT. 5,30 Decompressive craniectomy rates remained stable for both ischemic and hemorrhagic stroke patients. The hemorrhagic stroke group experienced a greater number of poststroke complications with higher rates of bleeding and need for blood transfusion, AKI and requirement for RRT, sepsis, and cardiac arrest. They also had more complex care needs including requirement for mechanical ventilation, longer lengths of stay, and associated higher health care costs. These complications in the acute phase have been shown to worsen brain injury and have been linked to worse outcomes after hemorrhagic stroke. 31,32 In-hospital mortality rates were higher among patients with ICH compared with those with ischemic stroke. Despite recent improvements in evidence-based therapies 33 and postacute stroke management including stroke units 34 in the past decade, mortality rates did not improve for either subtype during the study time period. For ischemic stroke, these findings somewhat conflict with the improved short-term fatality rates observed in other population-based studies. 12,35 However, this discrepancy may be attributable to the relatively narrow time window of this study or to the crude trends that we present in comparison with some of the former studies. Because only aggregate data were available, we were unable to produce age-standardized rates, where improvement for ischemic stroke mortality has been previously described. 12 In contrast, the high mortality rates that we report after ICH are similarly well-described in other contemporary cohorts, 36,37 likely related to the lack of effective therapeutic options 31 and the high complication rate as described above. 32 There have been no successful phase 3 clinical trials for acute interventions in ICH to date. Therefore, admission to dedicated stroke or neurocritical care units with intensive prevention or early detection of complications has been proposed as the best approach to improve ICH outcomes. 38 Cancer, LEAD, and CHF were all strongly predictive of inhospital mortality in both groups of patients. The association of cancer with poor survival after stroke has been well-described, 39 particularly in the setting of cryptogenic mechanisms, lung cancer, and systemic metastasis. 40 These factors are closely linked to hypercoagulability 41 and correction of coagulopathy in the setting of cancer appears to be protective for stroke survival. 42 Similarly, the presence of LEAD is known to increase the risk of secondary cardiovascular death and events, 43 including both ischemic and hemorrhagic stroke, 44 particularly if severe-that is, baseline ankle-brachial index (ABI) < 0.60 or history of amputation. 45 There is emerging evidence for the role of dual hemostatic blockade (aspirin and low-dose rivaroxaban) in such patients with polyvascular disease as it was associated with a lower risk of the composite outcome of cardiovascular death, stroke, or myocardial infarction in the COMPASS trial. 46 Stroke in patients with CHF has also previously been reported to be associated with worse outcomes and higher mortality, 47 underscoring the need for better prevention and treatment strategies in this vulnerable group.
This present study has many strengths resulting from the rigorous methodology applied to nationwide data, including consistency across cases. This is also a large and comprehensive sample of unselected participants who are also unbiased by insurance status, geographic location, or providing institution, and, therefore, the observed trends are high applicable to daily practice. High-quality, countryspecific epidemiologic data such as these are an important way to monitor the global stroke burden. 13 However, our study has several limitations. First, our analysis of administrative data may be subject to miscoding of ICD codes and despite rigorous abstraction processes, variation in reporting may still be present. Second, data acquisition was case-based rather than patient-based and, therefore, there could have led to a certain number of double-counted patients, for example, by hospital transfer. Third, hospitalbased data has some inherent limitations including selection bias related to the indication for admission which is influenced by stroke severity and prognosis, and variable access to specialized stroke units and center-specific services. As such, hospitalization data may not reflect the full burden of stroke in the population. However, because Germany has a health care system with a very high admittance rate for transient ischemic attack and minor stroke, this national data set is very inclusive and representative of all cases of acute stroke. Fourth, because data were limited to ICD codes, no medication, laboratory or imaging data were available for analysis. For this reason, we were also unable to capture discharge disposition information (i.e., home, rehabilitation hospital, or nursing facility) which would have enhanced the richness of this report. Finally, we were unable to distinguish first from recurrent strokes because of the case-based nature of the data.
In conclusion, we have presented a nationwide, administrative data analysis of all patients hospitalized with acute stroke in Germany from 2014 to 2019. There were favorable trends toward improved rates of hospitalization in recent years although these were still accompanied by high in-hospital mortality rates, particularly for patients with ICH, who require research prioritization to identify novel therapeutic tools and optimal strategies to prevent and treat early complications. Our study also highlighted the impact of multimorbidity on short-term stroke survival and the need to personalize management strategies to these comorbidities.
Study Funding
The study was conducted within the framework of the Gen-derVasc project (Gender-specific real care situation of patients with arteriosclerotic cardiovascular diseases) funded by The Federal Joint Committee, Innovation Committee (G-BA, Innovationsfond, number 01VSF18051). GenderVasc is a cooperation project with the AOK Research Institute of the AOK (WIdO). Dearbhla Kelly is an Atlantic Fellow for Equity in Brain Health at the Global Brain Health Institute (GBHI) and is supported with funding from GBHI, Alzheimer's Association, and Alzheimer's Society (GBHI ALZ UK-
|
v2
|
2022-11-06T16:20:01.138Z
|
2022-11-04T00:00:00.000Z
|
253355386
|
s2orc/train
|
Calcium Electroporation for Management of Cutaneous Metastases in HER2-Positive Breast Cancer: A Case Report
We report a case of successful treatment of cutaneous metastases in HER2-positive breast cancer with calcium electroporation (CaEP), in addition to trastuzumab, over a period of 5 years. CaEP is performed in local or general anesthesia, by injecting calcium chloride intratumorally and then electroporating cells in the area. Using a handheld needle electrode, a series of short, high-voltage electric pulses are delivered, which transiently permeabilizes cell membranes, causing toxic intracellular calcium levels. The treatment causes cancer cell death, while normal cells are less affected, making the treatment useful for local management of cutaneous lesions. This case presents a 66-year-old female, who had mastectomy surgery followed by adjuvant chemo- and radiotherapy for an ER-negative, HER2-positive breast cancer on her right side in 2003, and a mastectomy followed by endocrine therapy for an ER-positive, HER2 normal breast cancer on her left side in 2006. In 2015, the patient presented local cutaneous recurrence of the ER-negative, HER2-positive breast cancer. The patient was treated with trastuzumab alone, trastuzumab emtansine (TDM1), and a combination of trastuzumab and CaEP. TDM1 was found to have a slightly better effect on the cutaneous metastases than trastuzumab, but the side effects of TDM1 were not acceptable to the patient. The combination of continuous HER2-inhibition and intermittent CaEP, when needed, has been effective in keeping the cutaneous metastases under control for 5 years, and presumably more tolerable for the patient than chemotherapy. An interesting finding was local sparing of calcium electroporated skin from new recurrences, otherwise seen in the general area, which could be a sign of local immunity. This warrants further studies investigating local immunomodulation following CaEP. The patient reported appreciation of a treatment option without chemotherapy, and satisfaction with the outcome of the combination of HER2 inhibition and CaEP treatment. CaEP treatment is currently phase II treatment, and mechanisms and possible applications still need investigation. This novel anticancer treatment could potentially benefit many patients, due to its efficacy, low cost, and accessibility. This case provides observations, which may inspire future trials with CaEP for skin metastases of HER2-positive breast cancer.
Cutaneous Metastases from Breast Cancer -The Clinical Challenge
Breast cancer mortality is declining, partly due to earlier detection and partly due to improved treatment options [1]. Medical therapies, in the form of chemotherapy, endocrine therapy, and targeted agents, are often used either separately or combined, in the effort to keep systemic disease under control [2]. However, development of cutaneous metastases or locoregional recurrence in patients with breast cancer, even when antineoplastic treatment can control internal disease, is a relatively common phenomenon. A possible explanation could be lower drug concentration in the skin, especially in an area previously treated with radiotherapy, as radiation is known to cause hypoperfusion [3]. Breast cancer is more prone to metastasize to the skin than other cancers, and local treatments are often more efficient for management of cutaneous metastases than systemic treatments [4,5]. The incidence of cutaneous metastases from breast cancer ranges from about 5% to as much as 30% in latestage disease [4,6]. As such, cutaneous metastases often occur when distant metastases are widely spread in internal organs, and can be associated with poor survival rates [6].
Calcium Electroporation for Cutaneous Metastases
Treatment with calcium electroporation (CaEP) is a novel anticancer treatment for malignant cutaneous and subcutaneous tumors -both primary tumors and metastases [7,8]. In local anesthesia (or general anesthesia if indicated), calcium chloride is injected intratumorally (with a margin of normal tissue), which causes a high extracellular concentration of calcium. Immediately after injection, the area is electroporated with a handheld needle electrode inserted through the tumor tissue. Once inserted, a series of high-voltage electric pulses are sequentially applied (shown in Fig. 1) [9,10].
The cell membranes in the electroporated area transiently permeabilize, which allows diffusion of calcium ions into the cell. The cell membrane will reseal shortly after, which traps the calcium inside the cytosol and creates a toxic intracellular calcium concentration [9]. In normal cells, homeostasis is restored as calcium is chelated, compartmentalized into mitochondria and endoplasmic reticulum, or extruded through ATPases and exchangers [11]. In cancer cells, mutations may alter these mechanisms, and calcium regulation becomes less efficient, whereby the toxic intracellular calcium levels may induce cell death [12]. As normal cells are less sensitive to CaEP than cancer cells, the treatment can spare healthy tissues around treated tumors [12]. Previous clinical studies that have investigated CaEP treatment of cutaneous metastases from breast cancer ( Table 1) have shown that CaEP is a feasible, effective, and safe treatment option with only mild adverse events.
In one study, 6 breast cancer patients and a single patient with melanoma were treated. The patients had an objective response rate of 72% and the study proved CaEP non-inferior to electrochemotherapy (ECT) after 6 months of follow-up [8]. Another study, that included 1 breast cancer patient and 6 melanoma patients showed an objective response rate of 33%, with CaEP non-inferior to ECT after 1-year follow-up [13]. Calcium electroporation procedure. a The target area is defined as the area which is clinically visualized as tumor + a margin. b When performing the local anesthesia, it is important to provide coverage of the margin as well as a zone around the margin so that the electrodes may be inserted without discomfort. Adding further local anesthetic below the tumor can also be helpful, in particular when treating larger lesions. c The calculated i.t. dose of calcium is injected into the tumor in a parallel fashion throughout the tumor. The margin area is then supplemented with calcium until the calcium is evenly distributed throughout the entire target area. d The electrode is inserted so that needles reach just beyond the deepest part of the tumor, and a pulse sequence is applied. The electrode can then be subsequently inserted in a systematic way to cover the entire tumor volume, as indicated in (e). As can be seen, the treatment area then covers the tumor with treatment margin. From Vissing et al. [10], with permission.
Case Presentation
This paper describes a case of recurring breast cancer metastases managed with intermittent CaEP treatments. A 66-year-old woman presented with an ER-negative and HER2positive breast cancer in her right breast and lymph node metastases in 2003. The patient had a mastectomy on the right side with radical axillary lymph node dissection, followed by radiation therapy and chemotherapy ( Fig. 2 provides a treatment overview). A tumor was discovered in the patient's left breast in 2006, with an ER-positive and HER2 normal phenotype, which was managed with mastectomy and endocrine therapy. Disseminated disease was uncovered during routine follow-up in 2010. Biopsies verified a metastatic tumor in the periclavicular connective tissues, and an axillary lymph node metastasis with ER-negative and HER2-positive phenotypes, compatible with the initial tumor of the right breast. Standard treatment with chemotherapy and HER2-targeted antibodies (vinorelbine and trastuzumab) was initiated. After 11 months of chemotherapy, the metastatic tumors had fully remitted, and the treating oncologist discontinued vinorelbine, while treatment with trastuzumab was continued.
Heart failure is a well-known side effect of trastuzumab [14], and multigated acquisition scans were performed regularly. In 2015, her left ventricular ejection fraction (LVEF) had dropped to 45%, and trastuzumab treatment was paused.
A couple of months later, the patient presented with a pale red rash around the mastectomy scar on the right side. Although clinically described as nonsuspicious, a biopsy proved cutaneous metastases, concordant with the primary ER-negative and HER2-positive cancer of the right breast. The patient started treatment with trastuzumab emtansine (TDM1).
After 10 weeks of TDM1 treatment, the metastatic area on the chest had fully remitted, and it was recommended to continue treatment with a lower dose due to side effects, including fatigue and nausea. However, the patient decided to end the TDM1 treatment due to side effects.
A recurrence reappeared in the area after 5 months without treatment. The patient did not wish to resume chemotherapy, but agreed to resume trastuzumab treatment. Trastuzumab induced remission of the cutaneous recurrence, and over the course of the next 6 months, the appearance waxed and waned following trastuzumab treatments.
In 2016, the patient was referred to receive ECT due to the presence of cutaneous lesions. However, there was an ongoing study using CaEP for cutaneous metastases with promising results (Falk et al. [8]; see Table 1), and the patient requested CaEP instead of ECT and was treated outside protocol (CaEP-1). The target area measured 5.5 × 5.5 cm and was situated on the upper medial quadrant of the right chest (Fig. 3). After 11 weeks, the initial crusted wound had healed leaving a scar with no sign of residual tumor. The patient received trastuzumab treatment concurrently without side effects.
In 2018, new metastases appeared adjacent to the area treated with CaEP. The patient requested retreatment with CaEP for symptom relief from pruritus. The patient initially requested CaEP in order to avoid chemotherapy. As the patient had received CaEP once, standard treatment ECT was agreed upon with the patient, to see if this would give a longer response. The patient received ECT on four skin metastases. The metastases partly remitted, but the itching sensation persisted, and eventually, there was progression. Trastuzumab was paused concurrently due to a low LVEF. Since the cutaneous metastases kept growing, the patient resumed trastuzumab after a month, despite a declining LVEF. The patient never had clinical symptoms of heart failure, but began prophylactic medication with an angiotensin-II receptor-blocker and beta-blocker.
Four months after ECT, the patient was retreated with CaEP on part of the affected area (CaEP-2), upon her request. A larger area on the chest was treated two months later in general anesthesia (CaEP-3), covering an area of 6 × 8 cm (Fig. 4). In the meantime, the patient switched from trastuzumab treatment to TDM1, due to progression in the skin. Plastic surgeons offered the patient surgical removal and reconstruction of the affected skin, but she declined. After 2 months of TDM1 therapy, oncocardiologists recommended a dose reduction, due to declining LVEF. Instead, the patient requested substitution of TDM1 with trastuzumab due to side effects.
In 2019, cutaneous metastases reappeared, now involving the first area treated with CaEP, otherwise without recurrence for 3 years (Fig. 4). After a few months of observation, the area was retreated with CaEP with complete response (CaEP-4). Recurrence emerged after 6 months, and the area was retreated (CaEP-5). After about 6 months, recurrence appeared just outside the previously treated area. Shortly after, a routine follow-up CT scan showed a suspicious lymph node in the left lung, and the patient resumed TDM1 therapy (subsequently reduced to a 60% dose, due to side effects with fatigue and local infections). After observation for 6 months, the nonresponsive skin recurrence was treated with CaEP twice in late 2020 (CaEP-6 and CaEP-7), with complete response and pruritus relief. The patient had to reduce her TDM1 treatment dose further to 50%, due to severe fatigue and bleeding from the mucosa.
In August 2021, a more nodular skin recurrence was treated (CaEP-8). The patient discontinued TDM1 after 18 series, because of side effects and continuous development of cutaneous metastases. As the metastases were now her only sign of residual disease, she went back to receiving trastuzumab. One of the smaller cutaneous elements kept presenting with itching, which the patient recognized as metastatic activity, and the single tumor was therefore treated again shortly after (CaEP-9). As of September 2021, after 5 years with a combination treatment of continuous HER2 inhibition and CaEP when needed, the patient had no other manifestation of breast cancer disease, than recurrent skin metastases (Fig. 3).
Calcium Electroporation -The Procedure
The CaEP treatments were performed in local anesthesia using a Cliniporator pulse generator (IGEA, Italy), which delivers eight 0.1 ms pulses with an amplitude of 1 kV/cm and a frequency of 1 Hz (CaEP-3 was performed under general anesthesia, due to the size of the target area, and discomfort during the previous treatment). Calcium chloride 220 mmol/L was injected intratumorally, and linear electrodes were used for pulse application. A schematic overview of the procedure and a response to treatment are shown in Figures 1 and 4, respectively.
The ECT treatment was performed in local anesthesia using the Cliniporator pulse generator. A total of 1.8 mL bleomycin 1,000 IU was injected intratumorally and a linear electrode was used for pulse application.
Patient Experience
The patient claimed to appreciate a treatment option without chemotherapy. The patient described the CaEP treatment as mild and tolerable. She experienced slight discomfort during pulse delivery, and some soreness lasting a day after treatment, that resembled muscle pain experienced after exercising. The symptoms that followed CaEP and ECT treatment were similar. The patient expressed preference of CaEP treatment for her cutaneous disease over ECT, as the ECT did not relieve pruritus, and she experienced that remission from ECT had shorter duration. She felt comfortable monitoring progression of the disease, and requesting treatment when needed. The patient would recommend CaEP to others in a similar situation.
Discussion
Cutaneous metastases can be distressing for patients, causing symptoms such as pruritus, pain, and often ulceration of the skin [10,15]. Sometimes, progression in skin is treated with systemic agents, which target known immune markers such as HER2.
Researchers have previously hypothesized that TDM1 treatment is more effective than trastuzumab alone because the combination of trastuzumab and chemotherapy agents works through different mechanisms, thereby enhancing responses [5]. When treating metastases from HER2-positive breast cancers, CaEP could induce a local response, potentially enhanced when combined with trastuzumab.
Despite TDM1 being slightly more efficient than trastuzumab alone, TDM1 side effects were unacceptable to this patient. The cutaneous metastases have been successfully managed for 5 years with a combination of trastuzumab and CaEP, indicating that this could be an efficient treatment for metastases of HER2-positive breast cancer.
The long-term effects of CaEP are results of the toxic levels of cytoplasmic calcium inducing necrosis in the cancer cells, mainly in correlation with ATP depletion. A preclinical study has shown that CaEP has antivascular effects, which also play a part in the overall antitumor response [7,16]. In the two clinical studies, with 6-and 12-month follow-up, respectively, high response rates were seen, warranting studies with a longer follow-up period, both to investigate recurrence rate and duration of response, as well as to investigate the cumulative effect of successive CaEP treatments.
Studies have suggested that CaEP can lead to an immune response [8,17]. In this case, we observed that a previously treated area stayed clear of recurrence, even when new metastases arose outside the treated area. In a preclinical trial from 2017, mice were treated with CaEP and then rechallenged with either the same tumor cell type or a different tumor cell type, showing protective immunity toward the same tumor cell type, as well as an increase in immune-specific proteins and proinflammatory cytokines [17].
A clinical study from 2018 showed an indication of a possible systemic immune response, as a patient withmelanoma had complete remission of both treated and untreated tumors after CaEP [8]. Possible immune responses related to CaEP are described in a few cases, yet the specific mechanisms are unknown. Future studies should further investigate possible effects of CaEP on both local and systemic immunity.
As ECT is currently a standard therapy for cutaneous metastases, CaEP builds on experience with ECT. CaEP is easily accessible, as most facilities using ECT already have the necessary equipment. Calcium is readily available at low costs, does not require the same authorization for handling as chemotherapeutic agents as it is not cytotoxic by itself, and has been shown to be a safe and efficient treatment, with few side effects [7,8,10]. This patient expressed appreciation for a cancer treatment without chemotherapy and was satisfied with the outcome of CaEP combined with HER2-antibody treatment.
This case supports CaEP as an effective and tolerable treatment for patients with cutaneous metastases, specifically metastases of HER2-positive breast cancer. CaEP may have a potential synergistic effect when combined with other treatments.
Statement of Ethics
The patient gave written informed consent to publication of medical history and anonymized data, including clinical images. Ethical approval was not required for this case report, in accordance with local and national guidelines.
|
v2
|
2022-11-07T16:06:31.232Z
|
2022-11-04T00:00:00.000Z
|
253379041
|
s2orc/train
|
Differential expression of CCR8 in tumors versus normal tissue allows specific depletion of tumor-infiltrating T regulatory cells by GS-1811, a novel Fc-optimized anti-CCR8 antibody
ABSTRACT The presence of T regulatory (Treg) cells in the tumor microenvironment is associated with poor prognosis and resistance to therapies aimed at reactivating anti-tumor immune responses. Therefore, depletion of tumor-infiltrating Tregs is a potential approach to overcome resistance to immunotherapy. However, identifying Treg-specific targets to drive such selective depletion is challenging. CCR8 has recently emerged as one of these potential targets. Here, we describe GS-1811, a novel therapeutic monoclonal antibody that specifically binds to human CCR8 and is designed to selectively deplete tumor-infiltrating Tregs. We validate previous findings showing restricted expression of CCR8 on tumor Tregs, and precisely quantify CCR8 receptor densities on tumor and normal tissue T cell subsets, demonstrating a window for selective depletion of Tregs in the tumor. Importantly, we show that GS-1811 depleting activity is limited to cells expressing CCR8 at levels comparable to tumor-infiltrating Tregs. Targeting CCR8 in mouse tumor models results in robust anti-tumor efficacy, which is dependent on Treg depleting activity, and synergizes with PD-1 inhibition to promote anti-tumor responses in PD-1 resistant models. Our data support clinical development of GS-1811 to target CCR8 in cancer and drive tumor Treg depletion in order to promote anti-tumor immunity.
Introduction
The immune system has evolved complex cellular and molecular mechanisms to prevent deleterious immune-mediated disorders. A subset of CD4 + T cells, known as T regulatory (Treg) cells, that express the X chromosome-linked transcription factor forkhead box P3 (FOXP3), are central to the prevention of autoimmunity. 1 They act by suppressing aberrant immune responses against self-antigens, resulting in immune tolerance. 2 This tolerance mechanism can be co-opted by tumors when FOXP3+ Tregs, localized to the tumor microenvironment (TME), suppress anti-tumor immunity. Tumorinfiltrating Tregs can disrupt the function of tumor-specific T effector cells through direct cell-cell interactions by engaging inhibitory co-signaling molecules expressed on their cell surface or by secreting anti-inflammatory soluble factors such as transforming growth factor β (TGFβ); both mechanisms are known to contribute to the immunosuppressive TME. 3 The presence of tumor T regulatory cells has been associated with poor prognosis in cancer [4][5][6] and resistance to immune checkpoint blockade. 7 While immune checkpoint blockade targeting the co-inhibitory molecules cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and programmed cell death protein 1/programmed death-ligand 1 (PD-1/L1) has revolutionized the treatment of cancer, the majority of patients do not respond to checkpoint inhibition, and primary and secondary resistance to immune checkpoint blockade is common. Growing evidence suggests that targeting CTLA-4 and PD-1 is insufficient for the full elimination of Tregs within the TME, 8 and PD-1 blockade could even result in enhanced tumor Treg function. 9,10 Thus, new therapies are needed to preferentially deplete Tregs in the tumor while sparing the peripheral pool to ensure continued immune tolerance and minimize immune related toxicities.
A major challenge facing a Treg depletion strategy has been the identification of a specific marker for tumor-infiltrating Tregs. Daclizumab targets the interleukin-2 receptor alpha chain (CD25), which is abundantly expressed on peripheral Treg and T effector cells, leading to broad and relatively nonspecific depletion of T cells. 3 Mogamulizumab targets C-C motif chemokine receptor 4 (CCR4), which shows high expression on peripheral Tregs as well as expression on 20-30% of peripheral blood CD4 + T cells. 11 More recently, Plitas et al. identified C-C motif chemokine receptor 8 (CCR8) as a potential new target on tumor Tregs, reporting that CCR8 was highly expressed on human tumor-infiltrating as compared to peripheral Treg and T effector cells in the TME. 12 Analysis of gene expression data from cancer patient samples also found CCR8 expression to be highly correlated with FOXP3 in Tregs and preferentially expressed on tumor over peripheral Tregs. 13,14 CCR8 is a cell surface receptor that belongs to class A of the G protein-coupled receptor (GPCR) family. It contains seven transmembrane domains, three extracellular loop domains and three cytoplasmic domains. In addition to tumor Tregs, CCR8 has been reported to also be expressed on T-helper 2 (Th2) cells and innate lymphoid cells, 15 skin-resident memory T cells, 16 monocyte-derived dendritic cells and eosinophils. 17 In the tissues of healthy adults, CCR8-expressing T cells are reported to be unique to skin and not in the small intestine or colon, and only rarely found in peripheral blood. 18 CCR8 has four known ligands of which the best characterized is CCL1, which has been reported to be important for skin homing of T cells, as well as Treg survival and chemotaxis into tumors. [19][20][21] Thus, the relatively restricted expression pattern of CCR8 and its higher level of expression on tumor-infiltrating Tregs suggests it may be an optimal target for selective depletion of these cells.
Herein, we describe the development and characterization of GS-1811, a humanized monoclonal antibody (mAb) consisting of two identical afucosylated gamma 1 (IgG1) heavy chains and two identical kappa (Igk) light chains, that specifically binds to and inhibits human CCR8 and is designed to selectively deplete tumor resident Treg cells. We show that CCR8 gene expression is restricted to tumor resident Tregs, and that CCR8 protein is expressed at higher densities on tumor resident Tregs than on other tumor resident T cells or circulating peripheral T cells, thus providing a window for tumor-specific Treg depletion. In vitro, GS-1811 selectively depletes cells expressing CCR8 receptor at levels observed in human tumor Tregs. Targeting the murine CCR8 receptor results in antitumor efficacy in vivo, which is dependent on Treg depleting activity, across several tumor models, including tumor types where PD-1 blockade has no effect. Additionally, depletion of the murine CCR8+ tumor resident Tregs induces CD8 infiltration, proinflammatory responses and immunological memory. Finally, anti-CCR8 is a potent combination partner with anti-PD-1 in murine PD-1 resistant models, resulting in improved tumor regression. These data indicate the potential clinical utility of selective depletion of tumor resident T regulatory cells using GS-1811.
PBMCs, tissue samples and cell lines
Human PBMCs were purified from whole blood collected from healthy volunteers (Research Blood Components) by Ficoll-Paque Plus (GE Healthcare) density separation as previously described. 22 Fresh human tumor and normal adjacent tissue (NAT) samples from surgical resections were obtained through the NCI Cooperative Human Tissue Network (CHTN) and the National Research Disease Interchange (NDRI). Human skin samples from abdominoplasty procedures were provided by BioIVT. All human samples were acquired under approved vendor IRB protocols and were de-identified. The mouse cell line MC38 was kindly provided by Dr. James Allison (MD Anderson Cancer Center) and cultured in DMEM medium (Gibco Life Technologies) supplemented with glutamine and 10% Fetal Bovine Serum (FBS, Sigma). The B16-F10 mouse cell line was obtained from ATCC and cultured in RMPI medium (Gibco Life Technologies) supplemented with glutamine and 10% FBS. The mouse cell lines CT26 and Pan02 were provided by Lab2Pharmacy through ATCC. The mouse MBT-2 cell line was provided by Crown Biosciences. All mouse cells were tested and confirmed negative for Mycoplasma and viral pathogens. For tumor inoculation, cells were expanded for 2 passages after thaw and harvested by trypsin-EDTA treatment when 50-70% confluent. Cells were re-suspended in serumfree, phenol-free DMEM (Thermo Fisher Scientific) and viability >95% was confirmed by trypan blue staining. The human Hut-78 cell line was obtained from ATCC and grown in Xvivo15 medium (Lonza) supplemented with glutamine, 10% human serum, and 1% non-essential amino acids.
Mice and mouse tumor studies
For all animal studies conducted at Jounce Therapeutics, mice were maintained in accordance with the Jounce Therapeutics Institutional Animal Care and Use Committee (IACUC) protocol JT02-13-19 and were approved by the Jounce IACUC. 6-8 week old female C56BL/6 mice were obtained from the Jackson Laboratories and implanted subcutaneously on the right flank with 5 × 10 5 MC38 cells under isoflurane anesthesia. When tumor volumes reached approximately 100mm 3 , mice were randomized by tumor volume and assigned to treatment groups of 10 (for efficacy studies) or 5 (for immunophenotyping studies) animals per group. Mice were injected intraperitoneally twice weekly for a maximum of three weeks with 200 μg (approximately 10 mg/kg) of anti-mouse CCR8 mIgG2a, anti-mouse CCR8 mIgG1 or isotype control (all from Jounce Therapeutics) prepared in sterile phosphatebuffered saline (PBS). Tumor growth and animal body weights were monitored at least twice weekly. Mice were sacrificed when tumor volumes exceeded 2000mm 3 , tumor became ulcerated, body weight decreased by at least 20% or other signs of clinical distress were noted, in accordance with Jounce protocol JT02- [13][14][15][16][17][18][19]. For the tumor rechallenge study, 10 mice whose MC38 tumors had been previously eradicated by treatment with anti-mouse CCR8 mIgG2a antibody and that had remained tumor free for at least 20 weeks, and 10 agematched tumor-and treatment-naïve controls were implanted subcutaneously with 2.5 × 10 5 MC38 cells on the left flank and 1 × 10 5 B16-F10 cells on the right flank.
The MBT-2 tumor study was conducted by Crown Biosciences and was approved by their institutional IACUC. During the study, the care and use of animals was conducted in accordance with the regulations of the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC). 6-8 week old female C3H/He mice were inoculated subcutaneously on the right flank with 4 × 10 5 MBT-2 cells in 100 μL sterile 1x PBS. Mice were randomized into treatment groups of 10 animals each with an average tumor volume of 94mm 3 and dosed intraperitoneally twice weekly for three weeks with anti-mouse CCR8 mIgG2a (Jounce Therapeutics), InVivoMAb anti-mouse PD-1 RMP1-14 (BioXCell) or mouse IgG2a and rat IgG2a isotype controls (Jounce Therapeutics and BioXCell, respectively). Each antibody was administered at 10 mg/kg, and appropriate isotype controls were added so that the total antibody dose per each group was 20 mg/kg. Mice were checked daily for morbidity and mortality; tumor volume and body weight were measured twice weekly. Mice were sacrificed when tumor volume exceeded 3000mm 3 , tumor ulcerations reached 25% of the tumor surface, body weight loss reached 20% or other signs of clinical distress were noted.
The CT26 tumor study was conducted by Champions Oncology and was approved by their institutional IACUC. Prestudy mice were implanted subcutaneously into the left flank with 3 × 10 5 CT26 cells in 100 μL PBS. Tumor growth was monitored beginning 6-9 days after implantation using digital calipers. Treatment with anti-mouse CCR8 (10 mg/kg) or mouse IgG2a isotype control was started when tumors reached 100 or 250-350mm 3 . Dosing administration occurred every 3 days for a total of 4 doses. Mice were checked daily for morbidity and mortality; tumor volume and body weight were measured three times weekly. Mice were sacrificed when tumor volume exceeded 1500mm 3 , tumor ulcerations reached 50% of the tumor surface, body weight loss reached 20% or other signs of clinical distress were noted.
The Pan02 tumor study was conducted by Lab2Pharmacy. All procedures involving the care and use of animals in the study were reviewed and approved by the Stony Brook University Institutional Animal Care and Use Committee (Protocol 748435-3/2015-2182-NF-MI-5.18.18). Animals were inoculated by subcutaneous injection onto the right dorsal flank while under isoflurane induced anesthesia with 2 × 10^6 cells in 100 μL PBS. Tumor growth was monitored using digital calipers and treatment with anti-mouse CCR8 (10 mg/kg) or mouse IgG2a isotype control was started when tumors reached 100mm 3 . Dosing administration occurred every 3 days for a total of 4 doses. Mice were checked daily for morbidity and mortality; tumor volume and body weight were measured three times weekly. Mice were sacrificed when tumor volume exceeded 2000mm 3 , tumor ulcerations reached 25% of the tumor surface, body weight loss reached 20% or other signs of clinical distress were noted.
Treatment windows were chosen based on prior experience with each model, tumor growth kinetics and responsiveness to treatment. For MC38 and MBT-2 models, a standard twice weekly for 3 weeks dosing regimen was used, while for the more responsive CT26 model only 4 doses were administered over 2 weeks. For the Pan02 model, which is known to respond poorly to many IO and non-IO agents, 23 dosing was continued until all mice of the control group reached endpoint.
Tumor volume was calculated using the formula: V = (LxW 2 )/2, where L is the longest tumor dimension and W is width. Tumor growth and survival curves were obtained using GraphPad Prism. For average tumor growth graphs, the last measurement for mice that reached tumor volume endpoint was carried over until all mice in the same group were sacrificed. Survival p values were obtained using a pairwise Log-rank (Mantel-Cox) test.
Tissue sample dissociation
Human tumor and NAT samples were processed using the gentleMACS Octo Dissociator (Miltenyi Biotec). Briefly, samples were minced into small fragments in DMEM medium (Gibco Life Technologies) supplemented with 10% FBS (Sigma) on ice, then transferred to a GentleMACS C tube (Miltenyi Biotec) and dissociated using the Miltenyi Tumor Dissociation kit (cat#130-095-929). After a filtration through a 40 µm cell strainer, the cell suspension was centrifuged at 300xg for 5 min, and directly processed for flow cytometry or cryopreserved for later analysis.
Human skin samples (5-50 g, fat layer removed) were minced into approximately 1 mm 3 pieces as above and incubated in PBS + 20 mM EDTA at 37°C for 40 min with constant stirring. The suspension was passed through a 40 μM filter and centrifuged at 300xg for 5 min. Cell pellets were washed once in DMEM+10% FBS and transferred to ice until further processing. For some of the skin samples, the PBS/EDTA step was followed by an enzymatic treatment with 0.5 mg/mL hyaluronidase, 0.1 mg/mL liberase, 6 mg/mL DNaseI and 10 mM HEPES in DMEM for 1 hour at 37°C with constant stirring; this treatment did not increase immune cell yield and did not affect surface marker expression, therefore it was omitted for most of the samples.
Mouse tumor and spleen samples were divided into three sections, and for each tissue one section was processed for IHC (see below), a small piece (approximately 20 mg) was collected in RNAlater (Thermo Fisher Scientific) and processed for RNA extraction (see below). The remaining tissue was collected into 5 mL of DMEM+2% FBS on ice, and mechanically dissociated using a 70 μm cell strainer and the piston of a 5 mL syringe. Dissociated splenocytes were centrifugated at 360xg for 5 min at 4°C, and the cell pellets were resuspended in 1 mL ACK lysing buffer (Gibco) and incubated for 3 min at room temperature (RT) to lyse red blood cells. The reaction was stopped by adding 10 mL DMEM+2.5% FBS, cells were washed twice in 10 mL PBS+2.5% FBS (FACS buffer) and resuspended in FACS buffer for flow cytometry analysis.
Flow cytometry
To prepare human samples for flow cytometry, dissociated cells or PBMCs were washed once in ice cold PBS and stained with the Live/Dead Fixable Near-IR kit (Thermo Fisher Scientific, L34976), following manufacturer's instructions. 1:20) or GS-1811-Dylight 650 (Jounce Therapeutics, 0.8ug/200ul). Samples were then centrifuged at 400xg for 5 min at 4°C and washed once in FACS buffer. After washing, cells were fixed using the FoxP3/Transcription Factor staining kit (Thermo Fisher Scientific) and FoxP3 intracellular staining was performed following manufacturer's instructions for 90 min on ice with anti-human FOXP3-AF647 (Biolegend, 320114,1:20) or -BV421 (Biolegend, 320124, 1:20) antibodies. For quantitative flow cytometry, Quantum Simply Cellular anti-mouse IgG beads (Bangs Laboratories) were stained with anti-CCR8 or anti-CCR4 antibodies diluted in Brilliant Stain buffer for 30 min on ice. Following incubation, cells or beads were washed two times in FACS buffer and acquired on a BD LSRFortessa flow cytometer. Data were analyzed with FlowJo. Live CD45+ events were identified, then CCR8 or CCR4 expression was assessed in immune cell subsets. Tregs were identified as CD3+ CD4+ FoxP3+, while non-Treg CD3+ CD4 + FoxP3-cells were defined as Tconv. CD8 T cells were identified as CD3+ CD4-. Fluorescence-minus-one (FMO) and isotype controls were used to define positive and negative populations. Percentage of CCR8 or CCR4 positive events, and median fluorescence intensity (MedFI) of positive populations were established for each cell subset. MedFI values were used to calculate CCR8 and CCR4 median copies per cell values using standard curves of quantification beads and the manufacturer's provided excel sheet. Values were only reported if more than 50 cells from any specific immune cell subset could be clearly enumerated. Statistical comparisons between populations were calculated with unpaired twotailed t tests using GraphPad Prism software.
Immunohistochemistry and image analysis
Commercially sourced human skin samples were fixed for 24 hours in 10% neutral buffered formalin, then processed on the Leica ASP300S Tissue Processor through a graded series of alcohols (70% > 80% > 95% > 100%) prior to xylene, and then ASP Parablock paraffin wax. Once the tissue was processed, it was embedded in Paraplast Plus paraffin into a block, which was then sectioned at 5um onto glass slides to be used for IHC staining. FoxP3 (Abcam, clone 236A/E7) and CD8 (Dako, clone C8/144B) staining was performed on a Bond RX. Antigen retrieval, peroxide block, and washes preceded primary antibody incubation (FoxP3 or CD8, 1:100). Slides were again washed prior to secondary antibody application and subsequent Refine Red chromogen (Leica). Slides where finally incubated in Hematoxylin prior to coverslipping. After staining, each skin tissue sample was scanned into Halo. Annotation areas were created so that analyses of FoxP3 and CD8 IHC would take into account the epidermal and dermal layers. Once tissue annotations were complete, the Indica Labs -Cytonuclear v 1.6 module was used to create specific algorithms for the detection of FoxP3+ and CD8+ cells in all epidermal and dermal layers. Once image analysis was completed, all data were exported into .csv and various sample metrics were evaluated for percent positivity and density assessments.
Samples from mouse tumor studies were embedded and sectioned as above, then stained for FoxP3 (CST, clone D608R) and CD8 (Abcam, clone EPR20305) at 1:100. Staining was performed on a Bond RX. Antigen retrieval, peroxide block, and washes preceded primary antibody incubation (FoxP3 or CD8, 1:100). Slides were again washed prior to secondary antibody application and subsequent DAB chromogen (BOND polymer Refine Detection System). Slides were then incubated in Hematoxylin prior to coverslipping. Once stained, the slides were scanned into Halo and manual digital annotations were created to identify tumor regions where analyses were to be focused. For quantification, we set about to segment every cell within the analysis of tumor regions using both the signal from the IHC, as well as from the hematoxylin counterstain present in each image. This allowed for the identification of either CD8+ or FoxP3+ cells. Once image analysis was completed, all data were exported into .csv and analyses were performed using both Excel and Prism. ANOVA was used to compare the three treatment arms, while unpaired post-hoc t-tests, when appropriate, were used to evaluate differences between the individual treatment arms.
NanoString gene expression analysis of MC38 tumor samples
Samples from MC38 tumors were collected at day 3 after a single Ab dose (anti-mouse CCR8 mIgG2a, anti-mouse CCR8 mIgG1 and isotype control, n = 4 each) and stored in RNAlater (Invitrogen). RNA was extracted and quantified using the Maxwell SimplyRNA Tissue kit (Promega) and the Quantifluor RNA High Standard kit (Promega). Residual genomic DNA was removed using the Turbo DNA-free kit (Invitrogen). Gene expression was measured with the NanoString platform using the nCounter Mouse Immunology Panel. Quality control was conducted on field of view, binding density, positive and negative control probes, and estimated RNA amount. Raw gene expression values from RCC files were normalized to housekeeping genes, log2 transformed, and floored to the 95 th percentile of negative control genes. Differential gene expression analysis was conducted for comparison across treatment groups using unpaired t-tests, with false discovery rate (Benjamini-Hochberg procedure) p-value correction. Gene set enrichment analysis (GSEA) was performed using the Hallmark gene sets from the MSigDB collections. 24,25
Identification of Treg-specific targets using bioinformatics screen
To identify a list of potential targetable Treg-specific genes, a bioinformatics screen was conducted using gene expression profiles of human cancer patient tumors and peripheral blood of healthy donors. First, a list of genes was identified as correlated with a Jounce-generated Treg gene signature across RNAseq profiles of human tumor samples in TCGA. This list was further pruned to targets amenable to an antibody therapeutic modality with predicted cellular membrane bound or secreted localizations, and targets with preferential expression in tumor-infiltrating Tregs relative to tumor-infiltrating conventional T cells and other peripheral blood cell types. Additional functional qualifications were utilized to derive a list of 32 Treg genes and further select CCR8 as preferred target for therapeutic antibody discovery.
Analysis of CCR8 and Treg markers gene expression in human datasets
To analyze Treg markers gene expression, processed single-cell RNA-seq datasets from samples from melanoma, head and neck cancer, and hepatocellular carcinoma patients were downloaded from Gene Expression Omnibus, accession numbers: GSE72056, GSE103322, GSE98638. [26][27][28] All files were processed in R using Seurat with cells filtered using parameters: min cell per sample = 3, min genes per cell = 200, max mitochondrial genes = 5%. Gene expression measurements were normalized for each cell by total expression, multiplied by scaling factor (10,000 by default) and log transformed. Gene expression measurements were scaled to regress out sources of technical noise or batch effects including batch, cell alignment rate, number of detected molecules (nUMI), mitochondrial gene expression, and cell-cycle, where applicable. For combining datasets, log normalized gene expression values were then normalized against a housekeeping gene set across cells. 29 In post-processing, expression of cell-type markers or signatures (e.g. CD8A for CD8 T cells, FOXP3 for Tregs) were used to identify cell clusters. Gene expression of Treg markers were visualized using FeaturePlot with TSNE as the reduction method.
To analyze CCR8 gene expression in TCGA tumor vs adjacent normal tissues, TCGA RNA-seq gene-level expression data were downloaded from Omicsoft Array Studio from the TCGA_B37 data pulldown. Gene expression measurements were reported in log2(FPKM +0.01). All samples from primary tumors and adjacent normal tissues in solid tumor types are considered. Differential expression analysis between CCR8 levels in tumor samples compared to adjacent normal samples within each tumor type was conducted using unpaired, onetailed Mann-Whitney U test.
Generation and humanization of GS-1811
The parent antibody of GS-1811 was selected using mouse hybridoma technology from animals immunized with CCR8 + cells or with human CCR8 expression plasmid DNA.
Resulting clones were identified that demonstrated robust binding to CHO cell lines overexpressing human CCR8 but lacked binding to parental CHO cells or CHO cells expressing the CCR4 family member. Paired mouse VH and VL IgG DNA sequences were obtained from lead hybridomas for production. Chimeric versions of parental antibodies were generated by grafting mouse VH/VL regions onto human IgG1/kappa isotype backbone. Lead chimeras that specifically bound to CHO-CCR8 cells were humanized according to standard techniques. Briefly, the six complementarity determining regions (CDRs) were identified and grafted onto human IGHV/ IGKV germline frameworks identified via sequence and structural homology. Back mutations to the mouse germline were selected to maintain residues at the VH/VL interface and in the canonical loop structure. Humanized Fv sequences were cloned with human IgG1/kappa constant domains to produce a panel of humanized antibodies. GS-1811 was selected from this panel based on affinity and biophysical characteristics. Since a reduction in core fucosylation on amino acid N297 in the CH2 domain of the IgG1 Fc region has been shown to enhance ADCC function of antibody via improved CD16a binding, 30 GS-1811 was expressed in FUT8-/-CHO-GS cell line at Jounce and shown to be completely afucosylated as compared to an intact Ab standard (Waters Corp) by glycan analysis performed on UPLC with Glycan BEH Amide column following PNGase F enzyme treatment.
On-cell affinity assessment of GS-1811
To assess specific binding to GS-1811 to hCCR8, CHO-S cell lines overexpressing hCCR8, hCCR4 or parental CHO-S were plated at 1 × 10 5 cells/well in a 96-well plate and incubated with increasing concentrations (10-point 4-fold titration starting at 100 µg/ml) of GS-1811 or isotype control for 30 min at 4°C. After two washes in FACS buffer, cells were incubated for additional 30 min at 4°C with an anti-human IgG-AF647 secondary antibody (Biolegend, 409320, 1:100), washed twice in FACS buffer and acquired on a BD LSRFortessa flow cytometer. Each condition was run in triplicates. Data were analyzed with FlowJo and GraphPad Prism software.
The monovalent equilibrium dissociation constant (K D ) of purified GS-1811 antibody molecules to CHO-S cells expressing human CCR8 was determined using a solution equilibrium titration (SET) assay. Briefly, fixed concentrations (0.2, 0.1 and 0.05 nM) of GS-1811 were incubated for 3 hours at 37°C with increasing concentrations (11-point two-fold titration starting at 1.5x10 7 cells/mL) of CHO-S cells expressing ~300,000 copies of human CCR8 per cell. After equilibrium was reached, the samples were centrifuged at 500xg for 5 min and the supernatant, containing unbound antibody, was quantified on the Meso Scale Discovery (MSD) instrument (MESO QuickPlex SQ 120) using anti-IgG coated MSD plates and a SULFO-TAG anti-human IgG secondary antibody (MSD). The concentration of unbound antibody as a function of human CCR8-overexpressing CHO cell concentration was plotted using GraphPad Prism software. The signal from the MSD plate and known concentrations of the antibody and cells were then used to calculate the K D of the complex using a 1:1 binding model of the bimolecular equilibrium interaction. The fitting was done using a custom-built script on Mathematica software. Three independent replicate plates were tested and K D values were averaged together.
Assessment of GS-1811 antagonism of CCL1/CCR8 interaction
The DiscoverX PathHunter β-Arrestin assay (Eurofins) was used to determine GS-1811 antagonist activity. GS-1811 or isotype control were serially diluted starting at 30 μg/mL to form a 10-point 3-fold dilution curve in assay buffer and added to the PathHunter eXpress β-Arrestin GPCR cells expressing human CCR8. Following a 30 min incubation at 37°C, 5% CO 2 , human CCL1 (Peprotech) was added at the final concentration of 13.7 nM and incubated for additional 90 min. Signal was generated through a single addition of PathHunter Detection reagent cocktail, followed by a 60 min incubation at room temperature. Chemiluminescent signal generated was measured with the PerkinElmer Envision instrument. Each condition was run in triplicate. Data were analyzed with GraphPad Prism software.
ADCC assays
Human NK cells were isolated from fresh human healthy donor PBMCs by magnetic separation using the EasySep Human NK Cells Isolation Kit (StemCell Technologies) according to the manufacturer's protocol. Alternatively, purified frozen human NK cells were obtained from StemCell Technologies.
To evaluate GS-1811-mediated ADCC, a panel of recombinant CHO-S cell lines was generated expressing a range of CCR8 densities. CHO-S cells expressing either ~10,000, ~5,000, ~2,500, ~1,500, or ~600 copies of human CCR8 per cell, along with parental CHO-S cells, were stained with CellTrace Blue Dye (Thermo Fisher Scientific) according to manufacturing protocol, resuspended in RMPI Medium (Gibco) supplemented with Glutamine and 10% FBS (Sigma) and counted. 2 × 10 4 CHO-S cells were mixed with 1 × 10 5 purified human NK cells in 96-well plates, so that the final effector to target ratio was 5:1. GS-1811, its fucosylated version (either in chimeric or fully humanized formats) and isotype controls were serially diluted to generate a 10-point curve with 10-fold intervals starting at 10 µg/mL, added to designated wells and the plates were incubated at 37°C overnight. Each condition was run in triplicate for each NK cell donor. The next day, cells were washed in FACS buffer and prepared for flow cytometry. Cells were incubated for 30 min at 4°C with labeled antibodies against CD56 (1:100) to identify NK Cells and eFlour780 Viability dye (1:1000) to identify dead cells. After the incubation, plates were washed 3 times with FACS buffer and acquired on a BD LSRFortessa flow cytometer. Analysis was performed with FlowJo. Percent of specific lysis was calculated by normalizing values to isotype control at any given antibody concentration. Data were analyzed in GraphPad Prism to determine EC50 values.
ADCC assays with Hut78 cells, endogenously expressing CCR8 densities within the range of expression on tumorinfiltrating Tregs, were conducted as described above, but omitting the CellTrace Blue staining step and with a starting antibody concentration of 1 mg/mL. Three independent NK cell donors were tested, and each condition was run in triplicate. The flow cytometry panel included CD3-AF488 (Biolegend, 300320), CD4-BV421 (Biolegend, 317434), CD56-PE (BD Biosciences, 555516) at 1:100 dilution. Live Hut78 cells were identified from all events by gating on CD4+ and viability dye negative events. Percent killing compared to control was determined following the equation: (Isotype-GS-1811)/ Isotype*100, where "Isotype" is the average of the isotype control triplicates and "GS-1811" is one replicate of the GS-1811 condition. These values were analyzed in GraphPad Prism to determine EC50 values for ADCC potency.
CCR8 expression is highly restricted to tumor-infiltrating Tregs
To identify novel therapeutic targets that could enable specific depletion of tumor-infiltrating Treg cells, we conducted an insilico screen using The Cancer Genome Atlas (TCGA) followed by a series of custom designed filters. This approach (Figure 1a) led to the identification of 32 genes with preferential expression in tumor-infiltrating Tregs and with putative extracellular domains amenable to antibody targeting. Analysis of publicly available single cell RNA sequence datasets revealed that expression of one of these genes, CCR8, was exquisitely restricted to CD4+ CD25 (IL2RA)+ FoxP3+ cells as opposed to other Treg markers such as CD25 itself, CTLA-4, GITR (TNFRSF18), TIGIT, and CCR4 (Figure 1b and Supplementary Figure 1a). In TCGA, CCR8 gene expression was found to be higher in primary tumors than in normal adjacent tissues across many indications (Figure 1c). We assessed surface expression of CCR8 by flow cytometry and found preferential expression of CCR8 on tumor Tregs as compared to both peripheral Tregs and tumor and peripheral non-Treg CD4+ cells (referred to as T conventional cells, Tconv) (Figure 1d). We then expanded this analysis by comparing matched tumor and normal adjacent tissue samples from multiple indications, confirming highest CCR8 expression in tumor Tregs. Importantly, precise quantitative flow cytometry analysis showed that tumor CCR8+ Tregs express a significantly higher number of CCR8 molecules on a per-cell basis than any other T cell subsets, including CD8 Figure 1. CCR8 expression is highly restricted to tumor-infiltrating T regulatory cells in human tumors of different lineage. a) Schematic of the Treg bioinformatic screen used to identify putative targets for Treg depletion. b) t-SNE plots displaying single cell RNAseq gene expression profiles from 43 patients (HCC n=9, HNSCC n=18, melanoma n=19) including tumors, blood, LN and normal tissue that were combined and characterized by cell marker genes, as indicated. c) Gene expression analysis on bulk RNA-seq (TCGA) from primary tumor samples (red bars) and adjacent normal tissues (gray bars). The degree of statistical significance between tumor and adjacent normal samples is indicated below each pair of boxplots with tumor types (***: p<0.0001; **: p<0.01; *: p<0.05 from unpaired, one-tailed Mann-Whitney U tests). d) Flow cytometry analysis of peripheral blood mononuclear cells (PBMC) from healthy donors and indicated freshly dissociated human tumor samples. FoxP3 staining was used to identify T cell populations within CD45/CD3/CD4+ cells, and CCR8 expression was evaluated in Treg (FoxP3+, red histograms) and non-Treg (Tconv, FoxP3-, blue histograms) cells. Isotype control histograms are shown in black for total CD3+ T cells. Healthy PBMC data is representative of 3 independent donors. e) Percentage of CCR8+ cells (left) and median density of CCR8 molecules per cell (right) in Tconv (blue), Treg (red) and CD8 T cells (black) in freshly dissociated human tumor samples (Tumor) or normal adjacent tissue (NAT). Indications include breast, head and neck, lung, ovarian, colon, and bladder cancer. T cell populations were identified as in (d). Each data point represents a single sample. Error bars represent standard deviation from the average. The degree of statistical significance between populations indicated on top of the graphs (****: p<0.0001; **: p<0.01; *: p<0.05; ns: no significant difference).
T cells (Figure 1e). On the contrary, we found CCR4 to be broadly expressed on tumor and peripheral T cell subsets (Supplementary Figure 1b). As CCR8 has been identified as a marker of tissue-resident T cells in the skin, 16,31 we assessed its expression in skin samples, and compared it to CCR4 expression. Both CCR8 and CCR4 were expressed on all skin T cell subsets, at relatively high levels as compared to peripheral blood and other normal tissues (Supplementary Figure 2a). However, absolute numbers of T cells in the skin were low, with Tregs and CD8 T cells accounting for only 0.4% and 0.6% of total cells respectively, as determined by IHC analysis (Supplementary Figure 2b). The relatively restricted expression pattern of CCR8 and its higher level of expression on tumor-infiltrating Tregs suggests it may be a good target for selective depletion of these cells. Figure 3). We therefore sought to investigate the therapeutic effect of treatment with an anti-mouse CCR8 antibody (Ab) in a series of syngeneic mouse tumor models. In the MC38 model, administration of anti-mouse CCR8 Ab (murine IgG2a, mIgG2a) resulted in significant tumor growth inhibition with about 30-50% of mice showing complete tumor regression (Figure 2a and Figure 3a). Robust efficacy was also observed in the Pan02 tumor model (Figure 2b), which is reported to be poorly responsive to immuno-oncology therapeutics 23 and in the CT26 tumor model, where efficacy was maintained even when treatment was started following establishment of large tumors (Figure 2c). Lastly, we assessed the impact of anti-CCR8 mIgG2a treatment in combination with anti-PD-1 in the PD-1-resistant model MBT-2. Anti-mouse CCR8 mIgG2a monotherapy inhibited tumor growth compared to isotype control, while anti-PD-1 monotherapy did not. The combination of anti-mouse CCR8 mIgG2a and anti-PD-1 resulted in increased efficacy compared to anti-mouse CCR8 mIgG2a monotherapy with complete tumor regression observed in half of the animals receiving combination therapy compared to no complete responses in animals receiving either monotherapy or isotype control (Figure 2d). Of note, no signs of treatment-related toxicities or body weight loss were observed in any of these studies, even when anti-mouse CCR8 Ab was administered for an extended period of time, such as in the Pan02 study (Supplementary Figure 4).
Depletion of CCR8+ tumor-infiltrating Tregs is required for antitumor efficacy of anti-mouse CCR8 Ab
To demonstrate that Treg depletion is a major determinant of response to anti-mouse CCR8 Ab treatment, we took advantage of the fact that only antibodies with murine IgG2a and not murine IgG1 Fc regions are able to engage the FcγR and mediate antibody-dependent cellular cytotoxicity (ADCC). 32 We therefore compared the effect of both versions of the antimouse CCR8 Ab in the MC38 model. As shown in Figure 3a, only the mIgG2a version inhibited tumor growth, while the mIgG1 version did not show efficacy as compared to isotype control. We then assessed immune cell composition of MC38 tumors from mice treated with either one or two doses of anti-CCR8 mIgG2a or mIGg1 Abs by flow cytometry. Consistent with the proposed mechanism of action, only the Fc-competent version of the anti-mouse CCR8 Ab (mIgG2a) and not the mIgG1 version resulted in robust Treg depletion in the tumor (Figure 3b, left panel, and Supplementary Figure 5a). This observation was also confirmed by immunohistochemistry (IHC) on the same tumor samples (Figure 3c). Depletion of tumor Tregs was detectable from day 1 after administration of a single dose of anti-mouse CCR8 mIgG2a (Figure 3c) and was sustained up to day 7 (Figures 3b and 3c). Moreover, this effect was specific for tumor Tregs, as no depletion was detected in splenic Tregs (Figure 3b, right panel) or in other T cell subsets (Figure 4a and Supplementary Figure 5c), where CCR8 expression is low (Supplementary Figure 3). Depletion of CCR8positive Tregs could trigger compensatory changes in remaining Tregs or in non-Treg CD4 T cells subsets, resulting in a more immunosuppressive phenotype. We therefore verified that FoxP3 expression was not affected in tumor CD4 T cells upon treatment with anti-mouse CCR8 mIgG2a Ab (Supplementary Figure 5b). Gene expression analysis also showed no increase in expression of genes associated with suppressive Treg function, such as TIGIT, CTLA-4 and ICOS, in MC38 tumors from mice treated with either antimouse CCR8 mIgG2a or mIgG1 antibody (Supplementary Figure 6).
Anti-mouse CCR8 mIgG2a Ab treatment induces CD8 infiltration, proinflammatory responses, and immunological memory
We further characterized immune cell infiltration in MC38 tumors upon anti-mouse CCR8 Ab treatment by flow cytometry. In a subset of the tumors that showed decreased Treg infiltration (Figure 3b) we detected a concomitant increase in CD8 infiltration, which resulted in increased CD8-to-FoxP3 ratios (Figure 4a). No changes were observed in other immune cell populations, including Tconv, CD19 + B cells, or CD11b+ myeloid cells (Supplementary Figure 5c). Percentage of tumor CD8 T cells increased over time and IHC analysis demonstrated robust infiltration throughout the tumor tissue (Figure 4b). To more broadly assess changes in the tumor microenvironment, we then performed gene expression analysis in MC38 tumors after one dose of anti-mouse CCR8 mIgG2a or mIgG1 using a NanoString mouse immunology panel. Samples from animals treated with anti-mouse CCR8 mIgG1 and isotype control clustered together and were more similar to each other than to anti-mouse CCR8 mIgG2atreated samples using unsupervised clustering. No genes were significant in the differential gene expression analysis after correction for false discovery rate. However, genes with greater than one log fold-change in expression in anti-CCR8 mIgG2atreated compared to anti-CCR8 mIgG1-treated and isotypetreated samples showed upregulation of chemokine ligands and interleukins involved in pro-inflammatory responses (Figure 4c). Additionally, gene set enrichment analysis (GSEA) of anti-CCR8 mIgG2a-treated compared to isotypetreated samples revealed upregulated genes are enriched in inflammatory, interferon-α, and interferon-γ responses ( Table 1). These data suggest that treatment with Fccompetent anti-mouse CCR8 Ab can efficiently deplete tumor Tregs, resulting in increased CD8 T cell activity and anti-tumor responses. As CD8 T cell activation is expected to lead to establishment of immunological memory, we performed tumor re-challenge studies to assess the durability of response to anti-CCR8 mIgG2a treatment. Mice whose MC38 tumors had been previously eradicated by treatment with anti-mouse CCR8 mIgG2a Ab were re-implanted with MC38 and B16-F10 tumors on contralateral flanks. As a control, tumor-naïve and untreated age-matched mice were also challenged with MC38 and B16-F10 contralaterally. All naïve mice developed tumors on both flanks, with the exception of a single MC38 tumor that initially grew out and then regressed (Figure 4d). In the rechallenged mouse cohort, all mice remained MC38 tumor-free up to day 45 after tumor inoculation, with a single MC38 tumor growing by day 50. On the contrary, 8 out of ten rechallenged mice developed B16-F10 tumors (Figure 4d), albeit their growth appeared to be slower as compared to naïve mice, as previously reported for other therapeutic approaches (e.g.) 33
The anti-human CCR8 antibody GS-1811 specifically binds to human CCR8 and inhibits CCR8 downstream signaling
To discover anti-human CCR8 antibodies, we performed mouse immunizations and generated a panel of hybridomas that were selected for binding to CHO cells engineered to express human CCR8 but not to parental CHO cells or to CHO cells expressing the closely related CCR4 receptor (data not shown). The top candidate from this selection was then humanized to generate GS-1811, which was constructed with the human IgG1/kappa isotype backbone and expressed in a host cell line devoid of core fucosylation of N-linked glycans. As afucosylated hIgG1 shows higher FcγR binding affinity and can outcompete endogenous IgG for binding to CD16 for enhanced ADCC, 34 GS-1811 is optimized to selectively deplete tumor-infiltrating Treg cells. Similarly to its parental Ab, GS-1811 is specific for human CCR8, as determined by binding to human CCR8-expressing and not CCR4-expressing CHO cells ( Figure 5a) and is able to potently inhibit CCL1-induced signaling downstream of CCR8 (Figure 5b). On-cell affinity of GS-1811 was determined to be about 17pM (Figure 5c). We also confirmed that GS-1811 binds specifically to tumorinfiltrating Tregs isolated from human primary tumor samples, but only poorly to tumor Tconv or CD8 T cells, and to peripheral Treg or Tconv cells from healthy PBMCs (Figure 5d).
GS-1811 mediates ADCC of target cells expressing CCR8 at levels comparable to tumor-infiltrating Tregs
GS-1811 does not bind to mouse or monkey CCR8 thus limiting our ability to directly verify its effect in vivo. Therefore, to test the ability of GS-1811 to specifically deplete human tumor-infiltrating Tregs, while sparing peripheral Tregs and other T cell subsets, we generated a panel of CHO cells lines expressing increasing densities of human CCR8, from 600 to over 10000 molecules per cell. We then assessed in vitro ADCC activity of GS-1811 and fucosylated (fuc) GS-1811 toward this CHO cell line panel in the presence of freshly isolated NK cells from healthy human donors. While fucosylated GS-1811 depleting activity was low for target cells expressing less than 5000 CCR8 molecules, afucosylated GS-1811 robustly depleted cells expressing 2500 CCR8 molecules (Figure 6a, 6b), which is within the range of CCR8 expression in human tumors (Figure 1c). Importantly, despite its enhanced ADCC potential, GS-1811 was not efficient at data is presented as average +/-SEM. ANOVA was used to compare the three treatment arms, followed by Tukey's multiple comparison post hoc test to evaluate differences between the individual treatment arms. p values are shown for statistically significant differences (*: p=0.01 to 0.05; **: p=0.001 to 0.01; ***: p<0.001). c) NanoString gene expression analysis on MC38 tumors from mice treated with a single dose of anti-mouse CCR8 Ab in the mIgG2a or mIgG1 format or with isotype control; n=4 per group, evaluated at day 3 after dosing. Top: volcano plot showing differential gene expression between anti-mouse CCR8 mIgG2a and isotype groups; bottom: genes with greater than one log fold-change in expression in anti-CCR8 mIgG2a compared to anti-CCR8 mIgG1 and isotype samples. d)Tumor growth upon rechallenge in MC38 tumor-bearing mice that were fully cured by anti-mouse CCR8 antibody treatment (red) or in naïve mice as control (black). Mice were inoculated with MC38 and B16-F10 tumor cells on contralateral flanks. Curves represent average +/-SEM. depleting cells with CCR8 expression below 1500 molecules/ cells (Figure 6a, Table 2), within the range of peripheral T cell subsets or tumor-infiltrating Tconv and CD8 T cells. To evaluate the effect of GS-1811-mediated ADCC on cells endogenously expressing CCR8 at densities within the range observed on tumor-infiltrating Tregs, we performed ADCC assays using the Hut78 cell line. The average EC50 across three NK cell donors was 0.0017 ±0.0012 and average maximum killing was 67.5 ±2.69% (Figure 6c). Finally, we assessed ADCC in TILs from different tumor types using exogenously added NK cells. Results from three tumor types and two independent NK cell donors showed that percentages of Tregs were consistently reduced upon incubation with afucosylated GS-1811 as compared to isotype control (Figure 6d), indicating that GS-1811 can promote killing of tumor-infiltrating Tregs ex vivo.
Discussion
It is well established that the presence of tumor Treg cells is associated with poor prognosis in cancer, 4-6 likely resulting from a more immunosuppressive TME. [35][36][37] In this case, strategies to reduce tumor Tregs by their destruction via natural killer (NK) cells and by inhibiting their trafficking into the tumor would be expected to result in a less suppressive TME, 38,39 leading to anti-tumor immunity. Therefore, in this study we present evidence that tumor Tregs selectively express CCR8 and that targeting CCR8 in vivo results in improved survival with selective tumor Treg depletion. These data highlight the therapeutic potential of targeting CCR8, and we further present data on the generation of a selective and specific anti-human CCR8 antibody for clinical evaluation.
Through evaluation of TCGA and available single cell RNA sequence datasets, we found CCR8 expression to be highly restricted to FoxP3+ tumor-infiltrating Tregs, consistent with previous reports. 40,41 Expression of CCR8 on tumor Tregs is more selective than other well-known Treg markers, including CD25, CTLA-4, GITR, TIGIT, and CCR4. CCR8 expression is higher in primary tumors than in normal adjacent tissues and significantly higher than in peripheral Tregs and both tumor and peripheral CD4 + T conventional cells. We also show that, on a per-cell basis, tumor Tregs have the highest number of CCR8 molecules compared to any other T cell subsets. CCR8 + Tregs are also known to be highly immunosuppressive. 12,42 Thus, taken together, these data establish CCR8 as a definitive and highly restricted immunosuppressive marker for tumor Tregs. Importantly, this suggests CCR8 may be a good therapeutic target for selective depletion of these cells.
Of note, previous reports have detailed CCR8 expression outside of the tumor and, most notably, in skin. 18,19,43 Here we demonstrate that, while both CCR8 and CCR4 are expressed on all skin T cell subsets, the absolute number of Tregs and CD8 T cells in the skin is very low (<1%). Therefore, we would not expect significant systemic effects upon treatment with a CCR8-depleting Ab above the relatively mild and reversible effects reported with CCR4-depleting Ab mogamulizumab in clinical settings. 44 This conclusion is also supported by the lack of treatment-related toxicities observed in our mouse studies.
In this study, we sought to evaluate the therapeutic impact of depletion of CCR8+ tumor Tregs in cancer using syngeneic mouse tumor models, where genetic deletion of FoxP3expressing cells has been shown to inhibit tumor growth. 45 We observed robust anti-tumor response upon treatment with an anti-mouse CCR8 Ab in a variety of mouse tumor models, and efficacy was maintained also when treating larger tumors. In the MC38 model, anti-tumor response was dependent on the ability of the anti-mouse CCR8 Ab to engage with Fc receptors 32 and mediate efficient Treg depletion. Importantly, depletion was specific to tumor Tregs, and not detected in splenic Tregs or other T cell subsets. While we cannot exclude that CCR8 could be expressed transiently and/or on a small percentage of non-Treg cells, possibly leading to their depletion, our observation that lower levels of CCR8 lead to reduced potential for ADCC suggests that achieving a therapeutic window for effective and selective depletion of tumor-infiltrating Tregs could be feasible in clinical settings. In MC38 tumors, decrease in tumor-infiltrating Tregs was associated with an increase in CD8 infiltration, which resulted in increased CD8-to-FoxP3 ratio. Importantly, the increased CD8-to-FoxP3 ratio suggests that, concomitant with a loss of tumor Treg cells, there is a shift in immunological profile favoring an anti-tumor response, which is consistent with favorable prognosis in human cancer. 1,46,47 In fact, gene expression analyses revealed a shift toward upregulation of chemokine ligands and interleukins involved in proinflammatory responses associated with the tumor Treg cell loss seen upon treatment with the mIgG2a anti-CCR8 Ab. This was corroborated by GSEA, which revealed upregulated gene enrichment in inflammatory, interferon-α, and interferon-γ responses and further supports the notion that depleting tumor Tregs results in creation of a favorable anti-tumor immunological environment. In line with this hypothesis, as CD8 T cell activation is expected to lead to establishment of immunological memory, 48 we also demonstrated that mice whose MC38 tumors were previously eradicated by antimouse CCR8 mIgG2a Ab treatment were resistant to tumor rechallenge.
Tregs are highly malleable and can modulate their phenotype in response to extracellular signals in the tumor microenvironment. 3,35 We considered the possibility that depletion of CCR8-positive cells could promote a more suppressive phenotype in remaining Tregs, or that CCR8 Ab treatment could otherwise affect Treg conversion, 21 but did not detect changes consistent with these hypotheses in our studies. Moreover, the increased CD8 T cell infiltration in the tumor, increased expression of pro-inflammatory markers, and significant anti-tumor efficacy observed upon anti-mouse CCR8 mIgG2a Ab treatment, all suggest remaining Tregs are not highly suppressive in vivo.
While we could not directly assess anti-tumor response in the immunophenotyping studies due to the shorter duration of treatment, it is tempting to speculate that the subset of mice showing robust CD8 T cell infiltration upon anti-CCR8 Ab treatment are the ones that are developing robust anti-tumor immunity, which would eventually result in complete response. We therefore hypothesized that relieving tumor T cell inhibition through PD-1 blockade could synergize with CCR8+ Treg depletion and selected the PD-1 resistant MBT-2 tumor model to test this hypothesis. Of note, MBT-2 tumors are known to harbor a significantly greater proportion of CD4 + cells, compared with the PD-1 responsive MC38 model, 49 suggesting a greater degree of Treg infiltration as an underlying mechanism for PD-1 resistance. Indeed, we observed significantly increased anti-tumor efficacy with combination therapy as compared with either anti-PD-1 or anti-CCR8 monotherapy, clearly indicating that an anti-CCR8 Ab that depletes tumor Treg cells can synergize with established immunooncology therapeutic modalities to unleash robust anti-tumor responses. 50 Given the highly restricted expression of CCR8 to tumor Tregs, and the robust in vivo data, we sought to identify antihuman CCR8 antibodies for clinical use. Through a hybridoma screen, we identified a top candidate, GS-1811. GS-1811 is specific for human CCR8 and selectively binds to tumor Treg but not to tumor T conventional or CD8 T cells, nor to peripheral Treg or Tconv cells. ADCC has been shown to depend on target density and on the Ab ability to engage FcRs. 51 Since a reduction in core fucosylation in the CH2 domain of the IgG1 Fc region has been shown to enhance ADCC function of an antibody via improved FcγRIIIA binding, 30 we expressed GS-1811 in cell line devoid of N-linked glycan core fucosylation, which leads to an afucosylated IgG1 backbone. Using a carefully validated panel of target cells expressing increasing number of CCR8 molecules per cell combined with our precise quantification of CCR8 expression on a per-cell basis in tumor and peripheral T cell subsets, we were able to establish that the afucosylated GS-1811, but not a fucosylated version, efficiently depletes cells in which the density of receptor expression is greater than 2500 CCR8 molecules per cell, which is within the physiological level detected on human tumor Treg cells. Importantly, GS-1811 has low depleting activity toward cells expressing CCR8 within the range of peripheral T cell subsets or tumor-infiltrating Tconv and CD8 T cells. These data suggest that cells expressing CCR8 at levels similar to peripheral Tregs will not be subject to Fc-mediated ADCC by GS-1811 while cells expressing CCR8 at levels similar to tumor Tregs will be depleted via Fc-mediated ADCC in the presence of GS-1811. It should be noted that GS-1811 also potently inhibits CCL-1-induced signaling downstream of CCR8. While we believe that ADCC-mediated Treg depletion is the key mechanism of action of GS-1811, blockade of CCR8 function might also impact Treg cell survival and/or conversion, as demonstrated in mouse studies, 21,52 contributing to decreased Treg function and anti-tumor effects. Overall, our data support ongoing clinical development of GS-1811 (NCT05007782) to target CCR8 in cancer in order to drive tumor Treg depletion and to shift the immunological milieu toward promoting anti-tumor immunity.
Funding
The author(s) reported there is no funding associated with the work featured in this article.
Data availability statement
The single-cell RNA-sequencing data that support the findings in Figure 1b are openly available in Gene Expression Omnibus at https:// www.ncbi.nlm.nih.gov/geo/, reference numbers GSE72056, GSE103322, GSE98638. The RNA-seq gene expression data of CCR8 referenced in Figure 1c from QIAGEN OmicSoft OncoLand database TCGA_B38_GC33_20210915_v2 were used under license for this study. Data are available from the authors upon request with permission from QIAGEN Digital Insights. The mouse gene expression data referenced in Figure 4c are available from the authors upon reasonable request.
|
v2
|
2022-11-08T14:14:23.813Z
|
2022-11-04T00:00:00.000Z
|
253385860
|
s2ag/train
|
Oral mTOR Inhibition Limits And Reduces Actinic Keratosis And Cutaneous Squamous Cell Carcinoma In A UVB-Induced Mouse Model
Actinic keratosis (AK) is a skin disease that is characterized by clinical and subclinical cutaneous lesions in sun-exposed areas. It is a considerable burden due to its high occurrence in middle-aged and older populations, as well as its propensity to progress to invasive cutaneous squamous cell carcinoma. The mammalian target of rapamycin (mTOR) pathway is critical in carcinogenesis and tumor development, and it has been shown to be over-activated during skin tumorigenesis, particularly upon ultraviolet (UV) radiation exposure, the key risk factor for AK. However, the ability of mTOR inhibitors to treat AK is not well documented. Herein, we evaluated the effect of oral mTOR inhibitors in vitro and in vivo and found that mTOR inhibitors lower keratinocyte cell proliferation in vitro and both clear and prevent AK and cutaneous squamous cell carcinoma (cSCC) in a UV-B induced SKH1 hairless mouse model of disease. mTOR inhibition reduced the number and size of skin lesions and the frequency of cSCC, resulting in a considerable reduction in disease severity. mTOR inhibition prevented lesion occurrence in areas of field cancerization, without affecting epidermal thickness, keratinocyte proliferation in vivo, or the presence of p53+ cells. Our findings indicate that, when appropriately dosed, oral mTOR inhibitors provide a safe home-based systemic treatment alternative with significant benefits to patients over current topical treatment options.
|
v2
|
2022-11-08T14:14:37.135Z
|
2022-11-04T00:00:00.000Z
|
253385794
|
s2ag/train
|
Anti-HVEM mAb therapy improves antitumoral immunity both in vitro and in vivo, in a novel transgenic mouse model expressing human HVEM and BTLA molecules challenged with HVEM expressing tumors
Background TNFRF-14/HVEM is the ligand for BTLA and CD160 negative immune co-signaling molecules as well as viral proteins. Its expression is dysregulated with an overexpression in tumors and a connection with tumors of adverse prognosis. Methods We developed C57BL/6 mouse models co-expressing human huBTLA and huHVEM as well as antagonistic monoclonal antibodies (mAbs) that completely prevent the interactions of HVEM with its ligands. Results Here, we show that the anti-HVEM18-10 mAb increases primary human αß-T cells activity alone (CIS-activity) or in the presence of HVEM-expressing lung or colorectal cancer cells in vitro (TRANS-activity). Anti-HVEM18-10 synergizes with anti-PD-L1 mAb to activate T cells in the presence of PDL-1 positive tumors, but is sufficient to trigger T cell activation in the presence of PD-L1 negative cells. In order to better understand HVEM18-10 effect in vivo and especially disentangle its CIS and TRANS effects, we developed a knock-in (KI) mouse model expressing human BTLA (huBTLA+/+) and a KI mouse model expressing both human BTLA and human HVEM (huBTLA+/+ /huHVEM+/+ (DKI)). In vivo pre-clinical experiments performed in both mouse models showed that HVEM18-10 treatment was efficient to decrease human HVEM+ tumor growth. In the DKI model, anti-HVEM 18-10 treatment induces a decrease of exhausted CD8+ T cells and regulatory T cells and an increase of Effector memory CD4+ T cells within the tumor. Interestingly, mice which completely rejected tumors (± 20%) did not develop tumors upon re-challenge in both settings, therefore showing a marked T cell-memory phenotype effect. Conclusions Altogether, our preclinical models validate anti-HVEM18-10 as a promising therapeutic antibody to use in clinics as a monotherapy or in combination with existing immunotherapies (anti-PD1/anti-PDL-1/anti-CTLA-4).
|
v2
|
2022-12-21T16:18:40.436Z
|
2022-11-04T00:00:00.000Z
|
254918990
|
s2ag/train
|
Analysis of Health Belief Model regarding Human Papillomavirus Vaccination among Female Employees at Sukabumi Regency Government
The high rate of cervical cancer case confirms it as the second most common cancer in Indonesia which required specific management and primary prevention efforts for everyone. The appeal issued by the Ministry of Health regarding HPV vaccination which will be mandatory as an effort to reduce cervical cancer rates in Indonesia. This study aims to analyze the health belief model as the influential factor for the willingness to get HPV vaccination as an effort to prevent cervical cancer among female employees in Sukabumi Regency government in 2022. This was a quantitative study with cross-sectional design. Data were collected among 213 respondents who were selected using purposive sampling technique. The results showed that 50.7% of respondents were willing to get HPV vaccination. Meanwhile, 49,3% of respondents were not willing to get HPV vaccination. There was a relationship between perceived susceptibility, benefit, obstacle, cues to action, and self-efficacy with the willingness to get HPV vaccination. Meanwhile, the perceived severity was not related to the willingness to get HPV vaccination. In the multivariate analysis, it was found that cues to action (POR=5,477; 95% CI=2,6-11,2) had the most significant effect on willingness to get HPV vaccination. It is expected that the current study can be used as an input for the government of Sukabumi Regency to plan for health promotion programs for every agency regarding cervical cancer, especially HPV vaccination.
|
v2
|
2022-11-04T06:18:02.626Z
|
2022-11-08T00:00:00.000Z
|
253267447
|
s2ag/train
|
[Clinicopathological and molecular features of multinodular and vacuolating neuronal tumors of the cerebrum].
Objective: To investigate clinicopathological features of multinodular and vacuolar neurodegenerative tumor (MVNT) of the cerebrum, and to investigate its immunophenotype, molecular characteristics and prognosis. Methods: Four cases were collected at the General Hospital of Southern Theater Command, Guangzhou, China and one case was collected at the First People's Hospital of Huizhou, China from 2013 to 2021. Clinical, histological, immunohistochemical and molecular characteristics of these five cases were analyzed. Follow-up was carried out to evaluate their prognoses. Results: There were four females and one male, with an average age of 42 years (range, 17 to 51 years). Four patients presented with seizures, while one presented with discomfort on the head. Pre-operative imaging demonstrated non-enhancing, T2-hyperintense multinodular lesions in the deep cortex and superficial white matter of the frontal (n=1) or temporal lobes (n=4). Microscopically, the tumor cells were mostly arranged in discrete and coalescent nodules primarily within the deep cortical ribbon and superficial subcortical white matter. The tumors were composed of large cells with ganglionic morphology, vesicular nuclei, prominent nucleoli and amphophilic or lightly basophilic cytoplasm. They exhibited varying degrees of matrix vacuolization. Vacuolated tumor cells did not show overt cellular atypia or any mitotic activities. Immunohistochemically, tumor cells exhibited widespread nuclear staining for the HuC/HuD neuronal antigens, SOX10 and Olig2. Expression of other neuronal markers, including synaptophysin, neurofilament and MAP2, was patchy to absent. The tumor cells were negative for NeuN, GFAP, p53, H3K27M, IDH1 R132H, ATRX, BRG1, INI1 and BRAF V600E. No aberrant molecular changes were identified in case 3 and case 5 using next-generation sequencing (including 131 genes related to diagnosis and prognosis of central nervous system tumors). All patients underwent complete or substantial tumor excision without adjuvant chemoradiotherapy. Post-operative follow-up information over intervals of 6 months to 8 years was available for five patients. All patients were free of recurrence. Conclusions: MVNT is an indolent tumor, mostly affecting adults, which supports classifying MVNT as WHO grade 1. There is no tumor recurrence even in the patients treated with subtotal surgical excision. MVNTs may be considered for observation or non-surgical treatments if they are asymptomatic.
|
v2
|
2022-11-04T15:57:52.841Z
|
2022-11-23T00:00:00.000Z
|
253268751
|
s2orc/train
|
Podocyte lineage marker expression is preserved across Wilms tumor subtypes and enhanced in tumors harboring the SIX1/2 p.Q177R mutation
Wilms tumors present as an amalgam of varying proportions of three tissues normally located within the developing kidney, one being the multipotent nephron progenitor population. While incomplete differentiation of the nephron progenitors is widely-considered the underlying cause of tumor formation, where this barrier occurs along the differentiation trajectory and how this might promote therapeutic resistance in high-risk blastemal-predominant tumors is unclear. Comprehensive integrated analysis of genomic datasets from normal human fetal kidney and high-risk Wilms tumors has revealed conserved expression of genes indicative of podocyte lineage differentiation in tumors of all subtypes. Comparatively upregulated expression of several of these markers, including the non-canonical WNT ligand WNT5A, was identified in tumors with the relapse-associated mutation SIX1/2 p.Q177R. These findings highlight the shared progression of cellular differentiation towards the podocyte lineage within Wilms tumors and enhancement of this differentiation program through promotion of non-canonical WNT/planar cell polarity signaling in association with SIX1/2 p.Q177R.
Society of Paediatric Oncology protocols relapsed within five years of diagnosis and 95% of relapses were distant analysis of marker expression by immunofluorescence and in situ hybridization in developing hFK revealed overlap of NPC marker SIX2/SIX2 with MEIS1/MEIS1 as well as FOXD1/FOXD1, another marker used to distinguish interstitial progenitors 90 from NPCs in the mouse (Lindström et al 2018a). Overexpression of either SIX1 or SIX1-Q177R, alongside MEIS1-3xFLAG 91 in vitro followed by immunoprecipitation (IP) with a SIX1 antibody resulted in co-IP of MEIS1-3xFLAG (Supplemental Figure 92 1C). To our knowledge this is the first biochemical evidence of an interaction between SIX1 and MEIS1 in any context.
93
Further investigation is needed to characterize potential regulatory functions of complexes containing these proteins in 94 NPCs.
95
As the SIX1-Q177R ChIP-seq data was obtained from a single Wilms tumor (Wegert et al 2015), we sought to 96 interrogate the DNA binding preference of the mutant protein in the absence of potentially confounding variables including 97 tissue quality and chemotherapy-induced artifacts. To determine the effect on sequence specificity of the variant, we 98 expressed the reference allele and the Q177R mutant SIX1 homeodomains in vitro and assayed their specificities in 99 parallel by protein binding microarrays (PBMs) (Berger et al 2006). The primary and secondary motifs (Badis et al 2009) 100 recognized by both alleles are shown in Figure 2A; a replicate experiment on an independent array yielded qualitatively Nearly 30% of SIX1/2-Q177R tumors also harbor inactivating mutations in the miRNA processing genes DROSHA or DGCR8 (Wegert et al 2015, Walz et al 2015, Gadd et al 2017. Strikingly, co-occurrence of SIX1/2-Q177R and reduction in miRNA levels (Wegert et al 2015). Due to the scarcity of SIX1/2-Q177R tumors without mutations in miRNA 123 processing genes in this dataset and to best isolate the transcriptional effects attributable to SIX1/2-Q177R, tumors 124 harboring SIX1/2-Q177R with or without miRNA mutations were included in differential expression analysis (SIX1/2miRNA, Table 129 1). Of note, although SIX1/2-Q177R is most associated with blastemal histology (Wegert et al 2015, Walz et al 2015, three 130 of the SIX1/2miRNA tumors used in this analysis were classified as mixed histology and one tumor was classified as DAWT in SIX1/2miRNA tumors, 14 of which were also significantly upregulated compared to MIXED/ES tumors (log2 fold change > 1.5, adj. p <0.05) ( Figure 3B, Supplemental Figure 3A, Supplemental File 1). Remarkably, the expression levels of wellan augmentation of the pre-induction NPC transcriptional regulatory identity compared to tumors of other histology. CCND2, encoding cyclin D2, was significantly upregulated in SIX1/2miRNA tumors compared to both Blastemal and DAWT groups 154 and was upregulated 1.46 log2 fold compared to MIXED/ES tumors, similar to findings of other studies (Wegert et al 2015,
155
Walz et al 2015). Nine genes were significantly upregulated in SIX1/2miRNA tumors compared to all groups, including 156 known NPC markers TMEM100 and ITGA8. This suggests the sustained proliferation reflected by enhanced CCND2 sites may represent CREs. Moreover, these peaks did not appear in the wk17hFK dataset ( Figure 4A). Apart from WNT5A, proximal and/or distal SIX1-Q177R or shared tumor peaks were identified within candidate CREs as predicted by ENCODE 180 or GeneHancer for other upregulated SIX1/2miRNA genes including KDM2B, CDKN1C, TMEM100, ARHGEF3, and FOXC1 181 (Supplemental Figure 4A) (Abascal et al 2020, Fishilevich et al 2017.
182
To explore the potential regulation of gene expression by SIX1-Q177R through the WNT5A CREs, luciferase activity 183 was measured from minimal promoter constructs containing each putative regulatory element in MCF-7 cells while 184 overexpressing either SIX1 or SIX1-Q177R alongside the cofactor EYA1. As shown in Figure 4B, both wild type and mutant 185 SIX1 drove significant expression of luciferase from the proximal DNA element and both distal DNA elements. These data 186 support the regulation of WNT5A expression by both wild type SIX1 and SIX1-Q177R. Nevertheless, overexpression of the 187 proteins in these assays could mask subtle, yet biologically significant differences in their regulatory activities when 188 expressed at normal physiological levels.
189
The SIX1-Q177R mutation is almost exclusively heterozygous in Wilms tumors and in the heterozygous context, 190 both wild type SIX1 and SIX1-Q177R alleles were found to be expressed at similar levels (Walz et al 2015, Wegert et al gene expression in vivo. To interrogate this possibility, we carried out electrophoretic mobility shift assays (EMSAs) using 193 purified wild type SIX1 or SIX1-Q177R protein expressed in E. coli and biotinylated oligonucleotide probes containing the 194 putative DNA binding sequence found within the WNT5A proximal peak which is highly congruent to the primary SIX1-To relate our findings thus far to the normal human nephrogenic niche, wk14hFK and wk17hFK single cell RNAseq datasets (GSE112570, GSE139280, GSE124472 (only sample GSM3534656)) were integrated using Seurat to 209 generate a powerful reference with which to assess the expression and localization of the differentially expressed Wilms
219
Probing this dataset for the SIX1/2miRNA upregulated genes shown in Figure 3B
247
The hFK expression patterns of those genes that were expressed in all Wilms tumors analyzed and were also identified as
264
Subsequently, genes once characterized almost exclusively as pre-induction NPC markers are now known to be 265 expressed in differentiating structures including the PTA and epithelialized RV. This includes SIX2, CITED1, TMEM100, al 2018c). This raises the likelihood that SIX1 plays a meaningful functional role in the gene regulatory networks governing podocyte lineage specification. Thus, the persistent expression of these genes in Wilms tumors does not necessarily reflect in which cellular protrusions termed "foot processes" interdigitate and surround the glomerular capillaries to filter the 273 incoming blood. While it is well-understood that podocytes derive from the NPCs, the temporal relationship and . IGF2 was the most highly expressed gene in our differential expression analysis (log2 CPM = ~15). Therefore, 306 synergism between SIX1/2-Q177R and IGF2 overexpression or miRNA processing mutations may be required for 307 tumorigenesis.
Interrogation of the mechanism through which the podocyte lineage might contribute to Wilms tumor and be further 309 promoted by SIX1/2-Q177R will require an in vitro model system that recapitulates this developmental progression. Tran
322
Nonetheless, our use of stringent filtering, expression, and statistical thresholds increases the likelihood of biologically 323 significant findings resulting from of our differential gene expression analysis. Conversely, these same stringent thresholds 324 likely obscured some differentially expressed genes of biological significance. For example, NPC and podocyte marker 325 WT1, as well as CDH1 encoding E-cadherin, were excluded from this analysis due to several tumors not meeting the 326 expression threshold. Additionally, the SIX1-Q177R ChIP-seq data utilized in this study was derived from a chemotherapy-Q177R tumors used in the RNA-seq analysis. Therefore, inference of putative target gene regulation by Top, chemiluminescence images of blots from Electrophoretic Mobility Shift Assays (EMSA) using varying concentrations three times with PBS + 0.1% TWEEN-20, 5 minutes per wash. Secondary antibody solutions were made in PBS + 0.1% TWEEN-20 + 1% dry milk. Membranes were incubated in secondary antibody solution for 1 hour at RT. Membranes In vitro transcription/translation was carried out using PURExpress In Vitro Protein Synthesis kit, following the α-rabbit HRP, 1:5000) (Supplemental Figure 1D). Protein concentration was quantified using band intensities obtained 494 using the Gel tool in ImageJ (Schneider et al 2012). Standard curve was generated using known recombinant GST bands 495 in gel. Using Microsoft Excel, a logarithmic line of best fit was generated and used to solve for mass of the in vitro kDa, nuclease-free water was added to bring the molarity of each sample to 4.5µM. Aliquots were made and stored at - was generated between all tumor samples using all genes that passed limma-voom low expression filtering, excluding 519 duplicate genes. Log2 CPM values were first scaled using the scale() function in R and correlation coefficients were 520 generated using the cor() function. Unsupervised hierarchical clustering and heatmap generation was performed using Raw and processed data was obtained from three studies: GSE112570, GSE139280, GSE124472 (only sample GSM3534656). All subsequent data processing and analyses was performed using the Seurat package and following the analysis workflows outlined in vignettes "PBMC 3K guided tutorial" and "Introduction to scRNA-seq integration" TranWk17zone1 -nFeature_RNA = 1000-5000, mitochondrial counts < 5%. Each dataset was normalized independently 533 and variable features were identified independently. Integration features were selected and integration anchors were 534 identified. An integrated assay was then created, data was scaled, PCA and UMAP dimensional reduction were 535 performed using n=20 principle components/dimensions. Neighbors were found and clusters were found using resolution 536 = 0.5. Cluster markers were found using FindAllMarkers() function, min.pct = 0.15, logfc.threshold = 0.25.
537
AverageExpression() function was used to calculate average expression value for each cluster, used return.seurat = 538 TRUE to return SeuratObject with scaled and centered expression values generated from ScaleData() function. Dot plots 539 were generated using DotPlot() function, heatmaps were generated using DoHeatmap() function, and violin plots were 540 generated using VlnPlot() function. Ridgeline plots were generated using the ggridges package in R (Wilke, 2022).
543
• Cloning and plasmids 544 pBV-Luc was a gift from Bert Vogelstein (Addgene plasmid # 16539; http://n2t.net/addgene:16539; with NheI and HindIII, however this digestion removed the minimal promoter from the pBV-Luc vector. SIX1_enhancer 547 gBlocks (Supplemental Table 2) were PCR amplified and digested with NheI and HindIII restriction enzymes and 548 annealed with digested pBV-Luc. To re-insert the minimal promoter sequence, single-stranded DNA oligos containing the 549 minimal promoter sequence flanked by 5'-HindIII and 3'-NcoI restriction sites (Supplemental Table 2) were annealed in 1X 550 annealing buffer (10mM Tris Base, 50mM NaCl (Fisher Chemical), 1mM EDTA (Invitrogen)) and incubated on 551 thermocycler at 95C for 2 minutes followed by cooling to 25C at a rate of -0.1C/second. Annealed minimal promoter oligo 552 and SIX1_enhancer-pBV-Luc were then digested with HindIII and NcoI restriction enzymes and annealed. This plasmid 553 was then used for all subsequent cloning of WNT5A proximal and distal CRE luciferase constructs using NheI/HindIII 554 restriction sites (Supplemental Table 2).
For each biological replicate of each DNA element assayed, three wells would be transfected with 5 µg total DNA/well: one no protein control condition, one SIX1/EYA1 condition, and one SIX1-Q177R/EYA1 condition. Tris-HCl, 1 mM EDTA, 1 mM DTT, 1% Triton-X100 (Fisher BioReagents)) four times. 200 µL of PreScission Protease (Cytiva) was mixed with 9.8 mL cleavage buffer, 5 mL added to each column, inverted to mix, and incubated at 4C for 2 then rinsed thoroughly with H2O.
610
The entire volume of flow through was transferred to dialysis tubing, ends were clipped shut and incubated 611 overnight at 4C submerged in 1 L dialysis buffer (50 mM Tris-HCl, 1 mM EDTA, 0.8 mM DTT) with gentle stirring. Used 612 dialysis buffer was discarded, replaced with fresh 1 L dialysis buffer, and incubation continued at 4C for 2.5 hours. to each protein solution to bring concentrations to 2 mg/mL; aliquots were stored at -80C. Protein purification was 622 validated by SDS-PAGE of a dilution series of each protein solution, followed by Western blot using α-SIX1 antibody (Cell staining following kit manufacturer's protocol. Stained membranes were imaged using iBright FL1500 Imaging System.
|
v2
|
2022-11-04T18:26:31.565Z
|
2022-11-16T00:00:00.000Z
|
253299061
|
s2orc/train
|
Development and validation of novel nomograms to predict survival of patients with tongue squamous cell carcinoma
BACKGROUND There is no unified standard to predict postoperative survival in patients with tongue squamous cell carcinoma (TSCC), hence the urgency to develop a model to accurately predict the prognosis of these patients. AIM To develop and validate nomograms for predicting overall survival (OS) and cancer-specific survival (CSS) of patients with TSCC. METHODS A cohort of 3454 patients with TSCC from the Surveillance, Epidemiology, and End Results (SEER) database was used to develop nomograms; another independent cohort of 203 patients with TSCC from the Department of Oral and Maxillofacial Surgery, First Affiliated Hospital of Zhejiang University School of Medicine, was used for external validation. Univariate and multivariate analyses were performed to identify useful variables for the development of nomograms. The calibration curve, area under the receiver operating characteristic curve (AUC) analysis, concordance index (C-index), net reclassification index (NRI), and decision curve analysis (DCA) were used to assess the calibration, discrimination ability, and clinical utility of the nomograms. RESULTS Eight variables were selected and used to develop nomograms for patients with TSCC. The C-index (0.741 and 0.757 for OS and CSS in the training cohort and 0.800 and 0.830 in the validation cohort, respectively) and AUC indicated that the discrimination abilities of these nomograms were acceptable. The calibration curves of OS and CSS indicated that the predicted and actual values were consistent in both the training and validation cohorts. The NRI values (training cohort: 0.493 and 0.482 for 3- and 5-year OS and 0.424 and 0.402 for 3- and 5-year CSS; validation cohort: 0.635 and 0.750 for 3- and 5-year OS and 0.354 and 0.608 for 3- and 5-year CSS, respectively) and DCA results indicated that the nomograms were significantly better than the tumor-node-metastasis staging system in predicting the prognosis of patients with TSCC. CONCLUSION Our nomograms can accurately predict patient prognoses and assist clinicians in improving decision-making concerning patients with TSCC in clinical practice.
INTRODUCTION
Tongue squamous cell carcinoma (TSCC) is the most common malignancy of the oral cavity and pharynx and has a high risk of local invasion and lymph node metastasis[1-3]. Surgical resection is the first-line treatment, followed by adjuvant radiotherapy, chemotherapy, or chemoradiation therapy. Despite substantial improvements in diagnostic techniques and multimodal treatment in recent years, the survival rate of TSCC remains low [4,5].
Treatment strategies for TSCC and its prognosis are based principally on the tumor-node-metastasis (TNM) cancer staging system established by the American Joint Committee on Cancer (AJCC) [6]. However, the prognoses can vary among patients with the same TNM stage who are receiving similar treatments[7-9]. Such variation suggests that the TNM staging system does not adequately predict prognosis because it does not consider patient characteristics (e.g., age and marital status) or treatment ( e.g., type of surgery) [10,11]. Therefore, a new model that incorporates these variables is required to supplement the TNM staging system and accurately predict patient prognoses.
A nomogram is a graphical model that estimates the probability of a clinical event for an individual patient based on specific biological and clinical factors [12]. Nomograms are more accurate than the TNM staging system in predicting prognoses; they have been widely used to evaluate gastric [13][14][15], hepatocellular [16][17][18][19], and head and neck [20][21][22][23] carcinomas. However, there are few studies regarding the prediction of the prognosis of TSCC. Although Mair et al[24] predicted the prognosis of TSCC, the clinical utility of the prediction model (i.e., whether they facilitate decision-making and thus improve patient outcomes[12]) was not evaluated; thus, the model would be difficult to apply in clinical practice. Currently, individually predicting the prognosis of patients with TSCC remains insufficient.
Therefore, this study aimed to develop nomograms for predicting overall survival (OS) and cancerspecific survival (CSS) in patients with TSCC to externally validate the established nomograms (discrimination, calibration, and clinical utility) and to assist clinicians in improving therapeutic decisionmaking.
Patient selection
Patients diagnosed with TSCC between 2010 and 2015 were selected from the Surveillance, Epidemiology, and End Results (SEER) database using SEERStat 8.3.9.2. The inclusion and exclusion criteria are shown in Figure 1. Overall, 3454 cases were selected as the training cohort for the development of new nomograms. When performing the internal validation, it was assigned by the bootstrapping method. Another independent cohort that was diagnosed between January 2010 and December 2020 was obtained from the Department of Oral and Maxillofacial Surgery, First Affiliated Hospital of Zhejiang University School of Medicine. The National Comprehensive Cancer Network diagnosis and treatment guidelines for TSCC were followed. Using the same inclusion and exclusion criteria, 203 cases were selected as a the validation cohort to externally validate the established nomograms ( Figure 1). We retrospectively retrieved data regarding age, sex, marital status, ethnicity, tumor site, T stage, N stage, TNM stage, pathology grade, neck dissection status, and radiation treatment status. The tumor grading system of the 7 th edition of the AJCC Cancer Staging Manual was used. The subclassifications of each variable are shown in Table 1
Statistical analysis and nomogram development
First, descriptive statistics were generated for the demographic and tumor clinicopathological characteristics. Then, univariate and multivariate Cox proportional hazards models were constructed. Coefficients, hazard ratios, and 95% confidence intervals (CIs) were obtained for prognostic factors in the training cohort. Finally, nomograms that integrated significant independent risk factors were constructed based on the predicted 3-and 5-year OS and CSS in the training cohort. OS was defined as the time from surgery until death from any cause or the last follow-up. CSS was defined as the time from surgery until death from TSCC or the last follow-up.
Validation and evaluation of nomograms
Internal and external validation analyses were performed to assess the predictive accuracies of the nomograms for the training and validation cohorts. Discriminative ability was evaluated based on the concordance index (C-index) and area under the receiver operating characteristic curve (AUC). The Cindex and AUC values are often used interchangeably and range from 0.5 to 1 (no discrimination ability and perfect discrimination, respectively)[12]. Meanwhile, a C-index or AUC value of > 0.7 indicates satisfactory discrimination. The concordance between predicted and actual survival was assessed using calibration curves. The reference line is a 45° diagonal line that ideally includes both predicted and actual survival rates.
The clinical benefits and utility of the nomograms were compared with those of the TNM staging system using the net reclassification index (NRI) and decision curve analysis (DCA). The NRI is used to assess the predictive accuracies and utility of nomograms [25,26]. The DCA is used to estimate the clinical and net benefits of nomograms based on threshold probabilities [27,28]. A horizontal reference line indicates that no intervention was performed (i.e., there was no clinical benefit), while an oblique line indicates that all patients underwent the intervention (i.e., the clinical benefit was maximized).
R statistical software (ver. 4.0.5; R Development Core Team, Vienna, Austria) was used to perform all analyses. P values < 0.05 were considered statistically significant.
Clinicopathological characteristics
The clinicopathological characteristics of the SEER cohort and our cohort are described in Table 1. Most of the patients [training cohort, n = 1049 (30.4%); validation cohort, n = 65 (32.0%)] were aged 50-59 years, and approximately 60% patients were men. Overall, the proportion of married patients was significantly greater than that of unmarried patients; the proportion of married patients was greater in the validation cohort [n = 179 (88.2%)] than in the training cohort [n = 2098 (60.7%)]. Approximately 90% of patients in the training cohort were White, whereas all patients in the validation cohort were Asian. In both cohorts, the proportion of TSCCs located on the anterior 2/3 of the tongue was greater than that located on the base of the tongue (training cohort, 74.6% vs 25.4%; validation cohort, 82.3% vs 17.7%, respectively). In both cohorts, most TSCCs were stage T1 and T2 [training cohort, n = 2815 (81.5%); validation cohort, n = 186 (91.6%)]. Meanwhile, more than half of all TSCCs were stage N0 [training cohort, n = 1920 (55.6%); validation cohort, n = 133 (65.5%)], while a few TSCCs were stage N3 [training cohort, n = 48 (1.4%); validation cohort, n = 1 (0.5%)]. The proportion of TSCCs was evenly distributed across subclassifications of TNM stages. Approximately half of the TSCCs in the training cohort was moderately differentiated, whereas 69.5% of TSCCs in the validation cohort was well-differentiated. Most of the patients in both cohorts underwent neck dissection [training cohort, n = 2491 (72.1%); November 16, 2022 Volume 10 Issue 32 validation cohort, n = 194 (95.6%)]. The proportion of patients who did and did not undergo radiation after surgery was 49.2% and 49.2% in the training cohort, and 52.7% and 36.0% in the validation cohort, respectively.
Nomogram development
Eleven candidate variables associated with OS and CSS were evaluated by univariate and multivariate Cox analyses of the SEER cohort. Univariate analysis showed that age, marital status, ethnicity, tumor site, T stage, N stage, TNM stage, pathology grade, neck dissection status, and radiation treatment status were significantly associated with OS and CSS in (P < 0.05 for all; Tables 2 and 3). Multivariate analysis showed that age, marital status, tumor site, T stage, N stage, pathology grade, neck dissection status, and radiation treatment status were independently associated with OS and CSS (P < 0.05 for all; Tables 2 and 3). Based on the results of the multivariate analysis, eight prognostic variables (age, marital status, tumor site, T stage, N stage, pathology grade, neck dissection status, and radiation treatment status) were used to develop the nomograms. Figure 2 shows the OS and CSS predictions from the nomograms. N and T stages had the greatest effects on OS followed by tumor site and age. N stage had the greatest effect on CSS followed by T stage and tumor site. Generally, OS and CSS were better in younger patients with
Nomogram validation and evaluation
The results of the internal and external validation analyses are shown in Figure 3. In the training cohort, the internal calibration curves indicated excellent consistency between the predicted and actual 3-and 5year OS and CSS ( Figures 3A, B (Table 4). Overall, the nomograms exhibited satisfactory discrimination and calibration.
Comparison of clinical utility between the nomograms and the TNM staging system
The C-index values of the TNM staging system for OS and CSS were also estimated in both the internal and external validation analyses (Table 4). The C-index values of the nomograms were higher than those of the TNM staging system (Table 4). In terms of predictive accuracy, the AUC values for the nomograms were higher than those of the (Table 4). Notably, the nomograms performed significantly better than the TNM staging system in both the training and validation cohorts.
The DCA was used to compare clinical benefits between the nomograms and the TNM staging system. As shown in Figure 5, the nomograms exhibited greater net benefits than the TNM staging system at all threshold probabilities in the training cohort (i.e., they were better able to predict both 3and 5-year OS and CSS). For the 3-year OS and CSS in the validation cohort, the net benefits of the TNM staging system were generally equivalent to the nomograms, whereas the nomograms showed greater net benefits than the TNM staging system at almost all threshold probabilities for the 5-year OS and CSS.
DISCUSSION
We developed new nomograms to predict the 3-and 5-year OS and CSS in patients with TSCC, evaluated their discrimination and calibration abilities, and compared their clinical utilities with those of the TNM staging system. Our results showed that our nomograms accurately predicted both the OS and CSS of patients with TSCC. Additionally, the C-index and AUC values along with the calibration curves showed that the nomograms had satisfactory discrimination and calibration. Moreover, compared with the TNM staging system, the predictive accuracies of OS and CSS were higher for the nomograms, as revealed by the NRI values and DCA curves. Thus, the aforementioned results indicate that our nomograms exhibited satisfactory discrimination, calibration, and clinical utility.
In this study, age, marital status, tumor site, T stage, N stage, pathology grade, neck dissection status, and radiation treatment status were selected to develop nomograms to predict the 3-and 5-year OS and CSS of patients with TSCC. As an example, Figure 2 compares two patients with similar staging results November 16, 2022 Volume 10 Issue 32 but different treatments. The first patient was 60 years old, married, and with T2 and N1 stage cancer on the anterior 2/3 of the tongue that exhibited moderate differentiation; that patient underwent neck dissection and received postoperative chemotherapy. The second patient was 70 years old, unmarried, and with T2 and N1 stage cancer on the anterior 2/3 of the tongue that exhibited high differentiation; that patient underwent neck dissection but did not receive radiation treatment. According to the conventional TNM staging system, both patients had the same TNM stage and therefore should have similar OS. However, our nomograms predicted that the respective 3-and 5-year OS were 64% and 55% for the first patient, whereas they were 43% and 33% for the second patient. The inclusion of additional information regarding clinicopathological characteristics and demographics provides our nomograms with a more accurate prognosis prediction ability; we expect these nomograms to serve as a powerful supplement to the TNM staging system for predicting prognoses. The N stage had the greatest prognostic power followed by T stage, tumor site, and age ( Figure 2). Advanced T and N stages were associated with poor OS and CSS, consistent with findings in previous studies [4,9]. These results indicate that the prognosis of patients with TSCC is greatly affected by the T and N stages; the more advanced the T and/or N stage, the worse the OS and CSS. Meanwhile, the inclusion of age and radiation treatment status in our nomograms may be considered controversial. Previous studies revealed that age was independently associated with both OS and CSS; younger patients had better survival, whereas older patients had a significantly greater mortality risk [29][30][31]. Moreover, compared with younger patients, older patients with advanced tumor stages (III, IV) had a nearly two-fold greater mortality risk. Similar to radiation treatment, surgery alone is generally associated with a high risk of relapse, particularly in patients with advanced TSCC; adjuvant therapies are thus necessary [32]. Radiation treatment has been shown to improve locoregional control and survival in patients with TSCC after surgery, particularly in advanced cases [33][34][35][36]. Here we found that the ability of radiation treatment status for predicting OS and CSS was not inferior to that of pathology grade (Figure 2). Additionally, as shown in Tables 2 and 3, age and radiation treatment status were independent predictors of OS and CSS in patients with TSCC. Taken together, our results indicate that age and radiation treatment status have prognostic significance. It has been demonstrated that marital status is an independent prognostic factor in patients with TSCC[9]. Married patients had better OS and CSS than unmarried patients [37], which is consistent with our findings in this study. We found the independent and significant role of marital status as a prognostic factor of patients with TSCC. In addition to the above variables, our study identified tumor site, pathology grade, and neck dissection status as independent prognostic factors of patients with TSCC. The OS and CSS of patients with TSCC are affected by these factors, which are shown in Tables 2 and 3, and Figure 2.
Our nomograms accurately and effectively predicted the prognosis of patients with TSCC and exhibited high clinical potential. The satisfactory discrimination and calibration abilities of these nomograms were confirmed by the calibration and receiver operating characteristic curves as well as the C-index and AUC values. The C-index values in external validation were higher than that in the training cohort, which is consistent with that constructed by Lu and Zhang for predicting tongue cancer and low-grade endometrial stromal sarcoma, respectively [7,38]. These results may indicate the extensionality and applicability of the constructed model. Moreover, we also compared the clinical utilities of the established nomograms with that of the TNM staging system, with the NRI values indicating that our nomograms had significantly better predictive accuracy. Similarly, DCA revealed that the nomograms had more clinical benefits and were better able to predict survival compared with the TNM staging system. To reduce potential bias, we used multi-institution and multi-population data from the SEER database to develop our nomograms and to validate their discrimination and calibration abilities as well as their clinical utilities in both internal and external cohorts. Additionally, we adhered to the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis statement [39]. In summary, our nomograms were used to accurately determine the clinical prognosis of patients with TSCC.
Due to its retrospective nature, this study has some limitations. First, the depth of invasion (DOI) has been recognized as an independent predictor of survival [8,40]. Among the tumor parameters that were significant for prognosis, such as the tumor width, area, volume, and depth, the DOI was considered the most important [41]. Additionally, extranodal extension (ENE) has been widely recognized as a significant poor prognostic factor for patients with HNSCC [42,43]. Hence, the DOI and ENE were incorporated into the T and N classification, respectively, in the AJCC 8th edition of the cancer staging manual [44]. However, they were not available in the SEER database, thus not being included in our constructed model. Further improvements by incorporating these factors into the constructed nomogram should be undertaken in the future. Second, the current model only incorporates clinicopathological parameters to predict patient outcomes, which is nonsufficient for screening patients appropriate for adjuvant therapies, especially preoperative/postoperative adjuvant immunotherapy. More molecular markers should be incorporated into the constructed model to improve its clinical application value, such as PD-1[45-47], CD47[48], CXCL11 [49], and CXCR3[50], which have been reported to engage in tumor immunity and included in some efficient predictive models. Third, this retrospective study had an unavoidable risk of selection bias. Thus, prospective validation studies are needed before these nomograms can be used in clinical practice.
CONCLUSION
We used two databases to develop and validate new nomograms for predicting the 3-and 5-year OS and CSS in patients with TSCC. Compared with the TNM staging system, these nomograms exhibit greater accuracy, effectiveness, and clinical utility for predicting the prognosis of patients with TSCC. Thus, they are a strong complement to the TNM staging system in the prediction of patient prognosis.
|
v2
|
2022-11-04T18:43:15.961Z
|
2022-11-16T00:00:00.000Z
|
253300368
|
s2orc/train
|
Long survival after immunotherapy plus paclitaxel in advanced intrahepatic cholangiocarcinoma: A case report and review of literature
BACKGROUND Intrahepatic cholangiocarcinoma (iCCA) is the second most common primary hepatic malignancy worldwide. However, currently available systemic therapies are of limited effectiveness, and the median overall survival of patients treated with first-line standard chemotherapy is less than one year. Immune checkpoint inhibitors have been used to treat solid tumors. Clinical studies recently explored the combination of chemotherapy and immunotherapy for CCA. However, the clinical significance of predictive biomarkers for chemo-immunotherapy in CCA remains unclear. It is also worth exploring whether a combination of chemotherapeutic agents can increase the sensitivity of CCA immunotherapy. CASE SUMMARY This study reports a case of advanced iCCA in which clinical complete remission had been achieved using a programmed death 1 (PD-1) inhibitor and paclitaxel without known predictive biomarkers, but with BRCA1, KRAS, and NTRK3 mutations after rapid progression to first-line chemotherapy, and has remained in clinical complete remission for more than two years. This case suggests that chemo-immunotherapy is a potential therapeutic option for patients with iCCA and few known predictive biomarkers for immunotherapies as well as synergistic effect of the combination of paclitaxel and PD-1 monoclonal antibody. CONCLUSION The combination of paclitaxel and PD-1 monoclonal antibodyr can be explored in patients with advanced iCCA.
INTRODUCTION
As of 2020, primary liver cancer is the sixth most commonly diagnosed cancer and the third leading cause of cancer-related death worldwide [1]. Cholangiocarcinoma (CCA) is the second most common primary liver malignant tumor after hepatocellular carcinoma, accounting for 10%-20% of all primary liver cancers. According to anatomical location, it can be divided into intrahepatic CCA (iCCA), hilar CCA, and distal CCA. The incidence of iCCA has been increasing over the past few decades [2]. Owing to the characteristics of hidden onset and lack of typical clinical symptoms, iCCA is often diagnosed at an advanced stage. Only 30%-40% of patients receive surgical treatment after diagnosis, and the postoperative recurrence rate is high [3]. There is an urgent need to develop new strategies for the predictive diagnosis of biliary tract cancer (BTC) at an early, resectable stage. Liquid biopsy has received increasing attention over the years, given its promising application in cancer patients. In CCA the detection of circulating tumor cell, circulating free DNA and extracellular vesicles has tremendous potential applications in the early diagnosis of CCA and monitoring of treatment response [4,5]. But now due to the limitations of diagnostic tools, even after radical surgery, the 5-year overall survival rate is less than 40%, and the median overall survival (mOS) time is approximately 28 mo [6]. Also, local treatments include transarterial radioembolization, hepatic artery infusion, transarterial chemoembolization and radiofrequency ablation, which have been shown to improve the survival of ICC [7,8].
So far, chemotherapy is the primary treatment for patients with locally advanced or metastatic iCCA. A phase III clinical trial (ABC-02) reported that the mOS time and median progression-free survival time of patients with advanced CCA treated with the gemcitabine + cisplatin (GemCis) regimen were significantly longer than those treated with gemcitabine monotherapy (mOS time: 11.7 mo vs 8.1 mo; hazard ratio, 0.64; 95% confidence interval, 0.52-0.80; P < 0.001) [9]. Because of the ABC-02 results, the GemCis regimen was promoted as the standard first-line chemotherapy for advanced CCA. In addition, clinical trials reported the efficacy of albumin paclitaxel combined with gemcitabine as the first-line treatment [10,11]. Immunotherapy targeting programmed death 1 (PD-1), programmed cell death ligand 1 (PD-L1), and cytotoxic T-lymphocyte antigen-4 was recently approved for the treatment of different types of cancers, especially those with high PD-L1 expression, high tumor mutation load (TMB), or microsatellite instability (MSI) [12][13][14]. TOPAZ-1 is the first phase III clinical study using immunotherapy in combination with chemotherapy as the first-line treatment of CCA, with a significant improvement in mOS resulting in a breakthrough in 1-year survival in the chemo-immunotherapy group [15]. Considering the response of BTC to immunotherapy, reliable biomarkers of response to PD-1/PD-L1 inhibitor in BTC are still not identified and developed, clarifying the role of PD-L1 expression, MSI, mismatch repair, TMB and other emerging predictors [16]. However, after progression on first-line therapy quality evidence of second-line treatment is lacking, preventing standardized follow-up treatment after disease progression. The phase III clinical trial ABC-06 study showed that, as second-line treatment, the FOLFOX or Nal-IRI + 5-FU/LV regimen prolonged the survival of patients after the GemCis compared to palliative treatment (mOS time, 6.2-8.6 mo)[17,18], but the survival benefit was very limited.
Potentially actionable molecular alterations are identified in about 50% of iCCA cases [19,20]. Molecular profiling should be considered for all biliary tract cancer patients who may benefit from the discovery of a potentially actionable mutation, especially FGFR2 fusions or rearrangements and IDH1 mutations[20,21]. However, the benefit of using immunotherapy combined with chemotherapy as second-line therapy for these biomarker-negative iCCA patients is unclear, and the choice of chemotherapy regimen has a variable impact on the efficacy of immunotherapy. Here we report a case of metastatic iCCA treated with anti-PD1 monoclonal antibody combined with paclitaxel after first-line GemCis chemotherapy failure. After six treatment cycles, the best effect achieved was clinical complete remission with mild adverse reactions.
History of present illness
A 67-year-old man was admitted to the hospital with an elevated CA199 level (579.9 U/mL; reference range, 0-37 U/mL) but no discomfort. Enhanced computed tomography (CT) of the abdomen revealed a mass with a maximum diameter of 5.7 cm in S8 of the liver, and pulmonary CT showed multiple small nodules in both lungs ( Figure 1A and B).
History of past illness
No history of biliary stones, no history of hepatitis, no history of hepatic schistosomiasis.
Personal and family history
Denied smoking, drinking and history of epidemic disease. No family history of tumors.
Physical examination
The abdomen was soft, without pain, and no obvious masses were palpated in the abdomen.
Imaging examinations
Enhanced CT of the abdomen revealed a mass with a maximum diameter of 5.7 cm in S8 of the liver, and pulmonary CT showed multiple small nodules in both lungs.
FINAL DIAGNOSIS
Liver mass: iCCA considered.
Referring to the ABC-02 clinical trial, the patient was given first-line chemotherapy with GemCis for two cycles. But the patient presented with right-sided chest pain on breathing and no significant decrease in CA199, so an evaluation was performed upfront. However, the response evaluation suggested disease progression with CT showing an increased number and size of metastases in the abdomen and lungs ( Figure 1A and B). Genomic alteration testing was performed to explore the potential drug targets using the next-generation sequencing assay, which contains 520 genes that are related to cancer mechanism and targeted therapy. The results showed three somatic mutations, including BRCA1, KRAS, and NTRK, with MSI being stable and TMB 3.2 mut/Mb defined as TMB-low ( Figure 1C). PD-L1 expression was detected by immunohistochemical staining (Dako 22C3) with tumor proportion score (TPS) 0% and combined positive score 5.
The second-line treatment was changed to camrelizumab 200 mg in combination with paclitaxel 175 mg/m 2 every three weeks. The serum concentrations of CA199 and CA125 decreased to 452.3 U/mL and normal, respectively, and partial response was maintained based on CT scans after two cycles of Computed tomography (CT) of the abdominal and lung lesions throughout therapy. July 2019, CT images of lung and liver on presentation, benign proliferative lesion of lung was considered at that time; September 2019, CT images of lung and liver after 2 cycles of GemCis chemotherapy, the intrahepatic and intrapulmonary lesions were significantly larger than before, and the peritoneal lesions were increased and pleural effusion appeared; January 2020, CT images of lung and liver after 6 cycles of paclitaxel and programmed death 1 monoclonal antibody, lesions of liver and lung were clinical complete remission; C: Genomic alteration assay result; D: The Reactive cutaneous capillary endothelial proliferation developed in the oral gingiva after the third cycle of camrelizumab with paclitaxel and was surgically removed after 2 mo. iCC: Intrahepatic cholagiocarcinama cancer; PR: Partial response; PD: Progressive disease; cCR: Clinical complete remission; GemCis: Gemcitabine and cisplatin; IO: Immunotherpay; Pac: Paclitaxel; PD-1: Programmed death 1. combination chemotherapy and immunotherapy. After six cycles of therapy, both intrapulmonary and intra-abdominal lesions achieved clinical complete remission on CT ( Figure 1A and B). After eight cycles of combination therapy, the treatment was changed to camrelizumab 200 mg every three weeks until the completion of 1 year of anti-PD-1 antibody treatment.
OUTCOME AND FOLLOW-UP
This patient has now been followed-up for 32 mo (until May 2022), and no disease progression has been observed at regular follow-ups. Reactive cutaneous capillary endothelial proliferation (RCCEP) developed in the oral gingiva and facial skin with bleeding after the third cycle of chemo-immunotherapy. The RCCEP in the gingival area ( Figure 1D) was surgically removed after 2 mo, while that on the facial area self-exfoliated with no other serious adverse effects during treatment.
DISCUSSION
There is currently no standard second-line treatment for advanced iCCA. mFOLFOX has been recommend as the preferred second-line regimen, but its survival benefit is very limited, with an mOS time of 6.2 mo vs 5.3 mo, P = 0.031 [17]. Immunotherapy has shown relatively good efficacy against a variety of solid tumors, and data from small samples of cholangiocellular carcinoma suggest the efficacy of PD-1 monoclonal antibody in advanced cholangiocellular carcinoma [22,23]. Approximately 11% of cholangiocellular carcinomas have an immunoinflammatory phenotype, as defined by comprehensive genomic analysis, which may be predictive of the effectiveness of PD-1 or PD-L1 monotherapy [24]. Other recent studies have explored the molecular typing of CCA by genomics and proteomics, which show the different expression patterns of immune checkpoints, highlight the need to design personalized checkpoint inhibitors for use, and provide clues to explain the differential response of advanced iCCA to anti-PD-1 monotherapy [25].
In this case, MSI, TMB, and PD-L1 TPS expressions were low, but mutations in BRCA1 and KRAS were present. BRCA mutations were detected in approximately 3.6% of CCA samples (BRCA1, 0.6%; BRCA2, 3%) [26]. PARP inhibitors are expected to be the next category of targeted agents as breakthrough treatments for advanced CCA, but the research data are not sufficient [27]. The available genetic test results for this patient suggested that he was not sensitive to immunotherapy.
To improve the anticancer effect, many chemotherapeutic agents have been tested in combination with immunotherapy to modify the antitumor activity [28,29]. The TOPAZ-1 trial is the first phase III clinical study using immunotherapy plus a chemotherapy regimen of gemcitabine and platinum for the first-line treatment of biliary tract cancer to show a significant difference in mOS obtained in the treatment group [15]. Table 1 lists the reported cases of PD-1/PD-L1 inhibitors and chemotherapy in patients with CCA, and the chemotherapeutic drugs are mainly platinum and fluorouracil alone or in combination. Chemotherapeutic drugs can destroy tumor tissues and overcome immune rejection, causing antigen shedding and increasing tumor neoplastic antigens to improve immunotherapy efficacy. There is growing evidence that some chemotherapeutic agents, such as platinum and taxanes, induce apoptosis in tumor cells and cause immunogenic cell death, which activates the immune system [30,31]. The immunomodulatory effects of chemotherapeutic agents are summarized in Table 2. Chemotherapy-induced tumor cell death may release immunostimulatory signals that promote dendritic cell activation and induce T-lymphocyte-mediated immune tumor cell killing. Chemotherapy induces the upregulation of major histocompatibility complex I (MHC-I), which helps T cells recognize tumor cells. Furthermore, chemotherapeutic agents upregulate PD-L1 expression, and chemotherapyenhanced immunosuppression induced by PD-L1 upregulation can be abolished when immunotherapy is added to the treatment strategy, which is expected to produce a synergistic anti-cancer effect.
Based on these theories and the summary in Table 2, paclitaxel in combination with anti-PD-1 is a very favorable option for improving the activated immune response (upregulating MHC-I, dendritic cell maturation, and T cell effectors) and reducing immunosuppressive cells (myeloid derived suppressor cell, T-regulatory cells, and tumor-associated macrophage type 2). According to a preclinical study, most iCCA cell lines were resistant to platinum drugs, whereas most cell lines were sensitive to gemcitabine and paclitaxel [25]. Available clinical studies show that triple negative breast cancer patients who are expected to have greater BRCA mutations and urothelial cancer patients with BRCA1 mutations receive immune-combination chemotherapy regimens; PD-1 monoclonal antibody combined with paclitaxel has very good efficacy [32,33]. A hypnosis regarding these findings suggests that paclitaxel may be a better chemotherapeutic agent in combination with PD-1/PD-L1 immunotherapy for paclitaxel-sensitive solid tumors with or without BRCA gene mutations. This mechanism has been the focus of published research, where treatment with polyethylene glycol-sheddable nanodrugs containing paclitaxel and anti-PD-1 enhanced the tumor infiltration of cytotoxic T lymphocytes as well as local immune checkpoint blockade[34]. This case describes a potential clinical option for combination immunotherapy in patients with iCCA with or without BRCA1 mutations. This is the first study to investigate paclitaxel combined with a PD-1 inhibitor in patients with advanced iCCA, which signified the benefit of chemo-immunotherapy in advanced iCCA patients with low TMB, PD-L1, TPS, and microsatellite stability. Obviously, in iCCA and other types of cancer, predictive biomarkers are lacking for many patients. Thus, this case suggests that the paclitaxel and PD-1 inhibitor combination is a potential effective therapeutic option for the management of these patients.
In conclusion, chemo-immunotherapy offers a potential therapeutic option for patients with iCCA and few or no predictive biomarkers for immunotherapies, and the combination of paclitaxel as an effective chemotherapeutic agent with PD-1 monoclonal antibody may have a better synergistic effect. Future studies should better elucidate the therapeutic efficacy and potential mechanisms of action of chemo-immunotherapy in iCCA as well as the optimal combination strategy for immunotherapy.
CONCLUSION
For patients with advanced iCCA without predictive biomarkers, a regimen of immunotherpay combined with paclitaxel may be considered for treatment. And a more complex analysis will be performed to screen the population that really benefits from this treatment.
FOOTNOTES
Author contributions: He MY, Shen P designed the research study; He MY, Yan FF and Cen KL analyzed the data and wrote the manuscript; All authors have read and approve the final manuscript.
Informed consent statement: All study participants or their legal guardian provided informed written consent about personal and medical data collection prior to study enrolment.
Conflict-of-interest statement:
All the authors report no relevant conflicts of interest for this article.
|
v2
|
2022-11-04T18:46:50.391Z
|
2022-11-16T00:00:00.000Z
|
253286882
|
s2orc/train
|
Quality of life and symptom distress after cytoreductive surgery and hyperthermic intraperitoneal chemotherapy
BACKGROUND Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (CRS/ HIPEC) for peritoneal surface malignancy can effectively control the disease, however it is also associated with adverse effects which may affect quality of life (QoL). AIM To investigate early perioperative QoL after CRS/HIPEC, which has not been discussed in Taiwan. METHODS This single institution, observational cohort study enrolled patients who received CRS/HIPEC. We assessed QoL using the Taiwanese version of the MD Anderson Symptom Inventory (MDASI-T) and European Organization Research and Treatment of Cancer Core Quality of Life Questionnaire (EORTC QLQ-C30). Participants completed the questionnaires before CRS/HIPEC (S1), at the first outpatient follow-up (S2), and 3 mo after CRS/HIPEC (S3). RESULTS Fifty-eight patients were analyzed. There was no significant perioperative difference in global health status. Significant changes in physical and role functioning scores decreased at S2, and fatigue and pain scores increased at S2 but returned to baseline at S3. Multiple regression analysis showed that age and performance status were significantly correlated with QoL. In the MDASI-T questionnaire, distress/feeling upset and lack of appetite had the highest scores at S1, compared to fatigue and distress/feeling upset at S2, and fatigue and lack of appetite at S3. The leading interference items were working at S1 and S2 and activity at S3. MDASI-T scores were significantly negatively correlated with the EORTC QLQ-C30 results. CONCLUSION QoL and symptom severity improved or returned to baseline in most categories within 3 mo after CRS/HIPEC. Our findings can help with preoperative consultation and perioperative care.
INTRODUCTION
Peritoneal surface malignancy (PSM) is the spread of cancer cells inside the abdominal cavity, especially over the peritoneum, the membrane that covers the abdominal cavity. PSM was considered to be a terminal stage of cancer, and hence patients with PSM were often treated with palliative systemic therapies or supportive care [1][2][3]. PSM may cause abdominal distension, ascites, malnutrition, cachexia, and intestinal obstruction, which in turn can cause physical and mental discomfort, significantly reducing the quality of life (QoL) and shortening survival [1,[4][5][6].
However, cytoreductive surgery (CRS) combined with hyperthermic intraperitoneal chemotherapy (HIPEC) has become a treatment option beyond palliative treatment for patients with PSM [1,7]. Although CRS/HIPEC can prolong survival, it can also cause adverse effects such as postoperative ileus, wound infection, intra-abdominal abscess, bleeding, symptomatic pleural effusion, anastomotic leakage, and renal damage [7][8][9][10]. Although some of these adverse effects are short term, some may persist for a long time. The potential survival benefit must therefore be weighed against a possible reduction in QoL associated with the procedure and its complications. In addition, uncertainty of the November 16, 2022 Volume 10 Issue 32 illness and facing aggressive treatment may affect the emotional well-being of the patient [11]. Therefore, the QoL after CRS/HIPEC is an important issue [4,5].
In recent years, several Western studies have investigated the QoL after CRS/HIPEC. In a systematic review, Shan et al [4] reported that CRS/HIPEC for PSM could confer small to medium benefits for health-related QoL. However, the authors concluded that the results should be interpreted with caution due to the small studies and varying follow-up duration. Several studies have reported that the QoL of patients usually declines after surgery, but then recovers to baseline and improves in 3 to 6 mo [2,3,5,6,12]. However, most of these reported were retrospective QoL or clinical data studies. In addition, only two studies on Asian patients have been reported, and although they reported that QoL would recover in 6-18 mo after CRS/HIPEC, they both enrolled a small number of patients [2,13]. Taken together, these previous studies have all focused on the QoL 3 mo or later after surgery. Investigations of perioperative QoL and symptom severity after CRS/HIPEC are limited. However, perioperative psychological distress and changes in QoL are crucial, because they may decrease treatment acceptance by the patients and affect perioperative care by the physicians.
HIPEC has been reimbursed by the National Health Insurance system since 2019 in Taiwan, and the number of patients undergoing CRS/HIPEC has gradually increased. Consequently, the impact on QoL of this treatment has also gradually become more important due to socio-economic considerations. Contemporary cancer treatment focuses on both survival and the relief of symptoms to improve function and the QoL of patients. Thus, we conducted this prospective study to investigate changes in QoL in the perioperative stage after CRS/HIPEC, and explore the factors associated with these changes.
Study design
This was a prospective, single institution, cohort study in Taiwan. The inclusion criteria were: (1) Patients who planned to receive CRS/HIPEC at Chang Gung Memorial Hospital in Chiayi from September 1, 2018 to February 28, 2021; and (2) patients aged ≥ 20 years. The exclusion criteria were: (1) Patients who had psychiatric disorders; (2) patients unable to understand the questionnaires; or (3) patients who were not willing to complete all questionnaires. The participants were asked to complete the questionnaires at three time points (first visit, before CRS/HIPEC; second visit, the first outpatient follow-up after CRS/HIPEC; and third visit, the outpatient visit 3 mo after CRS/HIPEC). We defined the first visit as S1, second visit as S2, and third visit as S3. Data were collected using the Taiwan version of the MD Anderson Symptom Inventory (MDASI-T), and Traditional Chinese version of the Core Quality of Life Questionnaire compiled by the European Organization for Research and Treatment of Cancer (EORTC QLQ-C30). All questionnaires were completed in face-to-face interviews with the researchers and patients. The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Institutional Review Board of Chang Gung Medical Hospital (No. 201800726B0). The informed consent was obtained by all participants.
Survey measures
MD anderson symptom inventory: Symptom data were obtained using the MDASI-T [14], which contains 13 core symptom severity items and six interference items. Symptoms (pain, fatigue/tiredness, nausea, disturbed sleep, distress, shortness of breath, difficulty remembering, lack of appetite, drowsiness, dry mouth, sadness, vomiting, and numbness/tingling) were rated at their worst in the previous 24 h on a 0-10 scale, with 0 representing "not present" and 10 representing "as bad as you can imagine". The patients also rated the degree to which the symptoms interfered with various aspects of life during the past 24 h. Each interference item (general activity, mood, work [including both work outside the home and housework], relations with other people, walking ability, and enjoyment of life) was rated on a 0-10 scale, with 0 representing "did not interfere" and 10 representing "interfered completely" [15].
QoL Questionnaire
The health-related QoL was assessed using the EORTC QLQ-C30 [16]. The questionnaire contains a total of 30 questions and covers five functional scales (physical, role, cognitive, emotional, and social function), three symptom scales (fatigue, pain, and vomiting), six symptom single item scales (dyspnea, insomnia, loss of appetite, constipation, diarrhea, and financial status), and a self-perceived global health status scale. Except for questions 29 and 30, which are answered on a scale from 1 to 7 points, the options for the other questions range from 1 ("Not at all") to 4 ("Very much"). The scores are then converted into percent scores according to the questionnaire instruction manual. In the self-perceived global health status score and functional score, the higher the score, the better the patient's function or QoL. While in the symptom score and single selection, the higher the score, the more severe the symptoms, meaning poor QoL. November
Clinical data collection
Data on the patients' characteristics, operative details, postoperative outcomes, and pathology were evaluated by the MDT committee. The prospectively collected data of the patients included demographics, pre-existing co-morbidities (diabetes, hypertension, and hepatitis), Eastern Cooperative Oncology Group (ECOG) performance status, cancer type/disease status (primary or recurrence, histology type and grade, and peritoneal carcinomatosis index (PCI)), CRS/HIPEC parameters (chemotherapy regimen, duration, and completeness cytoreduction (CC) score [18], grade of postoperative complications according to the National Cancer Institute -Common Terminology Criteria for Adverse Events (NCI-CTCAE) v.5.0, and nutritional status according to the Patient-Generated Subjective Global Assessment (PGSGA) score.
Statistical analysis
The total sample size was calculated using Gpower version 3.1. The effect size was determined to be 0.25, and the study power and alpha value were set at 80% and 0.05, respectively. Based on these inputs, a minimum sample of 44 subjects was required. Demographic data and scale scores were reported with descriptive statistics, including number, percentage, mean (standard deviation) and median (range). The student's t-test, one-way analysis of variance (ANOVA), and Pearson's correlation coefficients were used to compare differences and correlations, respectively. Multiple regression analysis was used for inferential statistics. A two-sided P value of < 0.05 was considered to be statistically significant. All analyses were performed using SAS version 9.4 (SAS Inc., Cary, NC).
Patients
During the study period, 79 patients were screened preoperatively for enrollment into the study. However, 17 patients canceled the CRS/HIPEC procedure intraoperatively after the laparoscopic examination (13 because the disease was too extensive and cytoreduction could not be completed, and four who did not have PSM and refused to receive prophylactic HIPEC). After CRS/HIPEC, four patients withdrew from the study. Therefore, a total of 58 patients completed the study and were eligible for analysis ( Figure 1). However, three patients returned to their original hospitals for further salvage therapy and did not complete the third questionnaire. The basic and disease characteristics of the patients are shown in Table 1. The median (range) age of all patients was 60 (22-78) years, and the most common cancer type was gastric cancer (46.6%). The median length of hospital stay was 13 days. Fifty-two patients (89.7%) had postoperative complications, of which grade I complications were the most common (72.4%). Forty-two patients (85.7%) had a PGSGA score of A.
QoL and symptoms severity
The results of the EORTC QLQ-C30 and MDASI-T questionnaires are shown in Table 2. The average preoperative global health status scores at S1, S2, and S3 were 60.3, 56.6, and 64.4, respectively. The results showed a trend of a reduction in global health status after surgery and then an improvement at S3, however there was no statistical difference (P = 0.065). On the functional scale, there were significant decreases in the physical function (P = 0.001) and role function (P = 0.004) scores at S2, which then recovered to the preoperative baseline level at S3. In the symptom and multiple-item scales, fatigue (P = 0.004) and pain (P = 0.002) significantly increased at S2. The most significant improvement at S3 was in dyspnea (P = 0.041). In the MDASI-T questionnaire, there were no significant changes in the average scores for the severity of preoperative symptoms and the degree of interference with life between S1, S2, and S3 (Table 2). In the preoperative stage, the two symptom items with the highest scores were distress/feeling upset (2.2 ± 2.1) and lack of appetite (1.7 ± 2.4). After CRS/HIPEC, the two symptom items with the highest scores were fatigue (tiredness) (2.0 ± 1.8) and distress/feeling upset (2.0 ± 2.1) at S2, and fatigue (tiredness) (2.0 ± 1.6) and lack of appetite (1.7 ± 1.8) at S3. Regarding the interference items, the items with the highest scores were working (including housework) at S1 (2.1 ± 2.9) and S2 (2.2 ± 3.0) and activity at S3 (1.5 ± 1.5). Table 3 shows the relationships among the EORTC QLQ-C30 and its related factors using the student's t-test, one-way ANOVA, and Pearson's correlation coefficients. The severity score was significantly negatively correlated with preoperative global health status (r = -0.48, P < 0.001), emotional function (r = -0.34, P < 0.01), and cognitive function (r = -0.54, P < 0.001). The score of the degree of interference with life was significantly negatively correlated with preoperative global health status and all functional scales (r = -0.39 to -0.54, P < 0.01). At S2, the physical and social function scores of the patients who were ≥ 55 years old were significantly higher than those of the patients who were < 55 years old (P < 0.05). The symptom severity score was significantly negatively correlated with role function (r = -0.45, p < 0.001), emotional function (r = -0.49, P < 0.001) and social function (r = -0.33, P < 0.05). The degree of interference with life scores was significantly negatively correlated with global health status and all functional scales (r = -0.28 to -0.63, P < 0.05). At S3, the role function score in those who were ≥ 55 years old was significantly higher than in those who were < 55 years old (P < 0.05). The scores of global health status in the patients who received chemotherapy before surgery were significantly higher than in those who did not (P < 0.05). The symptom severity score was significantly negatively correlated with role function, emotional function, cognitive function, and social function (r = -0.48 to -0.72, P < 0.001), and the degree of interference with life score was significantly negatively correlated with global health status, role function, emotional function, cognitive function, and social function (r = -0.67 to -0.78, P < 0.001).
Determinants of QoL
The results of multiple regression analysis for the significantly correlated variables in Table 3 are shown in Table 4. The results showed that the important predictors were age ≥ 55 years old in emotional functioning at S2 (β = -0.40, P < 0.05), and ECOG performance status in preoperative physical functioning (β = 21.49, P < 0.05) and role functioning at S3 (β = 29.63, P < 0.05). Both the severity of symptoms and degree of interference with life in the MDASI-T were significantly correlated with QoL as measured using the EORTC QLQ-C30.
DISCUSSION
This is the first prospective study to investigate the QoL and symptom distress after CRS/HIPEC in Taiwan. The results of this study showed that most patients had a significant decline in physical and role function scores at S2, but that they returned to the preoperative status at S3. We also found that the most serious symptoms after surgery were fatigue and pain, and that pain returned to the preoperative status 3 mo after surgery. There was no significant decline in global health status after surgery. Both items in the MDASI-T were significantly negatively correlated with the EORTC QLQ-C30 results. We also found that the risk factors associated with a perioperative decline in QoL were an age < 55 years old and poor ECOG performance (ECOG = 2). Several studies have reported that patients' functional scales, especially physical and role functional scales, declined at 3 mo and then returned to the baseline level at 6-9 mo [1,2,5,6,19]. However, we found that the physical and role function scores were lower at the first outpatient follow-up visit after surgery and then recovered to the preoperative baseline scores within 3 mo. This result is similar to that reported by Alves et al [12]. We hypothesize that the patients may have felt a loss of role function under the care of family members after surgery, and that their physical function was also limited because of surgical wounds and pain. As the wounds gradually healed, their daily role functions were restored and the functional scale scores gradually increased.
In addition, the emotional and cognitive function scores of the patients in this study showed a tendency to increase after CRS/HIPEC. This result is similar to previous studies [1,2,8,13,20]. The reason may be due to a release of anxiety over uncertainty of the surgery, and because most of the patients recognized that the cancer was being well treated and that the treatment could prolong their life. In addition, patients with positive emotions or optimistic personalities tend to have a broader scope of cognition [21].
Of the symptom scales, fatigue and pain had the worst scores at the first outpatient follow-up visit after surgery. These symptoms may be caused by laparotomy wounds and the effects of HIPEC, and have been reported in other studies [6,22]. Chia et al [2] reported that other symptoms would recover in 6-12 mo after HIPEC/CRS as well as other major surgery. In this study, the pain scales returned to baseline at 3 mo after surgery, but the other symptoms did not. In addition, 90% of the patients in this study received adjuvant chemotherapy which may have begun within 3 mo postoperatively, and this may also have contributed to the persistent symptoms.
Previous studies have reported that high PCI score, poor ECOG performance status, high CC score, longer surgery duration, and postoperative complications were related to poor QoL, and that these factors were associated with the severity of disease, complicated surgery, and prolonged recovery [2,6,7,22,23]. However, we found that PCI score, CC score, surgical duration, hospitalization duration, and postoperative complications were not associated with QoL in the perioperative period after HIPEC/ CRS. This may be due to the strict clinical criteria used in this study (e.g., 94.8% had an ECOG score £ 1 and a median PCI score of 5.5 with some receiving adjuvant HIPEC who did not have PSM) to enroll the patients with CRS/HIPEC, and this may have contributed to a better baseline physical condition.
In this study, we found that younger age (< 55 years old) was a risk factor for a decline in perioperative QoL, which is similar to previous studies [24,25]. Younger patients may have greater socioeconomic stress, lower income, and weak family support, and these may contribute to a feeling of hopelessness and low QoL [25]. We also found that the younger patients (< 55 years) had poorer emotional functioning at the early post-operative visit (S2). However, further studies are needed to include these factors in prediction models and assess their effects on QoL.
There are several strengths to this study. First, all of the patients were enrolled after the consensus of the MDT committee, and CRS/HIPEC was performed by experienced team members. Thus, the quality of perioperative care was consistent and well documented. Second, the associated clinical data were prospectively collected. In addition, to make sure that the patients could understand the questions, the questionnaires were performed by a single well-trained case manager in face-to-face interviews with the patients, and this could minimize detection bias and missing data. Third, this study focused on measuring the change in QoL in the perioperative period after CRS/HIPEC, and this could minimize interference from the subsequent adjuvant therapy.
The major limitation was some patients transferred back to their original hospital for subsequent treatment when their condition after CRS/HIPEC had become stable, so it was difficult to collect longer term questionnaires. A minor limitation was that this study included patients with different types of cancer and cancer surgery. Moreover, subgroup analyses of patients with different treatment intent and preoperative status were not performed due to the small sample size.
The balance of treatment and QoL is often a controversial issue. Our findings showed that although CRS/HIPEC resulted in a short-term decline in the QoL of patients, most functions and the severity of symptoms returned to the baseline level within 3 mo after surgery. Understanding the clinical course may relieve the patients' anxiety over their disease. We also found that perioperative symptom severity and symptom interference with daily life in the MDASI-T were significantly correlated with the decline in specific functions. Therefore, it is important to continuously evaluate and provide timely care to improve the symptoms and symptom interference of patients undergoing CRC/HIPEC, and ultimately to improve their QoL.
CONCLUSION
Our findings of an association between younger age and poor preoperative ECOG performance status with a perioperative decline in QoL may help MDT members to identify patients undergoing CRC/HIPEC who are at high risk of perioperative symptom distress and decline in QoL. Patient counseling and perioperative support may be provided accordingly. The improvement or return to baseline in QoL and symptom severity after 3 mo highlight the importance of a MDT approach towards effective teamwork for CRS/HIPEC care.
Research background
Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (CRS/HIPEC) for peritoneal surface malignancy can effectively control the disease, however it is also associated with adverse effects which may affect quality of life (QoL).
Research motivation
Investigations of perioperative QoL and symptom severity after CRS/HIPEC are limited. The impact on
|
v2
|
2022-11-05T13:42:03.877Z
|
2022-11-05T00:00:00.000Z
|
253304993
|
s2orc/train
|
The g3mclass is a practical software for multiclass classification on biomarkers
The analytes qualified as biomarkers are potent tools to diagnose various diseases, monitor therapy responses, and design therapeutic interventions. The early assessment of the diverseness of human disease is essential for the speedy and cost-efficient implementation of personalized medicine. We developed g3mclass, the Gaussian mixture modeling software for molecular assay data classification. This software automates the validated multiclass classifier applicable to single analyte tests and multiplexing assays. The g3mclass achieves automation using the original semi-constrained expectation–maximization (EM) algorithm that allows inference from the test, control, and query data that human experts cannot interpret. In this study, we used real-world clinical data and gene expression datasets (ERBB2, ESR1, PGR) to provide examples of how g3mclass may help overcome the problems of over-/underdiagnosis and equivocal results in diagnostic tests for breast cancer. We showed the g3mclass output’s accuracy, robustness, scalability, and interpretability. The user-friendly interface and free dissemination of this multi-platform software aim to ease its use by research laboratories, biomedical pharma, companion diagnostic developers, and healthcare regulators. Furthermore, the g3mclass automatic extracting information through probabilistic modeling is adaptable for blending with machine learning and artificial intelligence.
www.nature.com/scientificreports/ predict class membership probabilities, such as the probability that a given value belongs to a particular class. The diagnostic applications of Bayesian analysis in medicine have been evolving over the past few decades [15][16][17] . We recently introduced the method for multiclass cancer classification on biomarkers by applying a Bayesian approach 18,19 . The proposed method allows the discovery of previously unknown groups with different levels of biomarkers by extracting information from molecular assay data using probabilistic modeling. The performance of this method has been validated over datasets of more than 300 clinical samples. In addition, it has been shown to improve the binary classification of clinical markers (HER2 and steroid hormone receptors). Given these successes and the need to improve disease diagnosis and personalized treatment options, we developed g3mclass, a practical Gaussian mixture modeling software for molecular assay data classification. It is intended for tests where the random variation of the parameter-of-interest is an essential component of the modeled situation. In this article, we introduce and employ g3mclass using a real-world example of human breast tissues. Samples were classified on clinical markers to help distinguish patients most likely to benefit from the Food and Drug Administration (FDA)-approved targeted therapies while sparing others that need different treatment.
Results
Overview of the g3mclass's functionality: or how it works. The g3mclass is a probabilistic modelingbased classification and visualization software purpose-built for analyzing laboratory assay data. Our development of g3mclass was motivated by the outstanding problem of classifying test results in the research laboratories and clinical settings, considering the prior knowledge of the reference. In this regard, Bayesian statistical methods to update pre-existing information about the likelihood of the change provided a robust system for our development of the g3mclass. The g3mclass core function requires two kinds of data entries for each analyte: test (e.g., disease, treated) and reference (e.g., healthy control, nontreated). Additionally, we built a complementary capability in this software to classify new data (e.g., suspected disease) obtained by the same assay but from an independent source. These incoming unknown data are queries. Thus, the g3mclass workflow includes data preparation and input, probabilistic modeling, automated classification of the test, reference, and optional queries, followed by the analysis and archiving of the results (Fig. 1).
The g3mclass learns the total test GMM and separates modes composing the mixture upon data input. The GMM learning is based on our semi-constrained algorithm, a modification of the original expectation-maximization (EM) algorithm 20 . To detect test classes distinct from the reference, we constrained the position (i.e., the mean value), spread (i.e., the standard deviation value), but not the weight of the test class 0 to be equal to the corresponding reference values. All other test classes have all parameters adjustable. While our semi-constrained approach preserves parameters of the reference class, it does not affect the convergence and high speed of the www.nature.com/scientificreports/ analysis regardless of the overlap between the reference and the test values. For example, using a laptop computer, the g3mclass runtime for classifying ten analytes with thousands of measurements takes less than 20 s. In the test classes labeled with negative (e.g., − 1, − 2, etc.) and positive (e.g., 1, 2, etc.) integers, the mean values are lower and higher than class 0. The greater the absolute integer value, the further the class is positioned relative to class 0, as illustrated in a plot created by the g3mclass software ( Fig. 2A). To learn the test model, the software initializes the semi-constrained EM algorithm with classes corresponding to peaks in the histogram calculated on the test sample. The software depicts the model as a probability density function (PDF) overlaid on a test histogram. Because information about the number of bins and their boundaries is not inherent to the data but will influence a GMM, g3mclass can vary binning parameters to initialize the EM algorithm. The model learning controls include the settings of the fixed number and the varying number of bins, as well as a threshold for fusing too close Gaussians, and a threshold for vanishing Gaussians with a too low number of values, as described in detail in the g3mclass user's manual on https:// g3mcl ass. readt hedocs. org site. Suppose several histograms are used for the semi-constrained EM algorithm initialization. In that case, g3mclass automatically selects the mathematically preferred model with the lowest Bayesian information criterion (BIC) that measures the model's fit to the test data and outputs the test model parameters. The classification of references and queries relies on the selected GMM for the test. Individual models could be optimized for each analyte across the different study populations. The g3mclass additional functionalities include the output of spreadsheets with summary statistics (the mean, median, and standard deviation values for each class) and classification heatmaps. The g3mclass performs consecutively three classification types-proba, cutoff, and stringent cutoff (a.k.a. s.cutoff) for a given test model to maximize the automated classification accuracy. The g3mclass proba utilizes the Bayesian approach to classify data. It classifies each value based on a maximum a-posteriori probability estimate of class membership. There is no a priori assumption about the number of classes. However, during this initial step, some analyte values in the test may be incorrectly assigned by the proba classification. This situation may occur when a component of GMM has a wide dispersion with its tails picking up values that otherwise belong to a different class (Fig. 2B). To directly address the separation of the GMM modes, g3mclass computes a set of cutoffs and autocorrects the potential proba misclassification. At cutoff classification step, data parsing is performed based on a minimal misclassification value with equal weights relative to adjacent classes. The consecutive s.cutoff classification relies on either the left or right interval values computed for a minimal misclassification cutoff that can be interpreted as tolerable intervals of the misclassification error rate (a tradeoff between misclassification of one type for misclassification of the other kind). During s.cutoff classification, more values are assigned to class 0 by the expansion of cutoff intervals. If the weight of class 0 in the test GMM is close to null, the proba classification of reference is invalid (Fig. 2C). However, cutoff and s.cutoff classification results may be valid for test and reference, depending on the degree of overlap for those samples. The lower overlap, the higher the accuracy of the g3mclass classification.
The g3mclass has a feature that allows users to evaluate parameter stability. The user can subsample the selected fraction from the original sample (reference, test, or both) up to 100 times. Resampling is done randomly without replacement to limit the risk of creating false classes based on repeated events from far-tailed distributions. The independent GMM is learned for each resample, and parameter estimates are provided. If, for example, a user selects the option of variable bin number, the software will apply it to each resample. The optimal number of classes for each resample will be based on BIC and may or may not differ from sample to sample. The user may compare the variability of the estimates of the biomarker classifier parameters: the number of classes, the mean values, and the diagnostic cutoffs separating the reference-like values from the disease-related test values (e.g., cutoffs between class 0 and class − 1; class 0 and class 1). Hence the user may base the judgment not only on the original model but also on the potential outcomes resulting from resampling. One way to conceptualize the utility of g3mclass in the field of biomarkers is to consider its value in updating the existing knowledge about biomarkers in reference samples (e.g., before disease or before treatment) and assessing the biomarker change in the test (e.g., in disease or post-treatment).
Multiclass classification on a single analyte: biomarker or therapeutic target. First, we demonstrated the capabilities of g3mclass to automatically classify samples on the gene expression data for the established diagnostic/drug-response biomarker, the ERBB2 encoding HER2. The mRNA levels were measured by a QuantiGene Plex 2.0 (QG2) assay in the intended-use population (as described in "Materials and methods"), where there is a regular need to differentiate among the patients with breast cancer. For the test and reference input, we used mRNA measurements from invasive breast carcinoma (IBC) and mammoplasties (noncancer), respectively. Additionally, we queried this gene expression data from the independent cohort of patients diagnosed with ductal carcinomas in situ (DCIS) on the pre-operative biopsy. To estimate the number and character- Considering all three models, we performed classification of the IBC, DCIS, and mammoplasty samples on ERBB2 mRNA (Table 1). The g3mclass automatically stratified heterogeneous populations into multiple classes with differential levels of ERBB2 expression, each of which was represented by Gaussian distribution. Apart from computing the proportions of each class as shown in Table 1, the g3mclass provides spreadsheet records on individual sample membership and summary statistics. For example, depending on the test GMM, an estimated 3-15% of ERBB2 mRNA values from the reference data belonged to class 1, whereas the majority of ERBB2 mRNA values from the test and query were classified into class 2 and higher that have not been present in the reference. Thus, it was reasonable to suggest that the software computed up-2 cutoff, which separated class 2 and higher from lower classes, was a diagnostic cutoff point associated with ERBB2/HER2 overexpression in the subset of breast cancers. www.nature.com/scientificreports/ The diagnostic performance of the g3mclass modeling solutions. Going a step further, we examined how well a HER2 status may be predicted from ERBB2 mRNA expression data using the g3mclass computed cutoffs because proteins are traditional therapeutic targets. We used the standardized statistical methods requiring biomarker dichotomization. For this purpose, we considered the current FDA-approved method for determining a binary HER2 status in breast cancer as a gold standard. We stratified HER2 positive (HER2+) and HER2 negative (HER2−) IBC based on the pathology reports. To test how well the measurements of ERBB2 mRNA predict HER2+, we dichotomized the ERBB2 mRNA data into groups using g3mclass-calculated cutoffs for three models. Choosing a model-predicted cutoff value, i.e., 186.6 (4-class GMM), 147.0 (5-class GMM), and 113.9 (8-class GMM), allows the transformation of multiclass into binary classification, at the same time eliminating equivocal results (Fig. 3B). When 186.6 is chosen as the ERBB2 expression cutoff, the sensitivity is 70%, and the specificity is 99% (Table 2). When the cutoff is decreased to 147.0, the sensitivity is increased to 90%, and the specificity is reduced to 97%. When the cutoff is further reduced to 113.9, the sensitivity increases to 93%, Additionally, using the g3mclass modeling solutions, we calculated the ERBB2 mRNA diagnostic test parameters for query-an independent cohort of patients diagnosed with DCIS ( Fig. 3C and Table 3). For 75 biopsies successfully profiled in QG2 assay, HER2 status by IHC could be assessed in 70 samples freshly cut from the same blocks of tissue that remain available after gene expression analysis. For the g3mclass selected 4-class GMM with the lowest BIC = 1650.6 and 186.6 as a cutoff, the accuracy, i.e., an overall probability that a patient's pre-surgical biopsy being correctly classified on a binary HER2 status from the ERBB2 mRNA expression data, was 92.86% (95% confidence interval 84.11-97.64%), the positive predictive value was 100%, while the negative predictive value was 91.53% (95% confidence interval 84.11-97.64%). Thus, the results of g3mclass data analyses show the accuracy of identifying breast cancer potentially sensitive to anti-HER2 therapy and the robustness of the software in making the correct classification without equivocal results in the test and independent query.
Multiclass classification on multiple biomarkers. The g3mclass
can easily be upscaled to analyze data from multiplexing assays. The data processing steps are the same as those for a single biomarker, i.e., preparation and entry of input data, modeling, classification, and analysis of the output results. We ran g3mclass to concurrently classify breast tissue samples on ERBB2, ESR1, and PGR mRNA measurements obtained in the validated and highly reliable multiplex QG2 assay ("Materials and methods"). These target genes encode HER2 and human steroid hormone receptors-estrogen receptor alpha (ER), and progesterone receptor (PR), abnormal presence of which defines treatments, such as anti-HER2 and hormonal therapies 21 . We obtained clinical markers' binary status (positive vs. negative) from pathology reports ("Materials and methods"). We input the mRNA expression data from the test (IBC), reference (mammoplasties), and two queries (DCIS and five human breast cancer cell lines) as one file into g3mclass. Instantaneously, the software selected and depicted each gene's mathematically favorable test model (Fig. 4A). It also performed three sequential classifications (proba, cutoff, and s.cutoff) for each model. Finally, it summarized data into spreadsheets and heatmaps. Using the resampling feature of the g3mclass, we found that diagnostic cutoff estimates for ESR1 were stable despite the appearance/ disappearance of far-tailed classes (Supplementary Table 1). To illustrate the essence of the tumor's classification on ESR1 in the context of the other two genes, we present heatmaps built for s.cutoff classification that improves the specificity (Fig. 4B-E).
As depicted in Fig. 4B, we found two groups of noncancerous breast tissues-with no activity of ESR1 (class − 1) and with physiological levels of the ESR1 transcript (class 0) based on a 5-class test GMM for ESR1 mRNA. In cancer, three other groups emerged with either slightly (class 1), moderately (class 2), or highly increased (class 3) levels of ESR1 mRNA (Fig. 4C, D). In our study populations, the ESR1's transcript levels were abnormally increased in 40% of IBC, 64% of DCIS, and 0% of noncancer, based on the up-1 cutoff separating class 0 from 1 and higher. The ERBB2 mRNA was abnormally high in 16% of IBC, 15% of DCIS, and 0% of noncancer, considering up-2 as a diagnostic cutoff. Thereby g3mclass automatically selected tumors potentially sensitive to endocrine and anti-HER2 therapy, while other cancers may need different types of treatments. Notably, the g3mclass estimates showed that about 24% of IBC and 19% of DCIS, scored as ER-positive in pathology reports had reference-like levels of ESR1 transcript. These cases are candidates for overdiagnosis. The potential underdiagnosis was estimated in 0% of IBC and 1% of DCIS. Concurrently, the g3mclass provided insights into the variability of PGR, encoding steroid hormone receptor PR, a marker recommended for testing in IBC but not in DCIS 22 . High levels of PGR mRNA (class 2) were found in 9% of IBC, 13% of DCIS, and 0% of noncancer. In sharp contrast, low/undetectable levels of PGR mRNA (class − 1) were in 48% of IBC, 25% of DCIS, and 21% of reference. Thus, the g3mclass revealed low/or loss of PGR mRNA expression in IBC and the upregulation of PGR mRNA in DCIS in our study populations. Finally, we queried an independent set of mRNA data from the human breast cancer cell lines with the know expression levels of HER2, ER/PR 23,24 and found them classified according to the established status (Fig. 4E). www.nature.com/scientificreports/ In short, we demonstrated the diagnostic accuracy of the g3mclass analysis of clinical biomarkers and therapeutic targets. Additionally, we showed the robustness of this software in the automated multiclass classification of the test and independent queries. We have also provided evidence of the scalability of the g3mclass software to classify and visualize classifications on multiple analytes concurrently. We demonstrated the software output's interpretability by showing how various valuable insights can be extracted from raw test and query data using the g3mclass. More importantly, we showed how the g3mclass helps analyze each person's cancer with a unique pattern of biomarkers.
Discussion
Modern biomedical science requires highly specialized but easy-to-adapt software. This article presents g3mclass, a practical stand-alone application for a general biomedicine task concerning molecular assay data classification. The g3mclass offers inference about the number of classes, the mean and spread levels of an analyte in each class, and the prevalence of each class in the study population. In oncology, this allows unraveling and taking full advantage of hidden unique patterns of biomarkers and targets in each person's cancer. In addition, it may help researchers in the early stages of pharmaceutical testing of new therapies and companion diagnostics to determine whether further, often expensive, studies are warranted.
In the present article, we demonstrated how g3mclass-assisted classification helps human experts quickly assess the biological variability of gene transcripts across the populations of women diagnosed with primary breast cancer without extensive and long-term data collection. We also provided how human experts may select among probabilistic GMMs automatically learned by the g3mclass software. GMM is often used for unsupervised clustering, mainly for data exploration. An example of such an approach is subgrouping cancers based on the similarity of gene expression patterns [25][26][27] . The g3mclass exploits the customized semi-constrained EM algorithm's ability to learn test models from known (provided by experts) and unknown (missing values) information. This computational approach is the opposite of supervised classifications requiring the predefined knowledge of the number of the mixture components. For example, a two-component mixture model sorts differentially expressed genes in microarray experiments 12 . The principal innovation of g3mclass is embedding pre-existing experts' knowledge of reference parameters into the test GMM. As a result, it substantially improves the differentiation of new-to-test versus reference-like values and provides biological and clinical context for interpretation outcomes. This approach defines the significant difference of g3mclass from other powerful software packages handling Gaussian finite mixture modeling as their clustering capabilities, including the most popular R package mclust 14 and the Addinsoft XLSTAT (https:// www. xlstat. com/ en/ compa ny). Overall, there are two critical applications of the g3mclass in the biomedical field. First, it enables the discovery of previously unknown groups with different levels of biomarkers, including those that are not part of the reference and thus are more likely linked with disease. Second, it allows individual patient classification in line with personalized clinical decision-making. www.nature.com/scientificreports/ This article focused primarily on validated biomarkers because the quality of HER2 and ER diagnostics affects millions worldwide. According to World Health Organization, with an estimated 2.3 million new annual cases reported globally, female breast cancer is the most diagnosed cancer type (https:// www. who. int/ news-room/ factsheets/ detail/ breast-cancer). Experts recommend that every primary IBC be tested for the presence of HER2 and ER and re-tested in subsequent recurrences and metastases by semi-quantitative immunohistochemistry (IHC) and/or fluorescence in situ hybridization (FISH) 22,28 . These tests may produce equivocal results that could not be interpreted as positive or negative. The challenge remains to define either at the protein 8,29 or the mRNA 30-33 the HER2/ER expression cutoffs that segregate patients who may derive meaningful clinical benefit from endocrine and targeted therapies from those who will not. We have previously demonstrated that dichotomization of structurally mixed mRNA data with a single cutoff, e.g., using a ROC model or two-component GMM, may result in the loss of reliable information about the patient groups, as well as misclassification of some individuals 19 . This article presented practical statistical software to help remedy such a problem and demonstrated how multiclass classification with g3mclass may help fine-tune stratification on clinical biomarkers. In our study cohorts, g3mclass automatically recognized cancers unlikely to be present in the reference, i.e., ERBB2 mRNA + (class 2 and higher) and ESR1 mRNA + (class 1 and higher). Likewise, recognizing by g3mclass the group of ERBB2 mRNA + (class 1) may help define HER2-low positive breast cancer in clinical trials 34 . Clinical studies adopting g3mclass are warranted to investigate whether the groups with differentially increased levels of ERBB2 mRNA and ESR1 mRNA have different sensitivity to the targeted therapy.
In clinical trial designs, g3mclass provides experts with a flexible diagnostic cutoff driven by the intended use where the sensitivity or specificity is more beneficial. HER2 and ER are the targets of the emerging therapies for breast cancer and other types of cancer 35,36 . The cutoffs necessary for testing the clinical benefits of new therapies are likely to differ across cancer types [37][38][39] . Using g3mclass, experts may tailor the biomarker cutoff for each disease. If a more standard binary classification is desired, experts may choose a data-driven approach and transform multiclass into binary classification while eliminating equivocal results. To further unravel the g3mclass capabilities in determining the biological and clinical value of candidate biomarkers, we provide evidence of the scalability of the g3mclass software to classify on multiple analytes and visualize concurrent classifications with heatmaps. This approach may help manage ER+/HER2+ cancers, as the dual implementation of the hormone and anti-HER2 therapies showed evidence of success in the clinical trials 8 .
In sum, the applicability of g3mclass may be easily extended beyond one biomarker, dataset, or disease. It provides a cost-effective and straightforward way to examine and deal with the variability of the molecular assay data. The analysis with the g3mclass does not depend on the computing environment, which ensures the research's reproducibility. In clinical settings, applying g3mclass promises more precise stratification that may help improve therapeutic sensitivity. Yet, the limitations for the g3mclass application exist, as outlined in this article and software documentation. To what degree this molecular classifier combined with the genomic classifier and conventional clinicopathological characteristics improves patient outcomes remains to be seen in clinical studies. The free dissemination of g3mclass paves the path towards such investigational studies and a personalized treatment approach.
Materials and methods
Human tissues and cell lines. The formalin-fixed paraffin-embedded (FFPE) human breast tissues were obtained from the Department of Pathology and Laboratory Medicine, Tumor Tissue and Biospecimen Bank, and the Cooperative Human Tissue Network at the University of Pennsylvania. The study was performed with 256 samples, including 34 mammoplasties from women with no history of breast cancer, 75 diagnostic biopsies of ductal carcinomas in situ (DCIS), 142 surgical excisions of primary invasive breast cancer (IBC), and 5 human breast cancer cell lines. The cell lines MCF-7T-47D, MDA-MB-231, SK-BR-3, and BT-474 were purchased from the American Type Culture Collection and cultured accordingly. MycoAlert Assay (Cambrex) confirmed that mycoplasma-free cells were used in the experiments. The tissue samples were accrued randomly from the same geographic region. Summaries of the characteristics of the study populations have been published 19,40 . Direct messenger RNA (mRNA) profiling. For mRNA data collection, we ran QuantiGene Plex 2.0 (QG2) assay (Genospectra/Panomics/Affymetrix/eBioscience/ThermoFisher Scientific, USA) and read on Flex-Map 3D (Luminex/Merck Millipore) according to manufacturers' protocol and as described in detail 19 . QG2 is a highly reliable and validated molecular assay that uses amplified branch DNA (bDNA) technology for parallel gene expression profiling 41,42 . We analyzed measurements of specific probes with the QG2 assay kits for quantitation of multiple target specific RNAs directly in lysates from FFPE tissue and cell lines. Our human Plex Set 12988 included 14 target-specific and two housekeeping gene probes described in detail 40 . Here, we analyzed mRNA for ERBB2 (probe set region 1203-1621), ESR1 (probe set region 5671-6292), and PGR (probe set region 2609-3194).
Clinical markers. The status of steroid hormone receptors (ER and PR) and HER2 were determined by IHC and/or FISH, the FDA-approved methods. The status of ER, PR for all tumors, and HER2 for primary IBC were obtained from surgical pathology reports. In addition, HER2 status in DCIS was assessed based on the FDAapproved method for IBC as described 40 .
Ethics approval. Studies were conducted in accordance with recognized ethical guidelines. We used approval from the University of Pennsylvania Institutional Review Board committee with a waiver of written informed consent to analyze patients' tissue and records.
Data availability
All data needed to evaluate the article's conclusions are present in the article or the Supplementary Information. Series record GSE214540 provides access to QuantiGene Plex 2.0 16-gene expression data submitted to the GEO repository. In addition, the g3mclass is available as a standalone application on https:// pypi. org/ proje ct/ g3mcl ass site.
|
v2
|
2022-11-06T06:17:53.988Z
|
2022-11-05T00:00:00.000Z
|
253350918
|
s2ag/train
|
Elucidating the neurological mechanism of the FLASH effect in juvenile mice exposed to hypofractionated radiotherapy.
BACKGROUND
Ultra-high dose-rate radiotherapy (FLASH-RT) affords improvements in the therapeutic index by minimizing normal tissue toxicities without compromising anti-tumor efficacy compared to conventional dose rate radiotherapy (CONV-RT). To investigate the translational potential of FLASH-RT to human pediatric medulloblastoma brain tumor, we used a radiosensitive juvenile mouse model to assess adverse long-term neurological outcomes.
METHODS
Cohorts of three-week-old male and female C57Bl/6 mice exposed to hypofractionated (2×10 Gy, FLASH-RT or CONV-RT) whole brain irradiation and unirradiated controls underwent behavioral testing to ascertain cognitive status four months post-treatment. Animals were sacrificed 6 months post-irradiation and tissues analyzed for neurological and cerebrovascular decrements.
RESULTS
The neurological impact of FLASH-RT was analyzed over a 6-month follow-up. FLASH-RT ameliorated neurocognitive decrements induced by CONV-RT and preserved synaptic plasticity and integrity at the electrophysiological (long-term potentiation), molecular (synaptophysin) and structural (Bassoon/Homer-1 bouton) levels in multiple brain regions. The benefits of FLASH-RT were also linked to reduced neuroinflammation (activated microglia) and a preservation of cerebrovascular structure, by maintaining aquaporin-4 levels and minimizing microglia colocalized to vessels.
CONCLUSIONS
Hypofractionated FLASH-RT affords significant and long-term normal tissue protection in the radiosensitive juvenile mouse brain when compared to CONV-RT. The capability of FLASH-RT to preserve critical cognitive outcomes and electrophysiological properties over 6-months is noteworthy and highlight its potential for resolving long-standing complications faced by pediatric brain tumor survivors. While care must be exercised before clinical translation is realized, present findings document the marked benefits of FLASH-RT that extend from synapse to cognition and the microvasculature.
|
v2
|
2022-11-06T06:17:54.457Z
|
2022-11-05T00:00:00.000Z
|
253351242
|
s2orc/train
|
Thoracoscopic “cut‐through” segmentectomy for small‐sized lung cancer in a deep central location
Abstract The use of segmentectomy and subsegmentectomy for the management of lung lesions is well established. However, the use of subsegmentectomy for deep seated lesions in the upper lobe is difficult because of sufficient surgical margins. Here, we present a patient whose lung lesion was in a deep central area and at the borders of three segments in the upper lobe of the right lung. We used combined subsegmentectomy (S1b + S3a) video‐assisted thoracoscopic surgery for this small‐sized lung cancer in a deep central location.
INTRODUCTION
Favorable results of combined subsegmentectomy and/or segmentectomy for small-sized lung cancer have been reported to generate favorable outcomes. [1][2][3] Most clinical studies of segmentectomy focus on its application to peripheral small-sized lung cancer. Nevertheless, some studies have reported the usefulness of segmentectomy for centrally located lung cancer, 3,4 and the indication for segmentectomy might expand in the future. However, it is essential to determine whether segmentectomy is difficult for deep lesions in the center of the right upper lobe. Here, we present a case of combined subsegmentectomy (S1b + S3a) for a small-sized lung cancer of deep central location that was successfully performed via video-assisted thoracoscopic surgery (VATS).
CASE REPORT
A 45-year-old man was noted to have a part-solid groundglass nodule (GGN) in the right upper lobe during a physical examination. Preoperative computed tomography (CT) revealed a 1.5 Â 1.1 cm tumor (solid component, 0.6 cm) in the S3. Although contrast-enhanced CT and three-dimensional CT (3D-CT) reconstruction with SYN-APSE VINCENT (Fujifilm Medical Co., Ltd.) revealed that the lesion was in a deep central area and at the borders of three segments in the lobe. Considering that the volume of S3 of this patient was larger than usual and S3b preservation would benefit the patient (Figure 1a-c), we planned a right S1b + 3a combined subsegmentectomy. Moreover, the surgical margins were determined to be greater than the tumor diameter using 3D-CT.
After opening the incomplete fissure between the upper and middle lobe, we identified the A3a, which we divided. Before dividing B3a, we performed the slip-knot method to create an inflation-deflation line of the segmental plane. 5 After dividing B3a, we identified A2a, 1a, and 1b by retracting the lung caudally. A1b was clipped and dissected. For B1b, we performed the slip-knot method again, and then divided it. After releasing the hilar structures, we used staplers to cut the intersegmental planes. The post removal view was similar to the cut-through road ( Figure 2). We sutured the segmental plane to prevent torsion, adhesion, and air leakage. Intraoperative segmental lymph node (no. 13) metastasis was absent; therefore, mediastinal lymph node dissection was omitted. The operation time was 213 min, and the total blood loss was 155 ml. The postoperative course was uneventful, and the patient was discharged 5 days after surgery. Histopathological examination confirmed a papillary adenocarcinoma (tumor diameter: 1.2 cm, solid component: 0.7 cm, papillary component: 50%, lepidic component: 50%, two #13 lymph nodes removed, margin: 1.2 cm, pT1aN0M0-IA1). We proposed a completion lobectomy; however, the patient refused to undergo additional surgery. The patient has been alive for 5 years postoperatively without recurrence, and the remaining lung in the upper lobe is well inflated.
DISCUSSION
The indications for anatomical segmentectomy have expanded because it reportedly yields relatively good clinical outcomes. [1][2][3][4] Our institution has reported comparable surgical and long-term outcomes for segmentectomy and/or subsegmentectomy. 2,3 We believe that the approach and resection concept of subsegmentectomy and segmentectomy are basically the same. 2,3,6 The problem is the higher local recurrence than lobectomy; in particular, segmentectomy reportedly has higher local recurrence for right upper lobe tumors than for those in other lobes. 7 If the tumor, such as pulmonary metastasis and pure GGN, is located deep in the center of the lobe, lobectomy might be preferred to ensure the adequate margins. We have reported a good outcome of segmentectomy or susegmentectomy even for those in the tricky location. 3 A segmentectomy for the right upper lobe must be performed cautiously to ensure adequate margins. Even with deep central lesions in the right upper lobe, as in this case, sufficient margins can be planned and performed by combining subsegmentectomy with 3D-CT.
Essential to the operation is the precise interpretation of the structure of the target lung segment and to review from various angles before the surgery using enhanced CT and 3D-CT. The key of this procedure is to release the hilar structures and open both intersegmental planes using staplers. The view post-segmentectomy is similar to a cut-through road. While there has been a perioperative report of similar method, 8 this study presented our anatomically combined subsegmentectomy (S1b + 3a) technique for small-sized lung cancer in the center of right upper lobe. Moreover, the patient had good long-term results. With this method, both raw cut segment surface fuse together after the operation, and dead space is small because of a large residual lung and this method might prevent air leakage, one of the serious complications of segmentectomy. Further, it might be advantageous in different cases, such as second primary lung cancer, to prevent excess adhesion in the future. F I G U R E 1 (a) Computed tomography (CT) reveals a lung tumor in the central area of the upper lobe. The tumor distance from the visceral pleura was approximately 3 cm. (b) The tumor is mainly located between B1b and B3a. B1b (arrow) and B3a (arrowhead). (c) We planned combined S1b + 3a segmentectomy using three-dimensional CT to keep sufficient margins. If sufficient parenchyma is to be preserved, S1bii + S3ai combined segmentectomy should be selected; however, the surgery is very difficult to perform. B1bii (arrow) and B3ai (arrowhead) F I G U R E 2 Segmental plane after combined S1b + 3a segmentectomy and schema. The central part of the upper lobe is resected. When the remaining segment was expanded, the remaining segment of the upper lobe became mountainous and looked like a cut-through road; hence, we named this resection method "cut-through" segmentectomy Although subsegmentectomy such as that reported here is rarely indicated, we believe that it has the potential to have a good outcome and we expect that it will be reported in prospective studies.
We demonstrated a combined subsegmentectomy via VATS for a centrally located lung cancer in the right upper lobe. "Cut-through" segmentectomy that preserves as much parenchyma as possible for deep central lesions in the upper lobe is a feasible technique.
|
v2
|
2022-11-06T14:44:25.431Z
|
2022-11-05T00:00:00.000Z
|
253352085
|
s2orc/train
|
Combined atezolizumab and nab-paclitaxel in the treatment of triple negative breast cancer: a meta-analysis on their efficacy and safety
Background Triple negative breast cancer (TNBC) is clinically aggressive breast cancer with a poor prognosis. Approximately 20% of TNBC has been found to express programmed death ligand 1 (PD-L1), making it a potential therapeutic target. As a PD-L1 inhibitor, atezolizumab is a recently approved immunotherapeutic drug for TNBC, this meta-analysis (MA) was aimed to review the randomized controlled trial studies (RCTs) of combined atezolizumab and nab-paclitaxel in the treatment of TNBC and synthesize the evidence-based results on its effectiveness and safety. Method We searched PubMed, Embase, EBSCOhost and ClinicalTrials.gov for the eligible RCTs which compared the efficacy and safety of combined atezolizumab and nab-paclitaxel with nab-paclitaxel alone. The outcomes analyzed included overall survival (OS), progression-free survival (PFS), objective response rate (ORR) and treatment-related adverse effects (AEs). Results A total of six RCTs were included in this MA. For efficacy, although OS was not significantly prolonged with combined atezolizumab and nab-paclitaxel (HR 0.90, 95% CI [0.79, 1.01], p=0.08), this combination therapy significantly improved PFS (HR 0.72, 95% CI [0.59, 0.87], p=0.0006) and ORR (RR 1.25, 95% CI [0.79, 1.01] p<0.00001). For safety, any AEs, haematological, gastrointestinal, and liver AEs showed no statistically significant differences between the atezolizumab and nab-paclitaxel combination group and nab-paclitaxel alone group. However, serious AEs, high grade, dermatological, pulmonary, endocrine, and neurological AEs were significantly lower with nab-paclitaxel alone compared to atezolizumab and nab-paclitaxel combined (p-value range from <0.00001 to 0,02). Conclusion Atezolizumab combined with nab-paclitaxel was associated with improved outcomes in the treatment of TNBC; however, this combination resulted in more toxicity compared to nab-paclitaxel alone. While nab-paclitaxel alone produced chemotherapy-related AEs, the combination of atezolizumab with nab-paclitaxel produced AEs, especially immune-related AEs such as haematological, pulmonary, endocrine, and neurological AEs. Trial registration This research work of systematic review has been registered on PROSPERO (Registration number: CRD42022297952). Supplementary Information The online version contains supplementary material available at 10.1186/s12885-022-10225-y.
Introduction
Triple negative breast cancer (TNBC) is a molecularly diverse breast cancer subtype which shows hormone receptor immunohistochemistry (IHC) stains of less than 1% for oestrogen and progesterone; and the absence of human epidermal growth factor receptor-2 Open Access *Correspondence: [email protected] (HER2) protein expression or HER2 gene amplification or both [1,2]. It is associated with earlier age of onset, aggressive clinical course and poor prognosis with worse survival [3].
TNBC is usually treated with a combination of surgery, radiotherapy and chemotherapy. The successes of atezolizumab which is a monoclonal antibody-based immune checkpoint inhibitor (ICI) against programmed cell death-ligand 1 (PD-L1) have been achieved in various cancers such as non-small cell lung carcinoma, renal cell carcinoma, urothelial carcinoma [4]. In March 2019, the Food and Drug Administration (FDA) approved atezolizumab for the treatment of people with TNBC. Efficacy of atezolizumab in combination with nab-paclitaxel has been further approved by IMpassion130 Investigators [5].
As atezolizumab is a recently approved immunotherapeutic drug for TNBC, systematic review (SR) and meta-analysis (MA) on its effectiveness and safety in combination with nab-paclitaxel, have not been studied properly in published primary studies. The results of some primary randomized control trial studies (RCTs) and clinical trials showed uncertainties and conflicting results which makes challenging for the clinician to integrate the data into clinical practice. Therefore, this study systematically reviewed the RCTs of combined atezolizumab and nab-paclitaxel in the treatment of TNBC and synthesize the evidence-based results on its effectiveness and adverse effects of it. This MA was also able to report the efficacy of the combined therapy group in association with PD-L1 immunohistochemistry positivity of TNBC.
Materials and methods
The SR and MA were done according to the updated guideline of the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) statement [6].
Identification of eligible studies
The systematic literature search was carried out in health-related electronic databases such as PubMed, Embase and EBSCOhost. Clinical trial studies were also searched at Clini calTr ials. gov. The search terms used were "immune checkpoint inhibitors", "PD-L1 inhibitor", "atezolizumab", and "triple negative breast cancer". The search was limited to original articles published in the English language up to October 2021. To find out additional studies, reference lists of the original articles were also screened.
Inclusion and exclusion criteria
Studies selection was based on criteria of PICOS format: (1) Participants: All female patients above 18-year-old who have been diagnosed with TNBC by histopathological and immunohistochemical methods; (2) Intervention: Intravenous Atezolizumab and nabpaclitaxel; (3) Comparisons: placebo and nab-paclitaxel or nab-paclitaxel alone; (4) Outcomes: primary outcomes were overall survival (OS) which is defined as the time from randomization to death from any cause and treatment-related adverse effects (AEs). The secondary outcomes included progression free survival (PFS) which is defined as the time from randomization to cancer progression or death from any cause and objective response rate (ORR). Studies included parallel RCTs reporting on the efficacy and or safety of combined atezolizumab and nab-paclitaxel in the treatment of TNBC. Review articles, case reports, editorials, and studies that used combinations other than nab-paclitaxel were excluded for this MA.
Literature search and study selection
Two researchers (TTW, SVK) conducted an independent literature search using healthcare electronic databases (PubMed, Embase and EBSCOhost) and ClinicalKey to search for related clinical trials. The articles obtained from the literature search were assessed by two researchers (TTW, SVK) independently. Any disagreements between both researchers were initially discussed between two researchers. If an agreement was not reached, two researchers discussed with a third researcher (SNA) to mediate and finalize. The articles were screened according to the PRISMA flowchart. As for the first stage, related articles were identified and grouped to create a total number of records. These records were screened leading to the removal of review articles and non-relevant articles. The articles that did not fit the inclusion criteria based on abstract and title alone were excluded. Finally, full-text articles were examined to obtain the included studies required for MA.
Data extraction
Two researchers (TTW, SVK) independently extracted the relevant data from the included studies using piloted data extraction sheet. Discrepancies were discussed thoroughly and finalized with a third researcher (SNA). The data that were extracted included study title, first author and year of publication, trial phase, and number of patients receiving combined atezolizumab and nab-paclitaxel, number of patients receiving nabpaclitaxel alone or combined with placebo, mean age of participants, PD-L1 positivity, a dose of atezolizumab and follow up duration. Outcome data included were OS, PFS, ORR, any AEs, serious AEs, high-grade AE, haematological AEs, gastrointestinal AEs, dermatological AEs, pulmonary AEs, liver AEs, endocrine AEs, and neurological AEs.
Quality assessment of the included studies
The Cochrane evaluation handbook of randomized controlled trials bias tool was used to evaluate the bias and quality of the included studies. It included the following assessment scopes: random sequence generation, allocation concealment, the blindness of participants and personnel, the blindness of outcome assessment, incomplete outcome data, selective reporting, and other biases. We rated each domain of the tool as having a 'low' , 'high' , or 'unclear' risk of bias at the study level and for each outcome if possible [7].
Statistical analysis
Among the outcomes, analysis of OS and PFS was estimated as the hazard ratio (HR) and analysis of other outcomes was estimated as the risk ratios (RR) for the treatment success of combined atezolizumab and nabpaclitaxel vs nab-paclitaxel alone. From the statistical analysis, heterogeneities were assessed using the chisquared test and the I 2 statistic. Pearson's chi-squared test was used to decide a statistically significant difference between the expected frequencies and the observed frequencies.
To avoid the heterogeneity, if the I 2 statistic is more than 50%, a random effects model was used; and if the I 2 statistic is less than 50%, a fixed effect model was used. We investigated the robustness of the review by performing sensitivity analyses when appropriate, such as performing fixed-effect models for selected outcomes and including only low risk of bias outcomes, according to the summary assessment of the risk of bias. A twotailed P value of less than 0.05 was considered statistically significant. A 95% CI was used to provide a range of values for the ORs obtained. MA has performed with Review Manager (RevMan 5.4) software.
Literature search results
A total of 3880 potential studies were identified using the preliminary search strategy; 2949 were from Pub-Med, 534 were from EMBASE, 392 were from EBSCOhost and 5 were clinical trials from ClinicalKey. A total of 2965 studies were compiled after removing the duplicates. Based on the titles, 88 review articles and 623 irrelevant studies were removed. After screening abstracts of the remaining 204 studies, 192 studies were excluded as they were not specifically related to our specific search criteria. Finally, 12 studies were accessed for eligibility by reading the full text. Among them, 6 studies were excluded with reasons based on inclusion and exclusion criteria. Therefore, the remaining 6 articles were included for SR and MA. A three-phase flow chart of the study selection process based on the updated PRISMA statement 2020 is illustrated in Fig. 1. A summary of the reasons for six excluded studies [8][9][10][11][12][13] is shown in Table S1.
Characteristics of the included studies
The characteristics of the six included studies [5,[14][15][16][17][18] are shown in Table 1. In all six included studies, the dosage of atezolizumab was 840 mg given intravenously on days 1 and 15, and chemotherapy was administered every week on days 1, 8 and 15: with a follow-up duration of 2 years. Five studies were phase 3 RCTs and 1 was phase 2 RCTs. These 6 studies include a total of 1078 study participants in the intervention arm and 886 participants in the control arm. Out of these participants, 837 had breast cancer cells which were PD-L1 positive. The first two studies by Schmid referred to the same study with a follow-up publication. The first article had data to assess the efficacy of the treatment whereas the updated article published in 2020 had more comprehensive data on the AEs which is used to assess the safety of this drug combination therapy [5,14].
Assessment of risk of bias and publication bias
The results of the assessment of the bias are shown in the risk of bias graph ( Fig. 2A) and the risk of bias summary (Fig. 2B). The overall risk of bias was evaluated as low risk and the quality of the studies was acceptable. Publication bias was assessed in six included studies. Begg's and Egger's test Funnel plot for publication bias is shown in Figure S1.
Efficacy of atezolizumab for TNBC
As included studies reported OS, PFS and ORR for all TNBC cases and PD-L1 positive cases, this MA performed a sub-group analysis on it. OS was reported by four studies [5,[16][17][18]. The HR for OS in the all-cases group, and in the PD-L1 positive subgroup were 0.93 and 0.79 respectively with HR 0.90, 95% CI [0.79, 1.01], p=0.08. There is no evidence of a difference in OS between combined atezolizumab with chemotherapy and chemotherapy alone (Fig. 3A). PFS was reported by four studies [5,15,17,18]. Combined atezolizumab with chemotherapy for both all cases and PD-L1 positive cases showed favourable PFS with HR 0.72, 95% CI [0.59, 0.87], p=0.0006 (Fig. 3B). ORR was reported by five studies [14][15][16][17][18], and favourable ORR was seen with combined atezolizumab and chemotherapy. Analysis on all-cases group and PD-L1 positive subgroup showed RR 1. 24 Fig. 3C).
Discussion
In recent years, the role of immunotherapy in cancer treatment has expanded and now it is the first choice of treatment intervention in many cancers [19]. ICIs have demonstrated favourable outcomes in a variety of refractory solid malignancies such as advanced non-small cell lung cancer, metastatic melanoma, and metastatic bladder cancer. Unlike these tumour types, most breast cancers are not inherently immunogenic, however, TNBC is the most immunogenic subtype [20]. TNBC has been found to have higher rates of cell surface PD-L1 expression compared to other breast cancer subtypes and higher PD-L1 expression suggests a greater potential benefit from the use of PD-1/PD-L1 targeted immunotherapy [20,21].
Atezolizumab which is anti-PD-L1 antibody was approved to be used in combination with nab-paclitaxel, mainly for PD-L1 expressed locally advanced or metastatic TNBC as determined by an FDA-approved test. The FDA also approved the VENTANA PD-L1 (SP142) Assay as a companion diagnostic device for selecting TNBC patients for atezolizumab [22].
In this MA, all included studies used VENTANA PD-L1 (SP142) immunohistochemical assay for the determination of PD-L1. Determination of PD-L1 status is indication-specific, and evaluation is based on either the proportion of tumor area occupied by PD-L1 expressing tumor-infiltrating immune cells (% IC) of any intensity or the percentage of PD-L1 expressing tumor cells (% TC) of any intensity. According to diagnostic indications for the VENTANA PD-L1 (SP142) Assay by ROCHE VENTANA, cutoff score for TNBC is ≥ 1% IC; whereas ≥ 50% TC or ≥ 10% IC for non-small cell lung cancer and ≥ 5% IC for urothelial carcinoma [23,24]. Routine testing for PD-L1 immune-cell expression in unresectable, locally advanced or metastatic TNBC with the approved companion diagnostic (VENTANA PD-L1 SP142 assay) should be used to identify patients who might benefit from treatment with atezolizumab plus nab-paclitaxel [5].
Efficacy combined atezolizumab and nab-paclitaxel for TNBC
Based on our best literature search, MA on the efficacy of combined atezolizumab and nab-paclitaxel versus nabpaclitaxel alone in the treatment of TNBC has not been published. All the MA on efficacy were done on all ICIs. In our MA, efficacy is significantly improved when atezolizumab is added to chemotherapy for the treatment of TNBC based on a pooled analysis of ORR, although there is no significant difference OS and PFS between the two groups. However, PFS in PD-L1 positive group showed slightly favourable PFS compared with all cases and ORR in PD-L1 positive group showed less favourable ORR compared with all cases.
The results of this MA are also concordant with other published MAs on combined ICIs. A MA on the efficacy of PD-1/PD-L1 inhibitors in TNBC reported significant anti-tumour effect with combined PD-1/PD-L1 inhibitors and chemotherapy proven by OS and ORR [25]. Efficacy of combined neoadjuvant chemotherapy and ICIs based on pathological complete response (pCR) and tumour characteristics were reported by a MA on early stage TNBC. It reported improved pCR and PFS in the group with combined ICIs and neoadjuvant chemotherapy [26].
In that study, response to ICIs depending on PD-L1 status was also assessed, and it was found that both PD-L1 positive and PD-L1 negative patients had no statistically significant difference in their treatment outcomes. One possible explanation for this could be that the study participants in this study were patients with early-stage TNBC. The immune cell dynamics in the tumour microenvironment in the early stages and late stages of TNBC are substantially different, and a subgroup analysis of the PD-L1 status is still important and valid [27]. Latest MA on atezolizumab and pembrolizumab in TNBC reported that ORR of atezolizumab/pembrolizumab plus chemotherapy was higher in the intention to treat arms than the placebo groups in TNBC. This study also collate evidence of atezolizumab/pembrolizumab as viable therapeutics among patients with TNBC with PD-L1 subgroups deriving higher benefits [28].
A clinical trial, IMpassion 130 which is a phase 3 trial reported favourable OS with combined atezolizumab and nab-paclitaxel in PD-L1 positive TNBC. However, this positive result could not be formally tested due to the prespecified statistical testing hierarchy [5]. It also prolonged PFS in both intention-to-treat population and PD-L1-positive subgroup. In IMpassion 130, the median PFS in PD-L1 positive patients was 7.5 months in the atezolizumab plus nab-paclitaxel group versus 5.0 months in the placebo plus nab-paclitaxel group [14]. However, these results were contradicted by the results of IMpassion 131 which reported that combining atezolizumab with paclitaxel did not improve PFS or OS versus paclitaxel alone in PD-L1 positive group [17]. Therefore, PD-L1 expression has not consistently correlated with response to immunotherapy in clinical trials, and it cannot yet be considered a meaningful predictive biomarker. Other biomarkers such as tumour mutational burden and tumour infiltrating lymphocytes are now being explored as PD-L1 independent predictive biomarkers for immune responsiveness of the tumour in immunotherapy of TNBC [20].
A MA on the efficacy of atezolizumab for non-small cell lung cancer (NSCLC) found that both high PD-L1 expression and negative PD-L1 expression subgroups had an improved efficacy with atezolizumab. Interestingly, in patients with low expression of PD-L1, OS showed no significant difference between atezolizumab treatment group and the placebo group though the results favoured atezolizumab in terms of PFS [29]. Therefore, atezolizumab combination therapy has its benefits regardless of the PD-L1 status. Although atezolizumab is a PD-L1 inhibitor which binds to PD-L1 on the tumour cells, it is evident that the treatment of cancers with atezolizumab is efficacious even when PD-L1 is absent. Currently, the US FDA has approved atezolizumab to be used with chemotherapy for PD-L1 positive patients in both NSCLC and TNBC [22,30]. These results could pave the path toward the drug approval of atezolizumab regardless of PD-L1 status in cancers including NSCLC and TNBC.
Studies on the use of atezolizumab as a combination therapy in urothelial carcinoma and hepatocellular carcinoma also showed that the efficacy of atezolizumab combination therapy is greater than chemotherapy alone or atezolizumab monotherapy [31,32]. This effect is due to the combination of atezolizumab with chemotherapy improving the immunological conditions in the tumour micro-environment which enhance the anti-tumour immune response [33].
Safety of atezolizumab for TNBC
ICIs can cause immune-related AEs as the mechanism of ICI action relies on the inhibition of the physiological brake of immune activation and they often have off-target effects resulting in immune-mediated inflammation of diverse organs or tissues [34].
In this MA, we analyzed common treatment-related AEs caused by atezolizumab plus nab-paclitaxel versus nab-paclitaxel alone. The AEs analyzed in this MA were any treatment-related AEs, serious AEs, high-grade AEs, haematological, gastrointestinal, dermatological, pulmonary, liver, endocrine and neurological AEs. There is a significant greater incidence of serious AEs, highgrade AEs, dermatological, endocrine, and neurological AEs in the group treated with combined atezolizumab and nab-paclitaxel group compared to chemotherapy alone. However, there are no significant differences in any treatment-related AEs, haematological, gastrointestinal, pulmonary, and liver AEs between the two groups of treatment.
The results on the safety of our MA are concordant with other MA. A MA on the safety of PD-1/PD-L1 inhibitors also showed a higher frequency of AEs in the atezolizumab combination therapy group compared to the placebo control group. Specific immunemediated AEs above grade 3 highlighted in this study were neurological AEs like peripheral neuropathy and haematological AEs like anaemia and neutropenia [25]. An RCT study on atezolizumab plus nab-paclitaxel in non-small cell lung cancer showed a higher incidence of high-grade AEs in this combination group compared with chemotherapy alone [35]. Other studies have also shown that most AEs of atezolizumab are immunerelated, mainly dermatological, gastrointestinal, pulmonary, endocrine and neurological AEs [36]. When comparing the safety of atezolizumab with chemotherapy, cytotoxic AEs are more serious with chemotherapy, producing chemotherapeutics-related deaths [37,38].
There were some limitations during the conduct of this MA such as a small sample size in some of the primary studies and ongoing trials without reported results. There was also a discrepancy among clinical trials. Although IMpassion 130 used nab-paclitaxel, IMpassion 131 used paclitaxel. In the MA of the efficacy of atezolizumab regarding PD-L1 status, we could manage to perform a sub-group analysis of overall cases and PD-L1 sub-group. Likewise for safety, subgroup analysis of the AEs in the PD-L1 positive population could not be assessed since the primary studies did not provide sufficient data. After the conduct of this MA, the manufacturer of atezolizumab (Roche) has withdrawn the indication for atezolizumab in combination with nab-paclitaxel in the treatment of TNBC following consultation with the FDA in the United States to reassess the statu s of syste mic thera py agent s grant ed accel erate d appro val [39].
In conclusion, there is evidence of increased efficacy in combining atezolizumab with chemotherapy for the treatment of TNBC. Furthermore, atezolizumab combination therapy had similar efficacy in the PD-L1 positive subgroup compared to all participants. However, the combination therapy of atezolizumab with chemotherapy tends to increase the incidence of treatmentrelated AEs, especially immune-related AEs. Therefore, treatment-related AEs of atezolizumab should be monitored carefully and cautiously. Future primary studies are needed to identify key predictors associated with immune-related AEs as well as to compare their incidences to help the clinicians personalize treatment for patients with TNBC, working to improve their prognosis. Although atezolizumab has been withdrawn recently to be used in TNBC in the United States, the results of this MA may help the clinicians and manufacturers in the reassessment of atezolizumab and consideration of alternative ICIs in the treatment of TNBC.
|
v2
|
2022-11-07T16:17:43.229Z
|
2022-11-05T00:00:00.000Z
|
253375523
|
s2ag/train
|
Antecedent infections, recent developments and future directions in Guillain-Barré syndrome
The Guillain-Barré syndrome is an autoimmune polyradiculoneuropathy causing symmetrical weakness of limbs. After poliomyelitis, it is the second most common cause of paralysis, with an annual incidence of 0.84-1.91 per 100,000 individuals. The syndrome affects both men and women, showing a male preponderance. Campylobacter jejuni, epstein-barr virus, cytomegalovirus, mycoplasma pneumoniae and hemophilus influenzae are amongst the most common causative agents of Guillain-Barré syndrome. Several immunological and genetic factors have been recognised as the risk factors. Human leukocyte antigen, cluster of differentiation 1, and tumor necrosis factor-alpha alleles are among the frequently investigated loci in Guillain-Barré syndrome. Genome-wide association studies have found no significant association of Guillain-Barré syndrome with common variants. Many vaccines against Campylobacter jejuni infection have been proposed, but there are concerns about the efficacy and safety of these vaccines. So far, there is no approved vaccine against Campylobacter jejuni.
|
v2
|
2022-11-07T06:19:19.344Z
|
2022-11-06T00:00:00.000Z
|
253370099
|
s2ag/train
|
NKG2D‐mediated cytotoxicity improves after primary surgery for high‐grade serous ovarian cancer
Tumors compromise the patients’ immune system to promote their own survival. We have previously reported that HGSC exosomes play a central role, downregulating NKG2D cytotoxicity. Primary surgery's effect on tumor exosomes and NKG2D cytotoxicity in HGSC patients has not been studied before. The overall objective of this study was to explore the effect of surgery on the exosome‐induced impairment of NKG2D cytotoxicity in HGSC.
|
v2
|
2022-11-07T06:19:19.410Z
|
2022-11-06T00:00:00.000Z
|
253370131
|
s2orc/train
|
Computational modeling studies reveal the origin of the binding preference of 3‐(3,4‐di hydroisoquinolin‐2(1H)‐ylsulfonyl)benzoic acids for AKR1C3 over its isoforms
Abstract As a key regulator for hormone activity, human aldo‐keto reductase family 1 member C3 (AKR1C3) plays crucial roles in the occurrence of various hormone‐dependent or independent malignancies. It is a promising target for treating castration‐resistant prostate cancer (CRPC). However, the development of AKR1C3 specific inhibitors remains challenging due to the high sequence similarity to its isoform AKR1C2. Here, we performed a combined in silico study to illuminate the inhibitory preference of 3‐(3,4‐dihydroisoquinolin‐2(1H)‐ylsulfonyl)benzoic acids for AKR1C3 over AKR1C2, of which compound 38 can achieve up to 5000‐fold anti‐AKR1C3 selectivity. Our umbrella sampling (US) simulations together with end‐point binding free energy calculation MM/GBSA uncover that the high inhibition selectivity originates from the different binding modes, namely “Inward” and “Outward,” of this compound series in AKR1C3 and AKR1C2, respectively. In AKR1C3/38, the tetrahydroquinoline moiety of 38 is accommodated inside the SP1 pocket and interacts favorably with surrounding residues, while, in AKR1C2/38, the SP1 pocket is too small to hold the bulky tetrahydroquinoline group that instead moves out of the pocket with 38 transitioning from an “Inward” to an “Outward” state. Further 3D‐QSAR and energy decomposition analyses suggest that SP1 in AKR1C3 prefers to bind with a rigid bicyclic moiety and the modification of the R3 group has important implication for the compound's activity. This work is the first attempt to elucidate the selectivity mechanism of inhibitors toward AKR1C3 at the atomic level, which is anticipated to propel the development of next‐generation AKR1C3 inhibitors with enhanced efficacy and reduced “off‐target” effect for CRPC therapy.
Partially owing to the availability of crystal structures of AKR1Cs, several kinds of selective AKR1C3 inhibitors have recently been reported, including derivatives of flufenamic acid (FLF), 35,[45][46][47] indomethacin, 36 naproxen, 34 cinnamic acids, 28,[48][49][50] and long-chain polyunsaturated fatty acids. 51 Of them, 3-(3,4-dihydroisoquinolin-2(1H)ylsulfonyl)benzoic acid 38 in Jamieson and colleagues' work showed 6.1 nM inhibitory activity and 5000-fold AKR1C3 versus AKR1C2 selectivity. 46 However, the origin of the selectivity of these inhibitors has not been systematically explored and thus remains elusive. Current explanation of the agents' binding preference is merely based on the comparison of the static crystal structures of AKR1C3 and its isoforms. For example, Adeniji et al. proposed that the SP1 pocket is lined by polar residues Ser118, Ser308, and Tyr319 in AKR1C3, while the corresponding residues in AKR1C1 and AKR1C2 are more hydrophobic, including Phe118, Leu308, and Phe319. 6,34 Additionally, FLF and its derivatives were found to adopt distinct binding poses in AKR1C3 and AKR1C2 by overlaying the co-crystal structures ( Here, we employed an integrated computational approach to elucidate the molecular mechanisms behind the inhibitory selectivity of 3-(3,4-dihydroisoquinolin-2 (1H)-ylsulfonyl)benzoic acids 46 toward AK1R1C3 over AKR1C2 (Table S1). First, our molecular docking and molecular dynamics (MD) simulations showed that this class of compounds adopts difference binding poses, "Inward" and "Outward," respectively, in the active site pockets of AKR1C3 and AKR1C2. Additionally, the endpoint free energy calculations revealed stronger interaction energies of the inhibitors with AKR1C3 than with AKR1C2.
Furthermore, umbrella sampling (US) simulations were performed to explore the binding landscapes of compound 38 in AKR1C3 and AKR1C2, providing an explanation for their preferential inhibition of AKR1C3 over AKR1C2. Finally, the crucial molecular features that determine the inhibitors' potency/selectivity against AKR1C3 were identified by 3-dimensional quantitative structure-activity relationship (3D-QSAR) and free energy decomposition analyses. We anticipate this work will provide valuable insights to facilitate the development of more effective and selective anti-AKR1C3 drug candidates.
| Sequence and structural comparisons of AKR1C3 with its isoforms
The sequence alignment and comparison between AKR1C3 and its isoforms (AKR1C2 and AKR1C1) were calculated using the Chimera software 53 ( Figure S1). The sequence of the full length AKR1C3 is 84.76% and 86.67% identical to AKR1C2 and AKR1C1, respectively, and the active binding site (within 10 Å of inhibitors) of AKR1C3 shares $75.9% sequence identity with AKR1C1 and AKR1C2. The high similarity explains the emergence of many pan-AKR1Cs inhibitors, 2,4,44 implicating the challenge of developing AKR1C3 selective inhibitors. As illustrated in Figures S1 and S2d,e, the AKR1C enzymes all possess a highly conserved canonical α 8 β 8 superbarrel core, capped at the N-terminus with an antiparallel hairpin. The active site is constituted by several loops (Loop A-C) located at the C-terminus of the central β-barrel, which contains a conserved catalytic tetrad consisting of Asp50, Tyr55, Lys84, and His117. 54 The binding site of AKR1C3 can be divided into five compartments: an oxyanion site (OX) comprising Tyr55, His117, and NADP + , a steroid channel (SC) formed by Trp227 and Leu54, and three subpockets including SP1, SP2, and SP3. Notably, the structural differences between the subtypes are located primarily in the active site loops, and most residues involved in substrate binding and catalysis are on the three loops A, B, and C ( Figures S2a-c). Specifically, the SP1 of AKR1C3 is lined by multiple non-conserved residues (the corresponding residues in AKR1C2 are given in parenthesis), such as His304 (Arg304), Phe306 (Leu306), Asn307 (Thr307), Ser308 (Leu308), Ser310 (Ala310), His314 (Gly314), and Tyr319 (Phe319) in Loop C together with Ser118 (Phe118), and Met120 (Val120) in Loop A ( Figure S2a vs. b), which gives rise to the noticeable structural difference of SP1 between two enzyme isoforms ( Figure S2d). Given the extremely high sequence identity (full-length: 95.9 %, active site: 95.7 %, Figure S1) and almost completely superimposable structures ( Figure S2e) of AKR1C2 and AKR1C1, we focus on the inhibitory selectivity mechanism toward AKR1C3 over AKR1C2 for a series of compounds with a common 3-(3,4-dihydroisoquinolin-2(1H)-ylsulfonyl)benzoic acid scaffold.
| Exploration of key structural features on ligand selectivity by 3D-QSAR
3D-QSAR models were built by molecular field-based QSAR tool in Maestro 11.2 (Schrodinger) 55 to investigate the crucial structural features determining the activity and inhibition selectivity of the studied AKR1C3 inhibitors ( Figure S3 and Table S2). The selected 3D-QSAR model (Table S3, PLS factor = 4) exhibited high prediction accuracy for both activity and selectivity, in view that the predicted pIC 50 Figure S4). In Table S4, the relatively high contributions of the steric (0.3108, 0.2799) and hydrophobic (0.2231, 0.2272) features indicate their major roles in governing ligand activity and selectivity, respectively. From the contour maps of both activity and selectivity 3D-QSAR models, the bulky tetrahydroquinoline moiety of the potent compounds (26, 28, 37, and 38) was found to locate in the favorable regions of the Gaussian-steric and aromatic ring fields (Figures 1a, 3a,f, and S5a,f), suggesting this rigid bicyclic fragment may be a dominant molecular feature responsible for ligand's activity and selectivity. This is further confirmed by the decreased activity and selectivity of compounds 89-94 with alternative aromatic groups as well as open chain compounds 56-64 (Table S1). Strikingly, we also observed favorable regions of steric and hydrophobic field distributed around the tetrahydroquinoline moiety (Figures 1a,c-f). Thus, we speculate that ligand potency can be further improved by introducing favorable groups into certain positions of this bicyclic moiety, such as the 6-position Br in 38 and 5-position NO 2 in 28. However, the added group cannot be too large, as evidenced by 44 and 54 with bulky substituents but increased IC 50 . Moreover, the contour maps reveal positive effects on ligand selectivity of incorporating negatively charged and H-bond acceptor groups, for example, carboxylate, into the three-position of the phenyl ring (Figures 1b, S5b, 1d, and S5d), while negative effects of hydrophobic and H-bond donor groups (Figures 1c,e and S5c,e). Although the 3D-QSAR models identified some critical structural features that contribute to inhibitors' selectivity toward AKR1C3 over AKR1C2, these models were solely based on ligand information, which may limit the scope and accuracy of the models. Thus, structure-based molecular modeling methods were subsequently carried out to provide a full understanding of the mechanism of ligand selectivity. AKR1C2, the compound could likely adopt a similar "Inward" binding mode in ARK1C2 as in AKR1C3. On the other hand, considering that compound 38 was derived from flufenamic acid (FLF), we predicted the AKR1C2/38 complex structure by docking 38 into the binding pocket of AKR1C2 bound with a FLF analogue, mefenamic acid (PDB entry: 4JQA), in view that Glide docking method exhibits strong ability to reproduce the crystal poses of the known AKR1C2/3 complexes ( Figure S6, RMSD: 0.19, 0.32, 0.46, and 0.32 Å for 4JQA, 4FAM, 4FAL, and 4FA3, respectively). Surprisingly, 38 was found to bind AKR1C2 in an outward posture (referred to as "Outward1" below) with the tetrahydroquinoline completely outside of SP1 (Figure 2b).
To discern which conformation, "Inward" or "Outward1," is the preferred binding pose of 38 in ARK1C2, MD simulations were carried out on the AKR1C2/38 complex in both binding conformations. Figure S7 shows that the system starting from the "Out-ward1" binding conformation reaches equilibrium at $220 ns. However, the "Inward" conformation does not reach equilibrium within 500 ns, and the MD trajectory shows a sudden RMSD increase at $450 ns. We thus extended the "Inward" simulation for another 500 ns. Visual inspection of the MD trajectory reveals that the sudden RMSD increase in the "Inward" AKR1C2/38 simulation corresponds to an "Inward" to "Outward" binding mode transition of 38 in AKR1C2 (Figure 3a-c). After F I G U R E 2 (a) Co-crystal structure of AKR1C3 in complex with 38 that adopts an "Inward" binding conformation (red), and computationally docked structures of AKR1C2 bound with 38 in (b) an "Outward1" (olive) and (c) an "Inward" binding mode (yellow) The root-mean-square deviations (RMSDs) of backbone heavy atoms as a function of time in the MD simulation of the AKR1C2/38-Inward complex. The "Inward" binding mode (b) remains until 400 ns, and then transitions to an "Outward2" binding mode (c-e) for the rest of the simulation time; (f) Structural overlay of the "Outward1" and "Outward2" binding poses of 38 in the active site of AKR1C2 the transition, the system reaches equilibrium for the rest of the 600 ns simulation, where 38 adopts a new outward binding conformation (referred to as "Outward2" below) with the tetrahydroquinoline moiety partially out of SP1 (Figure 3c-e). The "Outward2" binding pose derived from the MD simulation exhibits an RMSD of $1.5 Å from the "Outward1" conformation obtained in the docking calculation (Figure 3f). In the "Outward1" binding pose, 38 interacts with Ile 310 and Phe311 in SP1, Trp86, Val128, Ile129, and Trp227 in SP2, Tyr24, His222, and Glu225 in SP3, and the OX residue Val54 (Figure 4a), while the surrounding residues of 38 in the "Outward2" binding mode include His118, Val120, Ile137, Ile310, Phe311, Tyr317 in SP1, and Trp86, Ile129, and Trp227 in SP2 ( Figure 4b).
2.3.2 | "Inward" binding mode leads to higher binding affinity in AKR1C3.
As shown in Table 1, the predicted binding energy of 38 with AKR1C3 is much lower than that with AKR1C2, regardless of which "Outward" binding pose in AKR1C2 (À44.4, À34.1, and À33.4 kcal/mol for AKR1C3/38-F I G U R E 4 3D structure diagrams and key residues involved in protein-ligand interaction of (a) AKR1C2/38-Outward1, (b) AKR1C3/38-Outward2, and (c) AKR1C3/38-Inward (The residues in SP1, SP2, and SP3 are colored in gold, green, and blue, respectively) Inward, AKR1C2/38-Outward1, and AKR1C2/38-Out-ward2, respectively). The energy component analysis indicates that van der Waals and electrostatic interaction are both dominant factor for the stronger protein-ligand interaction in AKR1C3 than AKR1C2 (Table 1). In the "Inward" bound AKR1C3-38 structure (Figure 4c), the tetrahydroquinoline moiety is surrounded by multiple residues in SP1 as well as Asn167 and Trp227 in SP2, and thus forms more extensive interaction with the protein in AKR1C3 than in AKR1C2 for both "Outward1" and "Outward2" poses (À23.5 vs. À17.3 and À15.4 kcal/mol, Table S5). This may offer an explanation why the replacement of the phenyl ring in FLF by a tetrahydroquinoline fragment can dramatically improve the compounds' inhibitory selectivity toward AKR1C3 over AKR1C2. 45,46 However, given the small difference in the calculated binding free energies between the two different 38-bound AKR1C2 structures ("Outward1" and "Outward2"), it is impossible to determine which of the two outward binding poses is the preferred binding mode of 38 in AKR1C2. To this end, we performed US simulations to investigate how 38 is stabilized in the active site of AKR1C2.
2.4 | What leads to the distinct binding modes of 38 in AKR1C3 and AKR1C2 2.4.1 | The SP1 in AKR1C3 is spacious enough to accommodate the large tetrahydroquinoline fragment As shown in Figure S2d, the structural differences in SP1 are evident between AKR1C3 and AKR1C2 due to the presence of multiple non-conserved residues in Loops A and C ( Figure S2a vs. b). We first employed the Measure Volume and Area Module in the Chimera Software 56 to compute the volume of the SP1 pocket of AKR1C3 and AKR1C2 based on their crystal structures with bound ligand deleted (PDB entry: 4FAM and 4JQA, respectively), which are both larger than that of inhibitor 38's tetrahydroquinoline moiety (224.0 and 215.4 vs. 166.9 Å, Figure 5). However, the pocket size of protein's one conformation is not its really size in view of the inherent protein flexibility.
Thus, we extracted the bound ligand of 4FAM and 4JQA, and run 500 ns MD simulations for the apo- AKR1C3 and AKR1C2. Clustering method was employed to generate 20 representative structures from the trajectory of two MD simulations, respectively. The value comparisons of the SP1's volume suggest that the binding pocket of AKR1C3 is $20% larger than that of AKR1C2 (226.3 ± 21.73 vs. 185.3 ± 18.59, Table S6). Moreover, the SP1 volume of all AKR1C3 structures are larger than the size of the tetrahydroquinoline fragment, while 20% of AKR1C2's SP1 cannot hold the tetrahydroquinoline group because of their smaller size. The different size of SP1 can be explained by the non-conserved residues Ser118 and Ser308 in AKR1C3, which, in AKR1C2, were replaced by the bulky residues Phe118 and Leu308 ( Figure S8). Therefore, the tetrahydroquinoline moiety can snug into the more spacious SP1 of AKR1C3 without insurmountable steric hindrance, while the SP1 in AKR1C2 is too crowded to bind the bulky tetrahydroquinoline group.
2.4.2 | High flexibility of SP1 in AKR1C3 enables the "Inward" binding of 38 The root-mean-square fluctuations (RMSFs) of Cα atoms were calculated for the six simulated systems using the cpptraj module. As shown in Figure S9, the apo-AKR1C3 is more flexible than the apo-AKR1C2 in several regions, especially Loops A-C that constitute SP1, reflected by the larger RMSF values of the apo-AKR1C3 ( Figure S9a). The GNM-driven protein mobility analysis (Figures S9b,c) confirms the higher inherent flexibility of Loops B and C in AKR1C3 than AKR1C2 ( Figure S2). Additionally, the binding of 38 to AKR1C3 leads to strong interaction of the tetrahydroquinoline with and thus stabilizes the residues in SP1 of AKR1C3, particularly those in Loops B and C (Table S5, Figures 4 and S2). By comparison, docking of the tetrahydroquinoline into SP1 of AKR1C2 leads to unfavorable interaction with SP1 and dramatically amplifies the dynamics of Loop C (Figure 6c), while the 2.4.3 | The "Inward" to "Outward" conformational transition in ARK1C3 and AKR1C2 The MD simulation on the AKR1C2/Inward-38 system showed a spontaneous transition from an "Inward" to an "Outward2" conformation (Figure 3a-e). However, considering the limited conformation sampling in conventional MD simulations, the obtained "Outward2" conformation is unlikely the final binding mode of 38 in AKR1C2. Thus, US simulations were performed to quantitatively probe the thermodynamically stable binding conformations of compound 38 in ARK1C2 and ARK1C3. As shown in Figure S10, the computed PMF curves reached convergence after $6 ns US simulations, which were used for further analyses below (Figures 7 and 8). Figure 7f shows the PMF profile for the "Inward" to "Outward" transition in AKR1C2 by rotating the C7 S1 bond in 38. Different binding poses of 38 are captured along the PMF with the representative ones illustrated in Figures 7a-e and 7a 0 -e 0 , indicating the torsional angle C3-C7-S1-N1 is a sound reaction coordinate for the US calculation. As shown in Figure 7f, the "Inward" conformations corresponding to points d 0 and e 0 on the PMF curve (Figures 7d 0 ,e 0 . Figure 7b 0 ). In both complexes, the tetrahydroquinoline fragment of 38 is partially out of the SP1 cavity. The decreased PMF value of state b 0 relative to the "Inward" conformation d 0 provides an explanation for the spontaneous "Inward" to "Outward2" conformational transition of 38 in the MD simulation of AKR1C2/38-Inward. After flipping out of the SP1 pocket, the AKR1C2/38 system samples three representative states c, d, and e with the tetrahydroquinoline group completely out of SP1 (Figure 7c-e). Structural comparison reveals that state c that corresponds to the global minimum ($0 kcal/mol) of the PMF curve (Figure 7f) represents the "Outward1" binding pose predicted by the docking (Figures 2b vs. Figure 7c). It is evident from the PMF curve that the "Outward1" conformations of 38 in AKR1C2 are more stable than the "Inward" and "Outward2" bound states (Figure 7c vs. 7c 0 -e 0 and 7b 0 ), suggesting that the "Out-ward1" state is likely the preferred binding pose of 38 in AKR1C2. To further corroborate this, we performed an additional set of US simulations using AKR1C2/38-F I G U R E 8 (a-e, a 0 -e 0 ) Representative conformations sampled in the umbrella sampling (US) simulations of AKR1C3/38. (f) Potential mean force (PMF) profile as a function of the C3-C7-S1-N1 dihedral angle using AKR1C3/38-Inward (b 0 ) as the input structure Outward2 as the starting structure to explore the local transition between the two outward states. The computed PMF profile (Figure 7g) is broadly consistent with the corresponding region in the full PMF for the Outward2 to Outward1 transition, with "Outward2" corresponding to a local minimum g 0 ($3.7 kcal/mol), which is less stable than "Outward1" corresponding to the global minimum c ($0 kcal/mol). Therefore, we conclude that 38 adopts the "Outward1" binding mode in the active site of AKR1C2.
In contrast, as shown in Figure 8a-e and a 0 -e 0 , the true "Outward" states were not sampled in the US simulations of the AKR1C3/38-inward system in which the tetrahydroquinoline moiety of 38 stays strongly bound and does not escape from the SP1 cavity. Hence, states be can also be defined as "Inward" conformations, as the tetrahydroquinoline fragment rotates only inside the SP1 pocket (Figure 8b Figure 8f, which is also consistent with the strong binding interaction of 38's tetrahydroquinoline fragment with surrounding residues in SP1 of AKR1C3 ( Figure 4 and Table S5).
To sum up, the US simulations quantitatively confirm that compound 38 binds AKR1C2 and AKR1C3 in distinct poses, an "Inward" conformation in AKR1C3 versus an "Outward1" conformation in AKR1C2, consistent with our above results that the tetrahydroquinoline group of 38 can fit well in the SP1 pocket of AKR1C3, while AKR1C2's SP1 pocket is too crowded and rigid to accommodate this bulky group.
| DISCUSSION
The high selectivity of compound 38 toward AKR1C3 over AKR1C2 was shown to originate from its distinct binding modes, "Inward" and "Outward1," respectively, in the two isoforms. Specifically, in AKR1C3, the tetrahydroquinoline moiety of 38 can fit in SP1 and interact favorably with the surrounding residues. In AKR1C2, however, the tetrahydroquinoline moiety cannot be accommodated in SP1 and thus binds outside of the SP1 pocket, partially exposed to the solvent, which leads to weaker binding with AKR1C2 (À23.5 vs. À17.3 kcal/mol, Table S5). Therefore, modifications harboring a bulky group, such as tetrahydroquinoline or similar moieties, tend to enhance the compound's inhibition activity toward AKR1C3, while maintaining its low activity toward AKR1C2 (Table S1), as evidenced by the contour maps for the activity and selectivity of these compounds from the 3D-QSAR models (Figures 1 and S5).
In view of the dominant role of tetrahydroquinoline in compound potency and selectivity, we selected several analogues (28,38,61, and 85) with various modifications on the tetrahydroquinoline core to probe the binding mechanism with AKR1C3. Of them, analogues 28 and 38 with a NO 2 and Br group added at the 5-and 6-position of the tetrahydroquinoline core, respectively, are the most potent (IC 50 : 0.0061 and 0.0089 μM, respectively, Table S1) AKR1C3 inhibitors. In both 28 and 38 bound AKR1C3, the inhibitor's tetrahydroquinoline forms favorable van der Waals contact with SP1 residues Ser118, Met120, Phe306, Phe311, Tyr317, Pro318, and Tyr319, and SP2 residue Asn167 (Figure 9 and Table S5 and S7). Among them, the ΔG bind values of Phe306 and Phe311 are particularly strong, À5.64 and À5.26 kcal/mol for 28, and À5.74 and À5.78 kcal/mol for 38, respectively, due to favorable π-π stacking interactions, in line with the contour maps of the 3D-QSAR model in which the tetrahydroquinoline is located in the region favorable for aromatic ring interaction. Additionally, 28 and 38 form van de Waals interactions with Asn167, À4.08 and À4.02 kcal/mol, respectively. Furthermore, the 3D-QSAR analysis revealed an important role of electrostatic interaction, especially the negative charge carried by the tetrahydroquinoline, in governing both inhibitory activity and selectivity (Table S4, Figures 1 and S5). The energy decomposition calculations confirmed that the electrostatic component is a crucial part of the total proteinligand interaction energy with ΔE ele of À18.1 and À16.1 kcal/mol between SP1 residues and the tetrahydroquinoline moiety in 28 and 38, respectively. It can also be noted that minor changes on the tetrahydroquinoline core cause large changes in binding energy and potency (Figure 9 and Tables S5 and S7). For example, in 28, the electro-withdrawing NO 2 group in tetrahydroquinoline enhances its interactions with Ser118, Pro119, and Met120, while the 6-bromo analogue 38 shows stronger interaction with Tyr319 than 28, which can be attributed to the formation of a halogen bond between 38 and Tyr319. Therefore, introducing specific substituents at certain positions of tetrahydroquinoline can be a viable strategy to enhance the compounds' anti-AKR1C3 activity. However, despite the size-tolerance and plasticity of SP1 in AKR1C3, the added substituents still have a size limit to avoid steric clash, as exemplified by a 215-fold potency decrease of 44 that contains a large 4-trifluoromethoxyphenylacetylene group (Table S1).
In analogue 85, the two groups connected by the sulfonamide were partially reversed, with the tetrahydroquinoline and phenyl ring replaced by napthyl and piperidine, respectively. Despite alternative orientations of napthyl and tetrahydroquinoline, the interaction energies of 85's napthyl group with residues in SP1 do not change much, and those of its piperidine with Trp86 and Trp227 in SP2, and Tyr24 in SP3 increase slightly (Table S7 and Figure 9), which accounts for a moderate anti-AKR1C3 activity decrease of 85 (IC 50 : 0.032 μΜ). By contrast, for the open chain analogue 61, its anti-AKR1C3 potency is markedly reduced (IC 50 = 1.44 μM, Table S1), which can be explained by its far attenuated interactions with several critical residues Ser118, Pro119, Met120, Phe306, and Phe311 than its enhanced interactions with Asn307 and Ser308 (Figure 9 and Table S7), indicating that AKR1C3's SP1 prefers the binding of a rigid bicyclic moiety.
Furthermore, a wide range of IC 50 values due to the changes of the R 3 substituent in the phenyl ring (Table S1) highlights its importance for the inhibitor's activity. The contour maps of the Gaussian-based 3D-QSAR fields clearly show the negative effect of incorporating a hydrophobic group into the three-position of the phenyl ring (Figures 1 and S5c), which instead favors a positively charged and H-bond acceptor group (Figures 1b, S5b, 1d, and S5d). Consistently, as shown in Figure 10 and Table S8, unlike 38, both 67 and 68 undergo binding pose changes with the substituted NH 2 and CF 3 pointing toward the entrance of the active site pocket, resulting in reduced electrostatic interaction of the R 3 moiety with His117 in the OX site (À6.70, À3.00, and À2.82 kcal/mol in 38, 67, and 68, respectively). These structural changes also lead to decreased interactions of the compound with several other residues, such as Lys84, Tyr24, Leu54, and Phe311. A strong H-bond between His117 and the benzoic moiety seems to anchor the ligand in a proper binding pose to strengthen its interactions with AKR1C3-specific residues, supported by the high anti-AKR1C3 activity of analogues 28, 38, and 74 that all contains a H-bond acceptor, acid or its isostere tetrazole at the 3-position of the phenyl ring. Therefore, the property and position of the substituents on the phenyl ring are the molecular determinants dictating the compounds' binding conformation and affinity, and thus its anti-AKR1C3 potency and selectivity.
| CONCLUSIONS
In this study, an integrated computational approach was employed to offer molecular-level insights into the high selectivity mechanism of 3-(3,4-dihydroisoquinolin-2 (1H)-ylsulfonyl)benzoic acid derivatives toward AK1R1C3 over its close isoform AKR1C2. Of them, the most AKR1C3-selective inhibitor 38, up to 5000-fold, was chosen for in-depth analysis. The calculations showed that 38 interacts strongly with AKR1C3 in an "Inward" binding mode with its tetrahydroquinoline moiety stabilized in the sub pocket1 (SP1), while the crowded and rigid SP1 of AKR1C2 cannot accommodate the bulky tetrahydroquinoline moiety of 38, leading to an "Outward1" binding mode and accordingly a lower binding affinity of 38 in AKR1C2. The US MD simulations confirmed that the AKR1C3/38-Inward is the most stable binding pose in AKR1C3 while the AKR1C2/38-Outward1 corresponds to the lowest free energy (PMF) state in AKR1C2. Hence, 38's inhibitory preference toward AKR1C3 can be explained by its distinct binding modes in AKR1C2 and AKR1C3, which is in turn owing to the inherent size and flexibility difference between the binding pockets of the two isoforms. Moreover, 3D-QSAR combined with residue-based energy decomposition analysis revealed that SP1 in AKR1C3 prefers to bind with a rigid bicyclic moiety and the interaction of the R 3 substituent with His117 in the OX site is key for stabilizing the ligand in AKR1C3. Overall, this study represents a first attempt to elucidate the inhibitor selectivity mechanism between two AKR1C isoforms, which will pave the way for developing novel AKR1C3 inhibitors with improved efficacy while avoiding "off-target" effects for CRPC therapy.
Prior to MD simulations, all bound ligands, including selected 3-(3,4-dihydroisoquinolin-2(1H)-ylsulfonyl)benzoic acids and the coenzyme NADP + , were optimized by the Hartree-Fock (HF) method 62 at the level of B3LYP/6-31G* theory 63 with Gaussian 09. 64 Partial atomic charges for the optimized geometries of the ligands were calculated by the restrained electrostatic potential (RESP) algorithm. 65 The AMBER14SB force field 66 was used for the proteins, while the atom types and force field parameters of ligands were assigned using the general AMBER force field (gaff). 67 All the complexes were immersed into a periodic TIP3P water box 68 with at least 12 Å water padding for each side of the rectangular box, and appropriate numbers of counter-ions (Na + or Cl À ) were added to neutralize each system. The missing hydrogen atoms were added to the heavy atoms using the leap module. 69 A stepwise minimization was then followed using the pmemd program in Amber 18 70 : (1) 1000 cycles of steepest descent and 4000 cycles of conjugate gradient minimization with the non-hydrogen atoms restrained by 5 kcal/mol/Å 2 force constant; (2) 5000 cycles of minimization (1000 cycles of steepest descent and 4000 cycles of conjugate gradient minimization), where a 5 kcal/mol/(mol Å 2 ) force was used to fix the non-hydrogen of the complex; (3) 1000 cycles of steepest descent and 4000 cycles of conjugate gradient minimization with the heavy atoms of the complex restrained (2 kcal/mol/Å 2 ); and (4) the whole system was subject to 10,000 cycles of optimization without any restraints (5000 cycles of steepest descent and 5000 cycles of conjugate gradient minimization).
The minimized systems were gradually heated up to 300 K in the NVT ensemble with a restraint of 2 kcal/ (mol Å 2 ) on the backbone of each complex (100 ps). Each system was then equilibrated in the NPT ensemble (T = 300 K and P = 1 atm) with and without the 2 kcal/ (mol Å 2 ) restraints for 500 ps, respectively. Afterward, 1 ns additional equilibration in the NPT ensemble was carried out for each unrestrained system using the algorithms of Berendsen barostat 71 and Langevin thermostat. 72 Finally, all AKR1C2 and AKR1C3 systems were subjected to 500 ns NPT MD simulations. In particular, the MD simulation of the AKR1C2/38-Inward complex was extended to 1 μs. The Particle Mesh Ewald (PME) technique 73 was utilized to treat the long-range electrostatic interactions with a cutoff distance of 10 Å, and the same threshold value was used for the truncation of Lennard-Jones potentials. The SHAKE algorithm 74 was applied to constrain all bonds involving hydrogen atoms to their equilibrium length. The time step was set to 2 fs and the snapshots were saved every 1 ps for subsequent structural and energetic analysis.
| Structural motion analysis and pocket size calculations of AKR1C2 and AKR1C3
MD simulation trajectories were analyzed using the cpptraj module 75 in AmberTools18. The root mean square deviations (RMSDs) as a function of the simulation time (t) were computed to assess the stability of the protein structure in the simulation. The flexibility and dynamics of individual residues were evaluated by the root mean square fluctuations (RMSFs). Besides, the residue mobility were also quantified by the slowest GNM modes using the DynOmics ENM server. 76 For each system, 5000 conformations were evenly extracted from the 500 ns MD trajectory. Then, these conformations were clustered using the k-means clustering algorithm through iterative minimization of the sum of the pairwise root-mean-square displacements (RMSDs) between each conformation and its cluster centroid over all clusters. Twenty representative structures were generated for each trajectory of the apo AKR1C3 and apo AKR1C2 simulations. The pocket size was calculated as the average over the 20 obtained structures using the Computed Atlas of Surface Topography of proteins (CASTp) 77 software.
| Protein-ligand interaction energy calculations
The binding affinity of a drug with its target was estimated by the Molecular Mechanics/Generalized Born Solvent Area (MM/GBSA) methodology. [78][79][80][81] According to Equation (1), the total binding free energy (ΔG bind ) can be subdivided into several terms, including the van der Waals interaction (ΔE vdW ), the electrostatic interaction (ΔE ele ), the polar (ΔG GB ), and nonpolar (ΔG SA ) parts of the solvation free energy (ΔG solvation ), and the conformational entropy upon ligand binding (ÀTΔS). 80,81 Herein, the pmemd module in Amber 18 was used to compute the molecular mechanics gas-phase energy ΔE MM (ΔE int , ΔE vdW , and ΔE ele ), where ΔE int , change of the intramolecular energy upon ligand binding, can be neglected due to the use of the single trajectory strategy. 80 82 with the solute (ε in ) and solvent (ε out ) dielectric constants set to 2.0 and 80, respectively. ΔG SA was determined by the change of the solvent-accessible surface areas (ΔSASA) with the LCPO algorithm 83 : ΔG SA = γÂΔSASA + β, where γ and β were set to 0.0072 kcal/(mol Å 2 ) and 0 kcal/mol, respectively. ÀTΔS is not considered here since the studied compounds are analogues with high structural similarity, and the inclusion of the entropy term demands expensive computational cost but does not necessarily improve the prediction accuracy. 84,85 Binding free energy decomposition as implemented in MMPBSA.py 86 was used to identify and compare the residue-ligand interaction differences between AKR1C2 and AKR1C3. The interaction energy of individual residues with the ligand can be formulated as: ΔG Bind = ΔE vdW + ΔE ele + ΔG GB + ΔG SA . Except for ΔG SA , which was calculated by the ICOSA algorithm, 87 all other terms were calculated based on the same parameters used in the above MM/GBSA calculations.
| Dataset collection and 3D-QSAR model construction
A total of 77 derivatives of 3-(3,4-dihydroisoquinolin-2 (1H)-ylsulfonyl)benzoic acids (compounds 17-94, Table S1) that covered a reasonable range of chemical diversity and inhibitory activity were collected from Jamieson and colleagues' work. 46 The biological activities of these compounds against AKR1C2 and AKR1C3 (IC 50_C2 and IC 50_C3 , μM), together with their structures, were summarized in Table S1. The 3D structures of all molecules were sketched using Maestro and were then optimized using Ligprep module with the OPLS3e force field in Schrödinger 2018. 88 The flexible ligand alignment method was employed to align all the compounds using 38 as a template given its highest IC 50 value ( Figure S3). Herein, the IC 50 values of compounds were converted into pIC 50 . pIC 50_C3 and ΔpIC 50 (pIC 50_C3 -pIC 50_C2 ) were then used as the dependent variables in 3D-QSAR models for AKR1C3 activity and selectivity, respectively (Table S2). The dataset was divided randomly into the training set and test set with a ratio of 4:1. The extended Gaussian-based potential function 89 of Schrodinger 2018 Suite was employed to construct the 3D-QSAR models using Partial Least Squares (PLS) regression, 90 where six Gaussian fields, including steric, electrostatic, hydrophobic, hydrogen bond donor (HBD), hydrogen bond acceptor (HBA), and aromatic ring potential fields, were calculated to evaluate each type of interactions using Gaussian equations. Parameters involved in 3D-QSAR model building were set using the default values.
| The validation and analysis of 3D-QSAR models
The accuracy and usefulness of the Gaussian-based 3D-QSAR models were evaluated based on the internal and external validation parameters, especially test set (RMSE test , Q 2 test , and Pearson-r test ) and training set (R 2 training ) statistics. RMSE test indicates root-mean-square error in the test set predictions. Q 2 test is directly analogous to R 2 training (value of R 2 for the regression, correlation coefficient, 0-1), but based on the test set predictions. Pearsonr test indicates the correlation between the predicted and observed activity for the test set (0-1). In a validated 3D-QSAR model, the "field fractions" provide a general idea of the relative contribution of each field to the activity and selectivity of the molecules. For example, if the steric and hydrophobic Gaussian field fractions are much larger than all other types, then most of the binding energy comes from then hydrophobic interactions. Besides, the alignment of the studied compounds with the 3D-QSAR contour maps of each field can help identify the favorable (positive) and unfavorable (negative) structural features for ligand selectivity/activity.
| US simulations
US simulation, a common enhanced sampling approach, was used to investigate the interconversion between the "Inward" and "Outward" (Outward1 and Outward2) conformations of 38 in the binding cavity of AKR1C3 and AKR1C2, in which AKR1C3/38-Inward (Figure 2a), AKR1C2/38-Outward2 (Figure 2e) and AKR1C2/38-Inward (Figure 2c) structures were respectively used as the input structure. In US simulations, biasing potentials are imposed on the reaction coordinates (RCs) to drive the system from one conformational state to the other. [91][92][93] The transition of 38 between the "Inward" and "Outward" conformations can be realized through rotating 38's tetrahydroquinoline fragment around the C7 S1 bond by 360 , so the torsional angle C3-C7-S1-N1 was chosen as the RC. The full torsional angle range was divided into 36 windows (every 10 ), and each window was subjected to 10 ns US simulations. The harmonic potentials applied to each window for biased sampling in that window are shown in Equation (2).
where k i is the spring constant in window i (100 kcal/ mol rad 2 ), and θ ref i and θ denote the reference and instantaneous angle of window i, respectively. All US simulations were performed using Amber 18. The weighted histogram analysis method (WHAM) 94,95 was employed to estimate the potential mean force (PMF) curve along the RC. WHAM is an effective method to reconstruct unbiased probability distribution from biased probability distribution by umbrella sampling, which contains three steps: (1) calculate the unbiased probability density Pi (x) of each window, (2) obtain the overall P(x) of each RC by linearly combining the Pi(x) of each window, (3) compute the free energy of each RC point and finally draw the PMF curve.
ACKNOWLEDGMENTS
This work was supported by the startup fund from the Faculty of Environment and Life at Beijing University of Technology. Computation was performed using the computing resources at the Ohio Supercomputer Center and the Faculty of Environment and Life at Beijing University of Technology.
CONFLICT OF INTEREST
The authors declare no competing financial interest.
DATA AND SOFTWARE AVAILABILITY
Data and software used in this manuscript are all available.
The software packages utilized in this manuscript include ChemDraw Ultra 8.0, Glide, Gaussian 09, Amber18, UCSF Chimera 1.14, and OriginPro 2016. All input and output data files are made available upon request. The DynOmics ENM portal employed in this manuscript can be accessed at http://enm.pitt.edu/ or http://dyn.life.nthu.edu.tw/oENM/.
|
v2
|
2022-11-08T06:17:37.186Z
|
2022-11-06T00:00:00.000Z
|
253383514
|
s2ag/train
|
Effect of positive surgical margins at radical prostatectomy on cancer-specific mortality in high/very high-risk prostate cancer patients with Gleason Grade Group 4-5.
BACKGROUND
The effect of positive surgical margins (PSM) on cancer specific mortality (CSM) in high/very high-risk (HR/VHR) prostate cancer (PCa) with aggressive Gleason Grade Group (GGG) is unknown. We tested PSM effect on CSM in this setting, in addition to testing of radiotherapy (RT) benefit in PSM patients.
METHODS
We relied on Surveillance, Epidemiology, and End Results database (2010-2015), focusing on HR/VHR patients with exclusive GGG 4-5 at radical prostatectomy (RP). Kaplan-Meier plots and multivariable Cox regression models tested the relationship between PSM and CSM. Moreover, the effect of RT on CSM was explored in PSM patients.
RESULTS
Of 3383 HR/VHR patients, 15.1% (n = 511) exhibited PSM. Patients with PSM harbored higher rates of GGG 5 (60.1% vs. 50.9%, p < 0.001), pathologic tumor stage T3a (69.1% vs. 45.2%, p < 0.001) and lymph node involvement (14.1% vs. 9.4%, p < 0.001), relative to patients without PSM. PSM rates decreased over time (2010-2015) from 16.0% to 13.6%. Seven-year CSM-free survival rates were 91.6% versus 95.7% in patients with and without PSM, respectively. In multivariable Cox regression models, PSM was an independent predictor of CSM (hazard ratio = 1.6, p = 0.040) even after adjustment for age, prostate specific antigen, pathologic tumor stage and lymph node status. Finally, in PSM patients, RT delivery did not reduce CSM in either univariable or multivariable Cox regression models.
CONCLUSIONS
In HR/VHR PCa patients with exclusive GGG 4-5, PSM at RP adversely affect survival. Moreover, RT has no protective effect on CSM. In consequence, lowest possible PSM rates are crucial in such patients.
|
v2
|
2022-11-08T06:17:37.311Z
|
2022-11-06T00:00:00.000Z
|
253383398
|
s2ag/train
|
Melanoma surgery—An update
Cutaneous melanoma is the major cause of mortality from all skin cancers. The treatment has been revolutionized in recent years by introduction of immunotherapy and targeted therapy for melanoma patients Stages III and IV. Therefore, the role of surgery in melanoma treatment needs to be redefined. In this narrative review, we will focus on surgery for diagnosis, treatment of primary tumor, and metastases in the era of new and effective medical treatment options. Neoadjuvant therapy is currently investigated in several trials. Surgery for treatment‐resistant metastases is another field of interest. In conclusion, surgery remains a cornerstone for diagnosis and treatment of primary melanoma. Therapeutic lymphadenectomy has lost importance while surgery in sentinel lymph node diagnostics and metastasectomy are useful in a tailored individual approach of combined treatments. There is a trend to less invasive surgical procedures.
|
v2
|
2022-11-27T16:16:21.084Z
|
2022-11-06T00:00:00.000Z
|
253990078
|
s2ag/train
|
High Serum Level of TNF-α in Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis
BACKGROUND: Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis are severe cutaneous adverse drug reactions. Some immunological and genetic factors are believed to be involved in the pathogenesis of SJS/TEN, including tumor necrotic factor-alpha (TNF-α). Activated T-cells secrete high amounts of TNF-α and interferon-gamma that both cytokines lead to increased expression and activity of keratinocyte inducible nitric oxide synthase playing an important role in the apoptosis of keratinocytes.
AIM: This study aims to evaluate the serum level of TNF-α in SJS/TEN and the relation between it and the progress of SJS/TEN.
METHODS: This was a sectional descriptive study conducted at the National Hospital of Dermatology and Venereology, in Hanoi, Vietnam, from October 2017 to September 2019. Forty-eight SJS/TEN patients, 43 erythema multiforme (EM) patients, and 20 healthy controls (HCs) participated. TNF-α levels were measured using the fluorescence covalent microbead immunosorbent assay (FCMIA) (ProcartaPlex Immunoassay Panels kit, Thermo Fisher Scientific, USA). The Mann–Whitney U-test was used to compare serum TNF-α levels of two groups. The Wilcoxon tests were used to compare quantitative variables before and after the treatment. Differences were considered to be statistically significant at p < 0.05.
RESULTS: Nineteen SJS patients (39.5%) and 29 TEN patients (60.5%) participated in our study. The mean age was 49.3, range 19−77 years (47.9% of males and 52.1% of females). The most common causative drugs were traditional medicine (29.1%), carbamazepine (12.5%), and allopurinol (12.5%). On the day of hospitalization, the mean serum level of the SJS/TEN group was 32.6 pg/ml with a range from 1.3 pg/ml to 771.2 pg/ml. This level was significantly higher than that of the HCs group (p < 0.05) but not higher than that of the EM group. The mean serum level of TNF-α in the SJS/TEN patients on the day of hospitalization was 32.6 pg/ml, higher than that on the day of re-epithelialization (2.7 pg/ml) and the difference was statistically significant with p < 0.05.
CONCLUSION: Serum TNF-α levels are a good biomarker to evaluate the progress of SJS/TEN but it is not good to differentiate SJS/TEN from EM.
|
v2
|
2022-11-07T14:32:46.713Z
|
2022-11-07T00:00:00.000Z
|
253372257
|
s2orc/train
|
Case Report: MAP2K1 K57N mutation is associated with primary resistance to anti-EGFR monoclonal antibodies in metastatic colorectal cancer
Background We aim to identify the prevalence and the role of the MAP2K1 K57N mutation in predicting resistance to anti-EGFR agents in metastatic colorectal cancer (mCRC) patients. Methods We retrospectively reviewed tumor-based next generation sequencing (NGS) results from mCRC patients screened for enrollment in the GO40872/STARTRK-2 clinical trial between July 2019 and March 2021. Then, in patients harboring microsatellite stable (MSS) RAS and BRAF wild-type MAP2K1 mutant mCRC, we reviewed outcome to treatment with anti-EGFR monoclonal antibodies. Results A total of 246 mCRC patients were screened. Most of them, 215/220 (97.7%), were diagnosed with MSS mCRC and 112/215 (52.1%) with MSS, RAS and BRAF wild-type mCRC. Among the latter, 2/112 (1.8%) had MAP2K1 K57N mutant mCRC and both received anti-EGFR monotherapy as third line treatment. In both patients, MAP2K1 K57N mutant tumors proved primary resistant to anti-EGFR agent panitumumab monotherapy. Of interest, one of these patients was treated with anti-EGFR agents three times throughout his course of treatment, achieving some clinical benefit only when associated with other cytotoxic agents (FOLFOX or irinotecan). Conclusion We verified in a clinical real-world setting that MAP2K1 K57N mutation is a resistance mechanism to anti-EGFR agents in mCRC. Thus, we suggest avoiding the administration of these drugs to MSS RAS and BRAF wild-type MAP2K1 N57K mutant mCRC.
Introduction
Anti-epidermal growth factor receptor (EGFR) monoclonal antibodies (MoAbs) are recommended as standard treatment options for patients affected with RAS wild-type (WT), mainly left-sided, metastatic colorectal cancer (mCRC) (1). Anti-EGFR MoAbs prevent EGFR activation, thus blocking the downstream mitogen-activated protein kinase (MAPK) pathway and the resulting proliferative signal (2). Mutations occurring in KRAS and NRAS exons 2, 3, and 4 represent the main mechanisms of resistance to these anticancer agents, countering the MAPK pathway blockade and even entailing detrimental survival in RAS mutant mCRC patients receiving anti-EGFR drugs (3). Accumulating evidence also suggests BRAF mutations being a resistance mechanism to anti-EGFR MoAbs, even though treatment is allowed as per label (4). Over the years, several other molecular biomarkers beyond RAS and BRAF have been associated to primary and/or acquired resistance to anti-EGFR MoAbs, such as gene mutations of ERBB2, EGFR, FGFR1, PDGFRA, PIK3CA, PTEN, AKT1, and MAP2K1/MEK1, amplifications of KRAS, ERBB2, and MET, and fusions of ALK, ROS1, NTRK1-3, and RET (5)(6)(7)(8)(9)(10)(11). However, given the low prevalence of such alterations and the confounding effect of cytotoxic agents administered together with anti-EGFR MoAbs, no definitive consensus on therapeutic implications has been derived so far.
The increasing availability of next generation sequencing (NGS) panels in the clinical setting allows frequent detection of these alterations, thus requiring evaluation of whether available evidence is strong enough to preclude patients from a potential effective treatment option, therefore posing an emergent therapeutic challenge. Here we provide a molecular tumor board (MTB) discussion of two emblematic cases of uncommon MAP2K1 K57N mutated mCRC patients, commenting upon treatment decision making focused on anti-EGFR drug administration.
Methods
At Niguarda Cancer Center, we retrospectively reviewed tumorbased NGS results of mCRC patients screened for enrollment in the GO40872/STARTRK-2 clinical trial (NCT02568267) between July 2019 and March 2021. Through this process, we checked for mCRC samples harboring MAP2K1 mutations. After the identification of microsatellite stable (MSS), RAS and BRAF WT, MAP2K1 mutant mCRC, we retrospectively reviewed patients' records to evaluate treatment outcome to anti-EGFR MoAb.
The molecular results presented in this manuscript are based on FoundationOne CDx NGS (Foundation Medicine, Inc.) panel results performed on archival formalin-fixed paraphingembedded (FFPE) tumor tissue; data from NGS and allele frequency were made available by the sponsor of the GO40872/STARTRK-2 trial upon personalized request, following the approval by the local ethical committee. Concerning mismatch repair (MMR) assessment, results from FoundationOne CDx NGS panels were integrated by immunohistochemistry (IHC) testing performed on archival FFPE tumor tissue, if the NGS data was not available.
After MAP2K1 mutant mCRC patients' identification, previously collected plasma samples were retrospectively analyzed by looking for MAP2K1 K57N circulating tumor DNA (ctDNA). The ctDNA analysis was performed by droplet-digital PCR (dd-PCR).
Both patients consented to the submission and publication of the following article reports.
Case #1
In February 2015, a 41-year-old man was diagnosed with stage IV, RAS and BRAF WT, ERBB2 not-amplified, MSS rectal cancer with synchronous liver metastases. As presented in Figure 1 and Table 1, the patient underwent multimodal treatment with initial first-line therapy with FOLFOX and panitumumab, achieving partial response (PR) after 4 cycles, followed by short-course radiotherapy (RT) on the pelvis and radical surgery on both the primary tumor and liver metastases in June 2015. Thereafter, the patient received post-surgical treatment with FOLFOX for further 8 cycles. After 11 months of follow-up, a liver recurrence occurred and the patient underwent a second metastasectomy in November 2016. However, due to nodal relapse soon after surgery, FOLFOX and bevacizumab was then reintroduced as second line therapy, achieving stable disease (SD) as best response. Treatment had to be interrupted after 6 courses due to persistent G4 thrombocytopenia, that was interpreted as secondary to 5-fluorouracil infusion. Due to oligo-progressive disease (PD) in a single lymph node, the patient underwent stereotactic RT in October 2017, with no further PD until July 2018. Then, the patient received panitumumab monotherapy as a reintroduction strategy with PD after 6 cycles, developing a severe mediastinal syndrome requiring palliative thoracic surgery and RT. At this stage, aiming to extent the spectrum of druggable therapeutic targets, the patient was screened for the GO40872/STARTRK-2 clinical trial (NCT02568267) at Niguarda Cancer Center, Milan, Italy. Thus, a F o u n d a t i o n O n e C D x N G S a s s a y ( h t t p s : / / w w w . foundationmedicine.com/test/foundationone-cdx, Foundation Medicine, Inc.) was obtained from the archival formalin-fixed paraffin-embedded (FFPE) tissue derived from the first liver metastasectomy. The case was presented for MTB discussion, revealing no actionable targets. Of note, a somatic tumor MAP2K1 K57N mutation was found on tissue with 17.1% variant allelic frequency (VAF), and confirmed also on liquid biopsy (LB) at progression to anti-EGFR therapy, assessed as reported elsewhere (12) through droplet digital polymerase chain reaction (ddPCR), with a 16.9% minor allelic frequency (MAF). Further treatment with weekly irinotecan and cetuximab was administered, achieving SD as best response. Finally, regorafenib was initiated after PD but soon interrupted due to the deterioration of the clinical conditions. The patient died on March 31 th , 2020.
Case #2
In September 2019, a 62-year-old man was diagnosed with RAS and BRAF WT, ERBB2 not-amplified, MSS, right-sided mCRC with synchronous liver and bone metastases. Initially, the patient underwent FOLFOX first-line therapy for 12 cycles achieving PR, followed by a therapeutic holiday due to an occurring, nontreatment related intracranial hemorrhagic event. Following systemic PD, the patient received palliative bone RT and FOLFIRI as second line therapy, with further PD. Given the absence of available clinical trials and according to the molecular Timeline of treatment for patient #1 and #2. Keys: beva, bevacizumab; cmab, cetuximab; FU, follow-up; mCRC, metastatic colorectal cancer; NGS, next genome sequencing; OS, overall survival; pmab, panitumumab; PD, progressive disease; PR, partial response; RAR, rectal anterior resection; rego, regorafenib; RT, radiotherapy; SC, short-course; SD, stable disease. status, the patient started third line panitumumab monotherapy in October 2020. After 3 cycles, due to worsening clinical conditions, we performed a CT scan demonstrating dimensional and numerical PD. At this stage, aiming to expand druggable therapeutic targets, a FoundationOne CDx NGS panel was performed on archival FFPE tissue from the primary tumor biopsy undergone at baseline. The case was presented at MTB: although no druggable targets were retrieved, a tumor somatic MAP2K1 N57K mutation was found, with a 14.7% VAF, thus potentially entailing anti-EGFR resistance. This was confirmed by LB through ddPCR showing a 0.51% MAF. Subsequently, the patient was treated with FOLFOX reintroduction achieving PD as best response, and then regorafenib with SD. In July 2021, further palliative RT was provided for worsening gonalgia secondary to a knee bone metastasis. However, the patient died on August 20 th 2021 ( Figure 1 and Table 1).
Discussion
We present two cases of patients affected by RAS and BRAF WT, MSS mCRC, both harboring a MAP2K1 K57N mutation, and both treated with anti-EGFR monotherapy with early symptomatic PD. Particularly, patient #1 achieved clinical benefit from anti-EGFR agents only when administered in combination with cytotoxic agents (FOLFOX or irinotecan, respectively), while he rapidly progressed to anti-EGFR monotherapy ( Table 1 and Figure 1). Similarly, patient #2 had a right-sided mCRC that proved primary resistant to panitumumab monotherapy as third line of treatment. Altogether, overall survival was consistent with tumor sidedness and other molecular characteristics for both patients.
The presence of a MAP2K1 mutation was the only anti-EGFR resistance identified in both cases presented; indeed, no other MAPK pathway alterations known as putative mechanisms of resistance to anti-EGFR drugs were found (Table 1). For patient #2, we can allege that MAP2K1 activation exerted a primary mechanism of resistance, as the NGS analysis preceded the administration of the first anti-EGFR therapy. Following, LB confirmed the persistence of this alteration also at the time of PD. Differently, it was not possible to distinguish whether MAP2K1 K57N was primary or acquired in patient #1, since archival tissue from initial tissue biopsy was insufficient for NGS analysis; indeed, in this case the NGS analysis was performed on the surgical specimen collected after an initial anti-EGFR based therapy (FOLFOX plus panitumumab). Despite this limitation, we can speculate that it is unlikely that the MAP2K1 K57N mutation in patient #1 was acquired upon anti-EGFR exposure, as it has been reported that acquired MAPK pathway alterations are uncommon [6.8% of mCRC patients progressing to first line anti-EGFR containing regimens (13)] and that patient #1 received only 4 courses of FOLFOX and panitumumab (thus having a time-limited anti-EGFR exposure for resistance acquisition).
Notwithstanding, these data provide retrospective evidence that MAP2K1 K57N mutation is a potential mechanism of resistance to anti-EGFR MoAbs in mCRC patients. Hence, despite the lack of high level of evidence, MAP2K1 alterations should be taken into consideration in clinical practice and reviewed through MTB discussion. Indeed, the Mitogen-Activated Protein Kinase Kinase 1 gene, also known as MAP2K1, MAPKK1 or MEK1, is a well-known oncogene encoding for the protein kinase MEK, and exerting its function downstream of the RAS and BRAF proteins in the MAPK signaling cascade, thus conceivably capable of resistance to MAPK pharmacologic silencing (Figure 2) (14). The prevalence of MAP2K1 mutations in CRC is 1-2%, mainly enriched in RAS/BRAF WT tumors (https://www.cbioportal. org/). Most common alterations are found in the hotspots K57 (as in both the reported cases) and Q56 codons (15). A previously published mechanism study identified MAP2K1 K57N mutations as a class II category mutation, which are partially dependent on upstream RAF activation (16). In clinical practice, the role of MAP2K1 mutations is seldom addressed in mCRC patients, given the absence of specific targeted therapy. However, a contribution to carcinogenesis and resistance to EGFR inhibition has been highlighted in preclinical and translational studies, making MAP2K1 alterations relevant for MTB discussion before treating these patients with anti-EGFR drugs (at least as monotherapy) (6,11,17,18).
Even if the clinical characterization of patients affected by mCRC harboring this alteration is still partial and not conclusive, comparable cases were previously reported by other authors (15, 19). However, in this prior reports patients with MAP2K1 mutant tumors were treated with anti-EGFR drugs combined with other cytotoxic or targeted agents, thus hampering definitive clinical conclusions concerning resistance to anti-EGFR MoAbs (15,19). Besides, only 1 case of MAP2K1 K57N was reported in these studies (15,19). Differently, both our patients were affected by MAP2K1 K57N mutant mCRC experiencing PD to panitumumab monotherapy. Given the infrequent occurrence of these alterations, taking into account MAP2K mutations in clinical practice and then gathering further clinical evidence in form of scientific reports is needed to translate preclinical discoveries to real-world clinical practice and refine the spectrum of mCRC patients not benefitting from anti-EGFR MoAbs.
In conclusion, we discuss the role of MAP2K1 K57N mutation as a negative predictive factor of response and mechanisms of primary resistance to anti-EGFR MoAbs, occurring in 1.8% of RAS and BRAF WT, MSS mCRC. No assumption can be made on the prognostic value of these alterations given the restricted number of patients, although survival data were in line with expectations in both cases. Therefore, based on previously available literature and the present MTB discussion, anti-EGFR agents should be omitted in MAP2K1 K57N mutant mCRC, regardless of primary tumor sidedness, to spare toxicities in patients likely not benefitting from such treatment.
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
All patients consented to the submission and publication of the following article. This study respected ethical principles as established by the Helsinki Declaration and the Good Clinical Practice: Consolidated Guideline approved by the International Conference on Harmonisation (ICH); a study protocol was presented to the local ethics committee Milano Area 3 (Italy). In no way therapeutic changes took place secondary to the results of this study, in conformity with the aforementioned clinical guidelines for the treatment and management of vulnerable patients. The collection, recording, and reporting of data was accurate and ensured the privacy, health, and welfare of research subjects during and after the study. All collected data will be preserved anonymously and with respect to the patients' privacy.
Author contributions
GM, GP, and VG were major contributors in writing the manuscript; CL and EBono performed the pathology procedures that were preparatory to the NGS analysis; BM and AB performed the molecular analyses on plasma samples; GM, GP, VG, AA, KB, FT, EBona, AS-B, and SS provided clinical care to the patients; AS-B and SS supervised oncological care and critically reviewed this article. All authors contributed to the article and approved the submitted version. Publisher's note
|
v2
|
2022-11-07T14:32:56.907Z
|
2022-11-07T00:00:00.000Z
|
253372424
|
s2orc/train
|
CXCL12, a potential modulator of tumor immune microenvironment (TIME) of bladder cancer: From a comprehensive analysis of TCGA database
Background Tumor immune microenvironment (TIME) plays a significant role in the initiation and progression of bladder urothelial carcinoma (BLCA). However, there are only a few researches regarding the association between immune-related genes and tumor-infiltrating immune cells (TICs) in TIME of BLCA. Methods We calculated the proportion of immune/stromal component and TICs of 414 BLCA samples and 19 normal samples downloaded from TCGA database with the help of ESTIMATE and CIBERSORT algorithms. Differentially expressed genes (DEGs) were obtained from the comparison between Stromal and Immune Score and further analyzed by GO and KEGG enrichment analysis, as well as PPI network and COX regression analysis. CXCL12 was overlapping among the above analyses. Single gene analysis of CXCL12 was carried out through difference analysis, paired analysis and GSEA. The association between CXCL12 and TICs was assessed by difference analysis and correlation analysis. Results Immune and stromal component in TIME of BLCA were associated with patients’ clinicopathological characteristics. 284 DEGs were primarily enriched in immune-associated activities, among which CXCL12 was the most significant gene sharing the leading nodes in PPI network and being closely related with patients’ survival. Single gene analysis and immunohistochemistry revealed that CXCL12 was down-regulated in BLCA samples and significantly related with the clinicopathological characteristics of patients. Further analysis suggested that CXCL12 was involved in the immune-associated activities probably through its close cross-talk with TICs. Conclusions CXCL12 down-regulation could be a potential biomarker to predict the unbalanced immune status of TIME of BLCA, which might provide an extra insight for the immunotherapy of BLCA.
Introduction
Since the application of cisplatin-based chemotherapy in the mid-1980s, only a few advances have been made in the treatment of bladder urothelial carcinoma (BLCA) (1). Fortunately, immunotherapy has emerged as a novel potential therapy recently. Several clinical trials, such as IMvigor 210 study (2) and CheckMate 275 study (3), revealed that some BLCA patients benefited greatly from the treatment of immune-checkpoint blockade (ICB). Although multiple biomarkers have been associated with the prediction of immunotherapy effect, including the expression of programmed cell death protein 1 (PD-1), PD-L1 and tumor mutation burden (TMB), it is still less than satisfactory to select BLCA patients who are likely to benefit most from immunotherapy (4).
In recent years, the tumor microenvironment (TME) has drawn our attention as its important role in modulating the initiation and progression of cancers, including BLCA (5)(6)(7)(8)(9)(10). TME is composed of nonmalignant cells, vessels, lymphoid organs or lymph nodes, nerves, intercellular components and metabolites (11). In brief, stromal component and immune component constitute the TME (12). Furthermore, accumulating research found that tumor immune microenvironment (TIME) had great potential in influencing tumor initiation, predicting immunotherapeutic responsiveness and new therapeutic targets (13). Numerous studies manifested that immune-related genes (IRGs), which were obtained from TME, could predict the survival of cancer patients, including breast cancer (14), endometrial cancer (15), liver cancer (16), gastric cancer (5), bladder cancer (17) and so on. For example, Q. Ding, et al. (18) found that a nine-gene signature was closely related with immune infiltration in TME and the survival of ovarian cancer patients. Besides, tumor-infiltrating immune cells (TICs) in TIME, such as tumor-infiltrating lymphocytes (TILs) and tumor-associated macrophages (TAMs), had the potential to be biomarkers and predictors of multiple cancers (19)(20)(21). For instance, elevated level of TILs was associated with better overall survival (OS), higher pathological complete response (pCR) rate, lower risk of recurrence, and more benefit from trastuzumab treatment in breast cancer (22)(23)(24)(25). Deficient CD4+ T cells helped to suppress the response of cytotoxic T lymphocytes (CTLs), which means they could establish efficient and durable antitumor activity (26). TAMs acted to inhibit T cell recruitment and modulate the immunity of various tumors, thus affecting the response of immunotherapy (27). Galectin-9+ TAMs predicted prognosis and response to adjuvant chemotherapy in BLCA patients (28). However, there are few researches regarding the association between IRGs and TICs in the TME of BLCA. The exploration of the relationship between IRGs and TICs in TIME could provide us a new sight into the progression and immunotherapy of BLCA.
Fortunately, with the rapid development of transcriptome profiling based on functional genomics analysis, comprehensive analysis of IRGs and TICs in the TIME of BLCA has become possible. In our study, we applied ESTIMATE and CIBERSORT algorithms to calculate the proportion of immune and stromal component, as well as TICs proportion of BLCA samples from The Cancer Genome Atlas (TCGA) database. Next, we started with differentially expressed genes (DEGs) acquired from the comparison of immune and stromal component in BLCA samples, and found out that CXC chemokine ligand-12 (CXCL12) acted to be a potential biomarker and a promising modulator of TIME through its communicating with multiple TICs,
Difference analysis between scores and the clinicopathological characteristics
Difference analysis was conducted to learn the correlation b e t w e e n I m m u n e / S t r o m a l / E S T I M A T E S c o r e a n d clinicopathological characteristics, such as age, gender, pathology grade, stage and TNM classification. It was also carried out to find the association between the expression of CXCL12 and clinicopathological characteristics as well as the association between the expression of CXCL12 and TICs. Wilcoxon test was used to compare two groups, and Kruskal-Wallis followed by post-hoc Dunn test was used for multiple groups. P < 0.05 was considered to be statistically significant.
DEGs acquisition
414 BLCA samples were grouped into to two subgroups, including high Immune/Stromal Score group and low Immune/ Stromal Score group based on the comparison with the median. Package limma in R was applied for the analysis. A fold change (FC, log 2 (high score/low score) ) > 2 and false discovery rate (FDR) <0.05 were used to search the DEGs. Pheatmap package in R was used to plot the heatmaps of DEGs.
Intersection analysis
VennDiagram package in R was used to plot the venn diagram of DEGs.
Gene ontology and kyoto encyclopedia of genes and genomes enrichment analysis GO and KEGG enrichment analysis of the above DEGs were further carried out by using the ClusterProfiler, enrichplot, and ggplot2 packages in R. P< 0.05 was considered to be statistically significant.
Protein-protein interaction network and gene set enrichment analsis
The preliminary PPI network of 284 DEGs was acquired from the Search Tool for Retrieval of Interacting Genes/Proteins (STRING) database (version 11.0) and further reconstructed in Cytoscape (3.6.1). The confidence of interactive relationship of the nodes was >0.95. GSEA (4.1.0) based on the different gene sets, was applied to learn the specific functional profile of CXCL12. P< 0.05 was considered to be statistically significant.
COX regression analysis
Package survival in R was applied to conduct univariable COX regression.
Difference analysis and paired analysis of CXCL12
Package Beeswarm in R was used to assess the expression of CXCL12 in bladder tumor samples and normal samples. Packages Ggpubr and BiocManager were applied to learn the expression of CXCL12 in bladder tumor samples and the paired normal samples. Wilcoxon test was used for the comparison. P < 0.05 was considered to be statistically significant.
Survival analysis
Packages survival and survminer in R were used to carry out the survival analysis. Kaplan-Meier plot and log-rank tests were conducted to learn the associations between the expression of CXCL12 and the survival of BLCA patients. P < 0.05 was considered to be statistically significant.
TICs Profile
TICs profile in BLCA samples was evaluated by CIBERSORT algorithm.
Correlation analysis
Correlation analysis was carried out by using spearman's correlation analysis.
Tissue samples acquisition and Immunohistochemistry
BLCA and corresponding normal tissues were harvested from BLCA patients at Nanjing Jiangning Hospital. The study was permitted by the Ethics Committee of Nanjing Jiangning Hospital. Slides (4mm) of formalin-fixed paraffin-embedded tissue sections were incubated with CCL19 (1:200; Proteintech) antibody. The specific procedure was consistant with that described in our previous study (29).
Analysis process of the study
We downloaded the transcriptome profiling and clinical data of 414 BLCA samples and 19 normal samples from TCGA database and further analyzed the proportion of immune and stromal component in TME of each tumor sample through ESTIMATE algorithm. Difference analysis was conducted to find out the association between immune/stromal component and the clinicopathological characteristics of BLCA patients. A total of 284 DEGs were further acquired based on the immune and stromal component in TME of BLCA. GO and KEGG enrichment analysis of these 284 DEGs were performed to learn their biological functions. Finally, CXCL12 was found out through the intersection analysis of PPI network and univariable COX regression. Then, we focused on the expression of CXCL12 in BLCA samples and normal samples, the association between the expression of CXCL12 and survival and clinic-pathological characteristics of BLCA patients. GSEA of different gene collections was also carried out to learn the function of CXCL12. TICs profile in TIME of BLCA samples was calculated by using CIBERSORT algorithm. Difference analysis and correlation analysis were applied to find out the correlation between the expression of CXCL12 and TICs. The analysis process was shown in Figure 1.
Immune and stromal component in TME were associated with the clinicopathological characteristics of BLCA To learn the association between the proportion of immune/ stromal component in TME and the clinicopathological characteristics of BLCA patients, we set up the immune/stromal component evaluating system of TME in BLCA samples. Immune, Stromal and ESTIMATE Score represented the proportion of immune, stromal and the total component in TME of each tumor sample, respectively. The higher Score suggested the larger proportion of immune or stromal component in TME. The clinicopathological characteristics of BLCA patients, including age, gender, pathology grade, clinical stage and TNM classification, were concluded in Supplement Table 1. Difference analysis suggested that Stromal Score was associated with patients' age and gender, especially the pathology grade (Figures 2A-C, p=0.018; 0.032; <0.001). The higher Stromal Score predicted the higher pathology grade of BLCA ( Figure 2C, p<0.001). It was also found that Stromal Score was related with T, N classification and clinical stage of BLCA ( Figures 2D-G). Regarding Immune Score, it was significantly up-regulated in female patients ( Figure 2I, p=0.037) and the higher Immune Score suggested the higher pathology grade of BLCA ( Figures 2H-J). However, neither TNM classification nor clinical stage of BLCA was associated with Immune Score (Figures 2K-N, p>0.05). As for ESTIMATE Score, the results showed that it was up-regulated in female patients and predicted the higher pathology grade of BLCA ( Figures 2O-Q). In addition, ESTIMATE Score was connected with T classification and clinical stage of BLCA ( Figures 2R-U). From above, we could conclude that the stromal and immune component in TME was significantly associated with the pathology grade of BLCA, and partly related with TNM classification and clinical stage. This provided a new sight for us to explore the underlying mechanisms of the development and progression of BLCA.
DEGs were primarily enriched in immune-associated activities
In order to understand the underlying mechanisms of BLCA TME, we conducted difference analysis to acquire the DEGs profile in TME. BLCA samples were divided into two groups, including high Stromal/Immune Score group and low Stromal/ Immune Score group. Heat map shows the DEGs profile in TME of BLCA ( Figures 3A, B). To be specific, there were 537 DEGs obtained from the comparison between high Stromal Score and low Stromal Score, among which 496 DEGs were up-regulated and 41 DEGs were down-regulated (Supplement Table 2). Besides, there were 531 DEGs obtained from the comparison between high Immune Score and low Immune Score, among which 466 DEGs were up-regulated and 65 DEGs were downregulated (Supplement Table 3). In conclusion, a total of 284 DEGs were synchronously up-regulated or down-regulated in Stromal or Immune group ( Figure 3C; Supplement Table 4). Next, we carried out GO and KEGG enrichment analysis to assess the biological functions of these 284 DEGs. Go enrichment analysis revealed that 284 DEGs were primarily enriched in immune-associated activities, such as leukocyte proliferation, migration and chemotaxis, lymphocyte and mononuclear cell proliferation and so on (Figures 3D-F). Similarly, KEGG enrichment analysis suggested that these DEGs were mainly enriched in immune-associated activities, including complement and coagulation cascades, cytokinecytokine receptor interaction, B cell receptor signaling pathway, chemokine signaling pathway and so on ( Figures 3G-I). Therefore, we considered that DEGs acquired from Stromal and Immune group of BLCA were significantly associated with immune-associated activities in TME.
PPI network and univariable COX regression analysis of 284 DEGs
To learn the detailed reciprocity among these 284 DEGs, we plotted PPI network through STRING database and cytoscape software. The potential interactions among DEGs were shown in Figure 4A. DEGs, which shared more than seven nodes, were ranked in Figure 4B (Supplement Table 5). Univariable COX regression of 284 DEGs was conducted at the same time to find out the specific DEGs, which were significantly associated with BLCA patients' survival ( Figure 4C). MMP9, COMP, F13A1 and CXCL12 drew our attention ( Figure 4C, p<0.05; Supplement Table 6). Intersection analysis was finally carried out to search the DEGs who shared the leading nodes in PPI network and were significantly related with BLCA patients' survival ( Figure 4D). Fortunately, CXCL12 emerged.
CXCL12 was down-regulated in BLCA tissues and associated with the clinicopathological characteristics
The CXC chemokine CXCL12 participated greatly in multiple physiological and pathological processes through interacting with its receptors CXC chemokine receptor 4 (CXCR4) and atypical chemokine receptor 3 (ACKR3) (30). In our study, CXCL12 was significantly down-regulated in BLCA tissues compared with normal tissues ( Figure 5A, p<0.001). Besides, the paired analysis found that CXCL12 was obviously down-regulated in BLCA tissues compared with the paired normal tissues ( Figure 5B, p<0.001). In order to further validate the expression of CXCL12 in BLCA, we carried out IHC and the result was consistent with the above analysis ( Figure 5C). However, survival analysis suggested that the expression of CXCL12 was not significantly associated with the survival of BLCA patients ( Figure 5D, p=0.050). As for clinicopathological characteristics of BLCA patients, the results showed that the expression of CXCL12 varied with ages and higher expression of CXCL12 predicted the higher pathology grade of BLCA (Figures 5E-G; p<0.001, p<0.001). In addition, the expression of CXCL12 was partly related with T, N classification and clinical stages of BLCA ( Figures 5H-K). Hence, we concluded that CXCL12 had the potential to participate in the progression of BLCA. The analysis process of the study.
CXCL12 participated greatly in immuneassociated activities
On account of the potential role of CXCL12 in the progression of BLCA, we further explored the underlying mechanisms of CXCL12. BLCA samples were first divided into two groups, including high CXCL12 expression group and low CXCL12 expression group. GSEA suggested that for C5 collection, the gene ontology sets, genes in high CXCL12 expression group were enriched in cytokine binding, leukocyte signaling and so on ( Figure 6C, p<0.05). For C7 collection defined by MSigDB, the immunologic gene sets, genes were found to be enriched in multiple immune-associated activities, which were related with CD4 T cell, CD8 T cell, B cell and so on ( Figure 6D, p<0.05). As a result, according to GSEA, CXCL12 was significantly related with immune-associated activities and had the potential to regulate TIME of BLCA.
CXCL12 could communicate with TICs in TIME
Since GSEA suggested that CXCL12 was significantly correlated with immune-associated activities, we speculated that there could be underlying connections between CXCL12 and TIME of BLCA. In consequence, we calculated the proportion of TICs in BLCA samples through CIBERSORT algorithm. The specific ratio of 22 TICs in BLCA sample was shown in Figure 7A. And the association among these TICs was exhibited in Figure 7B. Difference analysis was carried out to learn the association between the expression of CXCL12 and specific TICs. The results showed that 10 kinds of TICs, such as Macrophages M2, naïve B cell, gamma delta T cells, CD4 naïve T cells, follicular helper T cells and so on, were significantly associated with the expression of CXCL12 ( Figure 8A; Supplement Table 7). The correlation analysis suggested that the expression of CXCL12 were related with 8 kinds of TICs, among which naïve B cells, macrophages M2 and resting mast cells were positively associated, and activated dendritic cells, resting dendritic cells, resting NK cells, CD4 naïve T cells and follicular helper T cells were negatively associated (Figures 8C-J, p<0.05; Supplement Table 7). Finally, intersection analysis between difference analysis and correlation analysis suggested that 6 kinds of TICs were significantly related with the expression of CXCL12 ( Figure 8B). In conclusion, CXCL12 was obviously related with various TICs in TIME of BLCA and had the potential to regulate the TIME of BLCA probably via communicating with multiple TICs.
Discussion
In this study, we focused on the associations between IRGs and TICs in the TME of BLCA. Firstly, we downloaded the transcriptome profiling and clinical data of 414 BLCA samples and 19 normal samples from TCGA database, and calculated the A B D C FIGURE 4 PPI network and univariable COX regression analysis of 284 DEGs. (A) PPI network was constructed by STRING database and Cytoscape software; interaction confidence value >0.95; (B) The top 10 genes which shared the leading nodes in PPI network; (C) Univariable COX regression analysis of 284 DEGs; p<0.05 was considered to be statistically significant; (D) Intersection analysis of DEGs which shared the leading nodes in PPI network and were closely related with BLCA patients' survival. p<0.05 was considered to be statistically significant.
proportion of immune and stromal component in TME of each BLCA samples with the help of ESTIMATE algorithm. Secondly, difference analysis found out that Immune/Stromal Score was associated with the clinicopathological characteristics of BLCA. Thirdly, DEGs were obtained through the comparison between high Immune Score and low Immune Score (or high Stromal Score and low Stromal Score), and GO and KEGG enrichment analysis suggested that these DEGs were mainly enriched in immune-related activities. Fourthly, PPI network and univariable COX regression analysis found that CXCL12 shared the leading nodes in PPI network and was significantly related with BLCA patients' survival. Fifthly, signal gene analysis was conducted and found out that CXCL12 was down-regulated in BLCA samples compared with normal sample and significantly related with the clinicopathological characteristics of BLCA. GSEA revealed that CXCL12 was significantly associated with immune-associated activities and could play an important role in regulating TIME of BLCA. Finally, CIBERSORT algorithm was applied to calculate the proportion of TICs in BLCA samples, and difference analysis as well as correlation analysis suggested that CXCL12 was obviously connected with multiple TICs in TME of BLCA and could modulate the TIME of BLCA via closely communicating with TICs.
CXCL12 was traditionally classified as a homeostatic CXC chemokine and took a great part in modulating kinds of physiological and pathological processes via binding to its receptors CXCR4 and ACKR3 (30,31). CXCL12/CXCR4 axis had been proved to be associated with the progression and therapy of cancers. For example, CXCL12/CXCR4 axis advanced the invasion and metastasis of pancreatic cancer through complex crosstalk with other pathways, and was correlated with the poor prognosis of patients (32). Treatment targeting the CXCL12/CXCR4 pathway increased the efficacy of radiotherapy of locally advanced cervical cancer (33). The role of CXCL12 in tumor development mainly depended on the specific microenvironment of tumors (34). In addition, CXCL12-CXCR4/CXCR7 axis had a great influence in gastrointestinal malignancies through immune resistance (35). Further researches suggested that the mechanisms of immunotherapy resistance could be associated with the CXCL12/CXCR4 axis (36). However, only a few studies explored the association between CXCL12 and tumorigenesis and progression of BLCA (37). Researches found that CXCL12 was down-regulated in BLCA tissues compared with normal bladder mucosal tissues, and positively associated with the differentiation degree and invasive depth of BLCA tissues (38). Additionally, CXCL12/CXCR4 promoted the invasion of BLCA cells through activating Stat3 (39). Furthermore, increasing studies have revealed that TICs played an important role in the progression and treatment of BLCA. For example, intratumoral TIGIT+ CD8+ T-cell abundance functioned A B D C FIGURE 6 GSEA for genes in high CXCL12 expression group. (A) GSEA for genes in high CXCL12 expression group in C5 collection, the gene ontology sets; (B) GSEA for genes in high CXCL12 expression group in C2 collection, the KEGG gene sets database; (C) GSEA for genes in high CXCL12 expression group in hallmark gene sets; (D) GSEA for genes in high CXCL12 expression group in C7 collection defined by MSigDB, the immunologic gene sets. FDR<0.05 and p<0.05 were considered to be statistically significant.
as a potential prognostic factor for patients' survival and a predictive biomarker for adjuvant chemotherapeutic effect (40). CD19+ tumor-infiltrating B-cells activated CD4+ tumor-infiltrating Tcells in the TMB of BLCA and acted as an independent prognostic factor for post-surgery survival and adjuvant chemotherapy benefits of BLCA patients (41). Besides, DC-SIGN+ TAM infiltration was significantly associated with a tumor-promoting TIME and functioned as a prognostic indicator and therapeutic target in the immunotherapy of BLCA (42).
However, there are few researches focusing on the associations between CXCL12 and TICs in TIME of BLCA.
In our study, we confirmed that CXCL12 was significantly down-regulated in BLCA tissues and associated with patients' clinicopathological characteristics. In addition, functional analysis revealed that CXCL12 participated in immuneassociated activities and could regulate TIME of BLCA probably through communicating with multiple TICs, such as macrophages M2, B cells naïve, T cells follicular helper, mast A B FIGURE 7 TICs Profile, and the difference analysis as well as correlation analysis between CXCL12 and TICs in TIME of BLCA. (A) The ratio of TICs in BLCA samples; (B) The correlation analysis among TICs; each spot indicated p value. cells resting, dendritic cells resting and T cells CD4 naive. Further researches should be carried out to clarify the accuracy of the above combined analyses, and focused on the underlying mechanisms of the communication between CXCL12 and TICs in TIME of BLCA.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Materials. Further inquiries can be directed to the corresponding author. (A) Violin plot displayed the association between the expression of CXCL12 and TICs; p<0.05 was considered to be statistically significant; (B) Venn plot showed 6 TICs shared by the difference test (A) and correlation test (C-J); (C-J) Scatter plots showed the association between the expression of CXCL12 and 8 TICs; p<0.05 was considered to be statistically significant.
Ethics statement
The studies involving human participants were reviewed and approved by The Ethics Committee of Nanjing Jiangning Hospital. The patients/participants provided their written informed consent to participate in this study.
|
v2
|
2022-11-07T14:33:05.429Z
|
2022-11-07T00:00:00.000Z
|
253372440
|
s2orc/train
|
Impact of intestinal dysbiosis on breast cancer metastasis and progression
Breast cancer has a high mortality rate among malignant tumors, with metastases identified as the main cause of the high mortality. Dysbiosis of the gut microbiota has become a key factor in the development, treatment, and prognosis of breast cancer. The many microorganisms that make up the gut flora have a symbiotic relationship with their host and, through the regulation of host immune responses and metabolic pathways, are involved in important physiologic activities in the human body, posing a significant risk to health. In this review, we build on the interactions between breast tissue (including tumor tissue, tissue adjacent to the tumor, and samples from healthy women) and the microbiota, then explore factors associated with metastatic breast cancer and dysbiosis of the gut flora from multiple perspectives, including enterotoxigenic Bacteroides fragilis, antibiotic use, changes in gut microbial metabolites, changes in the balance of the probiotic environment and diet. These factors highlight the existence of a complex relationship between host-breast cancer progression-gut flora. Suggesting that gut flora dysbiosis may be a host-intrinsic factor affecting breast cancer metastasis and progression not only informs our understanding of the role of microbiota dysbiosis in breast cancer development and metastasis, but also the importance of balancing gut flora dysbiosis and clinical practice.
Epidemiology and staging of breast cancer
The 2020 Global Cancer Statistics report shows that female breast cancer is the most common cancer worldwide, with the highest number of new cases annually (approximately 11.7% of all new cases in both men and women), having overtaken lung cancer (11.4%) (1). There are four main subtypes of breast cancer, approximately 75% of them are positive for ER and/or PR (2). The luminal A (ER and PR positive, HER2 negative, low Ki67) subtype accounts for approximately 40% of all cases; it is characterized by low invasiveness, a low recurrence rate, a high survival rate, and the best response to hormonal therapy (3). In turn, the luminal B (ER and PR positive, HER2 positive or HER2 negative, with high Ki67) subtype is responsible for 10-20% of all cancer cases, has a higher relapse rate, proliferative index, and lower recurrence survival (4)(5)(6). HER2 positive (non-luminal) were defined as HER2 overexpression or amplification, ER and PR absence, and survival rate significantly improvement with targeted therapy (7). Triple-negative breast tumors (TNBC) are defined as ER, PR, and HER2 negative. TNBC which makes up approximately 15% of all breast tumors and have a high risk of distant relapse in the first 3 to 5 years following diagnosis (8, 9).
With advances in early diagnosis and comprehensive treatment, the prognosis for patients with breast cancer has improved; however, the incidence of metastasis is also increasing (10). It has been reported that 20%-30% of patients with breast cancer can develop metastases after diagnosis and treatment of the primary tumor, with metastases being the cause of approximately 90% of deaths (11). Breast cancer shows a tendency to metastasize to a variety of organs, including bone, lung, liver, and brain, which is termed metastatic heterogeneity. Bone metastases account for approximately 75% of metastases (12), with an overall 5-year survival rate of 22.8% (13). Lung is the second most common site of breast cancer metastasis (14), with an overall 5-year survival rate of 16.8%. The liver is second only to lung as a metastasis site, but survival is poor relative to local, bone, and lung recurrences, with an expected 5-year overall survival rate of 8.5% (15). Brain accounts for approximately 15%-30% of metastatic sites in patients with metastatic breast cancer, limiting quality of life and a very short life expectancy (16)(17)(18).
The priority of metastasis varies from organ to organ, resulting in differences in prognosis and treatment response. A widely accepted model of metastasis is the "seed and soil" hypothesis proposed by Paget (19), which initially revealed that successful colonization of second organs depends on the intrinsic properties of the tumor cells and the compatibility and support of the microenvironment.
Intestinal flora dysbiosis 1.2.1 Gut microbiota composition in human health
A dynamic balance is maintained between the microbiota and the host, and this balance plays an important role in human health by influencing the physiological functions of the organism. A healthy intestinal microbiota is composed mainly of the phyla Firmicutes and Bacteroidetes, followed by the phyla Actinobacteria and Verrucomicrobia (20). The distribution of microorganisms in the gastrointestinal tract varies longitudinally from the esophagus to the rectum, Helicobacter is the dominant species in the stomach and determines the microbial status of the entire gastric flora. While H. pylori inhabits the stomach as a commensal, other genera constitute the rich diversity of the gastric flora (21, 22). Conversely, this diversity is reduced when H. pylori cause disease. Firmicutes and Actinobacteria are the most dominant phylum in the duodenum (23). The jejunum is dominated by the growth of Gram-positive aerobic and facultative anaerobes, including Lactobacilli, Enterococci and Streptococci. In the ileum, with predominance of aerobic species, while the distal ileum has a similar bacterial body to the colon, with anaerobes and Gram-negative organisms (23). The bacteriophage in the large intestine is dominated by the phyla Firmicutes and Bacteroidetes. Furthermore, there are other important pathogens in the human colon, such as Campylobacter jejuni, Salmonella enterica, Vibrio cholera and Escherichia coli (E. coli), and Bacteroides fragilis (24, 25). The abundance of the Proteobacteria phylum is significantly lower in normal humans, and its absence along with the high abundance of genera such as Bacteroides, Prevotella and Ruminococcus indicates a healthy gut microbiota (26).
Gut microbiota function
Intestinal flora homeostasis has an important role in maintaining normal body function, The gut microflora creates a stable mucosal barrier for the intestine to prevent the invasion of pathogenic microorganisms (27). Gut microbes break down non-digestible compounds through anaerobic fermentation to produce compounds of short-chain fatty acids (SCFAs), which have good anti-inflammatory and chemopreventive properties and act as barrier protectors (28,29) and are considered as tumor suppressors (30). Microorganisms containing Lipopolysaccharide (LPS), such as Salmonella and Escherichia coli, activate antigen presenting cells through pattern recognition receptors to produce cytokines, which together with endogenous glycolipid antigens and the major histocompatibility complex (MHC) class I-related glycoprotein CD1d activate Invariant natural killer T (iNKT) cells and participate in various immunomodulatory responses (31,32). In addition, many intestinal microbiota are involved in bone remodeling processes as immunomodulators, such as Lactobacillus acidophilus, Lactobacillus plantarum, Lactobacillus rhamnosus GG, Lactobacillus reuteri, Lactobacillus paracasei and Bacillus clausii (33). The gut microbiota regulates nutrient metabolism by regulating lipid metabolism, propionic acid in short-chain fatty acids reduces fatty acid levels in liver and plasma and reduces food intake (34), and the gut microbiota regulates intestinal and plasma Lipopolysaccharide (LPS) levels by modulating the intestinal endocannabinoid (eCB) system (35), which affects adipose tissue metabolism. Intestinal flora are involved in the production of gastrin, insulin (36) and glucagon-like peptide-1 (GLP-1) (37, 38) through a paracrine pathway produced by enterocytes, and it is also involved in the synthesis of bile acids, cholesterol, bound fatty acids (39) and vitamin (40), thereby regulation of endocrine levels and metabolic changes in the host. The gut microbiota synthesizes a number of neurochemicals, (e.g., gamma amino butyric acid (GABA): an inhibitory neurotransmitter), which influence central nervous and gut function (41). A gut-brain microbial axis exists between gut microbes, the gastrointestinal tract and the central nervous system (42), which links brain emotional centers to mechanisms such as gut function, gut neural reflexes and gut endocrine signaling to jointly coordinate organismal changes (43,44). Circulating SCFAs produced by gut microbiota metabolism affect the integrity of the blood-brain barrier (BBB) by increasing the production of tight junction proteins, and increased BBB integrity reduces the entry of undesirable metabolites into brain tissue and strengthens the defense mechanisms of the blood-brain barrier (45). Compounds produced by the metabolism of the gut microbiota, such as lipoproteins and lipopolysaccharides, affect autoimmune function by stimulating the release of cytokines from immune cells. These cytokines can cross the BBB and activate neurons, altering neurological function and leading to changes in mood and behavior (46), providing new ideas for the treatment of brain dysfunction.
Dysbiosis
Dysbiosis refers to a state in which the intestinal flora loses its normal "beneficial" function and is continuously disturbed, causing disease. It is associated with a large proportional change in the composition of the microbiota beyond the normal range caused by host-related and environmental factors (47). Dysbiosis is usually characterized by the following feature: Bloom of pathobionts (48), Loss of commensals (49) and Loss of alpha diversity (50-52), which can be present individually or simultaneously and mutually exclusive. Currently, dysbiosis has a causal relationship with the manifestation, diagnosis or treatment of specific diseases, from the perspective of the composition of the intestinal microflora, mainly originating from Infection and inflammation (53,54); Diet and xenobiotics (55, 56); Genetics (57) and Familial transmission (58-60) etc.
Link between dysbiosis and cancer
Dysbiosis states may negatively affect the organism leading to various disease states. The microbiota may have some tumor suppressive effects on the host, and deviations in flora balance may be associated with cancer development (61). Studies (62)(63)(64)(65)(66)(67) have identified direct and indirect roles of the gut microbiota in carcinogenesis, cancer treatment and prevention. including colon (66, 68, 69), gastric (70-73), lung (74,75), prostate (76-78) and breast cancers (79) (Tables 1, 2), and suggest that the gut microbiota and these cancers are interlinked through tumor suppression and tumor initiation factors. Modification of the composition and activity of the intestinal flora through the administration of prebiotics, probiotics and synthetics, providing benefits to patients with colorectal cancer, such as: modulation of immunity, improvement of bile acid metabolism and restoration of intestinal microbial diversity (68). H. pylori is one of the major causative factors of gastric cancer. Probiotics against H. pylori through various mechanisms, including: secretion of antibacterial compounds; inhibition of H. pylori colonization; action through stimulation of mucin synthesis; and modulation of host immune response, which provides new perspectives on gastric cancer prevention and treatment (100). It was found that memory T and NK cell profiles were increased in peripheral blood samples from patients with beneficial and diversity-rich gut microbes. This has important implications for predicting the response to anti-PD-1 immunotherapy in Chinese non-small-cell lung cancer patients (101). Gram-positive bacteria stimulate the production of specific subpopulations of "pathogenic" T helper 17 (pTh17) cells and memory Th1 immune responses, and the absence of these bacteria leads to reduced pTh17 responses and cyclophosphamide tumor resistance, demonstrating that the gut microbiota contributes to the formation of anti-cancer immune responses in lung cancer patients (102). However, the symbiotic gut microbiota promotes endocrine resistance in castration-resistant prostate cancer by providing an alternative source of androgens, implying that the gut flora may play a negative role in this process (103). Gut bacteria can regulate insulin-like growth factor-1 (IGF1) levels in the host via short-chain fatty acids, thereby promoting the proliferation of prostate cancer cells, then modulating the gut microbiota to influence the gut microbiota-IGF1-prostate axis may be beneficial in the prevention and treatment of prostate cancer (104). In addition, the use of gut microbiota analysis to predict patient response to immune check inhibition sites has emerged in cancer treatment, e.g., breast cancer (105). Currently, the role of gut microbes in the development of various cancers varies, and their variation may have implications for achieving more personalized precision medicine in oncology. 1.3 The role of microbiota in breast tumourigenesis
Estrogen and metabolism
The gastrointestinal microbiome regulates systemic estrogen, and the development of postmenopausal breast cancer is associated with disordered (high) levels of estrogen in the body (106). The metabolism of estrogen occurs in the liver, where the metabolites are conjugated and excreted into the gastrointestinal lumen within the bile. They are de-conjugated by b-glucuronidase-producing bacteria in gastrointestinal lumen, and then they are reabsorbed as free estrogens through the enterohepatic circulation to reach breast (107). All the genes in the gut flora that metabolize estrogen are collectively known as the estrobolome (108). A study found that difference s in urinary estrogen levels were associated with beta-glucuronidase activity in pre-and post-menopausal women, and that gastrointestinal flora could influence non-ovarian estrogen levels via the enterohepatic circulation (109). In addition, urinary estrogen levels in men and postmenopausal women were strongly correlated with all indicators of microbiota richness and diversity in faeces, with non-ovarian-acting systemic estrogens significantly associated with fecal Clostridium perfringens (including non-clostridial and three genera of the family Rhizobiaceae), and Gut microbiota may influence estrogen-related diseases in the elderly (109), such as Postmenopausal Breast Cancer. Many of the microbes associated with breast cancer have the b-glucuronidase enzymatic activity mentioned above, which prevents the binding of estrogen and other compounds and makes them biologically active, thus affecting local and systemic levels of estrogen and its metabolites (79, 110). During estrogen metabolism, the gut acts as an important site for estrogen reactivation and microorganisms act locally or distally to regulate disease development and homeostasis (111). When the balance of the intestinal environment is disrupted and the structure and ratio of the flora are imbalanced, excess intestinal bacteria, Lipopolysaccharides and pro-inflammatory cytokines are produced, and this change disrupts the integrity of the intestinal mucosa, which in turn triggers inflammation after bacterial translocation (112). In addition to those involving hormone metabolism (estrogen and progesterone),Studies in growing numbers are exploring the relationship between the gut microbiome and breast cancer risk via a non-estrogendependent pathway. Obesity, insulin resistance, dyslipidemia, leukocytosis, and elevated C-reactive protein (113) are associated with reduced gut microbial diversity, some of which are associated with breast cancer. Studies have demonstrated that metabolic health status (as defined by the homeostasis model assessment of insulin resistance [HOMA-IR] index, or fasting insulin level), but not obesity per se, may be an associated factor in the risk of postmenopausal breast cancer development, suggesting that hyper insulinemia is an important risk factor for breast cancer (114). Karen L Margolis et al. demonstrated an increased risk of invasive breast cancer in postmenopausal women with higher white blood cell counts (115), Nicholas J Ollberding et al. concluded that circulating C-reactive protein levels44 reflecting adipokines and systemic inflammation were associated with the risk of postmenopausal breast cancer, independent of Body fat rate (116), These further support the possibility that inflammation may be associated with the initiation, promotion and progression of breast cancer. In addition, breast cancer in postmenopausal women is significantly associated with the immune-recognised (IgApositive) and -unrecognised (IgA-negative) gut microbiota, the former possibly through immune-mediated pathways and the latter possibly through the enterohepatic circulation effects of estrogen (117). It was shown that the microbiota of breast tissue is different from that of mammary skin tissue, where bacterial species are more abundant than in skin tissue, and more operative taxonomic units (mostly low abundance) were observed in the breast tissue microbiota. These taxa with different abundance were from the phyla Firmicutes, Actinobacteria, Bacteroidetes, and Proteobacteria (88). A comparison of breast tissue from breast cancer patients and normal women revealed higher levels of Enterobacteriaceae and Staphylococcus and increased numbers of Bacillus in breast cancer patients (87). In contrast, Lactobacillus and Streptococcus were higher in healthy women and have anticancer properties that may play a role in the prevention of breast cancer (118). Prevotella, which produces SCFAs propionic acid and exerts benefits in the intestine, was higher in healthy women compared with breast cancer patients (119). Further study of bacterial metabolites and bacterially induced host metabolites would provide insight into the role of bacteria in the role of breast disease will provide important information.
The role of antibiotics
Indirect evidence suggests that the development of breast cancer is strongly associated with alterations in specific microbiota when taking antibiotics or probiotics. Through a large-scale analysis of nearly 4 million women, Simin et al. (120) showed a specific dose-dependent relationship between antibiotic use and breast cancer, with a different correlation between the type of antibiotic and breast cancer risk, such as blactams, macrolide (121). Irregular use or overuse of antibiotics may increase the risk of gut dysbiosis and decrease microbial diversity, and this effect may be long-lasting (122,123), For example: co-amoxiclav and clarithromycin, Cefprozil, Amoxicillin, etc. Also, overuse of antibiotics (penicillins, streptomycin, chloramphenicol, tetracyclines, erythromycin, cephalosporins and their analogues) decreases plasma levels of lignans-enterolactone, which can increase the risk of breast cancer by affecting the microbiota (124). A study has shown that the increased excretion of bound estrogens in the feces of patients treated with ampicillin suggests that the gut microbiota are actively involved in estrogen metabolism and can have some effect on the pathogenesis of breast cancer by altering the individual's microbial status (106). Antibiotics have been shown to disrupt the microbiota, leading to a reduced response by tumor cells to platinum-based chemotherapy and immunotherapy (106, 125, 126), suggesting that a stable microbiota is necessary for an optimal response to antitumor therapy.
Regulation of chronic inflammation and immunity
Microbiota may promote the risk of malignancy by inducing the persistence of chronic inflammation, disrupting the balance between cell proliferation and death in the body, and triggering uncontrolled innate and adaptive immune responses (127, 128). A putative inflammatory mechanism associated with breast carcinogenesis has been demonstrated to be the upregulation of cyclooxygenase 2 (COX2) and its product prostaglandin E2 (PGE), which would increase the expression of aromatase in adipose tissue, thereby promoting the conversion of androgen precursors to estrogens (129,130) and increasing the risk of breast carcinogenesis. Studies have demonstrated that a potential inflammatory biomarker, mucosal secretory immunoglobulin A (IgA) (131), can maintain the integrity of the mucosal barrier by regulating the composition of the intestinal microbial community, thereby attenuating the host's innate immune response. The link between breast cancer and the mucosal secretory IgA has been established (117). This mechanism places some limits on the participation of intestinal microbial antigens in the circulation of the body, and some limits on the invasiveness of potentially dangerous microorganisms (132). Certain specific microbiota may also maintain breast health by stimulating the host inflammatory response. For example, specific bacteria S. yanoikuyae are present in the breast tissue of healthy women and their abundance is significantly reduced in the corresponding tumor tissue. An increase in its abundance may lead to a decrease in bacterial-dependent immune cell stimulation in the body, resulting in a reduced environmental risk level for the development of breast tumors (80). Studies have also confirmed the role of microorganisms in regulating specific immune processes in the development of cancer (133), For example, Lactococcus spp. can activate important cells associated with tumor growth (murine splenic NK cells), maintain their cytotoxicity, and enhance cellular immunity (134). In another case-control study, Goedert and colleagues (117) investigated the role of immunity and inflammation in breast cancer risk and whether the gut microbiota differed in the composition of the immune recognition microbiota and found significant differences in the composition, abundance and alpha diversity of the microbiota between the IgA+ and control IgAgroups in cancer cases and correlated with changes in high and low estrogen levels. This suggests a significant association with IgA+ and IgA-gut microbiota in postmenopausal women with breast cancer, suggesting that the gut microbiota may influence breast cancer risk through altered metabolism, estrogen cycling and immune pathways.
Genomic stability and DNA damage
DNA damage may not be sufficient to promote cancer development, but microbes can trigger transformation by destabilizing genes, cell proliferation and death, and it has been demonstrated that microbes cause cancer development by damaging host DNA in order to survive (135). Urbaniak et al. (87) found that Escherichia coli (a member of the Enterobacteriaceae family) isolates and a Staphylococcus epidermidis isolate from normal adjacent tissues of breast cancer patients had the ability to induce DNA double-strand breaks, thus causing genomic instability (136). In addition, some bacterial species may eventually lead to genotoxicity by increasing the production of reactive oxygen species (137).
2 Current status on breast cancer progression, metastasis and microbiology
Enterotoxigenic bacteroides fragilis
Bacteroides fragilis is a common colonic colonizing enterobacterium (138) whose virulence is attributed to a 20 kDa zinc metalloprotease toxin known as B. fragilis toxin (BFT) (139). With reference to the effect of enterotoxigenic B. fragilis (ETBF) intestinal or ductal colonization on breast cancer progression in the mammary intraductal model, Parida S et al. (140) colonized BALB/c mice via the teats with ETBF or a nontoxic mutant B. fragilis (086Mut) that does not secrete BFT. The presence of BFT was found to be detected in the mammary glands of ETBF-carrying mice compared to controls, with a 3.9fold higher tumor volume than 086Mut controls, enhanced lung and liver metastases, and more proliferative tumors forming in the ETBF group, exhibiting a mesenchymal phenotype. Moreover, trichrome staining showing significantly higher stromal infiltration, demonstrating that ETBF intestinal or ductal colonization was associated with breast cancer progression and distant metastasis. Furthermore, significant differences in breast tissue structure were found in the ETBF group compared with the 086Mut control group (140), including extensive local inflammation and tissue fibrosis, Ki-67 and proliferating cell nuclear antigen staining showed increased epithelial cell proliferation, CD3 staining showed increased Tcell infiltration, and significantly altered expression of pankeratin, all indicating that BFT was associated with a significant increase in oncogenic cell activity and growth rate. The study also found that RNA-seq analysis of secondary tumors arising from breast cancer cells treated with BFT showed enrichment of the b-catenin pathway. The expression of several Notch-responsive genes was enriched in breast cancer cells suggesting that BFT also triggered activation of the Notch1 pathway. The results advance our understanding of the molecular mechanisms associated with ETBF/BFT and breast cancer progression (140), and point to a hypothesis that dysbiosis or disruption of the gut flora might be associated with breast cancer metastasis and progression, and that inhibition of manipulable key molecules or pathways could potentially reduce the impact of ETBF infection on breast cancer.
In looking at whether BFT affects the tumorigenicity of breast cancer cells, the team found that, compared with cells from the control group, BFT-pretreated MCF-7 and MCF-10A cell groups showed greater invasion and migration, with local tumor expansion and the formation of multifocal tumors resembling local metastases (140).However, it was not clear whether ETBF spread from within the gut to the breast or whether gut-infected mice acquired the mammary gland infection through environmental factors. Data from RNA-seq analysis of secondary tumors with limited in vivo formation showed higher expression of genes associated with migration, homing, and metastasis in the BFT pre-treatment group, suggesting that BFT production by ETBF intestinal colonization might be associated with the initiation of breast cancer metastasis; and breast cells exposed to BFT showed dramatic changes supporting cell motility, embryonic pluripotency pathways, expression of metastatic genes, and molecular mechanisms. However, it cannot be demonstrated that ETBF can be the sole driver directly triggering the transformation of human breast cells into tumor cells or interacting with other microbiota to show oncogenic activity.
Antibiotic-induced intestinal flora dysbiosis and the progression of breast cancer metastasis
To assess the effect of pre-established dysbiosis on the metastasis of hormone receptor-positive (HR+) breast cancer in a more aggressive and metastatic tumor model, Parida S et al.
(141) evaluated tumor spread to the lung and axillary lymph nodes in a highly metastatic MMTV-PyMT mouse model with reference to the poorly metastatic HR+ mouse breast cancer cell line BRPKp110. The results were similar to those observed in the BRPKp110 cell line: where the spread of tumor cells to the lung was significantly increased after commensal dysregulated of the intestinal flora due to antibiotic treatment, independent of tumor volume. Moreover, the tumors progressed with the same kinetics regardless of the symbiotic dysregulation status in the experimental mice, suggesting that symbiotic dysregulation has a significant and sustained effect on HR+ breast cancer dissemination and that the enhanced ability of cancer cells to spread in symbiotically dysregulated mice is independent of tumor growth kinetics. To confirm the impact of the flora-dysregulation-driven host-intrinsic differences in inducing propagation in a mammary tumor model, they tested the s ymbio ti c d ysregu lation using t he L -Stop -L-KRasG12Dp53flx/flxL-Stop-L-Myristoylated p110a-GFP+ induced mouse model of breast cancer (141), and found that consistent with that observed in the homozygous model, the lungs of mice with dysbiosis of the intestinal flora showed a higher frequency of disseminated tumor cells. No significant increase in GFP+ tumor cells was observed in the distal lymph nodes. Those results confirmed that dysbiosis is independent of primary tumor growth and is associated with enhanced tumor cell dissemination; they also suggest that the tumor dissemination enhancement is the result of host dysbiosis rather than of intrinsic differences in tumor aggressiveness. Macrophages in the mammary gland may promote the metastasis of mammary tumors in experimental animals (142). Parida S et al. (141) found that commensal dysbiosis influenced the frequencies and numbers of macrophages during early or advanced stages of mammary tumor progression. Macrophages are one of the most abundant cell types within the breast tumormicroenvironment (143) and are a significant prognostic indicator of reduced survival for patients diagnosed with HR+ breast cancer (144). They observed that the majority of myeloid infiltrates within the mammary tumor microenvironment were M2-like macrophages during at early and advanced stages of tumor progression based upon CD206 expression. Importantly, the number of infiltrating tumor-promoting M2-like macrophages was significantly increased in advanced tumors of mice in the dysbiotic mice compared to non-dysbiotic controls with equal tumor burden. These data suggest that systemic expression of inflammatory mediators is increased in mice with dysbiosis tumors and that commensal dysbiosis acts synergistically with developing tumors to enhance myeloid recruitment into mammary tumors. Enhanced interstitial density or dense breast tissue is a recognized risk factor for the development of breast cancer metastases (145) and intramammary pro-tumor inflammation (146). They found that pre-established dysbiosis was associated with significantly enhanced collagen deposition in normal adjacent mammary glands and in tumors, and that collagen accumulation was slightly increased in the lungs of advanced tumor-bearing mice with dysbiosis, suggests that enhanced local and distal fibrosis is a long-term consequence of dysbiosis during breast cancer. Parida S et al. to determine whether gastrointestinal dysbiosis is sufficient to enhance mammary tumor cell dissemination (141), and a fecal microbiota transplantation (FMT) method was used, Both the experimental and control group and control groups were BRPKp110 breast tumor cells, Mice receiving floradysregulated cecal contents by FMT also showed enhanced infiltration of inflammatory myeloid cells into the mammary tissue and increased accumulation of myeloid cells into tumor tissue. Similar effects were observed in the mammary gland and tumor tissue during the advanced stages of tumor progressionthat is, mammary gland tissue and tumors showed enhanced tissue fibrosis. Importantly, the spread of tumor cells to peripheral blood, lung, and distal axillary lymph nodes was also significantly increased in mice receiving dysbiosis flora (rather than "normal" FMT) by FMT, considering that a dynamically imbalanced microbiome is sufficient to enhance the metastatic spread of breast cancer. Moreover, it may be an independent correlate of the distant spread of tumor cells. Further supporting the idea that dysbiosis contributes to the evolution of breast tissue and/or tumors toward more aggressive and high-grade disease. regardless of the metastatic potential of the HR+ breast tumor model used in the study, dysbiosis of the gut flora was associated with enhanced dissemination and metastasis of breast tumor cells.
Changes in the gut microbiota also to effects in metabolites, and inflammatory signaling pathways can be amplified or inhibited. Using an in-situ mouse model of breast cancer, Kirkup et al. (147, 148) found significant differences in metabolic regulatory pathways across the tumor transcriptome in animals treated with broad-spectrum antibiotics, and singlecell transcriptomics revealed that the stromal cell population was altered in breast tumors from antibiotic-treated mice. The main form of the alteration was an increased number of mast cells, which accelerate tumor progression. The breast cancer model used a PyMT-derived ductal lumen cell line (PyMT-BO1) to investigate the role of gut microbiota in regulating the growth of primary mammary tumors (149). Disruption of intestinal microbiota by gavage administration of oral antibiotics (vancomycin, neomycin, metronidazole, amphotericin, and ampicillin [VNMAA]) prior to administration of tumor cells to animals, producing severe intestinal microbial changes (150, 151), and although no significant differences in tumor tissue structure were observed in those animals compared with a control group receiving plain water, significantly accelerated tumor growth was observed. Under a similar treatment regimen (152), enhanced growth resembling basal-like breast cancers was observed when spontaneously derived basal cells (EO771) were implanted in situ, suggesting that antibiotic-induced microbiota disruption can drive disease progression in multiple breast cancer subtypes. To determine the effect of the VNMAA mixture on the microbiota, microbial DNA was isolated from the cecum of control and VNMAA-treated animals on day 18 and subjected to birdshot macro-genomics analysis. The analysis revealed dramatic changes in the populations and overall diversity of the bacteria obtained from the animals that received VNMAA treatment, with the Shannon diversity index showing that the abundance of several microbial communities in the gut of antibiotic-treated mice was significantly reduced. In parallel, some communities (e.g., Fusobacterium nucleatum) persisted or overgrew. The composition of the gut microbiota was significantly altered in terms of species, abundance and overall diversity following the use of antibiotics, which was associated with accelerated tumor growth and an increase in mast cells in the tumor stroma. To determine whether mast cells affected tumor growth, Kirkup et al. (147) treated control and VNMAA-treated tumor-bearing mice with cromolyn (a mast cell stabilizer) and found that cromolyn inhibited tumor growth in the antibiotic-treated animals. Notably, the VNMAA-treated group without cromolyn treatment showed a significant increase in tumor size when compared to the control animals treated with cromolyn, and an increase in the number of mast cells was observed in sections of the EO771 tumor stroma taken from VNMAA-treated mice (147, 148). Those data suggessed the key role that mast cells play in tumor progression after antibiotic-induced microbiota disruption in mouse breast cancer: when vancomycin alone was used to induce microbiota disruption, effects similar to those already described were observed in a completely different model of breast cancer. Possibly, microbiota disruption was associated with increased homing of mast cells to, and/or increased proliferation within, tumor. However, given that mast cells in the control animals did not affect tumor progression, the protumor function observed was shown to be specifically regulated by the microbiota. Given confirmation that antibiotic disruption of the gut microbiota has a detrimental effect on breast cancer, antibiotic-induced dysbiosis of the flora and dysregulation of the associated metabolites could be hypothesized to promote tumor growth by reprogramming mast cell homing and/or function. Future studies might consider determining the changes that occur in mast cells and breast tumor cells in response to gut dysbiosis. Kirkup et al. (147, 148) used a mixture of vancomycin, neomycin, metronidazole, and amphotericin (VNMA) to assess DNA concentrations in feces after microbiota dysbiosis and found very low DNA concentrations in the feces of the experimental group compared to the control group (water treatment). Importantly, the rate of PyMT-BO1 and EO771 breast tumor growth was significantly increased after disruption of the gut microbiota in the treated animals compared with the control animals (water treatment). Transcriptomic analysis also revealed dramatic differences in the regulation of metabolic pathways after antibiotic-induced dysbiosis of the intestinal flora, suggesting that accelerated breast cancer tumor growth might be associated with metabolic reprogramming. Fecal metabolomics was confirmed by 1H NMR spectroscopy analysis, which showed that 8 metabolites were elevated and 9 were significantly reduced in the major components of fecal samples from antibiotic-treated animals compared to fecal samples from control animals (147). Several of these amino acids (among them alanine, histidine and aspartic acid) were significantly increased in the antibiotic-treated animals. In contrast, the SCFAs butyrate and acetate, but not the branched-chain fatty acid isovalerate, were significantly reduced. Microbiota-derived butyrate is readily absorbed from the gut and can play a role in inhibiting histone deacetylases (153) in a variety of diseases, including cancer. Inhibition by butyrate can sensitize cancer cells to reactive oxygen species-induced apoptosis, thereby inhibiting the proliferation of breast cancer cells (154), but its role in the organism is yet to be confirmed in clinical trials. The authors hypothesized that a decrease in the bioavailability of the intestinal flora metabolite butyrate plays a role in enhanced tumor metabolism. Metabolites from the gut can reach distant tissues and organs such as the breast via the circulation, where they might play a role in regulating cancer cell function. Kirkup et al. noticed that antibiotics associated with breast cancer (e.g., cefadroxil, which is widely used in the USA after mastectomy). C57BL/6 mice carrying PyMT-BO1 tumor cells and receiving a cefadroxil dose equivalent to that in human patients experienced a significant acceleration in tumor growth. Analysis of the gut microbiota of the animals showed that the microbiota aggregates in samples from the experimental and control animals were independent and clustered differently before and after treatment. The relative abundance of Lactobacillus decreased over time in the control and experimental groups, and this appeared to be replaced mainly by fecal genera in the control animals, however, this did not occur in the antibiotic-treated animals. The genus with the most significant change in the microbial composition of the animals in the experimental group compared to the pre-treatment samples was Lactobacillus, but there was no significant difference before and after the control group. We presume that the disappearance of Lactobacillus might be driven by tumor cells, tumormicrobiota interactions, or natural maturation of the microbiome rather than by cefadroxil administration. Further analysis revealed differences in the abundance of 11 genera after cefadroxil treatment: Mucispirillum, Marvinbryantia, Parabacteroides, Anaeroplasma, Bacteroides, and Paraprevotella were significantly higher, and Alloprevotella, Alistipes, Odoribacter, Faecalibaculum, and Anaerotruncus were significantly lower. When multiple comparisons were made, 8 genera were significantly altered after antibiotic treatment. the genera that were significantly lesser abundant in treated animals relative to the controls, several are known butyrate-producing bacteria (e.g., Odoribacter and Anaerotruncus) or genera carrying the genes required for butyric acid production (e.g., Faecalibaculum and Alistipes) (147), consistent with the significant reduction in butyrate production observed in the metabolomic analysis of feces. That observation suggests that the use of a single antibiotic associated with breast cancer causes significant changes in microbiota genera and aggregation, potentially correlating with the tumor growth rate, but without a direct link to accelerated growth of breast tumors.
Effect of changes in metabolites following microbial perturbation on breast cancer metastasis
A major signaling route between the microbiome and the host is the secretion of Microbial metabolites that enter the circulation and reach their target cells (155)(156)(157)(158). microbial metabolites synthesized in organs or glands (in this study, in the microbiome) function much like human hormones, in that they transfer to other anatomic locations and exert biologic effects (159). Microbial metabolites can enter the circulation and interfere with the steady-state of the intestinal and other local environments, acting as signaling mediators that influence the progression of breast cancer. SCFAs (160,161), Lithocholic acid (LCA) (162-165), cadaverine (166), and de-conjugated estrogens (96, 109,167), these metabolites have the ability to inhibit tumor-cell proliferation, the conversion of epithelial cells to mesenchymal cells, tumor metastasis, and cell migration and metastasis, and to induce antitumor immunity, to restructure cell metabolism, to induce senescence, and lower the number of tumor stem cells (164,166,168,169).
The finding that perturbations in the gut microbiome are associated with tumor propagation at a distance supports the idea that the gut microbiome can be considered to be an endocrine gland (159,170). Some metabolites associated with the activity of gut bacteria can enter the bloodstream and have been shown in vitro to affect the functioning of breast cancer and immune cells. Members of the microbiota can digest certain indigestible components of the human diet (e.g., dietary fiber), and SCFAs -for example, acetate, propionate, and butyrate-are components of metabolized dietary fiber (30, 171) and act as modulators of the host's immune response. Bioactive compounds such as metabolic polyphenols (172) promote the growth of beneficial bacteria such as Bifidobacterium and Lactobacillus and produce SCFAs (173, 174). Some studies have shown that microbially derived homologous receptors for SCFAs were associated with a reduction in the invasive potential of breast cancer cells, with the homologous receptor FFAR2 inhibiting the Hippo-Yap pathway and increasing the expression of the adhesion protein E-cadherin, and FFAR3 inhibiting MAPK signalling (175), particularly butyrate, which has anticancer effects, as demonstrated in cancer cell cultures (176,177) and animal models (160). Crucially, those microbial metabolites are produced after fermentation and/or metabolism of dietary components, and one of the key roles of the microbiota is to break down complex foods into simple bioactive compounds.
In the gut, disruption of the microbiota breaches the biologic barrier between it and the underlying tissue, leading to adverse physical contact between microbes and host cells, inducing paracrine production of bacterial metabolites (135). Changes in the microbiome have been associated with metabolic diseases such as obesity and type II diabetes (178), which are risk factors for certain cancers, including breast cancer (80,179). The intestinal flora is responsible for the conversion of primary bile acids to secondary bile acids (180), and changes in the intestinal microbiota can therefore directly affect changes in secondary bile acids. Edit Mikoáh et al. (169) studied three secondary bile acids-LCA, deoxycholic acid, and ursodeoxycholic acid. Of those three, LCA was found to exert a tumor-suppressive effect by reducing the growth of MCF7, SKBR3, and 4T1 breast cancer cells. They tested the cytostatic properties of LCA in mice transplanted with 4T1 breast cancer cells and found that the ability of the primary tumor to infiltrate surrounding tissues and metastasize was significantly reduced after LCA treatment. This study was the first to provide evidence for a mechanism of interaction between the microbiome and breast cancer by describing that LCA, a metabolite of microorganisms in the gut, is transferred to the breast via the bloodstream and might play an important role in promoting antiproliferative effects in breast cancer. However, LCA might be produced by the breast's own microbiota and not only by the gut microbiota. The ratio of those two sources (breast and gut) in terms of LCA abundance is unknown and requires substantial research and continued trials.
Cadaverine is produced through lysine decarboxylation by lysine decarboxylase (181). Shigella felis, Shigella sonnei, Escherichia coli, and Streptococcus are all capable of expressing it (182). Kovaćs et al. (166) explored the effects of cadaverine supplementation (500 nmol/kg) on mice homozygously transplanted with 4T1 breast cancer cells and found a reduction in the aggressiveness of the primary tumor. Histologic examination of the primary tumors after cadaverine treatment showed a reduced mitotic rate and heterogeneity of nuclear morphology in the mammary tumor cells. To assess whether cadaverine treatment could convert mesenchymal-like carcinoma cells into epithelial-like cells, increased cadaverine resistance was measured using ECIS (Electric Cell-substrate Impedance Sensing), which showed better cell adhesion. To verify that finding, cells stained with Texas Red-X phalloidin and observed under microscopy showed that, after cadaverine treatment, the fibroblast-like morphology of 4T1, MDA-MB-231 and SKBR-3 breast cancer cells had changed to a cobblestonelike morphology that is characteristic of epithelial cells, and the inhibition of matrix metalloproteinase 9 expression also confirms the decrease in tumour cell migratory properties. A cellular flux analyser assessed the metabolic changes induced by necrotropism and found a reduction in glycolytic flux, which is characteristic of breast cancer mesenchymal cells (183). Cadaverine exerts its anticancer effects by inhibiting epithelialmesenchymal transition, cell motility, chemotaxis, and metastasis. A further assessment of the "stemness" of 4T1 cells using an aldehyde dehydrogenase assay found that "stemness" was also slightly reduced (166). Dysbiosis of the intestinal flora (i.e., a change in the basal environment) leads to a change in the level and type of metabolites produced, which might have no effect on reducing the proportion of stem cells in breast cancer and slowing the rate of metastasis or might have the opposite effect, promoting malignant progression of the tumor. In the early stages of breast cancer in dysbiosis mice, bacterial cadaverine biosynthesis in the gut is reduced, leading to lower production of anti-cancer bacterial metabolites. We can speculate that in the presence of disturbed or slightly disturbed gastrointestinal flora, the metabolites produced act as signaling mediators and a specific crosstalk reaction may occur with the host, and this process may be directly or indirectly linked to the metastasis, migration and invasion of mammary tumors in mice.
Role of probiotics to block breast cancer spreading
A few studies have found that probiotic preparations are gaining in popularity for the improvement of health conditions such as antibiotic-induced diarrhea, irritable bowel syndrome, and obesity (184, 185). The use of probiotics can reduce or inhibit tumor growth, reduce tumor angiogenesis, tumor cell extravasation and lung metastasis (186). Long-term disturbance of the gut microbiome, which disrupts the probiotic structure and composition, may conversely increase the risk of breast cancer metastasis (187,188).
Lactobacillus casei, a type of probiotic, is a Gram-positive bacterium that is resistant to the body's defense mechanisms. After entering the human body, L. casei can survive in large numbers in the intestinal tract and can play a role in regulating the balance of intestinal flora, promoting digestion and absorption, among other processes (189). It is highly effective in lowering blood pressure (190) and cholesterol (191), promoting cell division and antibody immunity, enhancing human immunity, preventing cancer, and inhibiting tumor growth. Aragoń et al. (186) used milk fermented with L. casei CRL431 to evaluate its possible effects on tumor growth, tumor cell extravasation and lung metastasis in a mouse model. By comparing mice fed fermented milk (FM), mice fed regular milk and mice not fed any special food, it was found that the group fed FM showed an inhibition of tumor growth and a decrease in tumor vascular filling, tumor cell extravasation and lung metastasis. Khoury et al. (192) used kefir water, a fermented milk product containing probiotics, to treat BALB/c mice that had been transplanted with 4T1 mammary cancer cells and, in the treated mice, detected a significant reduction in tumor size and weight, a significant enhancement of helper T cells and cytotoxic T cells, a significant reduction in lung and bone marrow metastases. Zamberi et al. (193) found that kefir water (mix of Lactobacillus acidophilus, Lactobacillus casei, and Lactococcus lactis) exerted an anti-angiogenic effect on mouse mammary tumors by down-regulating the tumor-promoting invasive interleukin 1b and vascular endothelial growth factor (a key mediator of angiogenesis). In the above model, levels of the pro-angiogenic factor interleukin 6 were found to have declined (186,189,194,195) after probiotic treatment, suggesting that downregulation by Lactobacillus might affect the metastatic potential of cancer cells. some study (186,196,197) demonstrated that milk fermented with Lactobacillus casei CRL431 (probiotic fermented milk (PFM)) reduced the side effects of capecitabine and reduced intestinal mucositis and mortality in a mouse model of breast cancer by modulating the immune response, this suggests the potential of PFM as a probiotic as an immune adjuvant that may reduce tumor growth and metastasis without compromising the anti-tumor/antimetastatic effects of chemotherapy. They differentially regulate cancer-related signaling pathways in a cell-type-specific manner and play a suppressive role in the pro-tumor microenvironment (198)(199)(200). Conversely, disruptions in the intestinal flora might simultaneously or subsequently affect the probiotic environment, which could cause probiotics to lose their "dominant" role in the tumor environment, negatively affecting the control or inhibition of breast tumor cell growth or even accelerating the growth of tumor cells and promoting angiogenesis, becoming an indirect contributor to tumor metastasis. Yazdi et al. (201) demonstrated that seleniumnanoparticle-enriched L. brevis administered to mammary tumor-bearing BALB/c mice induced an effective immune response, resulting in reduced liver metastases and an increased lifespan, included increases in the T helper cytokines, interferon-gamma and interleukin 17, and enhanced natural killer cell activity. Hassan Z et al. Demonstrated that (202) Enterococcus faecalis and Staphylococcus hominis can significantly inhibit cell proliferation, induce apoptosis, and cell cycle arrest, and that they have no cytotoxic effect on normal cells, making them a good alternative drug for breast cancer treatment (Figure 1). Figure 1. The linkage between probiotic environmental homeostasis and breast cancer metastasis Probiotics have specific anticancer properties, and studies have shown that they can alter the expression of various genes involved in apoptosis (203), invasion and metastasis (204), maintenance of cancer stem cells (205), and control of the cell cycle (206). Probiotics have been highlighted as superior in the treatment of cancer. however, more pre-clinical and clinical studies are needed to determine which strains are beneficial during specific treatments before probiotic administration is considered safe and customisable for all individuals. The linkage between probiotic environmental homeostasis and breast cancer metastasis. A good diet (e.g. foods rich in dietary fiber, soy isoflavones, fucoxanthin and polyphenols) can reduce intestinal flora dysbiosis and thus harmonize the body to reduce the incidence and metastasis of breast cancer. Diet as an important factor in the stable composition of the host probiotic environment, through intestinal flora regulation. probiotic environmental homeostasis can play an adjuvant anti-cancer role in the progression and metastasis of breast cancer (lung, brain, liver, bone). (146) VNMA PyMT-BO1, E0771 Up Antibiotic-induced dysbiosis of microflora was associated with reduced expression of pro-apoptotic genes and increased expression of pro-survival genes (153) PyMT-BO1, E0771 Up Antibiotic administration was associated with dramatic differences in the regulation of microbial metabolic pathways and increased tumor growth rates in laboratory animals (146,148) cefadroxil PyMT-BO1 Up Gut microbial aggregation, genus differences, and accelerated tumor growth were observed in cefadroxil-treated animals (
Diet affects the likelihood of breast cancer progression
Although the correlations between BRCA risk and dietary intake have been intensively studied, the underlying associations or effector mechanisms remain poorly understood. Historically, increased risk of BRCA has been tied to high intake of red meat and animal fat (207,208), with decreased risk being concurrently linked to fruit and vegetables consumption (209). Changing dietary patterns affects the microbiome and Indirect affects the development of breast cancer. A case-control study in Japan showed that regular consumption of Lactobacillus casei Shirota and soy isoflavones from puberty onwards reduced the incidence of breast cancer in Japanese women (210); Newman TM et al. also indicated that the Mediterranean diet could prevent breast cancer, because of its inclusion of an abundance of plant-based foods and the lack of processed foods (211). Xue M et al. confirmed through experiments (212) that fucoidan increases the diversity of intestinal flora and can promote the intestinal barrier function, and he suggested fucoidan as a preventive agent for breast cancer. studies have shown that increased polyphenol intake is associated with higher levels of beneficial bacteria (such as Bifidobacterium and Lactobacillus) and SCFAs in humans (174), while also decreasing levels of bacteria that have been associated with disease, so-called pathobionts. Diet is an important factor in all microbiota studies and can help maintain the stability of gut microbes, which can influence the development of breast cancer. If dietary interventions are to be successfully used in future treatments, studies of diet and microbiota metabolites might have to be conducted in parallel. Indeed, recent studies have highlighted the personalised response to individuals (and the microbiota) to the same diet (213), which highlights the limitations and challenges for next-stage studies of this kind.
Alcohol consumption increases the risk of breast cancer, although alcohol itself is not a direct carcinogen, acetaldehyde, a product of alcohol metabolism, is a mutagen which can form adducts with protein and DNA, inducing gene mutation, DNA crosslinks and chromosomal aberrations (214)(215)(216). Many studies have also confirmed that alcohol consumption not only induces breast cancer development (217, 218) but also promotes the progression of existing breast tumors and induces a more aggressive phenotype (219)(220)(221)(222). There are no clear reports to confirm the correlation between alcohol, microorganisms and breast cancer metastasis, but there is no doubt that alcohol causes dysbiosis of the intestinal flora (223-225). It is not difficult to guess that there may be an alcohol-gut flora-breast cancer axis, which means that changing lifestyle habits could have profound implications for the prevention and prognosis of the disease, but the role of gut flora in this needs to be studied in depth.
Details of the following studies included in this review are summarized in Table 3.
Conclusions and future prospects
Globally, the number of factors affecting gastrointestinal dysbiosis is increasing and the gastrointestinal microbiome is emerging as an important player in the risk and progression of breast cancer. This provides an exciting new perspective on breast cancer metastasis, namely that the causes of intestinal dysbiosis are complex and variable, and that there may be a complex causal relationship between progression and metastasis of breast cancer. Therefore, treating the gut flora to stabilize the microenvironment may reduce pro-tumorigenic factors and their propagation in the tissue microenvironment, and establishing new strategies to balance these deleterious fluctuations is of interest in the treatment and prognosis of breast cancer. Given that several intrinsic and extrinsic factors are known and that the gut microbiota and breast cancer have an interactive relationship, future sequencing of the microbiota to capture metadata about dysbiosis and the selection of in vivo models are expected. Those steps will be informative and positive in reducing the risk of breast cancer progression and metastasis, and in guiding therapy for gastrointestinal symptoms or prognosis in patients with breast cancer. Future studies analyzing the gastrointestinal microbiota in patients with breast cancer should consider definitive stratification by histology and molecular science, which could require longer experience and a longer time frame. In addition, because of the large number of complex resident gut flora species, the difficulty of data collection and the unclear specific mechanisms of microenvironmental changes due to dysbiosis, studies and evidence linking the gastrointestinal microbiota to breast cancer metastasis and progression are currently relatively scarce and need to be validated by more specific and high-quality clinical trials and data, and there is an urgent need to combine different disciplines and microbiome studies and design new technical approaches.
Data availability statement
The current state of research and references in this article (review) are cited from the relevant references and the data are authentic and publicly available.
Author contributions
JZ, ZL, MD, XH and MY contributed to the conception and the drafting of manuscripts. GS, QX and YZ are responsible for coordinating and participating in the article revision. All authors contributed to the article and approved the submitted version.
Funding
The project was supported by the 2022 Qinghai Province Central Guide to Local Science and Technology Development Fund, the Breast Disease Treatment Centre of the Affiliated Hospital of Qinghai University manages this funding.
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
|
2022-11-07T14:35:33.832Z
|
2022-11-07T00:00:00.000Z
|
253372312
|
s2orc/train
|
Connective tissue disease-related interstitial lung disease is alleviated by tripterine through inhibition of the PI3K/Akt, apoptosis, and TNF-α signalling pathways
Background: Interstitial lung disease (ILD) is the major cause of morbidity and mortality in patients with various rheumatic diseases. However, more interventions need to be sought. Tripterine, an extract of Tripterygium wilfordii Hook. F, has been widely studied for its powerful anti-inflammatory effect. However, its mechanism of action in treating connective tissue disease-related (CTD)-ILD remains unclear. Purpose: To investigate the mechanism of tripterine in CTD-ILD treatment by combining network pharmacology and an in vivo experiment. Methods: The related targets of tripterine were obtained after searching the Traditional Chinese Medicine System Pharmacology Database and Analysis Platform, Comparative Toxicogenomics Database, GeneCards, Search Tool for Interacting Chemicals database, and SymMap database. Following this, Online Mendelian Inheritance in Man, GeneCards, Genebank, and DrugBank were used to screen the targets of CTD-ILD. A target-signalling pathway network was constructed using Cytoscape. Additionally, topological analysis was performed. Protein interaction analysis was performed using the STRING online analysis platform. Following this, Gene Ontology (GO) and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) signalling pathway enrichment analyses were performed. Subsequently, the molecular docking between tripterine and the core targets was verified. Finally, experimental verification was performed in bleomycin-induced model mice. Results: A total of 134 common targets and 10 core targets of tripterine, including signal transducer and activator of transcription 3, tumour necrosis factor (TNF), v-rel avian reticuloendotheliosis viral oncogene homolog A, protein kinase B (Akt) α (Akt1), mitogen-activated protein kinase (MAPK) 1, Jun transcription factor family, tumour protein 53, MAPK3, nuclear factor kappa B subunit 1, and caspase 8, were obtained. GO enrichment analysis revealed that, while treating CTD-ILD, tripterine was mainly involved in cytokine receptor binding, receptor-ligand activity, signal receptor activation, cytokine activity, protein ubiquitination, deoxyribonucleic acid transcriptase activity, etc. The KEGG pathway enrichment analysis revealed that the most significant signalling pathways were multiple viral infections and the phosphatidylinositol-3-kinase (PI3K)/Akt, TNF, and apoptosis signalling pathways. Molecular docking results revealed that tripterine had good docking activity with the core targets. Experimental studies also demonstrated that tripterine could inhibit the activation of PI3K/Akt, apoptosis, and TNF-α signalling pathways in lung tissue and significantly improve lung pathology and collagen deposition in the model mice. Conclusions: This study preliminarily revealed the potential molecular biological mechanism of tripterine while treating CTD-ILD might be related to inhibiting the PI3K/Akt, apoptosis, and TNF-α signalling pathways. Tripterygium wilfordii Hook. F. and its extract could be used clinically for treating CTD-ILD.
Introduction
Connective tissue disease (CTD), a common clinical autoimmune disease involving multiple organs and systems, usually affects the respiratory system. Interstitial lung disease (ILD) is one of its severe pulmonary complications (Spagnolo et al., 2021). Studies have indicated that approximately 40% of patients with ILD also have CTD (Mira-Avendano et al., 2019). By far, the aetiology and pathogenesis of CTD-related ILD (CTD-ILD) remain unclear; however, immune-mediated lung inflammation and subsequent fibrosis are the key steps (Spagnolo et al., 2021). Infiltration of inflammatory cells in the lung interstitium is thought to be responsible for the early stages of ILD. Disease progression causes collagen deposition and fibrocyte proliferation, resulting in the pathological manifestations of fibrosis in the advanced stage (Desai et al., 2018). Regulating various pro-inflammatory factors and restoring immune homeostasis primarily inhibits the progression from inflammation to fibrosis in this disease (Hoffmann-Vold et al., 2019). Therefore, in recent years, glucocorticoids combined with immunosuppressive agents have been mainly administered for clinically treating CTD-ILD. It is noteworthy that numerous immunosuppressants used for treating CTD-ILD have varying degrees of adverse reactions and limited efficacy (Gao and Moua, 2020).
Recently, it has been reported that transforming growth factor (TGF)-β1 (Celada et al., 2018;O'Leary et al., 2020), platelet-derived growth factor (PDGF) (Zhu et al., 2019), vascular endothelial growth factor (VEGF), and fibroblast growth factor (FGF) play a vital role in pulmonary fibrous tissue proliferation through their effects on leukocytes and angiogenesis (Chaudhary et al., 2007). A popular drug for treating ILD, Nintedanib, plays an antifibrotic role by targeting the above molecular mechanisms (Hoffmann-Vold et al., 2019); however, it is expensive and has several side effects. Hence, it may not be the drug of choice for patients. Other interventions, such as complementary therapies, are yet to be explored.
Our previous study reported that Tripterygium wilfordii Hook. F. polyglycoside tablets satisfactorily treated CTD-ILD; however, its mechanism of action remains unclear (Li et al., 2021). Tripterine is a natural pentacyclic triterpenoid compound extracted from the root bark of Tripterygium wilfordii Hook. F. It has good anti-inflammatory, antioxidant, antitumour, immunosuppressive, and neuroprotective biological properties (Chen et al., 2018). Animal experiments have confirmed that tripterine has a significant therapeutic effect on injured lung tissue ; however, most studies were mainly about rheumatic diseases, and research on its applicability for treating CTD-ILD and its molecular biological mechanism are scarce. Therefore, this study aimed to study the mechanism of tripterine, which is anticipated to be effective in treating CTD-ILD.
Network pharmacology, a comprehensive research method based on a disease-gene-drug action target network (Yang et al., 2013), provides a fresh perspective for studying the complex interaction between potential traditional Chinese medicine components and disease targets. Nevertheless, the predicted results need to be experimentally verified. Therefore, the network pharmacology method was adopted in this study to explore the molecular mechanism of tripterine in the treatment of CTD-ILD. Following this, the results were verified using molecular docking technology and in vivo experiments. The concrete research workflow is illustrated in Figure 1.
Screening of related targets
The eight databases exploited to obtain the related targets were as follows: Frontiers in Pharmacology frontiersin.org 03 'Tripterine' was used as the keyword to retrieve key targets from the TCMSP, Comparative Toxicogenomics Database, GeneCards, STITCH, and SymMap databases. "Connective Tissue Disease-related Interstitial Lung Diseases" was then used as the retrieval term in OMIM, GeneCards, Genebank, and DrugBank. The potential targets were obtained after eliminating the duplicated targets. Organism equal to Homo sapiens was limited.
Intersection target acquisition and protein-protein interaction (PPI) network construction R software (R 4.0.2) was used to map tripterine potential targets to CTD-ILD-related targets, and a Venn diagram was drawn to obtain an intersection of the two targets. The targets common to tripterine and CTD-ILD were uploaded to the STRING database (https://string-db.org), the species was selected "Homo sapiens," and the confidence was set at > 0.9. A PPI network was constructed by importing the results into Cytoscape 3.8.0.
Gene Ontology (GO) and the Kyoto Encyclopaedia of Genes and Genomes (KEGG) enrichment analysis R software (R 4.0.2) was used to analyse the molecular function (MF), biological process (BP), and cellular components (CC) to identify the interaction characteristics of the common targets in the gene function and signalling pathway. The KEGG pathway enrichment analysis was performed to analyse the possible tripterine interventional pathway in CTD-ILD.
Molecular docking
Molecular docking was performed between the screened target protein and tripterine. The three-dimensional (3D) structure of tripterine was obtained from PubChem (https:// pubchem.ncbi.nlm.nih.gov/) and imported into AutoDock Tools 1.5.6 saved in the Protein Data Bank, Partial Charge (Q), and Atom Type (T) (PDBQT) format. The 3D structure of the target protein was obtained from the Protein Data Bank database (https://www.rcsb.org), and the water molecule and the original ligand were eliminated using PyMOL. Following this, the target protein was imported into AutoDock Tools 1.5.6 for hydrogenation, charge calculation, and non-polar hydrogen bonding, the results of which were saved in the PDBQT format. The size of the grid box was 40 × 40 × 40. Finally, molecular docking was performed using the CMD command of AutoDock Vina.
Test animals and groups
Specific pathogen-free C57BL/6J male mice (18-20 g, 6-8 weeks old) were procured from Cytointelligen (Taizhou, China). All the mice were kept at room temperature and acclimatised for a week before initiating the experiment. All animal experiments involved in this study were approved by the Ethics Committee of the Nanjing University of Chinese Medicine. Thirty mice were divided into the following five groups: the normal, model, low-dose tripterine, medium-dose tripterine, and high-dose tripterine groups. The mice in the high-, moderate-, and low-dose groups were intraperitoneally injected with 2 mg/kg, 1 mg/kg, and 0.5 mg/kg tripterine, respectively, every other day. The same volume of normal saline was injected intraperitoneally in the mice belonging to the normal and model groups.
Construction of the pulmonary fibrosis (PF) model BLM intratracheal instillation was employed to simulate the CTD-ILD model. After the mice were anaesthetised with isoflurane in the supine position, the lumbar puncture needle was inserted into the trachea under light source irradiation. Following this, 50 μL of BLM solution was quickly injected into the trachea using a syringe. Finally, the mice were kept in an erect position and shaken to facilitate the even distribution of the solution in the lungs.
Haematoxylin-eosin (H&E) and masson staining
Lung tissue was fixed in 4% paraformaldehyde and then dehydrated with alcohol at different concentration gradients.
Frontiers in Pharmacology frontiersin.org Subsequently, the sample was embedded in paraffin and sectioned for H&E and Masson staining. The degree of alveolitis and PF was scored according to the method reported by Ashcroft et al. (Ashcroft et al., 1988).
Micro-computed tomography (CT) scanning
On the 28th day, the mice underwent micro-CT under isoflurane anaesthesia. Mouse lung imaging was performed using a Quantum GX Micro-CT scanner (PerkinElmer, Inc, Waltham, MA). Specific parameter settings are performed according to previous studies (Ruscitti et al., 2020).
Measurement of hydroxyproline content in the lung tissue
The hydroxyproline level in the lung tissue samples was determined using a hydroxyproline assay kit. Twenty milligrams of lung tissue samples were homogenised using radioimmunoprecipitation assay (RIPA) buffer and proteinase inhibitors, and the supernatant was aspirated for use. The assay kit was removed from the refrigerator and kept at room temperature for at least 30 min. The concentrated wash buffer was diluted with double distilled water. The blank control, sample, and standard wells were set up on the plate according to the manufacturer's instructions, and the corresponding reagents were added. A standard curve was plotted based on the measured optical density values, and the measured concentrations of the samples in each well were calculated based on the curve equation. These concentration values were then multiplied by the dilution factors to obtain the final concentrations.
Measurement of TNF-α in the serum and lung homogenate
The samples were removed from the refrigerator (−80°C), and dissolved at room temperature. After the standard substances were diluted, the samples were loaded and washed, the enzyme was added, and the solution was incubated following the instructions of the TNF-α enzymelinked immunosorbent assay kit. Following this, the linear regression equation was derived based on the absorbance value of the standard hole. The concentrations of the samples were calculated based on the absorbance values of the samples and the regression equation. Finally, the final concentration was obtained based on the product of the measured concentration and the dilution factor.
Immunohistochemistry
Paraffin blocks were sectioned for dewaxing and hydration. To block the activity of the endogenous peroxidases for 10 min, 3% hydrogen peroxide was added. After antigenic repair, the paraffin blocks were incubated with serum blocking solution for 30 min to seal the non-specific binding site. Following this, the serum was removed, and the sections were washed. Subsequently, the sections were incubated overnight with the primary antibodies at 4°C. The next day, the slides were washed, treated with secondary antibodies, and stained with diaminobenzidine and haematoxylin. Brownishyellow areas observed under light microscopy (×400 magnification) indicated a positive result.
Western blot
The lung tissues were lysed using the RIPA buffer comprising proteinase inhibitors. The protein content was measured, separated using 10% sodium dodecyl sulphate-polyacrylamide gel electrophoresis, and then electroblotted onto a nitrocellulose membrane. Non-specific antigens were incubated in 5% bovine serum albumin with the membrane for 2 h. Following this, the membrane was incubated overnight with primary antibodies at 4°C. The membrane was washed thrice in 0.1% Tris-buffered saline with 0.1% Tween ® 20 Detergent (TBST) for 10 min at room temperature.
The following day, the corresponding secondary antibody (1:2000) was added, and the mixture was incubated at room temperature for 1 h. The membrane was washed thrice with TBST three times for 10 min and then exposed to a colour-developing reagent under dark conditions. ImageJ 1.49 software was used to analyse the grey values.
TUNEL staining of lung tissues
The paraffin section was dewaxed twice for 5-10 min. This was achieved by treating the section with anhydrous ethanol for 5 min, followed by 90% ethanol for 2 min, 70% ethanol for 2 min, and distilled water for 2 min. A drop of 20 μg/ml DNase-free protease K was added, and the section was kept at room temperature for 15-30 min. The section was then washed thrice with phosphatebuffered saline. Following this, the reaction solution was added separately. Finally, the sections were sealed using an antifade mounting medium with 4′,6-diamidino-2-phenylindole.
Data analysis
Data analysis was performed using GraphPad Prism 8.0 software. Between-group differences were tested using oneway analysis of variance. Values are expressed as the mean ± standard deviation (* represents the difference between the BLM-Frontiers in Pharmacology frontiersin.org 05 treated and the other groups). Statistical significance was set as p < 0.05.
Acquisition of the main intersection targets
A total of 247 tripterine targets were obtained by searching TCMSP, Comparative Toxicogenomics Database, GeneCards, STITCH, and SymMap databases. A total of 2,860 disease-related targets were obtained after eliminating duplicates by searching the OMIM, GeneCards, Genebank, and DrugBank databases. A total of 134 intersection targets were obtained by mapping disease-related targets to potential drug targets (Figure 2A).
PPI network construction and topology analysis
A total of 134 target proteins were imported into the STRING 11.0 online database for protein correlation analysis, and the PPI network parameter of tripterine in the treatment of CTD-ILD was obtained. Subsequently, the screened PPI network was imported into Cytoscape 3.8.0 for topological analysis. The deeper the node colour, the higher the value, and the higher the core position of the target protein in the network ( Figure 2B). Based on the results, the following were the top 10 targets: signal transducer and activator of transcription 3 (STAT3), TNF, v-rel avian reticuloendotheliosis viral oncogene homolog A (RELA), protein kinase B α (Akt1), mitogen-activated protein kinase (MAPK) 1, Jun transcription factor family (JUN), tumour protein 53 (TP53), MAPK3, nuclear factor kappa B subunit 1 (NF-κB1), and caspase 8 (CASP8) ( Figure 2C). The details of the top 10 targets are presented in Table 1.
GO and KEGG enrichment analysis
A total of 2,719 items were obtained by the GO enrichment analysis. The first 10 items with p-values < 0.05 in the MF, BP, and CC categories were selected. The bubble size represented the number of targets during the GO analysis, and the bubble colour represented the enrichment significance. It mainly involved cytokine receptor binding, receptor-ligand activity, signal receptor activation, cytokine activity, protein ubiquitination, deoxyribonucleic acid transcriptase activity, ribonucleic acid transcriptase Ⅱ specificity, etc. ( Figure 3A). A total of 155 items, including multiple viral infections and the PI3K/ Akt, TNF, and apoptosis signalling pathways, were obtained from the KEGG pathway enrichment analysis. The first 30 related GO enrichment results and 10 KEGG enrichment analyses were visualised ( Figure 3B).
Molecular docking
Molecular docking was performed between the first five core targets and tripterine. Furthermore, we docked tripterine with four main targets of nintedanib in the treatment of CTD-ILD. The docking results suggested that tripterine could Figure 4). All binding energies were < −5 kcal/mol. In terms of molecular docking, the smaller the binding energy, the better the docking activity between ligand and receptor. Therefore, we reported that tripterine could interact with these core targets.
Effects of tripterine on body weight and lung histopathology
After intratracheal instillation of BLM or normal saline, the mice lost weight during the first week. However, the mice belonging to the control group gained weight gradually after the second week. In contrast, the weight of the mice treated with tripterine belonging to the model group was relatively stable ( Figure 5A). After 28 days of treatment, the endpoint weights of the mice in each group were evaluated. It was observed that the body weight of the mice in the BLM group significantly decreased, while the body weight of the mice receiving varying doses of tripterine was relatively stable ( Figure 5B). H&E and Masson staining results in each group were analysed using the Ashcroft score. The result revealed that the mice treated with BLM had a significantly increased Ashcroft score ( Figure 5C). Furthermore, the alveolar structure underwent destruction, and the alveolar cavity was infiltrated with inflammatory cells. However, after management with the administration of medium-or high-dose tripterine, the inflammatory cell infiltrate in the alveolar cavity significantly improved, and the alveolar structure underwent varying degrees of repair. Masson staining revealed that lung sections of the model group were filled with massive blue collagen deposits. The blue areas significantly decreased after the administration of tripterine ( Figure 5D).
Effects of tripterine on collagen deposition and Micro-CT
One of the main pathological manifestations as ILD progresses is the thickening of the alveolar septum caused by collagen deposition, which further affects lung diffusion. CT is widely used in the clinical Frontiers in Pharmacology frontiersin.org diagnosis and management of ILD, and micro-CT is a mature tool for evaluating lung imaging changes in small animals (Ruscitti et al., 2018). Therefore, pulmonary collagen deposition and pulmonary micro-CT findings were evaluated in this study. After BLM was administered to the model mice, significant collagen I and collagen III deposits were observed in the lung tissue. Interestingly, collagen deposition was alleviated to varying degrees after varying doses of tripterine were administered ( Figure 6A). Micro-CT results revealed typical lung consolidation in the model group. Lung consolidations significantly decreased after tripterine treatment, as observed on radiographs ( Figure 6B).
Effect of tripterine on the epithelialmesenchymal transformation (EMT) marker protein in lung tissues
α-SMA is a phenotypic marker of myofibroblasts in PF. Vimentin anchors and supports organelles in the mesenchymal cell cytoplasm. Both of these are involved in EMT (García-Cuellar et al., 2021). EMT-derived cells might also promote abnormal epithelial-mesenchymal crosstalk, thereby promoting fibre formation. Hydroxyproline is a non-essential amino acid found in collagen and is one of its main components. It is also an important marker of PF. Herein, we found that the α-SMA, vimentin, and hydroxyproline expression levels significantly increased in the model group and significantly decreased after 28 days of treatment with tripterine, particularly in the mediumand high-dose groups (Figures 7A-D).
Effects of tripterine on the PI3K/Akt and TNF-α signalling pathways
Our network pharmacology study reported that the PI3K/ Akt and TNF-α signalling pathways might be involved in the anti-CTD-ILD effect of tripterine. Therefore, we measured the key proteins of the PI3K/Akt signalling pathway and the TNFα content in the serum and lung tissues to identify the effects of tripterine on these pathways. Our results revealed that the phosphorylated PI3K and Akt levels significantly increased in Results are presented as the mean ± standard error of the mean. ###p < 0.001 vs the normal group; *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the BLM-treated group. BLM, bleomycin (5 mg/kg); Trip, tripterine.
Frontiers in Pharmacology
frontiersin.org the model group but decreased to varying degrees after tripterine administration ( Figures 8A,B). Similarly, TNF-α levels increased in the model group and decreased in the lung tissue of the model mice after tripterine administration. However, no significant improvement was observed in the serum TNF-α levels of the model mice ( Figure 8C).
Effect of tripterine on the apoptosis level of lung tissues
Apoptosis is a biological process closely associated with ILD (Nho et al., 2022). The apoptosis level of the lung tissues in each group was observed by immunofluorescence assay to validate the results of our KEGG analysis. Apoptotic cells significantly increased in the lung tissues of the model mice, based on the immunofluorescence assay results. We observed that the positive rate of apoptotic staining in the lung tissues decreased gradually with an increase in the tripterine dose (Figure 9).
Discussion
CTD-ILD is a heterogeneous disease characterised by diffuse alveolar inflammation and fibrosis (Atzeni et al., 2018), with progressive aggravation of clinical symptoms, ultimately leading to respiratory failure, high mortality, and limited therapeutic options. Due to the heterogeneity and interdisciplinary nature of ILD, a high-level scientific basis for this disease is scarce (Mira-Avendano et al., 2019;Johannson et al., 2021). It also lacks treatment and monitoring guidelines or an expert consensus. In the development and evolution of this disease, Immune imbalance is the main pathological factor responsible for the development and progression of ILD, and controlling the various pro-inflammatory factors and the restoration of immune homeostasis help inhibit the progression from inflammation to fibrosis in ILD (Kolahian et al., 2016). Currently, it is believed that the damage-repair process along with epithelial cell injury (Chambers and Mercer, 2015), innate and adaptive immunity (Kolahian et al., 2016;Desai et al., 2018), and humoral factors (De Lauretis et al., 2013;Wu et al., 2017) are involved in Data are presented as the mean ± standard deviation (SD), n = 3. (C) TNF-α expression in the lung homogenate and serum of mice were detected by enzyme-linked immunosorbent assay. Data are presented as the mean ± SD, n = 6. ###p < 0.001 vs. the normal group; *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the BLM-treated group. BLM, bleomycin (5 mg/kg); Trip, tripterine.
FIGURE 9
The apoptosis level of the lung tissues was analysed by immunofluorescence. Normal: normal group; BLM, bleomycin (5 mg/kg); scale bars: 100 μm.
Frontiers in Pharmacology frontiersin.org 12 the process of fibrosis. Hence, Anti-inflammatory and antifibrotic therapy are the hot spots in the treatment of this disease (Wells, 2021). Studies have demonstrated that tripterine could decrease lung tissue injury by reducing capillary permeability, inhibiting inflammatory cytokines or chemokines, and governing the expression of numerous inflammatory mediators (Law et al., 2011). The pathogenesis of PF is typically characterised by excessive extracellular matrix (ECM) deposition and EMT, during which several factors change, including excessive accumulation of type I and type III collagens, the main components of the ECM, and the up-regulation of mesenchymal markers, such as α-SMA and vimentin (Hsieh et al., 2022). Through in vivo experiments, we reported that tripterine could decrease lung pathology and collagen deposition in the lung tissues of the model mice. Increased levels of fibrosis markers, namely α-SMA and vimentin, were observed in the model group. The levels of these proteins in the lung tissue decreased to varying degrees after treatment with tripterine. These results suggest that tripterine could reduce BLM-induced fibrosis in model mice. Therefore, we used network pharmacology to investigate the related targets and pathways of tripterine in the treatment of CTD-ILD.
A total of 134 potential targets of tripterine were obtained for treating CTD-ILD, of which STAT3, TNF, RELA, Akt1, MAPK1, JUN, TP53, MAPK3, NF-κB1, and CASP8 were strongly correlated with CTD-ILD. Additionally, the above-mentioned core targets were closely associated with PF. KEGG pathway analysis revealed that the potential mechanisms of triptolide in the treatment of CTD-ILD might be attributed to multiple viral infections and the TNF, PI3K/Akt, and apoptosis signalling pathways. These pathways are involved in the pathogenesis of various inflammatory diseases and are crucial for antipulmonary fibrosis progression (Meng et al., 2014).
Firstly, viruses have been considered pathogens or propagators of ILD-related inflammation (Wootton et al., 2011), and multiple viral infections can significantly increase the risk of interstitial pneumonia (Sheng et al., 2020). Recent studies (Pehote and Vij, 2020) have reported that viral infection impedes autophagosome formation in patients with ILD, and autophagy function was damaged due to invasive accumulation, thereby promoting the senescence of lung epithelial cells and myofibroblast differentiation.
Secondly, according to the KEGG enrichment analysis results, apoptosis, TNF-α, and PI3K/Akt signalling pathways might be related to the mechanism of tripterine in CTD-ILD treatment, as verified by in vivo experiments. The TNF signalling pathway is a pro-inflammatory signalling pathway that plays a crucial role in joint and lung inflammatory diseases. Previous studies (Wu et al., 2019) have reported that anti-TNF treatment could selectively increase the monocyte and dendritic cell count in lung tissues, thereby achieving the therapeutic effect of alleviating joint and lung inflammation. Furthermore, fibroblast to myofibroblast transformation, which is considered to play a crucial role in CTD-ILD pathogenesis, can be enhanced by TNF-α (Spagnolo et al., 2021). Apoptosis is a key mechanism for regulating cell death (Sauler et al., 2019), which is beneficial for eliminating damaged, infected, or excessive cells and is associated with the occurrence of ILD (Mahavadi et al., 2010). We reported that tripterine could reduce the fluorescence intensity of TUNEL staining in lung tissues, indicating that it could reduce the occurrence of apoptosis in the lung tissues of model mice. Furthermore, a significant increase in TNF-α levels was observed in the model group. TNF-α levels in lung tissues significantly decreased after tripterine administration.
In addition, the PI3K/Akt signalling pathway has been confirmed to play a pathogenic role in ILD. For example, TGF-β can promote EMT, induce fibrosis through the PI3K/ Akt signalling pathway, and modulate fibroblast differentiation into myofibroblasts that regulate ECM accumulation (Huang et al., 2021). PI3K is considered to be the focus of collagen synthesis induced by most pathways, and currently, a nontargeting PI3K protein inhibitor for treating idiopathic pulmonary fibrosis is undergoing human clinical trials (Hettiarachchi et al., 2020). Akt is the downstream protein of PI3K. Blocking these protein targets also has a corresponding effect on inhibiting PF progression. Our experiment demonstrated that tripterine could inhibit the activation of the PI3K/Akt signalling pathway in the lung tissues of model mice.
These results were consistent with our predictions based on network pharmacology analysis. Our findings suggested that tripterine attenuates collagen accumulation and fibrosis in lung tissue, resulting in the inhibition of the PI3K/Akt, apoptosis, and TNF-α signalling pathways. This may partly explain why a lesser degree of inflammation and fibrosis were observed in the BLM-induced lungs. Indubitably, this study has some limitations, which need to be resolved by further research. First, various cells play an important role in the pathological development of ILD; however, specific cellular experiments were not performed to verify the above findings in this study. Secondly, whether the PI3K/Akt signalling pathway plays a role through the apoptotic pathway and the concrete downstream regulatory mechanism involved needs to be elucidated. Lastly, the BLM-induced ILD model undoubtedly does not completely mimic immune-mediated lung interstitial lesions, and whether tripterine has the same effect in ILD models induced by other factors likewise needs to be explored.
Conclusion
Tripterine can lower the degree of pathology and fibrosis in BLM-induced ILD in mice. The specific mechanism may be related to inhibiting the PI3K/Akt, apoptosis, and TNF-α signalling pathways. Tripterygium wilfordii Hook. F. and its extract can be employed clinically for treating CTD-ILD.
Frontiers in Pharmacology frontiersin.org
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 Ethics Committee of Nanjing University of Chinese Medicine.
Author contributions
WZ designed this study, performed the online database search, and finished the primary manuscript. YW performed the data collection, the online database search and did data analysis. CL and YW conducted the Methodology and Investigation. YL participated in the data collection and designed the study. YW finished the revision of this manuscript. All authors discussed the results and approved the final manuscript.
|
v2
|
2022-11-07T14:38:59.913Z
|
2022-11-07T00:00:00.000Z
|
253372245
|
s2orc/train
|
The efficacy of cognitive behavioral therapy for cancer: A scientometric analysis
Cognitive behavioral therapy (CBT) is one of the most recognized psychological interventions to improve the overall quality of life of cancer survivors. To analyze current research trends in the field of the link between CBT and cancer and to provide potential future research directions, we conducted the scientometric analysis. The study was conducted on all documents in this field from 2012 to 2022 retrieved from Web of Science. Then Biblioshiny, VOSviewer software, and CiteSpace software were used for getting the information of article postings and citations, countries, institutions, journals, authors, and keywords. The number of documents about the link between CBT and cancer from 2012 to 19 July 2022, was 619, with 476 of articles and 143 of reviews. The number of annual publications has been fluctuating, with the highest number of publications in 2020. The country with the maximum number of publications and citations was the US. The University of Houston was the organization with the highest quantity of publications and total link strength (TLS). Psycho-Oncology was the most active journal in the field and has the highest h-index. Zvolensky MJ was the author with the highest quantity of publications. The most cited keywords were “Quality-of-life,” “Cognitive-behavioral therapy,” “Depression,” “Cognitive therapy” and “Breast-cancer.” And as evidenced by the keyword citations, the focus of this research area has gradually shifted to the mental health of patients and the underlying pathogenesis. The impact of CBT in cancer treatment is now well established and has gradually evolved toward symptom-specific treatment. However, the relationship between CBT and cancer has not been further developed. Future research is needed to be further developed in the identification of a generic formula for CBT in cancer and the exploration of mechanisms of CBT and cancer.
Introduction
Cancer is one of the world's most incurable diseases and has the second highest mortality rate after cardiovascular disease (1). According to statistics of GLOBOCAN 2020, 19.3 million new cancer cases and almost 10 million cancer deaths were indicated in 2020. By assuming the cancer incidence rate remains unchanged as in 2020, there will be 28.4 million new cancer cases worldwide in 2040, an increase of 47% from 2020 (2). With the development of early screening and innovative therapies such as personalized medicine and immunotherapy to prolong survival, the survival rate for cancer has increased significantly. However, this also brings with other aspects of problems. Long-term ongoing cancer treatment can take a huge physical toll and psychological burden on patients (1). Survivors may face months to years of cancer triggering and/or treatmentrelated symptoms such as fright of cancer recurrence, fatigue, insomnia, obesity, etc. which can adversely affect the quality of life of survivors (1,3). Therefore, it is especially vital to monitor and ameliorate the adverse consequences of cancer treatment in the course of the chronic process of cancer therapy. Cognitive behavioral therapy (CBT) is a treatment approach that uses psychological and behavioral interventions to alter the patient's dysfunction. Evidence suggests that CBT is the most effective psychological intervention to improve tiredness caused by cancer therapy and can make the quality of life of cancer survivors better (4)(5)(6).
CBT is a psychotherapeutic intervention that focuses on problem. It is an integration of behavioral interventions such as behavioral stimulation, exposure treatment, emotion regulation and relaxation training (7). It has evolved over more than 60 years from traditional face-to-face therapy to a diverse range of therapies such as internet-facilitated cognitive behavioral interventions. As an evidence-based treatment modality, CBT has been proven to be used in the treatment of a variety of psychiatric disorders such as depression, anxiety disorders, personality disorders, schizophrenia and more (8). Recently, the combination of CBT and medication is becoming more common in clinical practice and achieve greater efficacy, which will be one of the future directions of research and treatment.
There is a growing body of clinical researches on CBT and cancer, and most of them have confirmed the relationship between CBT and cancer. Although CBT is effective in monitoring and improving prognosis for cancer survivors, most survivors reported not discussing psychological interventions for cancer with their treatment providers, much less using them (8). Therefore, we conducted a scientometric analysis of the relevant publications in the field and identified the current research. Our analysis aimed to get specific information on the publication profile of the field, such as publications and citations, top active countries, most productive authors, influential journals, hot topics and keyword analysis, identifying current issues that need to be addressed and predicting key directions for research in the future.
Data collection
We searched all documents about the link between CBT and cancer on the Web of Science Core Collection (WoS) online databases. All entry terms related to "CBT" and "cancer" were searched as search themes through the medical subject headings (Mesh) of the PubMed website. The search input included the following: #1, ((((ALL = ("Cognitive Behavior Therapy")) OR ALL = ("Cognitive Therapy")) OR ALL = ("Cognitive Psychotherapy")) OR ALL = ("Cognition Therapy")) OR ALL = ("Cognitive Behavior Therapy"); #2, (((((((ALL = (cancer)) OR ALL = (carcinoma)) OR ALL = (neoplasm)) OR ALL = (adenocarcinoma)) OR ALL = (melanoma)) OR ALL = (adenocarcinoma)) OR ALL = (sarcoma)) OR ALL = (osteosarcoma); #3, #1 and #2. Then, time span of these publications was filtered from 2012 to 2022. The research was conducted on July 19, 2022. We searched for a total of 692 documents. After refining the types of documents to articles and reviews and restricting the language to English, 619 documents were retrieved, including 476 articles and 143 reviews. In total, 73 publications were excluded during the screening process: 9 articles not in English, 3 reviews not in English, 42 meeting abstracts, 9 editorial materials, 5 letters, 3 corrections, 1 biographical-items, and 1 book review.
Data analysis
We applied the online analysis of WoS, bibliometrix, VOSviewer software, and CiteSpace software to perform a scientometric analysis of all retrieved articles. The online "Results Analysis" function of the WoS was initially used to access important information on the article topics, publication dates and types, research fields, countries, authors, institutions, journals, and free access for these publications. The WoS "Citation Report" function also allowed us to obtain additional information about the publications, like the total quantity of citations to articles, the specific quantity of citations per article, average quantity of citations per article and the trend in citations over time, etc. Biblioshiny is the online analytics site of bibliometrix with powerful features for data analysis (version 4.2.1), one of which could be applied to perform a comprehensive scientometric analysis and to generate visual image of results (9). Therefore, we imported the raw files of the retrieved documents into this website and obtained related information about these publications, including time span, annual growth rate, number of sources, documents, and references, types of documents, authors, collaboration of authors, etc. In addition, other information analyzed on the Bibliometrix website would also be included in our scientometric analysis, including growth trend of annual publication outputs, and citations analysis, countries and organizations analysis (production), journal and research category analysis (journal sources, impact and dynamics), author analysis (authors' contribution and impact), and keywords analysis. The three field diagrams showed the collaboration between different countries, authors and institutions, where more occupied areas proved a higher production of articles. Finally, the co-occurrence network of thematic evolution formed by keyword plus was an indicator to identify research hotspots.
VOSviewer (version 1.6.18), is a network analysis software, which can be used to construct network of authors or journals based on co-citation data and keyword network based on cooccurrence data. It can display bibliometric views of the results of different analyses of a set of data in four different views, such as label view, density view, cluster density view, scatter view (10). In our study, it was mainly used to demonstrate the strength of the links between different countries, different institutions and different authors and co-citation network of references and keywords.
CiteSpace V (version 6.1 R2) is a knowledge domain visualization tool, which can find the key points of the development of a research domain. It facilitates the identification of citation concentrations of literature and keywords in our analysis (11).
Analysis of publication and citation
The all number of publications about the link between CBT and cancer that we retrieved from WoS from 2012 to 2022 was 619. 2020 was the year with the largest number of publication output (84, 13.57%). Production of annual publications was on an increasing trend from 38 publications (6.19%) in 2012 to 84 publications (13.57%) in 2020, but the overall trend seemed to be unstable between 2012 and 2022 (Figure 1). The period 2012-2014 was a relatively stable period with a steady annual publications of around 38 documents. The quantity of publications has fluctuated since 2014. Over the 7 years from 2014 to 2022, annual production has repeatedly risen and fallen a total of three times, which indicated that the overall development of the field is not mature enough. The annual production showed an increasing trend during 2014-2015, 2016-2018, and 2019-2020, while it shows a decreasing trend during 2015-2016, 2018-2019, and 2020-2022. The annual growth rate of annual scientific output is 2.36%, the highest in 2015 (66.67%) and the lowest in 2018 (18.64%). In 2020, annual production reached a maximum and exceeded 80 documents.
An article's scientific impact can be reflected by the number of citations (12). According to the citation report of WoS, there were 12,506 citing articles among all the retrieved publications, including 335 self-citations and 12,171 without self-citation Supplementary Table 1 shows details of the top 10 most quoted papers within the study on CBT and cancer, which was integrated through the specific information from all the documents presented on the Biblioshiny website. Among them, the highly cited articles could indicate the current research focus in the field. These top 10 articles were mainly published between 2012 and 2017, with the most cited article being published in 2013. Of these ten articles, 30% were published in 2013 and 30% in 2014, and one in each of the remaining years. The most cited article was a clinical review on psychology (13), with a total of 998 citations (13). The article "How do mindfulness based cognitive therapy and mindfulness based stress reduction improve mental health and wellbeing? A systematic review and meta-analysis of mediation studies" published by Gu et al. (14) was cited 744 times (14). The work "cancer related fate mechanisms, risk factors, and treatments" written by Bower (15) was the third-ranked article (cited 579 times) (15). In total, 3 studies were cited more than 500 times, representing 0.50% of all papers; 6 papers (1.01%) were cited more than 300 times and 28 papers (4.70%) were cited more than 100 times. However, no article has been cited more than 1,000 times since its publication. Figure 3 is the co-citation network diagram of cited literature of these publications crafted by VOSviewer. After selecting 20 as the lowest number of citations for the cited literature, 66 references met the requirement. The size of the node was related to the quantity of reference citations with a larger node meaning more references cited ( Figure 3A). In Figure 3A, we could observe that there were three clusters of the cited references, red, blue and green, with 25 items in the red cluster, 24 in green and 17 in blue. A cluster with a larger number of items means it is more attractive in the research area. Annual article output of paper citations between 2012 and 2022 in this research area.
The citations of reference could also be seen from Figure 3B.
The yellower color indicated more co-citations. In Figure 3, the article of Zigmond and Snaith (16) The paper written by Cillessen (21) was the reference with the strongest burst (strength = 4.95), which had the citation bursts from 2020 to now. Additionally, 6 references (24%) were cited consecutively until 2022, while 5 references (20%) were separately cited until 2017 and 2020.
Analysis of countries and institutions
The analysis of countries and organizations is a comparative analysis of the quantity of papers and article citations for each country and institution, which can show the current distribution of articles in this field. The 619 documents we retrieved were mainly from 44 countries or regions. Table 1 shows the top 10 countries with the maximum articles in the area. The US was the most active country among these 44 countries or regions, with a total of 276 publications (44.59% of all). Next up were the Netherlands (89 publications, 14.38%), Australia (82 publications, 12.92%), Canada (65 publications, 10.50%) and England (62 publications, 10.02%). Likewise, The United States leaded the way in terms of citations of publications, with 8,194 citations. It was followed by Canada (2,661 citations), Netherlands (2,596 citations) and England (2,226 citations). We also could compare the quantity of publications in different countries by the size of the points in Table 2).
We launched an analysis of co-authorship relationships for organizations with more than five publications, yielding results for a total of 76 organizations. Then, we obtained a plot of the relationship between the TLS between different
Analysis of journal and research category
The 619 papers were published in 264 different sources of journals. The h-index, the largest number of articles published by a researcher or journal which have been cited h or more times, would be used to measure the academic achievements of a researcher or journal based on both quantity and impact Table 3). Psycho-Oncology, BMC Cancer, Cognitive Therapy and Research, Supportive Care in Cancer and Mindfulness have remained active in the field during the years, especially Psycho-Oncology, which has been number one in terms of cumulative publications since 2012. Cognitive Behavior Therapy was constantly enhancing its influence in this research area, which attempted to publish articles in this field after 2014 and had become the second largest journal of cumulative publications by 2022 (Supplementary Figure 2).
These 619 papers cover a total of 61 research categories.
Analysis of author
According to the main information of 619 documents, a total of 2,864 authors were involved, including 20 authors for single-author articles and 2,844 authors for multi-author articles. Zvolensky MJ was the most active author in this field, with 62 publications (10.02% of all), which far exceeded that of the latter authors. The authors immediately following were Bleijenberg G (n = 21, 3.39%), Garey L (n = 20, 3.23%), Knoop H (n = 20, 3.23%) and Hunter MS (n = 17, 2.75%) ( Table 4).
Only one author has published more than 50 articles. Four authors have published more than 20 articles, and 17 authors have published over 10 articles. We created a figure of the top ten most-published authors in the field for each year (as shown in Supplementary Figure 3), with the size of the dots representing the articles published in that year and the intensity of the color of the dots representing the total number of citations per year. Zvolensky MJ's publications were low until 2014 but began to increase rapidly after 2014 and remained more active in the latter 5 years. Bleijenberg G has published more in the first few years and less in the last 5 years, while Garey L has published less in the first few years and more in the last 4 years. The number of articles published by Knoop H and Hunter MS has maintained a steady trend in recent years.
As with the analysis of journal, the h-index was applied to analyze the scholarly achievement and value of each author's paper. Only three of the top ten authors had an h-index of more than 10. Both Bleijenberg G and Knoop H had the highest h-index of 12 and was followed by Zvolensky MJ (h-index = 11), Hunter MS (h-index = 9), Gielissen MFM (h-index = 9), and Goedendorp MM (h-index = 9). The m-index was used to eliminate the effect of age and experience between authors, taking into account the differences in the start of publication and the quantity of papers published by different authors (23). Gallagher MW was the author with the highest m-index (mindex = 1,250). Bleijenberg G and Knoop H ranked No. 2, with the m-index of 1,091. The rest of the authors had an m-index either equal to 1,000 or less than 1,000.
The top 8 authors with the most times cited of their publications in the research scope of CBT and cancer were listed in Table 5 Co-authorship network diagram between institutions in the area. Top 10 active research subject categories. Co-citation analysis of authors was performed in Figure 9 is a three-field diagram, which indicates the collaboration among different countries (middle), authors (left), and organizations (right). The occupied area of the region in the figure represents the number of documents issued by a country, institution or author, and the density of line indicates the closeness of cooperation between them. The United States was the country with the strongest collaboration between organizations and authors in the field. The University of Houston and the University of Texas MD Anderson Cancer Center were responsible for nearly half of the articles published in the field in the United States. And the author with the strongest collaboration with the U.S. was Zvolensky MJ. The Radboud University Nijmegen and the Vrije Universiteit Amsterdam were two of the organizations with the most cooperations with other countries. As for the authors, most of them preferred to work with one country, and only some had cooperations with more than one country or institution. The total number of cooperation between Zvolensky MJ and other countries was the largest, and Knoop H was the second.
Analysis of keywords
After selecting five times as the minimum number of occurrences of the keyword, we obtained 273 keywords. The clustering analysis function of VOSviewer software would be used to divide 273 items into several clusters. There was a total of seven different colored clusters in the network visualization map (Figure 10). The relevance of keywords in clusters of the same color was higher. The size of the node in the Figure 10A and the degree of yellow in the Figure 10B represented the frequency of keyword occurrences. The top 10 frequent words in the research scope of CBT and cancer were listed in Table 6 and their rate of growth and number of cumulative occurrences per year were showed in Supplementary Figure 4. There was a total of six keywords with the occurrence of more than 100 times. "Quality-of-life" was the most frequently used keyword with 193 occurrences and was followed by "Cognitivebehavioral therapy" (156 occurrences), "Depression" (149 occurrences), "Cognitive therapy" (127 occurrences), "Breastcancer" (123 occurrences), and "Randomized controlled-trial" (119 occurrences). In term of TLS, the top 5 keywords were "Depression" (TLS = 1,971), "Quality-of-life" (TLS = 1,689), "Cancer" (TLS = 1,506), "Cognitive therapy" (TLS = 1,387) and "Cognitive-behavioral therapy" (TLS = 1,239). Figure 10C is the overlay visualization map which shows the average time of keyword appearance. The change in color reflects the time of occurrence of the keyword. Figure 11 indicates the trends of keyword development over the years. The length of the line indicates the duration of the keyword, the year in which the dot appears means that the keyword appears most often at that time, and the larger the dot, the more often it appears. Through Figures 10C, 11, we could find out a gradual shift in the focus of the research field from the adverse effects and prognosis of cancer therapy and the role of CBT to the mental health of patients and the mechanisms of morbidity.
Using the data provided by the Bibliometrix website, we analyzed the keywords plus that appear in the literature cited in the article. Figure 12 shows a four-quadrant diagram with the different keyword degree of relevance and development. The bottom horizontal axis represents the degree of relevance (centrality), with centrality representing relevance of the keyword to the field of study. The leftmost vertical axis represents the degree of development (density), which represents the degree to which the keyword has been studied or developed in the field. The four quadrants are: (First quadrant) Motor themes: the themes in this region have the best development trend and are most relevant to the area. (Second quadrant) Niche themes: the development of themes in this region is good, but it is lack of connection with other articles in this field. (Third quadrant) Emerging or declining themes: these themes are less well developed and related in this field of study. It is possible that these themes have just appeared or are in the process of disappearing. (Fourth quadrant) Basic themes: the themes are strongly related to this field, but are hardly developmental. They are may be some basic concepts or knowledge. From Figure 12, the keywords were divided into numerous clusters with different color. We could find that the purple cluster's keywords ("Depression, " "Anxiety, " "Symptoms") were in the first quadrant, which meant they were closely related to the field and were particularly important, while the yellow-brown cluster's keywords ("Anxiety sensitivity, " "Distress tolerance, " "Substance use disorders") were in the third quadrant and were not indicative for the moment. Compared to the small number of clusters present in quadrant one and there, there were many clusters present in quadrant two and four. The clusters in the second quadrant, such as orange clusters ("Negative effect, " "Nicotine dependence, " "Smokers") and red clusters ("Vasomotor symptoms, " "Postmenopausal women, " "Hot flashes"), were highly developmental, but had little connection with articles in this area. The clusters in the fourth quadrant, like gray cluster ("Cognitive-behavioral 10.3389/fpsyt.2022.1030630 Co-citations network diagram between authors with more than 50 citations. The three-field diagram of the collaboration between different countries (middle), authors (left) and organizations (right) (Occupied area represents the number of articles issued and the density of line indicates the closeness of cooperation). therapy, " "Survivors, " "Women"), and pink cluster ("Cognitive therapy, " "Meditation, " "Mental-health"), had a strong link to the researches but were poorly developed. In addition, both the green and orange clusters were close to the center, indicating a high degree of keyword plus relevance in both clusters.
We conducted a citation burst analysis of the keywords by CiteSpaceV software. Keyword citation burst analysis allows for a more detailed study of keyword citation duration and citation burst intensity. Figure 13 displays the top 25 keywords with the strongest citation burst. The outbreak of keyword citations appearing in different years represents trends in the research field for that year. The citations burst of these keywords started from 2012 to 2020. Among the top 25 keywords, 28% (7/25) began to appear citation explosion in 2012, the second was 2017 (20%, 5/25), and the third was 2018 and 2020 (15%, respectively, 3/25). "Association" had the highest citation burst intensity at 4.69 and the citation burst lasted from 2019 to 2022. Other keywords with citation burst strength of over 3 were "United States" (Citation burst strength = 3.64), "Prostate cancer" (3.36), "Stress reduction: Mindfulness-based stress reduction (MBSR)" (3.34), "Support" (3.2), "Reduction" (3.14), "Functional assessment" (3.02). There were 5 keywords for which the citation burst lasts until 2022. The last 12 keywords have been consistently cited in recent years, proving that they were the focus of recent research in the field, particularly "Depressive symptom, " "Association, " "Reduction, " "Functional assessment" and "Anxiety sensitivity, " which will continue to be active beyond 2022.
Discussion
This is a scientometric analysis of articles on the link between CBT and cancer published in 2012-2022 by using Bibliometrix, VOSviewer software, and CiteSpace software. Trends in keyword development over the years. Top 25 keywords with highest citations bursts in this area.
number of cancer survivors has increased significantly. But cancer survivors may face months to years of cancer triggers and/or treatment-related symptoms such as fright of cancer recurrence, fatigue, insomnia, obesity, etc. after long-term cancer treatment. The application of CBT can be very effective in alleviating these problems (23-26). Cognitive behavioral therapies have evolved to target a specific symptom, such as CBT for insomnia (CBT-I) (27), CBT for fatigue (CBT-F) (28), and so on. However, this is only a small step forward in improving the precision of CBT. The relationship between CBT and cancer has still not been groundbreaking on this basis. In total, 44 countries or regions contributed to the 619 articles in the country and institution analysis. The US was the largest contributor to the field with 276 articles, accounting for nearly 45% of the total. The top two institutions in terms of number of publications, the University of Houston and the University of Texas MD Anderson Cancer Center, are both located in the United States, which could demonstrate the importance and contribution of the US to the field. The US was a pioneer in the field of CBT and cancer, but other countries were also making breakthroughs in the field. A prime example of this was China. It took only 4 years for China to go from publishing its first article to being the sixth most published country. The trend of Chinese publication themes showed the progress of Chinese research in this field, from the early days of studying the relationship between CBT and cancer, to exploring the refinement of CBT for specific adverse effects caused by specific cancer treatments, and focusing more on patients' psychological health and living conditions (29,30). In addition, the biggest breakthrough of China in this area was the inclusion of the most representative Chinese elements such as acupuncture and Tai Chi in the research process (31)(32)(33), and the study of the effectiveness of these approaches in cognitive behavioral disorders. This is a completely new direction and deserves further research.
Among the top five journals with the maximum number of publications, four were psychological. Psycho-Oncology was the most common theme of research in the field. From the beginning, the psychological theme and the cancer theme were not overlapping, but have now evolved into psycho-oncology, which combines both themes and ensures a high volume of publications in the field.
Zvolensky MJ was the most active author in the field, with about three times the number of articles published by the second-placed Bleijenberg G. He was also the third highest author in the h-index apart from Bleijenberg G and Knoop H. His articles focused on the role of the relationship between emotional states and psychopathology on certain human behaviors, such as smoking and suicide, and were mainly aimed at younger people (34-36). Although Zvolensky MJ's research has not focused on the link between CBT and cancer, his research on the link between psychopathology and behavioral and emotional regulation was very thorough and revealed the potential mechanisms behind the development of mental health problems such as anxiety, distress and fatigue, such as distress tolerance (DT) and anxiety sensitivity (AS) (34, 37), which could provide a basis for the role of CBT in the treatment of poor mental health caused by long-term cancer treatment. And this has greatly guided the direction of subsequent research in the field.
There were two relatively novel points of research in this field during the literature search. The first was the use of "Acute cancer CBT, " which was different from conventional CBT, and the other is that there was numerous research on the role of CBT in breast cancer, but very little on the role of CBT in other cancers.
The emergence of acute CBT is mainly due to the fact that conventional CBT is primarily indicated for the treatment of cancers with a chronic course and most cognitive behavioral studies are driven by established protocols where the patient is stable and has a stable course of treatment. In contrast, CBT for acute cancers can be very different. The instability and uncertainty of acute cancers tend to cause greater psychological shock to patients than chronic cancers. This impact can be understood as a breakdown of psychological defenses due to difficulties in accepting, over-interpreting, misunderstanding, etc. of cancer. The difficulty and challenge of CBT for acute cancer lies in its unpredictable acute cancer setting. CBT is primarily characterized by crisis response to acute cancer settings. However, the life-threatening urgency as well as the unpredictability and adverse effects of cancer treatment can add to the crisis atmosphere, thus making the implementation of CBT more difficult. Levin and Applebaum (38) developed a cognitive behavioral treatment framework applicable to the integration of acute healthcare settings and applied it to acute cancer environment. In summary, empathy, coping models and psychopharmacology are core elements of CBT for acute cancer (38,39). In addition, perceptions of death, and mortality, are important components of acute cancer CBT. Although the model of CBT for acute cancer is different from that for chronic cancer, it has been researched and developed over many years for clinical application and is recommended as an essential training skill for clinicians (39).
The link between breast cancer and CBT is also one of the main directions in the field at present. Breast cancer was the most prevalent cancer among women and the second leading cause of cancer death in women (40, 41). Evidence suggested that depression was one of the most common psychiatric disorders in breast cancer patients, with over half (58%) of patients within the female breast cancer population experiencing mild depressive symptoms and 38% experiencing major depression (42,43). Mindfulness-based cognitive therapy (MBCT) was currently an effective intervention for psychiatric disorders associated with cancer treatment. Luberto et al. (44) described the theoretical rationale for MBCT for cancer recurrence and provided case examples (44). MBCT could guide the emotional changes of breast cancer patients, reduce their psychological disorders, and then reduce the risk of anxiety and depression. In addition, in the article "Predictors and moderators of outcomes in MBCT intervention for early breast cancer patients" (45), the authors aimed to discover the predictors and moderators of outcomes in MBCT for early breast cancer patients and explore the optimal course of treatment that produces significant outcomes (45). It can be argued that the role of MBCT in the treatment of breast cancer is now becoming clearer and represents a vital milestone in the palliative care of breast cancer patients (46). Similarly, CBT has also played a significant role in the treatment of other cancers, such as ovarian and prostate cancers, colorectal cancer and others (47)(48)(49). However, research in this area is currently limited to these specific cancers, and the role of CBT in the treatment of other cancers is not yet known. Therefore, whether CBT has the same efficacy in other cancers should be one of the directions for future research. A general formula for CBT in cancer treatment will be established in the future.
From the results of the keyword analysis, current research has focused on the role of CBT in adverse effects such as fatigue, pain, anxiety and depression following long-term cancer treatment. And these studies have mainly focused on a few specific types of cancer, such as breast cancer. It was clear that the research of CBT in cancer is precise but not extensive, and it fails to involve more kinds of cancer. However, this deficiency should be remedied and a general formula for CBT in cancer will be available in the future. Almost every research study is inseparable from the exploration of mechanisms. This is also true for the field of CBT and cancer. However, there was currently little research in this area that involves cellular or molecular mechanisms, which is a relatively new direction of research. In addition, the frequent occurrence of keywords such as "Mental-health, " "Stress-disorder" and "Anxiety sensitivity" in recent years indicated that the current research has gradually shifted from pure disease treatment to treatment targeting patients' physical and mental health. Therefore, the exploration of the mechanisms between CBT and cancer, as well as the psychological state of cancer patients, will also be the focus of research in the future.
The article "The modulatory role of internet-supported MBCT on extracellular vesicles and psychological distress in people who have had cancer: a protocol for a twoarmed randomized controlled study" written by Pereira et al. (50) was an important milestone in the development of molecular mechanisms in this field (50). It revealed the possible molecular mechanisms by which psychological interventions can improve cancer prognosis. The extracellular vesicles and their contents promote cancer growth, clarity and metastasis through intercellular communication and signal transmission and can therefore be used for cancer screening and diagnosis. The authors evaluated the concentration of biochemical markers in the blood of the body, such as interleukin, tumor necrosis factor, C-reactive protein, etc., telomerase activity, which is a marker related to cancer recovery, and cancerrelated antigens, and then detected the impact of MBCT intervention on the overall immune response, finally measured the changes of extracellular vesicles in the central nervous system (50). This article started with the substances that promote the growth and metastasis of cancer, analyzed whether psychological intervention measures can affect the changes of these substances, and then revealed some potential molecular mechanisms of psychological intervention in cancer treatment. Although this is an initial attempt to investigate the molecular mechanisms in the field of CBT and cancer, it is a huge step forward in the development of the field. Future research into the cellular molecular mechanisms of CBT in cancer treatment is set to flourish.
Conclusion
The relationship between CBT and cancer has become increasingly clear over time, and its role in ameliorating adverse effects following long-term cancer treatment has become indispensable. This scientometric analysis was completed to analyze 619 documents on the scope of CBT and cancer over the period 2012-2022 through Bibliomerix, VOSviewer, and CiteSpace software. The study focuses on a detailed analysis of countries, institutions, authors, keywords, etc., providing a large number of visual figures and tables and revealing the relationships between them. These analyses also provide insights into the current focus of research in the field and possible future research directions. Most of the research in this area has focused on exploring the role of CBT in the treatment of fatigue, depression, anxiety, fear of cancer recurrence and other adverse reactions following long-term cancer treatment, and on refining CBT for specific symptoms. The role of CBT for more cancers, general formulations of CBT in cancer, and the exploration of mechanisms between CBT and cancer will be the focus of future research in this area.
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
CL, HT, and LC involved in the overall design of the research, critically examined, and improved the article. CL was the writing, data collection, and data analysis for this manuscript. HT and LC provided instruction on the use of relevant analytical applications, literature collection, and data interpretation. QY, JW, ZJ, DZ, ZL, and YX contributed to the discussion of the article results. All authors checked and agreed on the final submitted version and contributed to the completion of this article.
|
v2
|
2022-11-07T14:41:12.995Z
|
2022-11-07T00:00:00.000Z
|
253372319
|
s2orc/train
|
Risk factor score for the prediction of central compartment lymph node metastasis in papillary thyroid carcinoma and its clinical significance
Objective To establish the criteria for a risk factor score (RFS) for predicting the probability of central compartment lymph node metastasis (LNM) in papillary thyroid carcinoma (PTC) and to explore the clinical significance of the RFS. Methods The data of 412 patients with PTC who underwent surgical resection between May 2013 and July 2016 were retrospectively analysed and divided into two groups: a central LNM group and a non-central LNM group. In each group, the frequency of six risk factors was documented: sex, age, tumour size, extracapsular spread (ECS), tumour multifocality, and tumour location. The maximum likelihood method of discriminant analysis was adopted to calculate patient scores for the six risk indicators. In addition, the data of 104 patients with PTC admitted between July 2016 and December 2016 were prospectively analysed using this method and these six risk factors. A higher score represented one certain possibility that was the more appropriate for one patient. Results In the retrospective group, the result was as follows: 129 patients with positive (+) lymph nodes in the central compartment and 168 patients with negative (−) lymph nodes in the central compartment, which was in line with the actual results. In the prospective group, there were 28 patients with positive lymph nodes in the central compartment and 48 patients with negative lymph nodes in the central compartment. The coincidence rates using the RFS were 71.9% for the retrospective group and 73.1% for the prospective group. Conclusion By simple and quantitative analyses of the presence of central LNM, the RFS is of great significance when choosing surgical approaches and postoperative individual-based treatment plans, as well as when determining the prognosis of central compartment LNM in patients with PTC.
Introduction
Thyroid carcinoma is the most common endocrine malignancy and a common head and neck malignancy. Papillary thyroid carcinoma (PTC), specifically, is a malignant tumour derived from the thyroid follicular epithelial cells that accounts for about 85% to 90% of thyroid cancers (1). Although PTC has a low degree of malignancy and a good prognosis, it is prone to either cervical lymph node metastases (LNM) or distant metastases at the time of diagnosis, and the central lymph node metastasis rate can be as high as 80% (2)(3)(4). The most fundamental and effective method of treating PTC primary tumours and lymph node metastases is surgery. In terms of patients with thyroid cancer who have suspected or confirmed central compartment LNM, therapeutic central lymph node dissection (CLND) is the mainstay of treatment, but one controversial issue is whether routine prophylactic dissection should be performed in patients without evident cervical LNM (5). If CLND is not performed in carcinoma cases, residual tumours are induced. When reoperation is required, it will inevitably lead to an increased risk of surgical complications due to the formation of surgical scars and the disruption of the anatomical hierarchy. Currently, ultrasonography (US) is the optimal imaging method in thyroid and regional lymph node examinations, and it has a high sensitivity to benign and malignant thyroid nodules and high accuracy for identifying them (6). However, due to the anatomic location of the central lymph nodes and for certain technical reasons, the diagnostic sensitivity is low (merely 31.3%) (7)(8)(9)(10)(11)(12). For this paper, data of patients with PTC were retrospectively analysed using the maximum likelihood method to obtain the scores of factors associated with central LNM. The factors are sex, age, tumour diameter, extracapsular spread (ECS), multifocality, and tumour location. Sex: In thyroid cancer, female patients are more of a concern because of a higher incidence rate. However, several studies suggest that the rate of cervical LNM in men is higher than in women (3,13,14). Age: This has been an important factor in various staging systems for differentiated PTC (13,15,16). Tumour diameter and ECS: Both have always been considered to be important factors in the progression of PTC and are important criteria for evaluating treatment options and surgical scope, and larger tumour diameters are associated with higher cervical LNM rates and T staging (17). Multifocality: PTC often leads to intraglandular metastasis, and multifocality is one of its prominent features. According to literature reports, multifocal carcinoma accounts for 20.3%-33.5% of cases of PTC (14,16,18). Tumour location: In some digestive tract tumours, such as gastric cancer and colon cancer, LNM is closely related to the lymphatic flow path in the region where the tumour is located (19,20). Therefore, we speculate that PTC LNM is related to the lymphatic flow path of the thyroid region where the tumour is located. The criteria for the risk factor score (RFS) were used to prospectively evaluate another group of 104 patients diagnosed with PTC, as well as the clinical significance of the RFS.
Retrospective data
We enrolled 412 patients admitted to our hospital between May 2013 and July 2016 who were diagnosed by postoperative paraffin pathology as having PTC. There were 90 males and 322 females in the study, with a sex ratio of 1:3.6 and a mean age of 44.5 (range 14-81 years old). Tumour diameters ranged from 0.2 to 3.8 cm, with an average diameter of 1.22 cm. A total of 81 patients had ECS (19.6%) and 331 patients did not have ECS. In addition, 169 patients (41%) had papillary thyroid microcarcinoma (PTMC), and 197 patients had central LNM, with a metastasis rate of 47.9%.
Prospective data
A total of 104 patients admitted to our department between July 2016 and December 2016 were diagnosed by postoperative paraffin pathology as having PTC. There were 24 males and 80 females with a mean age of 46.6 (range 20-73 years old). The mean tumour diameter was 1.19 cm (range 0.3-3.8 cm). The number of patients with and without ECS was 9 (8.7%) and 95 (91.3%), respectively. A total of 48 patients had PTMC (46%) and 41 patients had central LNM (39.4%).
Inclusion and exclusion criteria
Inclusion criteria were as follows: (1) neck surgery for the first time; (2) ipsilateral CLND performed; and (3) complete medical records, including a B-ultrasound examination and postoperative paraffin pathology. Patients with the following diagnoses were excluded: (1) metastatic thyroid cancer; (2) thyroid cancer combined with other types of undifferentiated carcinoma; and (3) multiple foci in two or more different parts (the upper, middle, and lower poles, as well as the isthmus).
Clinical observation criteria
Six factors predictive of central LNM were described: (1) sex: both male and female; (2) age: <45 years old and ≥45 years old; and (3) tumour diameter: <0.7 cm, 0.7-1.0 cm, 1.0-1.5 cm, 1.5-2.0 cm, and ≥2.0 cm; (4) ECS: presence or absence; (5) multifocality: presence or absence; and (6) tumour location: the upper (upper 1/3 of the gland), the middle (middle 1/3 of the gland) and the lower pole (lower 1/3 of the gland), as well as the isthmus. A frequency table of count data was constructed in accordance with these six indicators based on data collected from the 412 patients in the retrospective study ( Table 1).
Treatment
Ultrasonography was performed by two sonographers with more than 5 years of experience. The central lymph node status was recorded. The extent of thyroidectomy was determined according to the clinical examination results or preoperative biopsy results. At least unilateral gland lobe plus isthmus resection and total thyroidectomy or near-total thyroidectomy should be performed if any of the following conditions are met: distant metastasis, the primary tumour is larger than 4 cm, the primary tumour invades the surrounding tissue, or a multifocal tumour or cervical LNM can be seen with the naked eye. The histology of frozen sections further helps surgeons determine the extent of surgery required.
A total of 412 patients in the retrospective group underwent standardised surgical treatment: thyroid lobectomy, isthmus resection, and ipsilateral CLND (n = 168); total thyroidectomy and ipsilateral CLND (n = 190); and total thyroidectomy plus bilateral CLND (n = 54). In the prospective study of 104 patients, the above operation methods were used in 46 cases, 28 cases, and 30 cases, respectively. All patients who were given oral levothyroxine sodium tablets underwent thyroid-stimulating hormone (TSH) inhibition treatment. One month later, thyroid function tests were performed, and the dosage of Euthyrox was tailored according to the TSH level. During this period, thyroid function tests were reperformed every 1 and a half months. B-ultrasound examinations of the thyroid and neck lymph nodes were conducted half a year after surgery and every half a year thereafter, as well as other tests such as physical examinations, thyroid function tests, ECGs, and chest x-rays. All patients were advised to see a doctor if necessary.
Postoperative follow-up
All patients were followed up for one to 4 years, with an average of 28.3 months, apart from three who were lost to the study due to changes in telephone numbers. Some patients had postoperative transient hypocalcaemia and hoarseness. Nonetheless, all of them recovered within 6 months of surgery. No other serious complications arose.
Statistical method
As given in Table 1, the frequency (P) of the six indicators for central LNM was calculated using the formula (X kj /Y g ), in which Y represents the type of the central lymph node: g = 1 indicates positive (+), and g = 2 indicates negative (−). X kj (k = 1,2,3,4,5,6; j = 1,2) is the jth classification of the kth indicator. The maximum likelihood method of discriminant analysis for qualitative data was used to estimate the Frontiers in Surgery probability of type g central lymph node in X kj patients: P g = P (X 1j /Y g )×P (X 2j /Y g )…×P (X 6j /Y g ). Take the logarithm of both sides: lgP g = lgP (X 1j /Y g ) + lgP (X 2j/ Y g )…+lgP (X 6j /Y g ). The score of each indicator was derived from [lgP (X kj /Y g )+1]×10 ( Table 2).
The scoring formula was as follows:
Statistical analysis
In this study, the above formula was used for discriminant analysis of the data from the 104 patients in the prospective group and for back substitution of the data from the 412 patients in the retrospective group. The Wilcoxon signed-rank test of two independent samples was performed for comparison between the retrospective and the prospective groups. The significance for all variables was set at α = 0.05. SPSS 21.0 software was used for statistical analyses.
Results
The age distribution and tumour diameter distribution of the two groups are shown in Figures 1, 2, and the score of each factor related to central LNM is given in Table 2. In the retrospective group, the discriminant analysis result was as follows: 129 patients had positive (+) lymph nodes in the central compartment and 168 patients had negative (−) lymph nodes in the central compartment, which was in line with actual results. The coincidence rates were 65.5% and 78.1%, respectively, and the overall coincidence rate was 71.8% (Table 3). However, in the prospective group, 28 patients had positive lymph nodes in the central compartment and 48 patients had negative lymph nodes in the central compartment. The coincidence rates were 68.2% and 76.2% respectively, with an overall coincidence rate of 72.2% (Tables 3, 4).
Discussion
PTC is a common disease in clinical settings. A Bultrasound examination is currently an effective method of distinguishing between benign and malignant thyroid nodules. Dan et al. (21) reported that the diagnosis rate of PTC was as high as 87.64% by high-frequency colour ultrasound, which was, however, not sensitive in the detection of central LNM [approximately 31.3% (7-12)], much lower than the detection rate of lateral neck metastasis (93.8%) (22). This may be related to the anatomical location of the central lymph nodes (adjacent to thyroid and gas-containing trachea) and available technology.
Central LNM is associated with a number of factors. The relevant factors involved in this study were as follows: (1) Age distribution of patients in prospective and prospective groups. (25) demonstrated an association between a younger age and a higher rate of LNM. In addition, patients 20 years old or younger had a metastasis rate of up to 50%. (3) Tumour diameter: According to National Comprehensive Cancer Network guidelines, a tumour diameter of >1 cm is a risk factor for cervical LNM (26), and patients with PTC with tumour diameters of >2 cm had a higher LNM rate in the central region and the lateral neck than those with tumour diameters of ≤2 cm (27). 0.5 or 0.6 cm has been recognised as the cut-off value of central LNM in PTC. In our previous studies (28) Extrathyroidal extension contributed to increased invasion associated with a weakened inhibitory effect of extracellular matrix on LNM in PTC and capsular invasion of the dense network in the thyroid (30). (5) In terms of multifocality, Kuo et al. (31) found that multifocal primary PTMC had a significantly increased metastasis rate compared with unifocal PTMC, consistent with the findings in our previous studies. (6) Tumour location: Zhang et al. (14) reported that a tumour located in the upper pole of the gland was associated with an increased risk of LNM in the lateral neck but a decreased risk of central LNM. Wang et al. (32) found that patients with tumours located in the middle and lower poles are more likely to experience central LNM. Based on the data in Table 1, the rates of central LNM in tumours located in the upper, middle, and lower poles and the isthmus were 35.1%, 48.8%, 55.9%, and 42.1%, respectively. There has been no international consensus regarding prophylactic CLND in patients with cN0 PTC. In this study, the frequency of each relevant factor was measured based on retrospective data. Likelihood function discriminant analysis was used in the prospective group, and samples were subject to back FIGURE 2 Tumour diameter distribution in patients in retrospective and prospective groups. In this study, statistical methods were used for specific and quantitative discriminant analysis of central LNM and benign lesions in patients with PTC. The accuracy (approximately 72%) was superior to the current discriminant analysis of B-ultrasound measurements, which is of clinical significance. Several studies have reported that central LNM is also associated with the coexistence of Hashimoto's thyroiditis and higher TSH levels in patients.
The present study has several limitations. First, the above factors have not been included, and a multicentre study has not been conducted, leading to a certain selection bias, which should ideally be avoided. Further study is, therefore, needed to standardise the scoring criteria. Second, the current follow-up timeframe has not been long enough compared with disease development. We will continue to follow up on these patients. Another limitation of the current study is the small number of patients, and further studies with larger samples are needed. Furthermore, the time of the study has a certain bias, and the results need to be further verified in a subsequent study at a different time. Last, we have not analysed the relationship between Hashimoto's thyroiditis and PTC, and we will add this feature as a part of the risk factor score.
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material, and further inquiries can be directed to the corresponding author/s.
Ethics statement
The studies involving human participants were reviewed and approved by the Second Hospital of Hebei Medical University. The patients/participants provided their written informed consent to participate in this study.
Author contributions
XP conceptualized and designed the article; QL provided administrative support; all authors provided study materials or patients all authors collected and assembled data; all authors conducted data analysis and interpretation; all authors wrote the manuscript; all authors gave final approval of the manuscript. All authors contributed to the article and approved the submitted version.
|
v2
|
2022-11-07T15:13:11.377Z
|
2022-11-07T00:00:00.000Z
|
253372985
|
s2orc/train
|
Leptomeningeal carcinomatosis in a patient with recurrent unresectable squamous cell carcinoma of the retromolar trigone—a brief report
Background The reported incidence of leptomeningeal carcinomatosis is 3–8% in patients with solid tumours. More commonly, it has been described in the setting of advanced cancers of the lung, breast and malignant melanoma. Case presentation A 50-year-old diabetic patient with recurrent unresectable squamous cell carcinoma (SCC) of the right retromolar trigone (rT4bN0M0) presented with severe low backache and weakness in bilateral lower limbs 20 days after the completion of concurrent chemoradiotherapy. Contrast-enhanced MRI of the spine showed multiple nodular enhancing leptomeningeal lesions at the lumbar level and an intramedullary T2/FLAIR-hyperintense longitudinal lesion involving the central cord from C2 to C7 vertebral levels, suggestive of leptomeningeal metastases. Cerebrospinal fluid (CSF) analysis revealed pleocytosis, elevated protein and markedly decreased glucose. The CSF cytology revealed scattered large atypical cells, suspicious for metastasis. Non-contrast MRI of the brain showed a T2/FLAIR-hyperintense lesion involving the right caudate nucleus suggestive of either an acute infarct with haemorrhagic transformation or a haemorrhagic brain metastasis. During assessment, he had high-grade fever and was started on empirical intravenous antibiotics (ceftriaxone, vancomycin and subsequently meropenem) in line with the management for acute bacterial meningitis. Gram staining of CSF did not demonstrate the presence of any bacteria and the specimen was sterile on culture. He did not respond to empirical antibiotics, had a progressive downhill course and eventually died due to aspiration pneumonia. Conclusion This brief report highlights the importance of awareness of leptomeningeal carcinomatosis as a possible cause of backache with sensorimotor deficit and autonomic dysfunction in a previously treated case of head and neck SCC.
Background
The common causes of severe low backache and paraparesis in a patient with cancer include metastasis to the spinal cord, the vertebrae or the lumbosacral plexus. In this context another probable cause is carcinomatous meningitis, also known as leptomeningeal carcinomatosis, which being a rare entity, is difficult to diagnose unless there is high suspicion for the same. Leptomeningeal carcinomatosis has been reported in less than 10% of patients with solid tumours, more commonly in the setting of advanced cancers of the lung, breast and malignant melanoma [1][2][3]. We herein describe a case of carcinomatous meningitis in a patient with recurrent unresectable squamous cell carcinoma (SCC) of the retromolar trigone and illustrate the key clinical and laboratory findings in support of the diagnosis.
Illustrative case
A 50-year-old north Indian type 2 diabetic male presented to the emergency department with complaints of acute pain in the abdomen, constipation, severe low backache radiating to bilateral lower limbs, weakness and inability to walk without support for the past 10 days. Two years prior to these events, he was diagnosed with well-differentiated SCC of the right retromolar trigone cT2N0M0, for which he underwent wide local excision of the tumour with segmental mandibulectomy, upper alveolectomy and modified radical neck dissection type III. The postoperative histopathology report revealed a pT2N0 tumour with close (0.4 cm) superolateral margin. He did not opt for post-operative radiation therapy and was lost to follow-up during the first wave of the COVID-19 pandemic and the consequent nation-wide lockdown. He presented to our centre 18 months after surgery with complaints of progressively increasing trismus and local pain at the operated site. On further evaluation, including a whole body 18F-FDG positron emission tomography/ computed tomography (PET/CT) scan, a 5.2 × 4.2 cm FDG-avid soft tissue mass in the tumour bed with lytic lesions in the mandible, suggestive of local recurrence (rT4bN0M0), was detected. In view of high infratemporal fossa involvement, salvage re-surgery was ruled out and he was subsequently planned for concurrent chemoradiotherapy to a dose of 65 Gray to the high-risk planning target volume (PTV) and 54 Gray to the low-risk PTV (bilateral neck nodal levels Ia, Ib, II, III and right-sided level IVa, Va and Vb) by simultaneous integrated boostvolumetric modulated arc therapy (SIB-VMAT) technique in 30 fractions over 6 weeks along with concurrent cisplatin at a dose of 40 mg/m 2 weekly. Twenty days after the completion of concurrent chemoradiotherapy, he presented with the aforementioned complaints to the emergency department. On physical examination, he was afebrile and haemodynamically stable. On neurological examination, he had decreased tone in the lower limbs bilaterally. The power was MRC grade 2 in the right lower limb and grade 1 in the left lower limb. There was decreased sensation below the level of the umbilicus (T10 spinal segment) and loss of bladder and bowel control. After eliminating sub-acute intestinal obstruction as a cause of abdominal pain and constipation, a contrast-enhanced magnetic resonance imaging (MRI) of the lumbo-sacral spine along with a screening of the whole spine was done to rule out compressive myelopathy and the patient was prophylactically started on injection dexamethasone 8 mg TDS. The MRI scan revealed decreased intervertebral disc space at L5-S1 and disc bulge at L4-L5 and L5-S1 levels causing thecal sac indentation and compression of the exiting nerve roots (Fig. 1). Also, thecal sac indentation was noted at C4-C5 and C5-C6 levels due to prolapsed intervertebral disc (PIVD). In addition, an intramedullary T2/FLAIRhyperintense longitudinal lesion involving the central cord from C2 to C7 vertebral levels and multiple nodular enhancing lesions along the leptomeninges of the spinal canal at the lumbar level were noted suggestive of leptomeningeal metastases (Fig. 1). A neurosurgical opinion was sought for the management of PIVD but any need for active intervention was ruled out. A contrast-enhanced computed tomography (CT) scan of the head and neck revealed hypodense foci in the right frontal lobe and the right basal ganglia, suggestive of infarcts and a stable primary tumour. He was started on aspirin and atorvastatin in order to prevent further episodes of cerebrovascular accident. Meanwhile, he had two episodes of high-grade fever and a lumbar puncture was performed. Cerebrospinal fluid (CSF) analysis revealed an elevated cell count (30/ mm 3 , 70% neutrophils), elevated protein (788 mg/dl) and markedly decreased glucose levels (32 mg/dl; corresponding serum glucose levels-168 mg/dl). He was started on empirical intravenous (IV) antibiotics (ceftriaxone, vancomycin and subsequently meropenem) in line with the management for acute bacterial meningitis. However, Gram staining of CSF did not demonstrate the presence of any bacteria and the specimen was sterile on culture. It was negative for both cryptococcal antigen testing and cartridge-based nucleic acid amplification test (CBNAAT) for tuberculosis. The CSF cytology revealed scattered large atypical cells, which were suspicious for metastasis (Fig. 2). He subsequently underwent a non-contrast MRI of the brain which revealed a T2/ FLAIR-hyperintense lesion involving the right caudate nucleus suggestive of either an acute infarct with haemorrhagic transformation or a haemorrhagic metastasis to the brain (Fig. 3).
In view of a high index of suspicion of carcinomatous meningitis and no clinical improvement on IV antibiotics, a lumbar puncture was repeated after four days. Analysis of CSF revealed markedly elevated protein (777 mg/dl) and low glucose levels (19 mg/dl; corresponding serum glucose levels-179 mg/dl). The repeat CSF cytology showed numerous polymorphs and occasional lymphonuclear cells with no evidence of atypical metastatic cells. Meanwhile, he had neurological deterioration and developed quadriparesis with lower motor neuron (LMN) type bilateral VIIth cranial nerve and left-sided IIIrd cranial nerve palsies. A repeat contrast-enhanced MRI of the brain and CSF cytology were planned for confirmation of the diagnosis but he became dyspnoeic due to aspiration pneumonia and was electively tracheostomized. He was put on mechanical ventilation but subsequently deteriorated and eventually succumbed to his illness.
Discussion
Carcinomatous meningitis is defined as an infiltration of the leptomeninges (the arachnoid membrane and the pia mater) by malignant cells. This condition, synonymous with 'leptomeningeal carcinomatosis' (LCM) and 'meningeal carcinomatosis' , is an uncommon manifestation of solid malignancies [1][2][3]. Eberth is credited with the first published observation of leptomeningeal metastases in lung cancer in 1870 [4]. The multifocality of signs and symptoms is attributable to multiple tumour deposits throughout the neuraxis. Tumour cells enter this neuraxial space through multiple mechanisms such as direct extension from parenchymal disease through the pia mater, via arterial and venous channels by haematogenous spread, extension from bone either directly or through veins, seeding from subependymal disease, spillage into CSF cavities from surgery and retrograde invasion along peripheral nerves or their lymphatics [5][6][7]. The diagnosis of LCM is associated with poor prognosis with a reported median overall survival (OS) of 2.3-4.7 months in modern series [8].
The diagnosis is often difficult to establish despite strong clinical suspicion. Conventionally, the establishment of a definitive diagnosis requires the finding of malignant cells in the CSF on cytological examination, but at least 3 lumbar punctures may be required to ascertain the diagnosis. A single CSF examination reveals positive findings in approximately 50% of patients and this percentage rises to 85 to 90% after 3 procedures [7,9]. Cytological results remain negative in some patients despite repeated lumbar punctures. These false-negative findings may result from the strong adherence of malignant cells to the leptomeninges or due to the presence of focal rather than widespread leptomeningeal tumour [9]. Other CSF markers such as elevated protein levels, raised cell count, low glucose concentration, raised opening pressure and elevated levels of tumour markers may give an indication of the presence of LCM [7,9]. In the illustrative case, the CSF analysis showed increased protein concentration, low glucose, and an elevated white cell count (predominantly neutrophils). The first CSF cytology revealed scattered large atypical cells, suggestive of leptomeningeal metastases.
Contrast-enhanced MRI is the radiographic modality of choice for the diagnosis of LCM. The entire neuraxis must be imaged as multifocal involvement is common. In most patients, the MRI will reveal leptomeningeal enhancement that is frequently associated with cranial nerve enhancement and gross tumour deposits [11]. It may additionally include contrast enhancement of the sulci, basilar cisterns, cauda equina and hydrocephalus [3]. The sensitivity of Gadolinium (Gd) enhanced MRI is equivalent to that of CSF analysis. However, the specificity of Gd-MRI (77%) is lower than that of the CSF examination (100%). Hence, MRI could be of diagnostic value, especially when CSF cytology is negative. In such an instance, it can differentiate between patients with a low or high risk of LCM. Nevertheless, a negative MRI following negative CSF does not exclude LCM [13].
There is no consensus on the optimal management of patients with LCM. This is mainly because of the lack of large published experiences, limited number of randomized trials, nonuniform treatment regimens in single institution experiences and inclusion of various primary tumour histologies in the clinical trials. However, an aggressive central nervous system (CNS) directed treatment plan comprising radiation therapy (whole brain radiotherapy and/or focal spinal radiation to symptomatic sites) and intrathecal chemotherapy (methotrexate, cytarabine or thiotepa) should be considered in patients with good performance status. The commonly used palliative radiotherapy regimens are 20 Gray in 5 fractions over 1 week and 30 Gray in 10 fractions over 2 weeks with the latter being preferred in patients with relatively more favourable prognosis. Craniospinal irradiation to the entire neuraxis may be considered in select patients with diffuse leptomeningeal metastases. Usually, CNSdirected therapy is given in conjunction with tumourspecific systemic therapy in fit patients. In this context, drugs that cross the blood-brain barrier (intravenous high dose methotrexate, ifosfamide, thiotepa and oral temozolomide, capecitabine, small molecule tyrosine kinase inhibitors, e.g. gefitinib, erlotinib and lapatinib) may be considered depending upon tumour sensitivity. The incorporation of novel biological agents targeting individual tumour-specific mutation in the systemic and CNS-directed therapy is an innovative approach. Despite the aforementioned treatment approaches, the overall prognosis remains poor. Without any treatment, the OS is approximately 6 weeks, and with appropriate treatment, the median OS increases to approximately 3-6 months [1,3,9,11]. The patient, discussed in this report, had Eastern Cooperative Oncology Group (ECOG) performance status (PS) 3 at presentation which rapidly evolved to ECOG PS 4, which precluded the use of anticancer treatment. The option of intrathecal methotrexate monotherapy after the planned 3rd lumbar puncture was discussed with the patient's relatives but the patient succumbed to his illness before the procedure.
Head and neck squamous cell carcinomas (HNSCC) are considered curable malignancies, if diagnosed in the early stage. In locally advanced HNSCC, patterns of failure are usually local, regional or locoregional. Though distant metastases are relatively rare, they are difficult to cure. The reported incidence of distant metastases in HNSCC varies widely in the published literature and is usually between 10 and 25% [14,15]. The most common site of distant metastases is the lungs, accounting for half to two-thirds of all distant metastases, followed by bones and liver [14,15]. Leptomeningeal metastases from HNSCC have been scarcely described in the medical literature (Table 1) [16][17][18][19][20][21][22]. In the context of HNSCC, LCM has been mostly reported in cancers involving the lip, paranasal sinus and nasopharynx, due to the propensity for perineural spread and intracranial extension through the cribriform plate and skull base foramina [16,17,[20][21][22]. To our knowledge, we have described the first case of carcinomatous meningitis in a patient with SCC of the retromolar trigone. Retrograde perineural spread of cancer cells along the mandibular (V3) nerve from the recurrent tumour in the post-op bed, right masticator space and infratemporal fossa could represent a likely pathway of LCM in this patient.
Conclusion
Although locoregional failure is the most common pattern of failure in HNSCC in general, the incidence of distant metastasis is slowly increasing due to more effective locoregional disease control with the advancement of multimodal treatment strategies including surgery, radiotherapy and chemotherapy as well as improvement in diagnostic imaging. In the context of backache with sensorimotor deficit and autonomic dysfunction in a previously treated case of HNSCC, compressive myelopathy due to spinal metastasis is an important differential diagnosis. However, this brief report underpins the importance of awareness of leptomeningeal carcinomatosis as another likely possibility in this setting. Despite the grave prognosis and limited survival, early diagnosis of LCM may provide the cancer physicians a window to offer CNS-directed treatment, e.g. cranial or spinal RT and intrathecal chemotherapy for the palliation of symptoms and improvement of health-related quality of life.
|
v2
|
2022-11-08T06:17:39.841Z
|
2022-11-07T00:00:00.000Z
|
253382180
|
s2ag/train
|
ZBTB2 links p53 deficiency to HIF-1-mediated hypoxia signaling to promote cancer aggressiveness.
Aberrant activation of the hypoxia-inducible transcription factor HIF-1 and dysfunction of the tumor suppressor p53 have been reported to induce malignant phenotypes and therapy resistance of cancers. However, their mechanistic and functional relationship remains largely unknown. Here, we reveal a mechanism by which p53 deficiency triggers the activation of HIF-1-dependent hypoxia signaling and identify zinc finger and BTB domain-containing protein 2 (ZBTB2) as an important mediator. ZBTB2 forms homodimers via its N-terminus region and increases the transactivation activity of HIF-1 only when functional p53 is absent. The ZBTB2 homodimer facilitates invasion, distant metastasis, and growth of p53-deficient, but not p53-proficient, cancers. The intratumoral expression levels of ZBTB2 are associated with poor prognosis in lung cancer patients. ZBTB2 N-terminus-mimetic polypeptides competitively inhibit ZBTB2 homodimerization and significantly suppress the ZBTB2-HIF-1 axis, leading to antitumor effects. Our data reveal an important link between aberrant activation of hypoxia signaling and loss of a tumor suppressor and provide a rationale for targeting a key mediator, ZBTB2, to suppress cancer aggressiveness.
|
v2
|
2022-11-08T06:17:40.107Z
|
2022-11-07T00:00:00.000Z
|
253382430
|
s2ag/train
|
New treatment options for hormone receptor positive breast cancer in 2023.
PURPOSE OF REVIEW
Hormone receptor positive, human epidermal growth factor receptor 2 negative (HR+/HER2-) breast cancer does respond to chemotherapy but can be addressed with a better therapeutic index by using biologically modified endocrine therapy. The most pronounced recent successes were reached by antibody drug conjugates (ADCs).
RECENT FINDINGS
In early HR+/HER2- disease, adjuvant treatment escalations have taken place for high-risk patients using abemaciclib for the HR+ BRCA- subset and olaparib for HR+ BRCA+ patients. In metastatic spread, among all CDK (cyclin-dependent kinase) 4/6 inhibitors used for first-line therapy, only ribociclib improved overall survival in pre and postmenopausal patients. Palbociclib failed to demonstrate overall survival benefits. New options come up with oral selective oestrogen receptor degraders (SERDs) such as elacestrant, which will replace fulvestrant and is clinically important in combination therapies. ADCs, together with new patient categories such as HER2low or HER3+, enlarge the treatment portfolio and challenge the need of supportive care. The antitrophoblast antigen 2 (TROP2) ADC sacituzumab govitecan improves overall survival in heavily pretreated HR+/HER2- patients by 3.2 months. The best improvement of overall survival was shown bý trastuzumab deruxtecan in less pretreated HER2low (HER2 1+ or HER2 2+/no gene amplification) patients with a gained life span of 6 months.
SUMMARY
Real-world data on the sequence of different ADCs with similar payloads are needed to establish best possible treatment algorithms. All these new agents will find their place after CDK4/6 inhibitor-based endocrine combination therapy.
|
v2
|
2022-11-08T06:17:40.289Z
|
2022-11-07T00:00:00.000Z
|
253382761
|
s2ag/train
|
Numerical and experimental evaluation of low-intensity transcranial focused ultrasound wave propagation using human skulls for brain neuromodulation.
BACKGROUND
Low-intensity transcranial focused ultrasound (LI-tFUS) has gained considerable attention as a promising non-invasive neuromodulatory technique for human brains. However, the complex morphology of the skull hinders scholars from precisely predicting the acoustic energy transmitted and the region of the brain impacted during the sonication. This is due to the fact that different ultrasound frequencies and skull morphology variations greatly affect wave propagation through the skull.
PURPOSE
Although the acoustic properties of human skull have been studied for tFUS applications, such as tumor ablation using a multi-element phased array, there is no consensus about how to choose a single-element focused ultrasound (FUS) transducer with a suitable frequency for neuromodulation. There are interests in exploring the magnitude and dimension of tFUS beam through human parietal bone for modulating specific brain lobes. Herein, we aim to investigate the wave propagation of tFUS on human skulls to understand and address the concerns above.
METHODS
Both experimental measurements and numerical modeling were conducted to investigate the transmission efficiency and beam pattern of tFUS on 5 human skulls (C3 & C4 regions) using single-element FUS transducers with 6 different frequencies (150 kHz∼1500 kHz). The degassed skull was placed in a water tank and a calibrated hydrophone was utilized to measure acoustic pressure past it. The cranial computed tomography scan data of each skull was obtained to derive a high-resolution acoustic model (grid point spacing: 0.25 mm) in simulations. Meanwhile, we modified the power-law exponent of acoustic attenuation coefficient to validate numerical modeling and enabled it to be served as a prediction tool, based on the experimental measurements.
RESULTS
The transmission efficiency and -6 dB beamwidth were evaluated and compared for various frequencies. An exponential decrease in transmission efficiency and a logarithmic decrease of -6 dB beamwidth with an increase in ultrasound frequency were observed. It is found that a >750 kHz ultrasound leads to a relatively lower tFUS transmission efficiency (< 5%) while a <350 kHz ultrasound contributes to a relatively broader beamwidth (> 5 mm). Based on these observations, we further analyzed the dependence of tFUS wave propagation on FUS transducer aperture size.
CONCLUSIONS
We successfully studied tFUS wave propagation through human skulls at different frequencies experimentally and numerically. The findings have important implications to predict tFUS wave propagation for ultrasound neuromodulation in clinical applications, and guide researchers to develop advanced ultrasound transducers as neural interfaces. This article is protected by copyright. All rights reserved.
|
v2
|
2022-11-08T06:17:40.310Z
|
2022-11-07T00:00:00.000Z
|
253382696
|
s2ag/train
|
Adjuvant Abemaciclib Combined with Endocrine Therapy: Efficacy Results in monarchE Cohort 1.
The monarchE Cohort 1 patient population was enrolled based on high-risk clinicopathological features that can easily be identified as part of routine clinical breast cancer evaluation. Efficacy data from Cohort 1 demonstrate substantial evidence of benefit for adjuvant abemaciclib+ET in patients with HR+, HER2- early breast cancer at high risk of recurrence (ClinicalTrials.gov: NCT03155997 [monarchE]).
|
v2
|
2022-11-08T14:02:25.453Z
|
2022-11-07T00:00:00.000Z
|
253385264
|
s2orc/train
|
Changes in anticancer treatment plans in patients with solid cancer hospitalized with COVID-19: analysis of the nationwide BSMO-COVID registry providing lessons for the future
Background Solid cancer is an independent prognostic factor for poor outcome with COVID-19. As guidelines for patient management in that setting depend on retrospective efforts, we here present the first analyses of a nationwide database of patients with cancer hospitalized with COVID-19 in Belgium, with a focus on changes in anticancer treatment plans at the time of SARS-CoV-2 infection. Methods Nineteen Belgian hospitals identified all patients with a history of solid cancer hospitalized with COVID-19 between March 2020 and February 2021. Demographic, cancer-specific and COVID-specific data were pseudonymously entered into a central Belgian Society of Medical Oncology (BSMO)-COVID database. The association between survival and primary cancer type was analyzed through multivariate multinomial logistic regression. Group comparisons for categorical variables were carried out through a Chi-square test. Results A total of 928 patients were registered in the database; most of them were aged ≥70 years (61.0%) and with poor performance scores [57.2% Eastern Cooperative Oncology Group (ECOG) ≥2]. Thirty-day COVID-related mortality was 19.8%. In multivariate analysis, a trend was seen for higher mortality in patients with lung cancer (27.6% versus 20.8%, P = 0.062) and lower mortality for patients with breast cancer (13.0% versus 23.3%, P = 0.052) compared with other tumour types. Non-curative treatment was associated with higher 30-day COVID-related mortality rates compared with curative or no active treatment (25.8% versus 14.3% versus 21.9%, respectively, P < 0.001). In 33% of patients under active treatment, the therapeutic plan was changed due to COVID-19 diagnosis, most frequently involving delays/interruptions in systemic treatments (18.6%). Thirty-day COVID-related mortality was not significantly different between patients with and without treatment modifications (21.4% versus 20.5%). Conclusion Interruption in anticancer treatments at the time of SARS-CoV-2 infection was not associated with a reduction in COVID-related mortality in our cohort of patients with solid cancer, highlighting that treatment continuation should be strived for, especially in the curative setting.
INTRODUCTION
The COVID-19 pandemic has caused unprecedented strains on health care systems worldwide. Early medical responses and policy making relied on incomplete knowledge about the disease and its risk factors. Scientific efforts amidst this crisis, however, have since helped to improve insight so that policies can and could be adapted accordingly. Specifically in the field of oncology, it is now known that the pandemic impacted screening programs as well as diagnostic and therapeutic care planning, resulting in delayed care for many patients. 1,2 On an individual patient level, current or previous history of cancer increases the risk of SARS-CoV-2 infection, hospitalization with COVID-19 and even subsequent mortality, as recently reviewed extensively elsewhere. 3 Some factors that have been identified as poorly prognostic in the general population also apply to patients with solid cancer, such as male sex, older age, history of smoking, lower overall performance status and higher number of comorbidities. [4][5][6][7] Many risk factors, however, distinguish oncology patients from the general population as well. Patients with lung cancer, and in some reports patients with lung metastases, have been found to have higher mortality rates from COVID-19 compared with patients with other primary solid tumour types. 4,5,8,9 This might reflect how their disease and/or risk factors (e.g. smoking) negatively affect respiratory capacity. Likewise, systemic cancer treatments such as chemotherapy may disrupt the immune response to infectious diseases. 3 The role of cancer immunotherapy is less clear. On top of this, seroconversion after vaccination for SARS-CoV-2 appears to be lower in cancer patients, and even more so in those with metastatic disease and those treated with chemotherapy, steroids or cyclin-dependent kinase 4/6 (CDK4/6) inhibitors. 10 Nationwide and even worldwide registries focusing on patients with cancer and COVID-19 specifically, such as the COVID-19 and Cancer Consortium (CCC19) database, TERAVOLT and ESMO CoCare initiatives are essential to fill the remaining knowledge gaps. As data collection on outcomes in patients hospitalized with COVID-19 in Belgium was streamlined through the national health institution Sciensano, the Belgian Society of Medical Oncology (BSMO) decided to contribute to that growing knowledge by setting up a population-based nationwide study on the interplay between cancer and COVID-19. The first results of that effort focused on the comparison between patients with and without solid cancer and highlighted an important increase in in-hospital mortality among patients with solid cancer (31.7% versus 20.0%, respectively; adjusted odds ratio 1.34; 95% confidence interval 1.13-1.58). 11 Here we present the results of the second part of the study, where additional information was collected on patients with cancer and COVID-19 hospitalized in Belgium. Our aim was to investigate possible differences in outcomes between patients with different primary tumour types, between patients under active treatment versus those without, according to treatment received, and according to changes in the oncological treatment plans due to COVID-19.
Patient identification
Adults (individuals aged 18 years) hospitalized with COVID-19 in Belgium between 1 March 2020 and 1 February 2021 registered in the Sciensano COVID-19 database and identified by each participating institution as having a prior or current solid cancer were eligible for inclusion in the study. Patients with a haematological malignancy, but without a solid tumour and patients with only non-melanoma skin cancer were not included. Diagnosis of COVID-19 was based on a molecular test (PCR) and/or chest computed tomography imaging.
Data collection
A central electronic case report form (eCRF) was created using REDCap®. Data on patient demographics, comorbidities, primary cancer diagnosis, anticancer treatments previously received, anticancer treatment changes due to COVID-19 and survival outcome were retrospectively registered by the participating institutions in this central database. Standardized data reported by the hospitals on COVID-19 baseline characteristics (signs, symptoms, laboratory values) and COVID-19 disease evolution, treatment and outcomes during hospitalization were sent back by Sciensano directly to each reporting hospital from their central database. 12 Within each hospital, data were pseudonymized by the participating investigators, and then merged with the REDCap® eCRF.
Statistical analysis
Descriptive analyses of baseline patient, cancer and treatment characteristics were carried out and are provided as frequencies with percentages for categorical variables or mean with standard deviation/median with interquartile range (IQR) for continuous variables. Survival outcomes defined for analysis were overall mortality at 30 days after COVID-19 diagnosis (30-day mortality) and overall mortality at 3 months after COVID-19 diagnosis (3-month mortality), and analogously defined COVID-19-related and non-COVID-19-related 30-day and 3-month mortality. For subgroup analyses, an 'active cancer' subgroup was defined including patients undergoing active anticancer treatment at or within 90 days before diagnosis of COVID-19, as well as patients under palliative care only (without specific anticancer treatment). Active anticancer treatment was defined as any form of locoregional or systemic treatment given with the purpose of treating cancer and/or preventing cancer relapse. Subgroup analysis was also carried out for the first and second wave of the COVID-19 pandemic in Europe separately, with a cut-off on 30 June 2020.
Group comparisons with regards to categorical variables were carried out by means of a Chi-square test. The association between (non-)COVID-19-related death with cancer type, correcting for possible confounders was analyzed by means of multivariate multinomial logistic regression models including age, gender, Eastern Cooperative Oncology Group (ECOG) performance status, body mass index, Charlson comorbidity index (which includes the presence or absence of metastatic cancer disease at COVID-19 diagnosis), haematological malignancies and ICU admission as covariates. All P values reported are twosided, and P values <0.05 were considered to reflect statistical significance. All analyses were carried out using SAS software (SAS Institute, Cary, NC) (version 9.4 of the SAS System for Windows).
RESULTS
A total of 19 Belgian hospitals involved in cancer care received approval for the study by their local ethics committee and subsequently participated in patient registration. In total, 928 patients with solid cancer hospitalized with COVID-19 during the predefined study period were registered and available for subsequent analysis upon database lock on 26 April 2022. A descriptive analysis of baseline patient characteristics can be found in Supplementary Table S1, available at https://doi.org/10.1016/j. esmoop.2022.100610. Median age at COVID-19 diagnosis was 73 years (IQR 64-81 years) and most of the patients were of male sex (57.1%). ECOG performance status was 2 in 57.2% of patients, and median Charlson comorbidity index was 7 (IQR 6-9). Almost half of the patients (n ¼ 457, 49.3%) had received any type of anticancer treatment within 3 months of COVID-19 diagnosis. This included chemotherapy in 43.1%, endocrine treatment in 23.0%, immunotherapy in 17.1% and targeted treatment in 11.4%. Most of these patients (62.3%) received treatment in a noncurative setting. The main reason for hospitalization was the clinical status of the patient in 68.0% (n ¼ 631), whereas 27.6% (n ¼ 256) of patients were hospitalized for another reason but concomitantly got diagnosed with COVID-19 and 3.1% (n ¼ 29) were hospitalized out of precaution.
Thirty-day overall and COVID-19-related mortality in all patients assessable for survival analysis (n ¼ 895) were 26.8% and 19.8%, respectively (Table 1). COVID-19-related 30-day mortality was numerically higher in the first compared with the second wave of the pandemic, though this observed difference did not reach statistical significance (22.1% versus 17.1%, P ¼ 0.13). In multivariate analysis, primary tumour type did not influence COVID-19-related mortality (Supplementary (Figure 2). Most frequent treatment changes were delay/interruption (20.1% and 16.0% for noncurative and curative setting, respectively) and full cancellation (10.6% and 6.1% for non-curative and curative setting, respectively) of systemic treatment. Within the subgroup of patients receiving chemotherapy at the time of COVID-19 diagnosis, almost half of them experienced changes in their chemotherapy treatment plan (93/195, 47.7%). This involved mainly delays/interruptions due to COVID-19 diagnosis (n ¼ 62), with interruptions for 1 month in almost half of all patients where chemotherapy delay time was known, and even full cancellation in 29 patients. For patients already under systemic treatment in the curative setting at the time of COVID-19 diagnosis, cancellation of systemic treatment was observed in only 9.3% (10/108). Fourteen patients under active treatment were planned to undergo surgery, whereas only six received it as planned. Seventy-eight patients were receiving immunotherapy at the time of COVID-19 diagnosis, and this treatment was interrupted in 10 and cancelled in 9. The main reasons for changes in treatment plan were COVID-19related complications, followed by fear for/existence of anticancer treatment-related toxicity such as neutropenia (in 77.6% and 14.3% of patients with treatment changes, No active treatment <3 months Curative treatment <3 months Non-curative treatment <3 months respectively). Neither 30-day COVID-19-related mortality nor 3-month non-COVID-19-related mortality were significantly different between patients with and without treatment modifications (21.4% versus 20.5%, P ¼ 0.18 and 18.8% versus 17.0%, P ¼ 0.89, respectively) ( Table 3).
DISCUSSION
The impact of COVID-19 on patients with cancer and their treatments has been ubiquitous and unprecedented. Knowledge on how the disease course is influenced by the patients' oncological setting is needed to guide treatment decisions; however, patients with malignancies have largely been excluded from prospective clinical trials in this setting. We thus rely on retrospective and real-world data analysis to learn for the future. The BSMO-COVID database is a large (w1000 patients) nationwide database of patients with solid cancer hospitalized with COVID-19, set up to help answer some of the remaining questions on this topic. To enhance the leverage of our database, our data were transferred to the ESMO CoCare initiative on 12 April 2022. While prognostic analyses presented in this paper are thus limited, the results from analyses on that merged dataset will be reported separately.
The European OnCovid and American Society of Clinical Oncology (ASCO) registries, including patients with a history of solid or haematological cancer and a diagnosis of SARS-CoV-2 infection, already reported major differences in mortality rates between what we here defined as the first and second waves of the pandemic. 13,14 The same trend was seen in our hospitalized population, likely reflecting an improvement in diagnosis and management of patients with COVID-19, thanks to the increasing knowledge on the disease. Our database only covered patients hospitalized up until February 2021, expected to be infected mainly with alpha and delta variants of the SARS-CoV-2 virus. We assume mortality rates will have continued to evolve favourably for patients with cancer, as the omicron variant has become dominant later and vaccines have become available to most of our patients (primary course vaccination campaign in Belgium started in the first quarter of 2021, with coverage in adults now being 89%). 15 Of note, our database included mostly patients with poor ECOG performance scores and many of older age, as would be expected in a population hospitalized mainly because of their clinical status upon infection, so mortality rates should not be generalized to ambulatory and fitter patients.
Many early reports and guidelines have focused on delivering cancer care in challenging pandemic times. [16][17][18][19][20][21][22] The goal of these guidelines was to maintain high-quality oncological management while mitigating the risk of hospital-acquired SARS-CoV-2 infection for both patients and caregivers. Our focus was different, as for patients under active anticancer treatment who do have COVID-19, no clear guidelines for treatment modifications exist, but mostly expert recommendations. Although many patients experienced delays or even cancellations of their anticancer treatment plans, no significant difference in 30-day COVID-19-related mortality was observed between those with and without treatment changes in the population included in our study. This is in line with the observations in the OnCovid dataset, where only permanent treatment discontinuation resulted in increased risk of medium-term death, but regimen adjustments did not. 23 In contrast with our database, OnCovid also included ambulatory patients and haematological patients, and did not look at COVID-19-related mortality separately. Although treatment delays thus did not reduce COVID-19 mortality in our study, it is expected and modelled by others that these delays could result in a late increase in cancer-related mortality. 24 Our findings are therefore reassuring for the future, highlighting that treatment discontinuations upon COVID-19 diagnosis should be avoided where possible. This is of utmost relevance for the subsets of patients where cancer diagnosis influences their long-term prognosis, but not necessarily their COVID-19-related mortality. Factors to take into account in assessing that risk balance are the following: firstly, for most primary solid tumour types excluding lung cancer, no association has been proven with increased COVID-19-related mortality. 3 Secondly, some non-immunemodifying treatment regimens, such as endocrine treatment, have not been shown to influence COVID-19 outcomes. 3 Thirdly, seroconversion rates after SARS-CoV-2 vaccination tend to be lower in patients with solid cancer compared with the general population. 10 Lower vaccine efficacy is thought, however, to be restricted to only some subgroups of oncology patients. Although these subgroups are not yet well defined, rapidly advancing knowledge might teach us which patients are as protected against severe COVID-19 disease course as patients without a malignancy after full vaccination. And lastly, for patients in the curative setting long treatment discontinuation or even cancellation might jeopardize their chances of cure and impact long-term outcomes.
ESMO Open
Executive Board for their full support of this research; and the Belgian National Health Institution Sciensano for providing the required data back to the institutions. This study was an investigator-initiated trial conducted by the Belgian Society of Medical Oncology as legal sponsor, independently of the funding sources.
ROLE OF THE FUNDER
The funders of the study had no role in the design and conduct of the study, data collection, data management, statistical analysis and interpretation of the data, the preparation, review or approval of the manuscript and the decision to submit the manuscript for publication.
DISCLOSURE
MB: speaker honoraria and travel grants from Roche/GNE; research grants to her institution from Roche/GNE, AstraZeneca, GlaxoSmithKline (GSK)/Novartis and Servier. JC: advisory board and lectures from Amgen, Servier, Bayer, Novartis, Pfizer, Celgen, Ipsen (paid to institution); travel grants from Roche, Pfizer, Amgen, Novartis. CVM: consulting fees from Chrysalis Biomed (paid to institution); support for attending meetings and/or travel from Pfizer, Roche, Merck; participation of institution on advisory board for Eli Lilly and AstraZeneca. WD: payment or honoraria for lectures, presentations, speakers' bureaus, manuscript writing or educational events from Merck, Pfizer. CF: support for attending meetings and/or travel from Eli Lilly, Gilead, PharmaMar, Pfizer, Roche; participation on advisory board for Eli Lilly, GSK. SA: consulting fees from Galapagos; payment or honoraria for lectures, presentations, speakers bureaus, manuscript writing or educational events from Merck Sharp & Dohme (MSD), Sanofi, Roche, Bristol Myers Squibb (BMS); support for attending meetings and/or travel from MSD, Sanofi, Roche, BMS, Pfizer; participation on advisory board from MSD, Sanofi, Roche, BMS; receipt of equipment, materials, drugs, medical writing, gifts or other services from BMS. HVDB: participation on advisory board from AstraZeneca. MI: grants from Roche, Pfizer, Natera Inc. (paid to institution); consulting fees from Novartis; payment or honoraria for lectures, presentations, speakers bureaus,
|
v2
|
2022-11-08T14:02:46.207Z
|
2022-11-07T00:00:00.000Z
|
253386209
|
s2orc/train
|
Editorial: Neuroinflammation in acquired epilepsy
In spite of astonishing advances in epilepsy research and treatment over the past few decades, epilepsy remains one of the most common and devastating brain diseases and still affects approximately 65 million people globally (Devinsky et al., 2018). In addition to their wide-ranging side effects, antiseizure drugs (ASDs) are not effective in controlling seizures in more than 30% of patients who have pharmacoresistant epilepsy (Janmohamed et al., 2020). It is rather unfortunate that current medications provide merely symptomatic relief and have not been demonstrated to prevent epilepsy in people at risk or modify the disease progression (Galanopoulou et al., 2021). Therefore, there remains an urgent need for alternative antiepileptic treatments, despite the rapid expansion of modern ASDs that emerged during the first 2 decades of this century (Varvel et al., 2015; Löscher and Klein, 2020; Yu et al., 2022). Mounting lines of evidence support an essential role for proinflammatory mediators in the brain in acquired epileptogenesis, a pathogenic process that is proposed to transform a normal brain to one generating seizures following various brain insults, such as de novo status epilepticus (SE), brain infections, traumatic brain injuries, brain tumors, and strokes (Yu et al., 2019; Korgaonkar et al., 2020; Terrone et al., 2020). It has been widely proposed that modulating key proinflammatory mediators might disrupt the epileptogenic processes and lead to modification and/or even prevention of epilepsy. In the current Research Topic, we bring together a diverse collection of primary research and review articles that highlight the roles of brain inflammation in epilepsy of various etiologies and pathogeneses. Both inflammation and oxidative stress are well known for their pathophysiological roles in the epileptic brain. However, they are often studied as separate entities despite the evidence that the redox-based signaling cascades and inflammatory reactions have extensive crosstalk (Fabisiak and Patel). Recent studies have uncovered a variety of mechanisms whereby oxidative stress and neuroinflammation greatly influence each other in the context of epilepsy. For instance, neuroinflammation can be regulated by OPEN ACCESS
Editorial on the Research Topic Neuroinflammation in acquired epilepsy
In spite of astonishing advances in epilepsy research and treatment over the past few decades, epilepsy remains one of the most common and devastating brain diseases and still affects approximately 65 million people globally (Devinsky et al., 2018). In addition to their wide-ranging side effects, antiseizure drugs (ASDs) are not effective in controlling seizures in more than 30% of patients who have pharmacoresistant epilepsy (Janmohamed et al., 2020). It is rather unfortunate that current medications provide merely symptomatic relief and have not been demonstrated to prevent epilepsy in people at risk or modify the disease progression (Galanopoulou et al., 2021). Therefore, there remains an urgent need for alternative antiepileptic treatments, despite the rapid expansion of modern ASDs that emerged during the first 2 decades of this century (Varvel et al., 2015;Löscher and Klein, 2020;Yu et al., 2022).
Mounting lines of evidence support an essential role for proinflammatory mediators in the brain in acquired epileptogenesis, a pathogenic process that is proposed to transform a normal brain to one generating seizures following various brain insults, such as de novo status epilepticus (SE), brain infections, traumatic brain injuries, brain tumors, and strokes (Yu et al., 2019;Korgaonkar et al., 2020;Terrone et al., 2020). It has been widely proposed that modulating key proinflammatory mediators might disrupt the epileptogenic processes and lead to modification and/or even prevention of epilepsy. In the current Research Topic, we bring together a diverse collection of primary research and review articles that highlight the roles of brain inflammation in epilepsy of various etiologies and pathogeneses.
Both inflammation and oxidative stress are well known for their pathophysiological roles in the epileptic brain. However, they are often studied as separate entities despite the evidence that the redox-based signaling cascades and inflammatory reactions have extensive crosstalk (Fabisiak and Patel). Recent studies have uncovered a variety of mechanisms whereby oxidative stress and neuroinflammation greatly influence each other in the context of epilepsy. For instance, neuroinflammation can be regulated by This is an open-access article distributed under the terms of the Creative Commons Attribution License (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.
Frontiers in Cell and Developmental Biology frontiersin.org transcription factors such as NF-κB and nrf2 that are activated by reactive oxygen species (ROS). Neuroinflammation in turn can induce the expression and activity of NADPH oxidase (NOX), fostering a highly oxidative environment. Moreover, the oxidative and proinflammatory mediators can moulate distinct intracellular pathways expressed in different cell-types, exemplified by NOX-2 dependent increase in ROS in neurons and astrocytes triggered by SE, and myeloid differentiation primary response 88 (MyD88) dependent glial activation through Toll-like receptors (TLRs) (Almeida et al.). The reviews presented in this collection highlight how signaling crosstalk between neuroinflammation and oxidative stress and their cell type specific roles may be leveraged for novel therapeutic strategies for epilepsy. Neuroinflammatory processes triggered by acute brain insults such as SE are highly regulated and show time-and age-dependency. Using a rat model of kainate-induced SE, Erisken et al. demonstrate prolonged induction of many key inflammatory genes, particularly those associated with stressactivated protein kinases, p38 and JNK signaling pathways, uniquely in adult brains. In contrast, many of the same genes show relatively transient expression in developing brains under similar experimental conditions, suggesting that the immature brains might be more resistant to SE-induced cell death and neuropathology. In line with findings in adult animals, hippocampal tissues from mesial temporal lobe epilepsy patients showed upregulation of inflammationrelated genes. These results highlight the association between uncontrolled neuroinflammation and epileptogenesis and suggest that epileptic seizures might result from prolonged activation of neuroimmune processes beyond the homeostatic threshold.
Brain infection is a leading cause of epilepsy, but the underlying molecular mechanisms are poorly understood. Patel et al. use Theiler's murine encephalomyelitis virus (TMEV) infection to generate acute brain inflammation and the subsequent spontaneous seizures in mice. They show TMEV infection-induced seizures likely due to impaired GABAergic inhibition, secondary to alterations in neuronal intracellular chloride regulation. Their results further suggest that the brain-derived neurotrophic factor (BDNF) might contribute to the development of brain infection-triggered seizures by reducing the expression of K + /Cl − cotransporter 2 (KCC2). This has the potential to enhance accumulation of intracellular chloride and increases excitability by rendering GABA depolarizing instead of hyperpolarizing as observed in chemoconvulsant models of SE (Pathak et al., 2007;Yu et al., 2013). Notably, the upregulation of brain BDNF observed in TMEV-infected mice has also been found in chemoconvulsant models of SE (Zhu et al., 2012;Thomas et al., 20162016;Yu and Jiang 2020), and is believed to contribute to acquired epileptogenesis by acting on its high-affinity receptor, the tropomyosin related kinase B (TrkB) (Lin et al., 2020).
Exposure to diisopropylfluorophosphate (DFP), a structural analog of type G chemical warfare agents (e.g., sarin and soman), is well known to induce SE, gliosis, neuronal death, and eventually the development of spontaneous recurrent seizures in rodents. Gage et al. show that both male and female rats which experience DFP-induced SE develop unique regions of glial scarring in the piriform cortex and amygdala, but not in the hippocampus. DFP-induced cortical glial scars are characterized by a massive clustering of reactive microglia, with increase in Iba1-and CD68-positive cells, surrounded by hypertrophic astrocytes and a decrease in NeuN-positive neurons in the scar core. Although female rats have been shown to require a higher dose of DFP to induce SE when housed in a room with only females, Rao et al. demonstrate that when both sexes are housed in the same room and administered the same DFP solution, SE severity was not different between sexes. These results reinforce the importance of sex as a key biological variable in experimental design and suggest that housing animals of both sexes together and using the same batch of test reagents will reduce experimental variability.
The benefits of low-intensity physical exercise to the CNS have been shown in animal models and patients with neurological diseases, such as Alzheimer's disease, Parkinson's disease, stroke, epilepsy, multiple sclerosis, anxiety and depression (Allendorfer and Bamman, 2018). Jia et al. demonstrate that the conventional ASD, valproate, combined with low-intensity exercise can reduce seizures and associated comorbidities in kainate-treated mice. The reduction in seizure burden appears to be correlated with the suppression of inflammatory cytokines (IL-1β, IL-6, and TNF-α) and the immune receptor TLR4 in the hippocampus. Given that TLR4 is involved in epileptogenesis of diverse etiologies (Maroso et al., 2010;Korgaonkar et al., 2020), these findings suggest that the non-pharmacological intervention like lowintensity exercise might reduce neuroinflammation and provide an adjunctive strategy to enhance efficacy of conventional ASDs to treat epilepsy.
Developing preventive treatment for epilepsy is challenging because it is currently impossible to identify individuals that will develop epilepsy after initial precipitating brain insults. Theoretically, biomarkers that identify "at risk" individuals would facilitate the development of potential antiepileptogenic treatment (Simonato et al., 2021). By reviewing data from 60 human patients with focal epilepsy of autoimmune etiology, Sakamoto et al. propose a diagnostic algorithm that might help to predict the underlying autoimmune etiology of epilepsy before antibody testing results become available. Over 30% of epilepsy patients suffer from pharmacoresistant seizures associated with cognitive and psychiatric co-morbidities. Analyzing a microarray dataset from the Gene Expression Omnibus database, Min et al. identify 25 genes differentially expressed in the peripheral blood of patients with valproate resistance in epilepsy and significantly enriched in T-cell Frontiers in Cell and Developmental Biology frontiersin.org receptor recognition. While the potential confound posed by the differential seizure burden between valproate sensitive and resistant groups needs to be considered, these findings suggest that the peripheral blood T-cells and the differentially expressed genes could serve as biomarkers for refractory epilepsy. Identification of reliable biomarkers for diverse types of epilepsy and pharmacoresistance could facilitate both early diagnosis and development of new therapies, needed to achieve the ultimate goals of "no seizures, no side effects, and no co-morbidities" in epilepsy treatment. The series of articles presented here address diverse ways in which neuroinflammation could shape acquired epilepsy and offers insights into how these processes may be leveraged to inform mechanisms of epileptogenesis, identify biomarkers, and to develop novel strategies for disease modification and treatment in epilepsy.
|
v2
|
2022-11-09T06:16:56.924Z
|
2022-11-07T00:00:00.000Z
|
253397057
|
s2ag/train
|
Delayed oral feeding reduces pharyngocutaneous fistula formation after open surgical treatment of primary hypopharyngeal cancer: A case-control study.
OBJECTIVES
Pharyngocutaneous fistula (PCF) formation following open surgical treatment of hypopharyngeal cancer (HPC) is a common and troublesome complication. To date, the postoperative protocol of restarting oral intake is not clear, and vast discrepancies exist in the literature and among institutions. This study aimed to explore the impact of a postoperative protocol of restarting oral intake on PCF formation after open surgical treatment of primary HPC, and its impact on overall survival (OS) and swallowing function based on the functional outcome swallowing scale (FOSS).
MATERIALS AND METHODS
This was a prospective observational study of 42 patients who received open surgical treatment for primary HPC at Beijing Friendship Hospital between April 2019 and August 2021. This cohort included two groups: patients who restarted oral intake on the 10th postoperative day (Group 1), and those who started on the 20th (Group 2). The Chi-square test and Fisher's exact chi-squared test were used for comparing qualitative data among the groups.
RESULTS
Group 1 (n = 27) and Group 2 (n = 15) were comparable in clinical characteristics. PCF occurred in 7 (25.9%) patients in Group 1, while none occurred in Group 2 (P = 0.038). The 2-year OS of all 42 patients was 75.6%; 65.8% and 93.3% for Groups 1 and 2, respectively (P = 0.07). The swallowing function was satisfactory (FOSS Grades 0-III) for 19 (70.4%) patients in Group 1 and 15 (100%) patients in Group 2 (P = 0.035). Laryngeal preservation was achieved in 25 (59.5%) patients, while decannulation was successful in 22 (88.0%) patients.
CONCLUSIONS
Delayed oral feeding significantly reduces PCF after open surgical treatment of primary HPC, and improves the swallowing function outcome without jeopardizing the OS.
|
v2
|
2022-11-09T14:16:59.645Z
|
2022-11-07T00:00:00.000Z
|
253399508
|
s2orc/train
|
Radiation therapy induces immunosenescence mediated by p90RSK
Radiation therapy (RT) to the chest increases the patients’ risk of cardiovascular disease (CVD). A complete understanding of the mechanisms by which RT induces CVD could lead to specific preventive, therapeutic approaches. It is becoming evident that both genotoxic chemotherapy agents and radiation induce mitochondrial dysfunction and cellular senescence. Notably, one of the common phenotypes observed in cancer survivors is accelerated senescence, and immunosenescence is closely related to both cancer risk and CVD development. Therefore, suppression of immunosenescence can be an ideal target to prevent cancer treatment-induced CVD. However, the mechanism(s) by which cancer treatments induce immunosenescence are incompletely characterized. We isolated peripheral blood mononuclear cells (PBMCs) before and 3 months after RT from 16 thoracic cancer patients. We characterized human immune cell lineages and markers of senescence, DNA damage response (DDR), efferocytosis, and determinants of clonal hematopoiesis of indeterminant potential (CHIP), using mass cytometry (CyTOF). We found that the frequency of the B cell subtype was decreased after RT. Unsupervised clustering of the CyTOF data identified 138 functional subsets of PBMCs. Compared with baseline, RT increased TBX21 (T-bet) expression in the largest B cell subset of Ki67–/DNMT3a+naïve B cells, and T-bet expression was correlated with phosphorylation of p90RSK expression. CD38 expression was also increased in naïve B cells (CD27–) and CD8+ effector memory CD45RA T cells (TEMRA). In vitro, we found the critical role of p90RSK activation in upregulating (1) CD38+/T-bet+ memory and naïve B, and myeloid cells, (2) senescence-associated β-gal staining, and (3) mitochondrial reactive oxygen species (ROS) after ionizing radiation (IR). These data suggest the crucial role of p90RSK activation in immunosenescence. The critical role of p90RSK activation in immune cells and T-bet induction in upregulating atherosclerosis formation has been reported. Furthermore, T-bet directly binds to the CD38 promoter region and upregulates CD38 expression. Since both T-bet and CD38 play a significant role in the process of immunosenescence, our data provide a cellular and molecular mechanism that links RT-induced p90RSK activation and the immunosenescence with T-bet and CD38 induction observed in thoracic cancer patients treated by RT and suggests that targeting the p90RSK/T-bet/CD38 pathway could play a role in preventing the radiation-associated CVD and improving cancer prognosis by inhibiting immunosenescence.
Radiation therapy (RT) to the chest increases the patients' risk of cardiovascular disease (CVD). A complete understanding of the mechanisms by which RT induces CVD could lead to specific preventive, therapeutic approaches. It is becoming evident that both genotoxic chemotherapy agents and radiation induce mitochondrial dysfunction and cellular senescence. Notably, one of the common phenotypes observed in cancer survivors is accelerated senescence, and immunosenescence is closely related to both cancer risk and CVD development. Therefore, suppression of immunosenescence can be an ideal target to prevent cancer treatmentinduced CVD. However, the mechanism(s) by which cancer treatments induce immunosenescence are incompletely characterized. We isolated peripheral blood mononuclear cells (PBMCs) before and 3 months after RT from 16 thoracic cancer patients. We characterized human immune cell lineages and markers of senescence, DNA damage response (DDR), efferocytosis, and determinants of clonal hematopoiesis of indeterminant potential (CHIP), using mass cytometry (CyTOF). We found that the frequency of the B cell subtype was decreased after RT. Unsupervised clustering of the CyTOF data identified Introduction Radiation therapy (RT) is an important component of managing primary thoracic malignancies, e.g., lung cancer, breast cancer and lymphoma. The delayed effects of RT on the cardiovascular system causes substantial morbidity and mortality in cancer survivors (1). Cardiovascular disease (CVD) is a leading cause of premature morbidity and mortality among breast cancer (2)(3)(4), Hodgkin or non-Hodgkin lymphoma (5)(6)(7)(8), and lung cancer (9, 10) survivors more than 5 years after their diagnosis and treatment. RT to the chest in Hodgkin lymphoma increases the patients' risk of CVD by about sevenfold, based on the results of multivariate analyses controlling for other risk factors (11). The incidence of major coronary events increases linearly by 7.4% per Gy in breast cancer survivors (1). Modern RT techniques significantly decrease the doses of radiation to the heart, but in breast cancer patients, the heart may still receive doses of 1-5 Gray (Gy) (1). A more complete understanding of the mechanisms by which RT induces CVD could lead to specific preventive therapeutic approaches.
Risk factors for CVD are generally greater for cancer survivors than the general population, with higher values for mean body mass index, C-reactive protein and low density lipoprotein cholesterol (12). One of the common phenotypes observed in cancer survivors is accelerated senescence (12)(13)(14). It is becoming evident that genotoxic chemotherapy agents and radiation induce mitochondrial dysfunction and cellular senescence, i.e., stress-induced premature senescence (SIPS). DNA damaging agents can cause SIPS in a relatively short period (usually 3-10 days) with or without significant telomere shortening (15). Most DNA damage is repaired by DNA damage response (DDR) mechanisms within 24 h after stress (16), but, in contrast, telomeric DNA damage persists for months (17). Therefore, telomeric DNA damage-induced SIPS may be a feasible explanation for the late or delayed CVD effects triggered by cancer treatments.
Although cellular senescence can lead to cell cycle arrest, senescent cells remain metabolically active and secrete a variety of factors, including inflammatory cytokines, chemokines, growth factors, angiogenic factors, nitric oxide (NO), prostaglandin E2 (PGE2), and reactive oxygen species (ROS), constituting a senescence-associated secretory phenotype (SASP). To maintain tissue homeostasis, removing senescent cells promptly is crucial. The SASP cytokines can recruit immune cells to clear the senescent cells by efferocytosis. On the other hand, "immunosenescence" refers to the decline in the immune response with aging (18), including impaired clearance of senescent cells. The critical role of immunosenescence in regulating both tumorigenesis (19) and CVD (20) has been suggested, but the exact regulatory mechanism remains unclear.
As cells age or are exposed to mutagens, they may acquire mutations that lead to clonal expansion without malignant transformation. In the hematopoietic system, this process is known as clonal hematopoiesis of indeterminant potential (CHIP). In over 70% of people with CHIP, somatic mutations of TET2, DNMT3A, ASXL1, and JAK2 are commonly observed, and all but one (JAK2) are loss of function mutations (21,22). CHIP is associated with a more than twofold increase in risk of CVD (22,23), which is likely mediated through inflammatory pathways (22,24). Among survivors of solid malignancies, CHIP-associated mutations are common and are detectable in around 30% of individuals in their 7th decade of age, and may adversely impact overall survival (25). Oncologic therapies, such as RT, are frequently doselimited based on toxicity to normal organs-organs that may already have impaired function in the presence of CHIPassociated mutations (22,(26)(27)(28). It is undefined whether the presence of CHIP in patients undergoing RT for thoracic malignancies contributes to an increased risk of radiotherapyinduced CVD.
The critical role of NAD + depletion during aging in regulating metabolic dysfunction is well established (29). Previously, we reported that NAD + was depleted after IR via PARP activation, triggering persistent PISP in myeloid cells (30). In addition to PARP activation, CD38 may be involved in NAD + depletion during aging (31). CD38 is a NAD glycohydrolase, and the levels and activity of CD38 were significantly increased by aging, leading to NAD + depletion (31). CD38 is expressed mainly in immune cells, and recently, Chini et al. have reported that senescence-induced inflammation promoted CD38 accumulation and caused NAD + depletion (32). Taken together, these reports suggested that both PARP activation and CD38 expression promote NAD + depletion during the aging process.
T-bet (official gene name: TBX21) is an immune cellspecific member of the T-box family of transcription factors. In B cells, antigen stimulation activates B cell proliferation and induces affinity maturation and class switch recombination to produce high-affinity antibodies, which show different immune effector functions (33). T-bet induces Iγ2A transcripts and promotes IFNγ-mediated class-switching to the IgG2a isotype, which plays a critical role in the pathogenesis of humoral autoimmunity and protection against pathogens (34). Furthermore, T-bet promotes the survival of antigen-specific IgG2a + /CD38 hi memory B cells by upregulating mature B cell receptor transcription (35). Also, T-bet drives the migration of memory IgG2a + B cell to sites of inflammation by increasing CXCR3 expression (36). Importantly, the age-associated B cell (ABC) is characterized by T-bet expression and increases overall inflammatory state (37). Although systemic T-bet depletion inhibited atherosclerosis formation (38), the role of T-bet in B cells on atherosclerosis formation remains unclear. Previously, we reported the role of p90RSK activation in SASP after RT. IR increases p90RSK activation in myeloid cells and forms a nucleus-mitochondria positive feedback loop with p90RSK-mediated ERK5 S496 phosphorylation, leading to persistent mitochondrial (mt) ROS production and subsequent telomere DNA damage. We found that poly (ADP-ribose) polymerase activation induced by telomere DNA damage promoted nicotinamide mononucleotide (NAD + ) depletion and consequent mitochondrial dysfunction, which elicited mtROS production, activated redox-sensitive kinase of p90RSK, and formed a positive feedback loop. Therefore, p90RSK activation played a crucial role in inducing SASP.
Although RT may induce immunosenescence (39), the mechanisms remain incompletely defined. Therefore, in this study, we use multiparameter mass cytometry (CyTOF) to explore the association of RT with various senescence-related factors and determinants of senescence including DDR -related markers, efferocytosis, and CHIP-associated drivers in PBMCs from thoracic cancer patients before and 3 months after RT.
Patients selection
The study was approved by the Institutional Review Board of the University of Texas MD Anderson Cancer Center (#PA16-0971). The inclusion criteria of this study were as follows: (1) ≥ 18 years of age, (2) Karnofsky Performance Scale ≥ 70, (3) To receive radiation therapy (RT) with computed tomography (CT)-based treatment planning that will involve delivery of dose to the heart, (3) An estimated mean heart dose of at least 6 Gy and mean left ventricular dose of at least 5 Gy, as estimated by the treating radiation oncologist at the time of simulation.
The exclusion criteria of this study were as follows: (1) Unable or unwilling to give written informed consent, (2) Previous history of RT to the thorax or breast, (3) Allergy to gadolinium, (4) Implanted device that is non-MRI compatible, including cardiac devices and neurostimulators, (5) Pregnant or breast-feeding, (6) Atrial fibrillation or frequent ventricular or atrial premature beats, (7) GFR < 60 mL/min according to the Modification of Diet in Renal Disease equation, (8) Personal history of coronary artery disease or myocardial disease, (9) Personal history of hypertension, requiring > 1 antihypertensive agent to maintain blood pressure < 140/90, (10) Valvular stenosis or regurgitation of > moderate severity, (11) Heart failure (baseline NYHA > 2), (12) Systolic BP < 90 mmHg (15) Pulse < 50/min, (16) History of pulmonary hypertension or elevated right ventricular systolic pressures by echocardiogram, (17) Renal failure necessitating dialysis, (18) Suspicion or diagnosis of amyloidosis, (19) Suspicion or diagnosis of hemochromatosis, (20) Unable to obtain an MRI scan with gadolinium for any other reason that is not listed above.
Panel design and metal conjugation of antibodies
We generated a CyTOF panel which includes antibodies against CHIP drivers, SASP-related proteins, and various cell surface markers for PBMCs such as the markers described in the previous report (40). We used 37 antibodies, Ir DNA-Intercalator (Cell-ID Intercalator-Ir; Fluidigm #201192A) for nucleus staining and Rh DNA-intercalator (Cell-ID Intercalator-103Rh; Fluidigm #201103A) for viability staining, as listed in Supplementary Table 1. We designed the panel using Maxpar Panel Designer software (Fluidigm). The antibodies listed in Supplementary Table 1, except for the metalconjugated antibodies that were purchased from Fluidigm, BioLegend, BD Biosciences, and Miltenyi, were conjugated to lanthanides using the Maxpar X8 Multimetal Labeling Kit (Fluidigm) or cadmium using the Maxpar MCP9 Antibody Labeling Kit (Fluidigm) according to the manufacturer's protocol. To determine optimal antibody concentrations, three different concentrations of each metal-labeled antibody were tested by CyTOF before being used in analyses.
The 90 kDa ribosomal S6 kinases (p90RSKs) are a group of serine/threonine kinases consisting of 4 RSK isoforms (RSK1-4). The expressions of p90RSK1-4 isoforms in human are ubiquitous, and all of them express in immune cells. 1 The p-p90RSK antibody recognizes the phosphorylation of p90RSK1 S380 site, which has a homology between p90RSK1-3 at LFRGFSpFVA, and p90RSK4 at SpFVA. Therefore, at least this antibody can recognize p90RSK1-3 isoforms.
Cytometry sample preparation and staining
Whole blood was collected at two different time points, before (baseline) and 3 months after RT, from patients (n = 16) with thoracic cancer treated at The University of Texas MD Anderson Cancer Center (Supplementary Table 2). PBMCs were isolated using Ficoll-Paque Plus (Fisher #45-001-750) according to the manufacturer's protocol and were cryopreserved in liquid nitrogen until analysis. Cryopreserved PBMC samples were thawed in RPMI containing 5% pooled human serum (PHS), 10 mM HEPES, 0.2% gentamicin, and 2.5 µ/mL benzonase nuclease, washed once with the same medium and repelleted by centrifugation at 300 g at 4 • C). One million cells were applied to the viability staining with Cell-ID Intercalator-103Rh and washed with Maxpar Cell staining buffer (Fluidigm #201068). For cell fixation, the cells were resuspended in 1x Fix I Buffer (Fluidigm #201065) and start the barcoding step using Cell-ID 20 Plex Pd Barcoding Kit (Fluidigm #201060). All barcoded samples were combined, and the staining procedure was performed according to protocol, i.e., Maxpar Phosphoprotein Staining with Fresh Fix (Fluidigm PN400278A4 protocol). The cells were blocked by Human TruStain FcX (BioLegend #422302), followed by cell surface staining for cell-type identification. After the wash with Maxpar Cell staining buffer, the cells were chilled on ice and incubated at 4 o C with methanol for permeabilization, followed by two washes with Maxpar Cell staining buffer. The cells were incubated with metal-labeled antibodies against cytoplasmic proteins followed by two washes. For the final fixation, the cells were incubated with 1.6% formaldehyde, followed by the incubation with Cell-ID Intercalator-Ir for nucleus staining. After washing twice with Maxpar Cell Staining Buffer and once with Maxpar water (Fluidigm #201069), the cells were filtered to make a single cell suspension and centrifuged. After adding EQ beads (Fluidigm #201078), CyTOF data were acquired on a Helios mass cytometer (Fluidigm).
Data acquisition for cytometry
CyTOF data were acquired on a Helios mass cytometer (Fluidigm) and were analyzed using the Astrolabe Cytometry Platform (Astrolabe Diagnostics, Inc.). Single-cell data have been clustered using the FlowSOM R package (41) and labeled using the Ek'Balam algorithm (42). Cell subset definitions were followed as previously described (43,44). Cluster labeling, method implementation, and visualization were done through the Astrolabe Cytometry Platform (Astrolabe Diagnostics, Inc.).
Irradiation of human peripheral blood mononuclear cells followed by flow cytometry
Human peripheral blood mononuclear cells (hPBMC) were seeded into a 48-well plate in culture medium (RPMI 1640 with 10%FBS, 10 mM HEPES, 1 × non-essential amino acid mix, 1 mM sodium pyruvate, 100 mg/mL streptomycin, 0.55 mM 2-mercaptoethanol) at a density of 2 million cells per well. Treatments were performed in triplicate for each group. We pre-treated the cells with FMK (RSK inhibitor Fmk, Soluble in DMSO, Axon Medchem Cat#1848) or DMSO for 1 h before irradiation. The cells were irradiated at 2 Gy while the non-irradiation control was set aside. Twenty-four hours after irradiation, we harvested the cells of each group and analyzed them on an LSRII instrument (Becton Dickinson). Flow CyTOF data were analyzed using FlowJo software. The following Cell surface markers were used: human CD3
Senescence associated β-galactosidase activity in cell cultures
After 24 h of ionizing radiation, PBMCs were incubated with 33 mM C12FDG (Cayman chemicals) for 30 min in a 37 • C water bath. Cells were centrifuged at 500 g for 5 min, washed with Cell Staining buffer (BD Biosciences), and stained with anti-CD19-APC and anti-CD14-PerCPCy5.5 at 4 • C for 30 min. Cells were then washed in Cell Staining buffer and resuspended in Cell Staining buffer and analyzed by a BD LSR II Flow cytometer. Gates were set as follows: Dead cells and debris were excluded based on forward scatter and side scatter measurements, B cells (CD19 + ), Monocytes (CD14 + ) and analyzed for C 12 FDG fluorescence. Data were analyzed with FlowJo software (v.10).
Statistical analysis
The multidimensional scaling (MDS) map was generated using the cmdscale R package (45). Differential abundance analysis was done using the edgeR R package (46, 47) as previously described (48). Differential expression analysis was done using the limma R package (49), as described previously (50). The frequency data within -log(FDR) > 0.5 and log [fold change (FC)] < -1 or 1 < log(FC) were considered as significant. The expression data within adjusted P-value (adj. P-Val) < 0.05 were considered as significant.
For the adj. P-Val, limma R package is using the Benjamini-Hochberg correction.
We determined differences between 2 independent groups using the Student's t-test (2-tailed) and, when applicable, 1-way analysis of variance followed by Bonferroni post-hoc testing for inter-group comparisons using GraphPad Prism (GraphPad Software, San Diego, CA, USA). We used Welch's analysis for evaluating the equal variance of each group. P-values < 0.05 were considered statistically significant.
The responses of immune cells after radiation treatment in cancer patients
Previous studies showed that ionizing radiation and various cancer treatments induced a pro-inflammatory senescent phenotype in myeloid cells (30). To characterize how RT can change immune cell phenotypes in cancer patients using multiparameter mass CyTOF, we enrolled 16 patients with esophagus or lung cancer who were about to start RT (Supplementary Table 2). Blood was collected at the time of enrollment and 3 months after the initial radiation treatment, and peripheral blood mononuclear cells (PBMCs) were collected. We then evaluated the phenotypic changes of immune cells by multiparameter mass CyTOF with general cell surface markers and various senescent markers as described in the methods. Analysis using the Astrolabe platform (40) automatically labeled canonical immune cell subsets (Figure 1A and Supplementary Table 2). Astrolabe identified eleven T cell subsets [including CD4 + and CD8 + T cells, naïve T EMRA (effector memory cells re-expressing CD45RA), effector memory, and central memory cells], three B cell subsets, five myeloid cell subsets, two NK cell subsets, and granulocytes. We found that only B cells significantly decreased after RT ( Figure 1B). However, when we looked at each immune cell subset, we found four immune cell subsets: CD4 + T EMRA cell, B cell (Memory), T cell (unassigned), and naïve B cell (CD27 − ) were significantly decreased after RT ( Figure 1C). Our data suggest that RT may play a significant role in regulating B cell functions in both memory and naïve B cells, including antigen producing and presentation. Furthermore, CD4 + T EMRA cells can show a cytotoxic phenotype with upregulation of CX3CR1 expression (51-54), which is related to CD4 and CD8 T cell cytotoxic potentials. Therefore, these data suggest the potential role of RT in regulating cytotoxic abilities of T cells.
Additional subsets revealed using functional antibodies
We further included the staining data with functional antibodies, and the Astrolabe Platform derived 138 functional profiling subsets with unsupervised clustering by FlowSOM algorithm, which were labeled by the markers that provided the most significant separation among them (Figure 2 and Supplementary Figure 1). We found that the frequencies of 5 functional profiling subsets were significantly upregulated, and seven functional profiling subsets were downregulated after RT (Figures 2B,C, 3A and Supplementary Figure 2). Among 12 functional profiling subsets, we found that only three subsets, CD27 − /Ki67 lo /CD38 lo /DNMT3a hi B cells, CD4 + /CD25 lo /IL-7Ra lo /TET2 lo T Naive cells and CD4 + /Il-7Ra lo /PD-1 lo /CD25 lo T EMRA cells, showed > 1% frequency among total cells and were decreased after RT. These data suggest that the effects of RT on immune cells were relatively specific to a few unique functional subsets of naïve B cell and CD4 + T cell at 3 months after RT.
cell subset in cancer patients
Astrolabe revealed five functional subsets of B cells (CD27 − ). Among all the B cell subsets, CD27 − /Ki67 lo /CD38 lo /DNMT3a hi (Ki67 − DNMT3a + naïve B cell) was the largest, and CD27 − /Ki67 lo /CD38 lo /DNMT3a lo was the second-largest subset (Figure 3A). Although B cells did not highly express T-bet compared to other cell types in the baseline (Supplementary Figure 3) and the T-bet expression in Ki67 − DNMT3a + naïve B cell subset was not exceptionally high compared to different B cell subsets in the pre-RT (baseline) samples ( Figure 3B and Table 1), we found that a significant increase of T-bet expression after RT was observed only in Ki67 − DNMT3a + naïve B cell subset ( Figure 3B). Since it has been reported that most of the age-ABCs express T-bet (55), and we previously reported the role of p90RSK activation in regulating senescence (30,56), we hypothesized that p90RSK activation could up-regulate T-bet expression. First, we investigated the relationship between the intensities of p90RSK S380 phosphorylation (p-p90RSK) and T-bet expression. p-p90RSK expression in B cells showed no significant difference with other cell types except NK cells (Figure 4A), and p-p90RSK expression did not increase after RT in the Ki67 − DNMT3a + naïve B cell subset (Figure 4B). But interestingly, we found a good correlation between p-p90RSK and T-bet in the samples from pre-RT, post-RT, and the combination of both ( Figure 4C). Next, we compared T-bet expression with the intensities of total p90RSK (t-p90RSK). Although we could not find any significant correlation between T-bet and t-p90RSK in the samples from pre-RT, we found a significant correlation between T-bet and t-p90RSK in post-RT samples and a combination of both pre-and post-RT ( Figure 4D). These data suggest that p90RSK activation and total p90RSK expression may play a role in T-bet expression.
Astrolabe identified all protein expression levels examined in this study in each cell type ( Figure 5A) and cell subtype ( Figure 5B) and found that TOP2b and CD38 expression were increased after RT in the total B cells after RT (Figure 5A), and in the subsets of CD27 − B cells (Figure 5B). A significant Figure 4A and Supplementary Figure 2) in light green. Once the frequency data within -log(FDR) > 0.5 and log[fold change (FC)] < -1 or 1 < log(FC) were considered as significant, 31 profiling subsets were changed by RT. Then 12 profiling subsets were picked out from the 31 subsets by student's t-test.
increase in CD38 expression was also observed in CD8 + T EMRA cells ( Figure 5B). These data suggest a contribution of T-bet and CD38 expression to the induction of senescent phenotype after RT. Since TOP2b can protect against DNA damage, the increase of TOP2b expression after RT may contribute to DNA damage repair after RT.
Radiation-induced T-bet + memory and naïve B cells and myeloid cells by 90 kDa ribosomal S6 kinase activation
We next investigated whether p90RSK activation played a role in ionizing radiation (IR)-induced increase in CD38 + Tbet + cells. PBMCs from the healthy human subject were pre-treated with vehicle or FMK-MEA (a specific inhibitor of p90RSK), then exposed to IR (2 Gy). Previously, we have confirmed the specificity of FMK-MEA by using the Ambit/DiscoverX platform to screen 443 kinases (57). Twentyfour hours following the irradiation, the numbers and frequencies of naïve and memory B cells were decreased (Figures 6A,B). However, we found a significant increase of T-bet + expression in memory and naïve B cells and this effect was inhibited by the p90RSK inhibitor FMK-MEA (Figures 6C-F). Since we have reported the role of SASPinduced CD38 expression in myeloid cells (32), we also investigated the impact of IR on CD38 +/ T-bet + myeloid cells, and we found that pre-treatment of FMK-MEA also FIGURE 4 The relationship between T-bet and p90RSK in Ki67 − DNMT3a + naïve B cell subset. (A) p-p90RSK expression in each cell types. Mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001. Statistical significance was assessed by Dunn's multiple comparisons test. significantly inhibited IR-mediated CD38 + /T-bet + myeloid cell induction (Figure 7). These data suggest that IR increases CD38 + /T-bet + cells induction in both B-cells and myeloid cells even in the early phase after IR, dependent on p90RSK activation.
Immunosenescence after radiation therapy and 90 kDa ribosomal S6 kinase activation As shown in Supplementary Table 2 and Figure 7E, we found a significant decrease in lymphocytes number after RT.
In contrast, myeloid cells numbers were relatively unaffected by RT, which resulted in relative myeloid skewing in the RT patients. To determine if RT induces immunosenescent in immune cells, we stained senescence-associated β-gal (SAβ-gal) staining and mitochondrial ROS (mtROS) production using healthy human PBMCs. Interestingly, CD19 + B cells and CD14 + myeloid cells in the PB showed significant increase of SA-β-gal + cells and mtROS production after IR (Figures 7F,G). The increase of SA-β-gal + cells and mtROS production was clearly blocked by pre-treatment of p90RSK inhibitor, FMK-MEA, in both B and myeloid cells. These results also suggest the crucial role of IR-induced p90RSK activation in immunosenescence.
FIGURE 5
Only TOP2b and CD38 expression were significantly increased after RT at the cell subset level. (A) Only TOP2b and CD38 expression in B cells were increased after RT, and no significant change of studied molecules in cell types after RT compared to pre-RT level was observed. (B) CD38 and TOP2b expression in B cell (CD27 − ) and CD8 + T EMRA subset pre-and post-RT. Mean ± SD, *adj. P-Val < 0.05, **adj. P-Val < 0.01, ***adj. P-Val < 0.001.
Discussion
In this study, we investigated the effects of RT on various senescence markers, DDR, efferocytosis, and clonal hematopoiesis drivers in immune cells. At 3 months after RT we observed a significant reduction in several immune cell subsets, particularly B cells.
Ki67 − DNMT3a + naïve B cells were the largest B cell subset. Although the frequency of this B cell (CD27 − ) subset was reduced after RT, we found a significant increase of T-bet expression that was correlated with phosphorylation of p90RSK in Ki67 − DNMT3a + naïve B cells. Furthermore, we also observed an increase in CD38 expression in B cells after RT.
In vitro evidence suggested that p90RSK activation played a crucial role in the induction of CD38 and T-bet expression by IR in memory, naïve B, and myeloid cells. These data link RTinduced activation of p90RSK to upregulation of CD38 + and T-bet + which are reported to have a role in immunosenescence (31, 32, 58). The analysis using Astrolabe Platform of the CyTOF data revealed that the Ki67 − DNMT3a + naïve B cell subset was the largest subset of B cells, and had a unique response to RT. CD27 was used as a marker of memory B cells in humans based on the correlation of CD27 expression with somatic hypermutation in IgM + /IgD + cells, and CD27 was constitutively expressed in approximately 40% of peripheral blood B cells in humans (59). CD27 is a member of the TNF-receptor family with the ligand of CD70, which is expressed on the surface of activated T cells. Activation of CD27 signaling plays a crucial role in maintaining long term immunological memory against T cell development antigen (60) by activating B cell expansion, differentiation, and antibody production (61-63). Therefore, the CD27 − B cell is recognized as a naïve B cell.
We examined various senescence and aging gene markers in various blood subsets using CyTOF, hoping to see the correlation between the irradiation and some changes of theses markers. The definition of the various subsets of T-and Bcells are listed in Supplementary Table 3. We focused B cells because only B cells showed significant change of some of these markers (CD38, TOP2β). In addition, the roles of DNMT3A in B cells have been reported. For example, DNMT3 regulates B cell lineage specification in the mouse model (64). Importantly, the depletion of DNMT3a in the B cell lineage results in the leukemic transformation of B1a cells in the CLL mouse model (65), and CD38 is a marker of human CLL B cell activation (66). Therefore, we assume that the B cell subset represented by lower expression of CD38 and higher expression of DNMT3a[i.e., (CD27 − ) Ki67 lo CD38 lo DNMT3a hi cells] are a less proliferative B cell subset in accordance with low Ki67 expression. This was the dominant B cell subset before RT in thoracic cancer patients ( Figure 3A). After RT, CD38 expression was increased regardless of Dnmt3a expression, suggesting that the residual B cells after RT were activated.
ABCs are antigen-experienced B cells that are characterized by a T-bet-mediated transcriptional program (67). ABCs plays a critical role in regulating immune response to infectious agents mediated by IgG2a/c and inflammatory cytokines (68, 69). However, the excessive accumulation of ABCs can trigger auto-immune reactions in response to self-antigen and may cause detrimental effects by inducing inordinate inflammation (70). In our current report, we found that T-bet expression was increased in the Ki67 − DNMT3a + naïve B cell subset, suggesting RT differentiated naïve B cells to ABC and induced inflammatory and senescence pathways. Also, the increase of T-bet expression in this largest but less active B cell subset (Ki67 − DNMT3a + naïve B cells) may significantly enhance the pre-mature aging process after RT by differentiating naïve B cells to ABC. Since systemic T-bet depletion inhibited atherosclerosis formation (38), it is possible that the increase of T-bet in B cells plays a role in elevating cardiovascular events after RT. Further investigation will be necessary.
It has been reported that CD38 expression in immune cells is increased in response to infection and inflammation, which subsequently induce inflammatory response (71). In rheumatoid arthritis (RA), the inhibition of CD38 activation by anti-CD38 monoclonal antibody has been effective to reduce RA symptoms and disease progression. In systemic lupus erythematosus (SLE) patients, the upregulation of CD38 expression in marginal zone-like IgD + CD27 + B cells has been reported (72), and the depletion of CD38 in pristaneinduced lupus mouse model significantly improve the symptom (73). Thus, CD38 upregulation in B cells seems to be related to inflammatory status. In our current report, we found an increase in CD38 expression in B cells after RT in cancer patients. Since the essential role of CD38 in B cells in inflammatory responses has been reported (68, 69), these data suggest that NAD + depletion induced by CD38 expression in B cells may initiate SASP. The induction of SASP in B cells may contribute to CVD in cancer patients after RT.
Lastly, we found a correlation between T-bet and p-p90RSK in the Ki67 − DNMT3a + naïve B cell subset. Furthermore, we found a crucial role of p90RSK activation in IR-induced CD38 + /T-bet + memory and naïve B cells and myeloid cells in vitro. Taken together, these data suggested that p90RSK activation induced by RT was involved in immunosenescence via upregulating CD38 and T-bet expression. Notably, the contribution of p90RSK activation and T-bet expression in immune cells to the process of atherosclerosis formation has been reported (38,56), and to the best of our knowledge, this is the 1st report to show that p90RSK activation is required for T-bet induction. Interestingly, the potential role of T-bet on CD38 induction has been reported (74). Taken together, these data suggest that the activation of p90RSK induced by radiation plays a crucial role in T-bet-CD38 induction, resulting in immunosenescence and consequent atherosclerosis formation. Further investigation will be necessary.
Data availability statement
The original contributions presented in this study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author/s.
Ethics statement
The Institutional Review Board of the University of Texas MD Anderson Cancer Center approved the clinical study protocol (#PA16-0971). The patients/participants provided their written informed consent to participate in this study.
|
v2
|
2022-11-09T16:03:33.009Z
|
2022-11-07T00:00:00.000Z
|
253405055
|
s2ag/train
|
Apoptotic-based topotecan-loaded superparamagnetic drug delivery system: an in vitro study in MCF7
Biological barriers could be overcome using nano-biotechnology, which promotes the development of nanomaterial-based delivery systems. The primary objective of the present investigation focuses on superparamagnetic iron oxide nanoparticle (SPION) production for the delivery of topotecan to human breast cancer cells (MCF-7). The XRD results confirm the formation of pure SPION. The FTIR spectra indicate the functional groups related to aminopropyl trimethoxy silane (APTS) as a coating agent and topotecan. Topotecan-loaded magnetite nanoparticles with an IC50 of approximately 156 µg/mL exhibited dose-dependent cytotoxicity. The PCR method also proved that, in the mentioned cell line, topotecan-loaded SPION could increase the Bax/Bcl2 ratio and P53 gene expression. Annexin V/PI detection assay was done in order to detect the induction of apoptosis. According to the results, the nanoparticles inhibitively influence the survival of the MCF-7 breast cancer cells via boosting apoptosis, which helps to slow the growth of tumor cells.
|
v2
|
2022-11-09T16:22:41.834Z
|
2022-11-07T00:00:00.000Z
|
253412871
|
s2ag/train
|
Prevention of Hyponatremia After Transsphenoidal Surgery: A Systematic Review
Context: Pituitary adenomas are amongst the most common tumors with a low mortality rate compared to other intracranial malignancies. Delayed hyponatremia (DH) is a common finding after transsphenoidal resection (TSS), which is the basis for the management of these tumors. Although DH is one of the leading causes of readmission after TSS, no unified guidelines exist with regard to the prevention of this electrolyte disturbance. Objectives: This study aims to evaluate and compare existing preventive protocols for DH in order to identify and signify their common grounds. Methods: After a search in electronic databases, including PubMed (NCBI), Embase, Scopus, and Google Scholar with the keywords of “pituitary adenoma,” “hyponatremia,” “transsphenoid surgery”, “water electrolyte balance,” “patient readmission", six original articles were included in the study. Results: We found that a protocol that both identifies groups susceptible to DH (males, older individuals, and those with a lower BMI) and consists of fluid restriction, sodium supplementation, and regular serum sodium monitoring could be utilized to prevent DH in patients with pituitary adenomas after TSS. Conclusions: Further studies with a larger sample size must be conducted to compare existing protocols for preventing DH and also investigate post-surgery optimal fluid-restricted diets and corticosteroid therapy in these patients.
|
v2
|
2022-11-09T16:22:41.882Z
|
2022-11-07T00:00:00.000Z
|
253407933
|
s2ag/train
|
Psychometric Properties and Clinical Utility of the Distress Thermometer in Caregivers of Persons With Multiple Sclerosis
Caregivers of persons with multiple sclerosis (MS) report high levels of distress. The National Comprehensive Cancer Network Distress Thermometer (DT) is used extensively in cancer patient and caregiver populations but has not been tested in nononcology caregivers. The purpose of this study was to examine the psychometric properties and clinical utility of the barometer portion of the DT in caregivers of persons with MS.
A secondary analysis was performed of data from a randomized trial comparing the effectiveness of 2 interventions aimed at reducing psychological outcomes associated with caregiving. The DT and the 4-item Patient-Reported Outcomes Measurement Information System Anxiety and Depression scales, which were administered at baseline, were used for all analyses. Construct validity (known groups) and convergent validity (interscale correlations) were evaluated. Receiver operating characteristic curve analysis was used to evaluate clinical diagnostic test evaluation.
The DT had good construct validity supported by strong correlations for known-groups analyses and good convergent validity (r = 0.70–0.72). The DT also demonstrated good discrimination for anxiety (area under the curve [AUC] = 0.83) and depression (AUC = 0.80). The optimal screening cut point on the DT was 4 for anxiety and 5 for depression.
The barometer portion of the DT demonstrates good psychometric properties and clinical utility in caregivers of persons with MS. This is the first examination of the DT in MS care partners.
|
v2
|
2022-11-09T16:26:48.493Z
|
2022-11-07T00:00:00.000Z
|
253415809
|
s2ag/train
|
Network Analysis of Gut Microbiota Including Fusobacterium and Oral Origin Bacteria and Their Distribution on Tumor Surface, Normal Mucosa, and in Feces in Patients with Colorectal Cancer
Introduction: Fusobacterium and several bacteria are reported to be associated with colorectal cancer (CRC). However, their relationship and whether they cause CRC or are just adapted to the cancerous environment is not known. We approached this subject by investigating the correlation and distribution of the bacteria throughout the colon in patients with CRC and elucidated the relationship between microbiota and CRC. Methods: Twenty-five patients with CRC who underwent colonoscopy for endoscopic submucosal dissection or surgery were prospectively enrolled. Fecal samples were taken before bowel preparation, and mucosal samples were collected from three sites (tumor surface, tumor-adjacent mucosa, and cecum) during colonoscopy using a cytology brush. The microbiota was identified and analyzed by sequencing of the 16S rRNA gene of the V3–V4 region. We evaluated the correlation between the bacteria based on network analysis and the distribution of Fusobacterium in the colon. Results: A network consisting of many bacteria was found in all sites; especially, oral origin bacteria including Fusobacterium formed a positively correlated network on tumor surface. Streptococcus showed a significantly higher relative abundance on tumor surface than in feces. The relative abundance of Fusobacterium had significant positive correlations between tumor surface and feces, tumor-adjacent mucosa, and cecum. Conclusion: In patients with CRC, many bacteria were correlated with each other, and Fusobacterium and oral origin bacteria formed a positively correlated network on tumor surface. Fusobacterium was equally distributed on tumor surface and throughout the lumen and mucus in the colon. In the colon where Fusobacterium is widely distributed, Fusobacterium would adhere to the tumor surface and be correlated with oral origin bacteria to make a microenvironment that is favorable for CRC.
|
v2
|
2022-11-09T16:56:19.152Z
|
2022-11-07T00:00:00.000Z
|
253413500
|
s2orc/train
|
Biomarkers for anti-vascular endothelial growth factor drugs
Angiogenesis is regulated by interactions between vascular endothelial growth factors (VEGFs) and VEGF receptors. VEGF-A, VEGF-D, placental growth factor (PlGF) and plasminogen activator inhibitor-1 (PAI-1) have tumor angiogenic activity. VEGF-A and PAI-1 levels in the blood may impact the activity of bevacizumab, and VEGF-D levels may similarly diminish the efficacy of ramucirumab. However, the dynamics of these angiogenic biomarkers for anti-VEGF therapy have not been well established; therefore, they were evaluated in this retrospective study, which included two cohorts. Cohort 1 included patients who were treated with cytotoxic agents and bevacizumab as first-line chemotherapy, and Cohort 2 comprised patients who were treated with cytotoxic agents and anti-VEGF drugs (bevacizumab, ramucirumab or aflibercept) as second-line chemotherapy. VEGF-A, VEGF-D, PlGF and PAI-1 levels were measured before starting chemotherapy and were re-assessed every 1–2 months until disease progression. Bevacizumab had reduced benefit as a first-line chemotherapeutant in patients with very low or very high levels of VEGF-A. Bevacizumab increased VEGF-A and PlGF levels, but not VEGF-D or PAI-1. Anti-VEGF drugs offered the greatest benefit to patients with high PAI-1 before first- and second-line chemotherapy. PAI-1 levels were not affected by anti-VEGF drugs. Since ramucirumab increased VEGF-D, it offered less benefit to patients with high VEGF-D in second-line chemotherapy. Conversely, aflibercept offered greater benefits to patients with high VEGF-D, without increasing VEGF-D. These biomarkers may be useful for the prediction of drug efficacy and may predict resistance to anti-VEGF drugs.
Introduction
Fluorouracil-based chemotherapy (combined with oxaliplatin or irinotecan) plus anti-epidermal growth factor receptor/vascular endothelial growth factor (anti-EGFR/VEGF) therapy is the standard first-line treatment for metastatic colorectal cancer, with an overall median survival of about 30 months (1)(2)(3). Progression-free survival (PFS) of first-line treatment is about a year (3,4); thus, second-or third-line treatment assumes great importance. Cetuximab (5), bevacizumab (6), ramucirumab (7) and aflibercept (8) show survival benefits when used as second-line chemotherapeutants. RAS is a predictive marker for anti-EGFR therapy; however, no promising biomarkers have been established for anti-VEGF therapy.
Angiogenesis is regulated by interactions between vascular endothelial growth factors (VEGFs) and VEGF receptors (VEGFRs) and is essential for cancer growth and metastasis (9)(10)(11). VEGF-A is the central regulator of tumor angiogenesis, endothelial proliferation, and survival (12,13). VEGF-A binds with high affinity to two structurally similar tyrosine kinase receptors, VEGFR-1 and VEGFR-2, both of which are expressed in tumor vasculature (14). Blockade of the VEGF-A/VEGFR-2 interaction inhibits tumor angiogenesis and growth. Plasminogen activator inhibitor-1 (PAI-1) has angiogenic activity and contributes to tumor progression, tumor invasion, and metastasis (15). High levels of PAI-1 degrade prognoses of patients with various types of cancers (16), including colorectal cancer (17).
At present, three anti-VEGF drugs are available to block the VEGF pathway in different ways. Bevacizumab is a humanized monoclonal antibody that binds to VEGF-A and blocks its activation (4). Ramucirumab is a humanized IgG1 monoclonal antibody that recognizes VEGFR-2, preventing binding of agonists, VEGF-A, VEGF-C, and VEGF-D, and blocking VEGFR-2 activation (7). Aflibercept is a recombinant fusion protein containing a VEGF-binding domain, and it antagonizes the activity of VEGF-A, VEGF-B, and placental growth factor (PlGF) (8). Bevacizumab is used for first-line to third-line treatment, and ramucirumab or aflibercept in combination with FOLFIRI is an effective second-line treatment for patients with metastatic colorectal cancer. However, dynamics and contributions of angiogenic biomarkers to anti-VEGF therapy have not been well established. In this retrospective study, we evaluated those dynamics and contributions to anti-VEGF therapy.
Materials and methods
Patients and study design. We conducted a retrospective study of patients with metastatic colorectal cancer who were treated with anti-VEGF-drugs between May 2015 and July 2021. This study included two cohorts. Cohort 1 comprised patients who were treated with cytotoxic agents and bevacizumab as first-line chemotherapy, and Cohort 2 included patients who were treated with cytotoxic agents and anti-VEGF drugs (bevacizumab, ramucirumab, or aflibercept) as second-line chemotherapy. We included patients who participated in a bio-bank project at our institution. This project was approved by local ethics review boards (28-03-738) and written informed consent was obtained from all patients who participated in this project. Inclusion criteria were: histologically confirmed adenocarcinoma of the colon or rectum, patients 20-80 years of age, Eastern Cooperative Oncology Group performance status of 0-1, and adequate organ function (white blood cell count ≥3.0x10 9 cells/l, ≥1.5x10 9 neutrophils/l, platelets ≥100x10 9 /l, hemoglobin ≥10.0 g/dl, serum bilirubin ≤1.5x upper limit of normal; alanine aminotransferase and aspartate amino transferase ≤2.5x upper limit of normal, and serum creatinine ≤1·5x upper limit of normal), known RAS and BRAF status (mutant or wild-type), and blood samples stocked in the bio-bank. The presence of at least one measurable reference lesion following the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 was also required. Patients with a history of another malignancy within the past 5 years were excluded. All Cohort 1 patients were chemo-naïve. Eligible patients of Cohort 2 had to have experienced disease progression within 6 months of the last dose of first-line combination therapy with oxaliplatin and a fluoropyrimidine for metastatic disease and had to have received at least one cycle of doublet therapy. Exclusion criteria included brain metastases, poorly controlled hypertension, or any arterial thrombotic or thromboembolic events within 12 months prior to starting chemotherapy. The study was conducted according to ethical guidelines of the Declaration of Helsinki, and the protocol was approved by local ethics review boards (B-2021-467). Information about the right to opt-out was posted on the websites of Main hospital of Nippon Medical School Sample collection. In the bio-bank project, blood samples (10 ml in BD Vacutainer EDTA tube: Becton Dickinson) were obtained from participants every 1-2 months during chemotherapy. Blood samples were centrifuged at 1,900 g for 10 min and the upper layer of each sample was transferred to another tube and stored at -80̊C until analysis.
Measurement of VEGF-A, VEGF-D, PlGF, and PAI-1. Plasma samples stored from the start to the end of chemotherapy every 2 months were used for measurement of angiogenic factors. VEGF-A, VEGF-D, PlGF, and PAI-1 were measured using commercially available enzyme-linked immunosorbent assay (ELISA) kits (VEGF-A: Human VEGF Quantikine kit, VEGF-D: VEGF-D Duoset ELISA kit, PlGF: Human PlGF Quantikine kit, PAI-1: Human Serpin E1/PAI-1 Duoset ELISA kit. All kits were from R & D Systems, Minneapolis, MN, USA) and were used according to the manufacturer's protocols. To measure PAI-1 concentrations, samples were diluted 200-fold.
Evaluation of clinical responses. Tumor responses were assessed by computed tomography (CT) following RECIST 1.1 criteria, 3 months after starting chemotherapy. After the initial assessment, CT was performed every 3 months until disease progression. Patients who achieved complete responses (CR) or partial responses (PR) were categorized as responders, and those who achieved stable disease (SD) or progressive disease (PD) were considered non-responders. Carcinoembryonic antigen (CEA) and CA19-9 were assayed monthly throughout chemotherapy. The normal CEA level was <5.0 ng/ml and the normal CA19-9 level was <37 U/ml. Statistical analysis. Statistical analysis was performed using R version 4.1.2 (The R Foundation for Statistical Computing, Vienna, Austria). The Mann-Whitney U test was used to compare differences in each angiogenic factor. Multiple comparisons of the dynamics of each angiogenic factors were tested with the Dunn's test after the Kruskal-Wallis test.
To evaluate impacts of VEGFs on the cytoreductive effect, patients were divided into high, medium-high, medium-low, and low groups with quartile values for each angiogenic factor. Clinical responses were tested using Fisher's exact tests. To evaluate impact of VEGFs on survival, patients were divided into two groups, high and low, with the median as the cut-off for each angiogenic factor. Progression free survival (PFS) and overall survival (OS) were tested using Kaplan-Meier analysis followed by the log-rank test and two-stage test.
Results
Patients. Thirty-one patients were included in Cohort 1 and 40 patients in Cohort 2 ( Fig. 1). Twelve of 31 Cohort 1 patients had not received second-line chemotherapy. Four Cohort 1 patients were continuing first-line chemotherapy. Two patients were administered anti-EGFR agents as second-line chemotherapy, and the remaining 13 patients were included in Cohort 2. Patient characteristics are shown in Table I and adverse events are listed in Table SI. In Cohort 1, 17 patients had one metastatic site, 9 patients had 2 metastatic sites, 5 patients had three or more metastatic sites total numbers of cases; liver: 18 cases, lung: 9 cases, peritoneum: 11 cases, others such as lymph node or bone: 12 cases). In Cohort 2, 26 patients had one metastatic site, 10 patients had 2 metastatic sites, and 4 patients had three or more metastatic sites (total number of cases: liver: 23 cases, lung: 12 cases, peritoneum: 9 cases, others: 12 cases). Twenty-two patients (71.0%) with RAS mutations were included in Cohort 1 and 22 more (55.0%) in Cohort 2. No patients with BRAF mutations were included in either cohort. Among 40 patients belonging to Cohort 2, 8 patients received bevacizumab, 18 received ramucirumab, and 14 received aflibercept. In first-line chemotherapy, 21 of the 40 Cohort 2 patients received bevacizumab and 19 received anti-EGFR drugs or no molecular target drugs.
VEGF-A, VEGF-D, PlGF, and PAI-1 levels. At the start of first-line chemotherapy, there were no correlations between the four angiogenic factors.
Median VEGF-A level before treatment was 56.7 pg/ml (IQR: 131.0). There was no relationship between VEGF-A level and tumor RAS status, or the number of metastatic sites and tumor location (right or left side). VEGF-A level increased significantly one month after starting chemotherapy and continued to rise during treatment (Fig. 3A). At the end of chemotherapy, VEGF-A level (median: 566.8 pg/ml, IQR: 335.9) was significantly higher than before chemotherapy (P<0.0001, Fig. 3B).
Median VEGF-D level before treatment was 340.0 pg/ml (IQR: 534.0). There was no relationship between VEGF-D level and tumor RAS status, or the number of metastatic sites and tumor side. Bevacizumab had no impact on VEGF-D (Fig. 3C), and VEGF-D level at the end of therapy (median: 429.4 pg/ml, IQR: 543.0) was the same as before chemotherapy (P=0.29, Fig. 3D).
Median PlGF level before chemotherapy was 8.3 pg/ml (IQR: 5.1). There was no relationship between PlGF level and tumor RAS status, or between the number of metastatic sites and tumor side. PlGF level increased significantly one month after starting chemotherapy and continued to rise during chemotherapy (Fig. 3E). PlGF levels at the end of therapy (median: 18.6 pg/ml, IQR: 11.3) were significantly higher than before (P<0.0001, Fig. 3F). Median PAI-1 level before treatment was 16.1 ng/ml (IQR: 14.6). It had no relationship with VEGF-A level, tumor RAS status, the number of metastatic sites, or tumor side. Bevacizumab had no impact on PAI-1 levels (Fig. 3G) and PAI-1 levels at the end of therapy (median: 14.2 ng/ml, IQR: 19.7) were unchanged (P=0.81, Fig. 3H).
Impact of VEGF-A, VEGF-D, PlGF, and PAI-1 levels on the cytoreductive effect and survival. With regard to VEGF-A levels, 72.2% (13/18) of responders (CR or PR) belonged to the medium-low and medium-high quartiles. Conversely, 23.1% (3/13) of non-responders (SD or PD) belonged to the medium-low and medium-high quartiles (Table II, P=0.05). VEGF-D, PlGF, and PAI-1 levels did not influence the cytoreductive effect (Table II). There were no differences in dynamics of those four angiogenic factors between responders and non-responders (Fig. S1).
VEGF-A, VEGF-D, PlGF, and PAI-1 levels. Median VEGF-A, VEGF-D, PlGF and PAI-1 levels of all Cohort 2 patients before treatment were 342.9 pg/ml (IQR: 682.1), 480.0 pg/ml (IQR: 610.2), 16.9 pg/ml (IQR: 12.4) and 17.8 ng/ml (IQR: 16.5). VEGF-A and PlGF levels of patients before second-line chemotherapy treated with bevacizumab during first-line chemotherapy were significantly higher than those of patients treated without bevacizumab (P<0.00001, P<0.0001 Fig. 5A and C). Conversely, VEGF-D and PAI-1 levels before second-line chemotherapy of patients treated with bevacizumab in first-line chemotherapy were equal to those of patients treated without bevacizumab (P=0.86, P=0.2, Fig. 5B and D). In 5 of 7 patients with high levels of VEGF-A at the start of second-line therapy, VEGF-A level decreased at the end thereof. In all patients with low levels of VEGF-A at the start of second-line therapy, VEGF-A level increased until the end of treatment (Fig. 5E).
In patients treated with ramucirumab, VEGF-D increased over time (Fig. 5F); however, in patients treated with bevacizumab or aflibercept, there was no obvious increase. Also, PlGF increased gradually in patients treated with ramucirumab and aflibercept ( Fig. 5G and H). Only 8 patients received bevacizumab as a second-line. In these patients, PlGF increased; however, the difference was not significant.
Discussion
In the present study, there were three valuable findings. First, bevacizumab increases VEGF-A and PlGF levels, but not VEGF-D or PAI-1 levels. Conversely, anti-EGFRs have no impact on levels of these four growth factors. Second, anti-VEGF drugs have greater benefit for patients with high PAI-1 levels before starting first-and second-line chemotherapy, and PAI-1 level is not affected by anti-VEGF drugs. Third, VEGF-D may be a useful biomarker for drug selection in second-line chemotherapy.
It is noteworthy that VEGF and PlGF levels increase one month after starting bevacizumab. It has been reported that VEGF-A and PlGF levels are high in patients who had received prior bevacizumab (18). However, the present study shows that VEGF and PlGF levels increase one month after starting bevacizumab and maintain high levels during drug administration. Bevacizumab inhibits angiogenesis by binding VEGF-A (19). Thus, bevacizumab is less beneficial for patients with low VEGF-A levels before chemotherapy, as indicated in Cohort 1. However, the sustainable increase of VEGF-A induced by bevacizumab may provoke acquired resistance to bevacizumab. Indeed, most responders had medium-low or medium-high levels of VEGF-A before starting chemotherapy, indicating that adequate VEGF-A levels provide benefits for patients treated with bevacizumab. There appears to be no benefit if the initial VEGF-A levels are too high or too low. Conversely, high VEGF-A levels may not provide a benefit in patients treated with ramucirumab. This hypothesis is supported by the fact that ramucirumab has better outcomes in bevacizumab-naïve patients than in patients whose first-line treatment included bevacizumab (20). We believe that the VEGF-A increase induced by bevacizumab has an unfavorable impact on efficacy of ramucirumab. Anti-VEGF drugs have greater benefit for patients with high PAI-1 levels before chemotherapy in either the first or second line. This is the first study showing that high PAI-1 levels are a favorable prognostic factor for patients receiving second-line chemotherapy with ramucirumab or aflibercept. It has been reported that high PAI-1 levels are a poor prognostic factor for stage I-IV colorectal cancer patients who are not receiving chemotherapy (17). Previous studies have already reported that high PAI-1 levels are an unfavorable prognostic factor for patients receiving bevacizumab (21)(22)(23), in contrast to results of the present study. Tumor angiogenesis requires PAI-1 (24), and PAI-1 has a dose-dependent effect on tumor angiogenesis (25).
Inhibition of PAI-1 limits tumor angiogenesis (26), indicating that patients with high PAI-1 levels have hyper-vascular tumors that can respond to anti-VEGFs. However, patients with high PAI-1 levels had better PFS, but the same OS during the first line. Conversely, in the present study, patients with high PAI-1 levels had better OS, but similar PFS in the second line. Thus, we need further studies to clarify or resolve this contradiction.
In the present study, we measured PAI-1 repeatedly and showed that bevacizumab had no impact on PAI-1 levels.
In previous studies, bevacizumab decreased PAI-1 levels in patients with lung cancer (21), metastatic solid cancers (22), or colorectal cancer (23); however, PAI-1 levels were measured only once after starting chemotherapy in these studies. PAI-1 antigen is mainly detected in fibroblasts and endothelial cells (27), indicating that PAI-1 levels are strongly affected by the microenvironment of cancer cells; thus, an increase or decrease of PAI-1 has a complex mechanism. In the present study, tumor progression or shrinkage had no effect on PAI-1 levels and there was no association between VEGF-A and PAI-1 levels. Thus, PAI-1 inhibitors are accepted as anti-cancer drugs by virtue of their anti-angiogenic effects.
VEGF-D may be a useful biomarker for drug selection in second-line chemotherapy. Ramucirumab has less benefit for patients with high VEGF-D levels and ramucirumab increases VEGF-D. Conversely, aflibercept has greater benefit for patients with high VEGF-D levels without increasing VEGF-D. VEGF-D has no effect on benefits of bevacizumab and is not affected by bevacizumab. In the present study, patients with low VEGF-D levels had significantly better PFS and non-significantly better OS. Tabernero et al (28) reported that ramucirumab has a favorable impact on patients with high levels of VEGF-D (>115 pg/ml) before chemotherapy, but no other studies have reported an association between ramucirumab and VEGF-D levels. In the present study, VEGF-D level was ≤115 pg/ml in only one of 18 patients who were treated with ramucirumab. As with the association between bevacizumab and VEGF-A level, too high a level of VEGF-D may restrict the efficacy of ramucirumab. Ramucirumab increased VEGF-D levels one month after starting chemotherapy and sustained the elevation during the second line; however, bevacizumab and aflibercept did not. Although no studies, including that by Tabernero et al (28), reported VEGF-D dynamics after starting chemotherapy, including ramucirumab, this increase is easy to understand in that VEGF-D elevation caused acquired resistance to ramucirumab. The fact that VEGF-A level did not show a distinctive trend after administration of ramucirumab, supports this hypothesis. Interestingly, patients with high levels of VEGF-D had greater benefit from aflibercept. In the biomarker study, VELOUR, VEGF-D was not measured (18); thus, the present study is the first to report a clear association between VEGF-D level and efficacy of aflibercept. PlGF had no impact on the effect of aflibercept, similar to the results of a previous study (18).
Results of the present study suggest that moderate levels of VEGF-A are necessary for a favorable effect of bevacizumab, and moderate levels of VEGF-D are necessary for reasonable efficacy of ramucirumab. Bevacizumab inhibits angiogenesis by blocking VEGF-A, and ramucirumab inhibits it by blocking VEGF-D. Not surprisingly, bevacizumab has little effect in patients with low VEGF-A levels and ramucirumab does not help patients with low VEGF-D levels. However, it is surprising that bevacizumab may be less effective for patients with very high levels of VEGF-A, and ramucirumab, likewise, may be less useful for patients with very high levels of VEGF-D.
This study had several limitations. This is a retrospective, single-center study that included a small number of patients. Thus, regimens of chemotherapy other than anti-VEGFs were not standardized. It is unclear whether VEGF-A, VEGF-D, PlGF, or PAI-1 impacted the prognosis of patients receiving anti-EGFR therapy, because those were the only patients we included. The present study failed to identify an optimal level of VEGF-A and VEGF-D because it included small numbers of patients; thus, further study is needed.
In conclusion, there is an optimal level of VEGF-A for a favorable effect of bevacizumab and also of VEGF-D for positive outcomes with ramucirumab. VEGF-A and PlGF levels are increased by bevacizumab and VEGF-D levels are increased by ramucirumab. Anti-VEGF drugs have benefits for patients with high PAI-1 levels and PAI-1 levels are not affected by anti-VEGFs. These biomarkers may be useful for predicting drug efficacy and interpreting resistance to anti-VEGF drugs. Presently, two prospective studies (the Brave Ace study and the Ukit study) which evaluate the utility of biomarkers, including VEGFs, in patients treated with anti-VEGF drugs in second-line chemotherapy, are ongoing. These two studies have larger sample sizes compared with our study; thus, they may provide additional information on the efficacy of VEGFs.
|
v2
|
2022-11-10T17:20:07.927Z
|
2022-11-07T00:00:00.000Z
|
253440332
|
s2orc/train
|
Pattern of malignant head and neck tumors among children in a Nigerian Teaching Hospital
: Background: The pattern of head and neck cancer in children are well documented among Caucasians and the Orien-tals but this cannot be said among African children especially in developing country like Nigeria. Aim: To evaluate the pattern of malignant head and neck tumours among children in a Nigerian Teaching Hospital. Method: A retrospective study of cases of head and neck childhood malignancies at the University of Benin Teaching Hospital, Benin-City, Nigeria over a 12-year period, from January 2009 to December 2020. Results: A total of 127 children with head and neck malignant tumours were seen in this period. The mean age of children was 5.27±4.72 years (age range, 0.3 to 18 years). There were 83(65.4%) males and 44(34.6%) females. The most frequently seen tumour was retinoblastoma (44.1%) and this was followed by rhabdomyosarcoma (18.1%) and Burkitt’s lymphoma (16.5%). Apart from Burkitt’s lymphoma that was commonest in the 6-12 years group, all other cancers were most frequent during the 0-5 years. The peak incidence of cases was seen in 2015 followed by 2016 and 2019. Regarding the of the of the patients died of their disease while just only 1 (0.8%) was discharged against medical advice Conclusion: Retinoblastoma followed by rhabdomyosarcoma and Burkitt’s lymphoma were the most common tumors in study locality.
Introduction
Worldwide, cancer is one of the leading causes of morbidity and death. 1 It is estimated that by 2020, the number of new cases of cancer will increase to more than 15 million, with deaths increasing to 12 million, and the burden of incidence, morbidity and death will be greater in developing countries. 2 Because of the complex anatomy and development of the head and neck, neoplasms during infancy and childhood arising at this site pose diagnostic and management challenges in clinical practice. 3,4 It is important to emphasize that tumors of the head and neck in children show considerable differences in their behavior, histology and management from those of the adults. 5,6 It is also notable that they are not as rare as many clinicians assume. 7 They represent 5-10 % of all childhood solid tumors and 2-3% of all head and neck tumours 4 .
Cancers are relatively rare in children unlike in the older age groups; nonetheless cancers have become a major cause of paediatric deaths worldwide even in developed countries. Children constitute a significant proportion of the population in many developing countries, accounting for up to a third of the population in some instances. In Nigeria, children constitute more than half of the population, 52.4%, hence emphasizing the need to focus appropriately on this segment of the society. It is reported that the incidence of malignant tumors is higher during the first five years of life than it is during the subsequent ten years. 8,9 This probably reflects the embryonic nature of certain tumors encountered during this time. Such embryonic neoplasms tend to mimic structures normally present during organogenesis. 10 They may be present at birth or arise postnatally from such immature cells. It is these group of neoplasms that characterize the greatest differences between tumors in childhood and in adults. 11 Most malignant tumors of the head and neck in infancy and childhood manifest themselves by the presence of a solid mass, which may attain a large size quite rapidly. 12 It is therefore sensible to regard all such masses in an infant or child as malignant until proven otherwise.
There is a need to document data on childhood cancer in each country. Documentation is essential for planning medical/treatment services, resource allocation and policy formulation. Studies on the profile of head and neck childhood malignancies are well documented among the Caucasians and Orientals. [13][14][15][16][17] However, few studies are Pattern of malignant head and neck tumors among children in a Nigerian Teaching Hospital Edetanlen Benlance E et al reported in sub-Saharan Africa. 18,19,20 Hence, this present study evaluated the pattern of malignant head and neck tumors among children in the south-south geopolitical region of Nigeria.
Patients and method
This was a retrospective study of all head and neck childhood malignancy seen in the paediatric ward at the University of Benin Teaching Hospital, Benin-City, Nigeria. Cases of malignant tumors diagnosed from January 2009 to December 2020, were included in the study. Being a retrospective, with negligible risk, we sought for an exemption from ethical approval from the Institution's Ethical and Research Committee. Data was collected from the case notes and clinic registers of patients. The data obtained were age of children at presentation, gender, histological diagnosis, year of diagnosis, treatment given, history of discharge against medical advice (DAMA) and outcome of treatment. In descriptive statistics, continuous data were summarized as range, means and standard deviations while categorical data were summarized as frequency and percentages. Data were presented as tables or charts where necessary. All data were analyzed using Statistical Package for Social Sciences version 20.0 (IBM corp., Armonk, NY, USA).
Results
A total of 127 patients with head and neck childhood malignant tumors were seen in this 12-year period given a hospital-based incidence of head and neck cancers of 13 cases per year. The mean age was 5.27±4.72 years (age range, 0.3 to 18 years). The distribution of head and neck childhood malignancy by gender is shown in Table 1. There were 83(65.4%) males and 44(34.6%) females. The pattern of head and neck childhood malignancy and age-group of the children is presented in Table 2. More than half of the patients were less than 5 years. The most frequently seen malignancy was retinoblastoma (44.1%) and this was followed by rhabdomyosarcoma (18.1%) and Burkitt's lymphoma (16.5%). The distribution of head and neck malignancies by age-group is shown in Table 3.
Apart from Burkitt's lymphoma that was commonest in the 6-12 years group, all other cancers were most frequent during the 0-5 years. The trends in years of occurrence are shown in Figure 1. The peak incidence of cases was seen in 2015 followed by 2016 and 2019. All the children with retinoblastoma had chemotherapy with orbital enucleation. Burkitt's lymphoma patients had solely chemotherapy while rhabdomyosarcoma and other malignancies had chemotherapy with or without radiotherapy. Regarding outcome, 8(6.3%) of the patients died of their disease while just only 1 (0.8%) was discharged against medical advice.
Discussion
With rising incidence of head and neck childhood malignancy in resource-poor countries, measuring the baseline occurrence of such tumors is imperative to provide much needed allocation of scarce resources. Incidence of head and neck tumors is relatively rare in children compared to adults, but it is on the rise. [21][22][23] In the United States, approximately 1 in 333 individuals between the ages of 0 and 20 years will be newly diagnosed with cancer each year, affecting a total of nearly 7,500 children under the age of 15 years and another 3,500 adolescents between 15 and 20 years of age. Five percent of all childhood cancers are head and neck malignancies, thereby affecting approximately 550 children every year. [21][22][23][24] The overall annual incidence of cancer in children under 15 years of age rose from 11.22 cases/100,000 person-years in the time period of 1973-1975, to 14.03 cases/100,000 person-years in 1994-1996 -an increase of 25%; an even larger increase in the incidence of pediatric head and neck malignancies. 24 In this subset, the incidence rate increased from 1.10 to 1.49 cases/100,000 person-years in the same time frame, an increase of 35%. 24 Studies of head and neck malignancies among the adult patients are well reported globally [25][26][27] . Most of the studies reported squamous cell carcinoma as the most frequent of all the malignancies. Retinue of studies is available in the world literature about childhood malignancies. [28][29][30] The most reported sites were head and neck followed by the abdomen. 29 The most frequently reported tumours were Burkitt's lymphomas. 28 Albright et al 15 reported that the incidence of head and neck malignancy among children younger than 15 years in the United States from 1973 through 1996 increased at a greater rate than childhood cancer in general 15 . Similarly in our study, greater number of children less than 5 years was more affected compared to the older ones.
The hospital based incidence of head and neck cancer of 13 cases per year seen in this study is lower than those reported in previous studies in Nigeria 18,19 . The duration of study, cultural and life style differences could be the reasons for the difference in incidence seen in this study and in previous studies carried out in the western region of Nigeria18,19. In the present study, there was a clear male predilection for head and neck childhood cancers. This is seen in all ages and for all type of lesion considered. Similar findings were reported by previous studies 13,14,[16][17][18][19] however, some reports from the United States suggest girls were more affected than boys. 15 The possible reason for this male preponderance is unknown.
Retinoblastoma and rhabdomyosarcoma are the leading causes of head and neck childhood malignant neoplasm seen in our study. Same findings were reported by previous studies. 15,19 The prevalence of retinoblastoma was fairly consistent throughout the study period. Retinoblastomas have a high genetic undertone and thus tend to develop earlier in those with the genetic aberration. This may account for the high prevalence found in this study. Rhabdomyosarcoma is a common neoplasm in the pediatric age group; it is the commonest form of soft tissue sarcoma in children. It was listed among the top five common cancers in children from previous reports. 7,18,19,22 Like in previous studies [13][14][15][16][17] , majority of rhabdomyosarcoma in children in this study are seen in those less than 5 years unlike what was documented by Adeyemo et al 19 who reported higher frequency among children aged 6-18 years. The reason for the high incidence in those less than five years could be due to the fact that rhabdomyosarcoma is an embryonal tumor which tends to occur more in children under the age of five years.
In Africa, the association of cancers with infectious diseases has been noted to be responsible for the high incidence of non-Hodgkin's lymphomas (NHL); particularly Burkitt's lymphoma. 31 Burkitt's lymphoma is a B-cell NHL which has been associated with Epstein Barr virus and falciparum infection. 32 It has been reported to account for nearly 90% of paediatric lymphomas and half of all paediatric cancer cases in these high risk area. 33 In the current study, Burkitt's lymphoma was the third commonest tumor contrary to what has been documented that Burkitt's lymphoma is the commonest malignant tumor in children in Africa. The fact that Burkitt's lymphoma is not as common in this study as previously reported may signify a changing pattern of malignancies in the sub-region. It may also be due to the fact that this study focused on head and neck malignancies and not on overall malignancies in children. Also, for head and neck tumors, only jaw Burkitt's lymphoma would be captured, thus, this may account to the low prevalence reported.
Limitation of this study was the lack of documentation on the exact site of the lesion making it impossible to assess the distribution of head and neck childhood malignancy according to the site of lesion
Conclusion
Of the head and neck childhood malignancies, retinoblastoma followed by rhabdomyosarcoma and Burkitt's lymphoma are the most common tumors in the southern geographic area of Nigeria.
Recommendation
Awareness of a potential malignancy and careful followup of children with suspicious head and neck cancers is mandatory so that more and more head and neck cancers in children are brought to treatment before it is too late. In resource limited settings where diagnoses depend majorly on clinical acumen, an awareness of predictors of a disease can shorten the time spent on arriving at a working diagnosis and guide the immediate choice of investigations and treatment.
|
v2
|
2022-11-16T16:12:35.923Z
|
2022-11-07T00:00:00.000Z
|
253545151
|
s2ag/train
|
DRUG UTILIZATION STUDY IN A RADIOTHERAPY UNIT OF A TERTIARY CARE TEACHING HOSPITAL IN RURAL WEST BENGAL, INDIA
Objective: Drug utilization studies lay special emphasis on the medical social and economic consequences of use of medications in special settings. This study was undertaken to identify the pattern of drugs prescribed frequently among patients attending the radiotherapy department.
Methods: This is a prospective study undertaken between January 1, 2018, and December 31, 2018. Prescriptions and patient records were reviewed and analyzed using the World Health Organization (WHO) indicators for drug utilization studies.
Results: We encountered a total of 618 patients during the study period. Among them, 340 (55.01%) were female. The most common age groups presenting were between 21 and 60 years. Carcinoma breast was the most common type encountered (total cases 181, 29.28%), followed by carcinoma lung (total cases 92, 14.88%), carcinoma cervix, hematological malignancies, carcinoma prostate, and carcinoma rectum. Total number of drugs prescribed was 3008 in total 618 prescriptions making it 4.86 drugs per prescription on average. Among them on average per prescription, 2.82 drugs were cytotoxic drugs (1745 total), whereas 2.04 drugs were supportive or adjunct drugs (1263 total). Among the drugs prescribed, 96.24% were in generic names, 6.95% prescriptions contained antibiotics, and 96.44% (596) prescriptions contained injections. About 85.23% of drugs were prescribed from essential drug list. Average consulting was 8.2 min and dispensing time for adjunct drugs was 4 min on average. On average, 52.42% of patients (324) had complete correct knowledge of the dosage and schedule prescribed. Adverse drug reactions were common, out of 618 patients, 542 (87.7%) experienced ADRs most common being gastrointestinal and dermatological ADRs. The most common implicated drug was cisplatin. Six serious adverse events were encountered.
Conclusions: This study provides a clear picture of drug use in this special clinic in rural Bengal and paves the way for larger and long-term study.
|
v2
|
2022-11-22T06:17:22.844Z
|
2022-11-07T00:00:00.000Z
|
253733410
|
s2ag/train
|
[Clinical analysis of 12 cases of laryngeal neuroendocrine carcinoma].
Objective: To investigate the clinical and pathological features, treatments and prognosis of laryngeal neuroendocrine carcinoma (LNEC). Methods: We conducted the retrospective analysis of the clinical data of 12 patients with LNEC admitted to the Department of Otorhinolaryngology Head and Neck Surgery, Second Hospital of Shanxi Medical University from May 2014 to December 2021, including 9 males and 3 females, aged 50-77 years. There were 4 cases of typical carcinoid tumour (highly differentiated), 5 cases of atypical carcinoid tumour (moderately differentiated) and 3 cases of neuroendocrine small cell carcinoma (hypofractionated). The clinical features, diagnosis, treatment and prognosis of LNEC were analysed. Results: The clinical manifestations of LNEC varied according to the tumour type but did not correlate with the pathological types. The supraglottic type was characterized by sore throat, foreign body sensation in the pharynx, coughing, obstructive sensation when eating and choking on water. The treatments were determined according to the pathological types, lesion location and invasion scope. Of 12 patients 4 underwent horizontal partial laryngectomy plus elective lymphatic dissection plus postoperative radiotherapy/chemotherapy, 4 underwent vertical partial laryngectomy (3 of them with cervical lymphatic dissection), 3 underwent supported laryngoscopic plasma laryngectomy for laryngeal cancer, and 1 abandoned for treatment. With the follow-up of 8 -78 months, 5 patients were alive, 1 died from chemotherapy reactions, 3 died from other diseases, 1 died from lung metastasis, 1 died from lung infection and 1 was lost to follow-up. Conclusion: LNEC is clinically rare, the clinical manifestations are less specificity, diagnosis relies on pathological and immunohistochemical examinations, and treatment modalities and prognoses are closely related to the pathological subtypes of LNEC.
|
v2
|
2022-11-22T16:05:39.294Z
|
2022-11-07T00:00:00.000Z
|
253753370
|
s2ag/train
|
CHITOSAN NANOBUBBLES DEVELOPMENT AND EVALUATION FOR THE DELIVERY OF SUNITINIB-AN ANTICANCER AGENT
Objective: In the current study, we introduced a novel method for creating Sunitinib nanobubbles by incorporating it into chitosan-shelled nanobubbles.
Methods: The Design Expert® programme randomly assigned around 13 experiments, and multiple regression analysis was used to statistically examine the data. The effect of the amount of sunitinib, amount of chitosan, amount of Epikuron 200, amount of palmitic acid and stirring speed, on percent encapsulation efficiency and drug load while maintain minimum particle size of nanobubbles as considered through a definitive screening plan. By placing limitations on the response parameters, the optimum formulation was created using a numerical optimization approach. The three improved formulations (Batch1 through Batch3) were assessed.
Results: The findings show that the nanobubbles particle size of 78.56-82.42 nm with an encapsulation efficiency of 68.48-69.56 % and loading capacity of 23.88-25.02%. The quantity of sunitinib released from nanobubbles was much larger (96.52 percent) than that from the sunitinib solution within 24 h, according to an in vitro release profile of the medication using ultrasonography. The hemolytic activity of the blank nanobubbles and sunitinib-loaded nanobubbles was measured to assess their safety up to a concentration of 10 mg/ml. With erythrocytes, drug-loaded nanobubbles had a good safety profile. FTIR, DSC studies indicated no chemical interactions, TEM images revealed nanobubbles size of 70-100 nm and stability studies shows no significant changes.
Conclusion: For contrast-enhanced tumour imaging and subsequent therapeutic administration, nanobubbles were found to be superior.
|
v2
|
2022-11-25T16:33:24.358Z
|
2022-11-07T00:00:00.000Z
|
253861301
|
s2orc/train
|
Prognostic value of pretreatment systemic inflammatory markers in patients with stage I endometrial cancer
Objective: Evaluate the prognostic value of monocyte-lymphocyte ratio (MLR) in patients with stage I endometrial cancer. Method: Data from 225 patients with stage I endometrioid endometrial cancer who underwent surgical resection between January 2010 and December 2020 were reviewed. The receiver operating characteristic (ROC) curves were generated for the neutrophil-lymphocyte ratio, platelet-lymphocyte ratio, and MLR. Optimal cut-off values were determined as the points at which the Youden index (sensitivity + specificity - 1) was maximal. Based on the results of the ROC curve analysis, the patients were grouped into high MLR and low MLR groups. Recurrence rate, and disease-free survival were compared between the two groups. The prognostic factors were investigated using univariate and multivariate Cox proportional hazards model. Results: The optimal cut-off value of MLR was 0.220 (AUC, 0.835; p < 0.001). Significantly more patients in the high MLR group experienced recurrence (20.3% vs. 1.9%, p < 0.0001). In multivariate analysis, grade, depth of myometrial invasion, adjuvant RT, and high MLR were independent prognostic factors for disease-free survival. Conclusion: Elevated MLR was significantly associated poor clinical outcomes in patients with stage I endometrioid endometrial cancer. Our findings suggest that MLR may be clinically reliable and useful as an independent prognostic marker for patients with stage I endometrioid endometrial cancer.
Introduction
Endometrial cancer (EC) is the most common gynecologic cancer affecting women in developed countries [1]. Approximately 65,950 new cases and 12,550 deaths related to EC are expected to occur in the United States in 2022 [2].
About 70% of patients with endometrioid EC were diagnosed with stage I disease, and 5-year survival rate was nearly 90% [2]. The primary standard treatment for stage I endometrioid EC is surgery. After surgery, adjuvant treatment is recommended based on the patient's adverse risk factors [3]. Traditional prognostic factors for EC include initial stage, grade, histologic subtype, age at diagnosis, tumor size, and lymphovascular space invasion (LVSI) [4][5][6]. However, since stage I endometrioid EC has an excellent prognosis with a low rate of recurrence, these conventional risk factors are not sufficiently accurate to predict survival outcomes. A small but substantial number of patients with stage I endometrioid EC experience recurrence of disease and poor survival [7]. Thus, novel approaches for pre-treatment assessment to identify probable recurrence are crucial.
Peripheral blood cells, including neutrophils, lymphocytes, and monocytes, are biomarkers of tumor immunity and can reflect the cancer-related Ivyspring International Publisher inflammatory microenvironment [8]. Earlier studies have reported that systemic inflammatory responses play important roles in carcinogenesis, progression, and prognosis [9][10][11]. The neutrophil-lymphocyte ratio (NLR), platelet-lymphocyte ratio (PLR), and monocyte-lymphocyte ratio (MLR) are the currently available markers of the systemic inflammatory response [12]. These markers have been clarified to show prognostic significance in solid cancers, including gynecologic cancer [13][14][15][16][17][18]. However, the prognostic value of these ratios in patients with stage I endometrioid EC is unclear. Therefore, this study aimed to evaluate the prognostic value of NLR, PLR, and MLR for patients with stage I endometrioid EC.
Materials and Methods
This retrospective study was approved by the Institutional Review Board of the Catholic University of Korea. The requirement for informed consent was waived owing to the nature of the study. The study was conducted in accordance with the principles of the Declaration of Helsinki.
We reviewed our institution's cancer registry and identified patients who underwent primary surgical treatment for EC from January 2010 to December 2020. The retrospective review included all patients who were diagnosed as having EC. Thus, data from 338 patients were recorded in a single database.
We excluded patients who did not receive primary surgery; showed non-endometrioid histology; had stage II, III, or IV disease, inflammatory disease, hematological disease, or autoimmune disease; or had no preoperative complete blood cell count data or complete blood cell count data obtained within 2 weeks before surgery. Patients with incomplete clinicopathological or follow-up data were also excluded. The remaining 225 patients were included as the study population.
Systemic lymphadenectomy included pelvic and para-aortic lymphadenectomy; however, the latter could be omitted when the pelvic lymph nodes were disease-free. Postoperatively, patients were treated with adjuvant radiation therapy according to the disease risk factors.
NLR and PLR were defined as the absolute neutrophil count and platelet count, respectively, divided by the absolute lymphocyte count. Similarly, MLR was defined as the absolute monocyte count divided by the absolute lymphocyte count. Disease-free survival (DFS) was measured from the date of diagnosis of EC to the date of first disease progression. If the patient showed no recurrence, the observation was censored at the date of death or the last follow-up. Overall survival (OS) was measured from the date of initial diagnosis to the date of cancer-related death or the last follow-up. The primary endpoint was DFS, and the secondary endpoint was OS.
Receiver operating characteristic (ROC) curves of DFS were generated for NLR, PLR, and MLR. The optimal cut-off values of NLR, PLR and MLR were determined as the points at which the Youden index (sensitivity +specificity -1) values were maximal. Based on the results of ROC curve, patients were divided into high-MLR and low-MLR group. DFS and OS were analyzed by the Kaplan-Meier method, and the curves were compared using the log rank test. We performed univariate and multivariate analyses using Cox proportional hazards model to evaluate effects of prognostic factors. All statistical analyses were performed using Statistical Package for the Social Science (SPSS) statistical software package, version 22.0 (SPSS Inc., Chicago, IL, USA), and p < 0.05 was considered statistically significant.
Results
Overall, 225 patients were included in the final analysis. The baseline characteristics of the patients are presented in Table 1. Next, we defined the optimal cut-off values of NLR, PLR, and MLR by ROC curve analysis for our patient population (Figure 1) 0.637; 95% CI: 0.521-0.753, p = 0.09). Median MLR was 0.191 (range, 0.037-0.755). The optimal cut-off value of MLR for DFS was 0.220 (AUC: 0.835; 95% CI: 0.764-0.906, p < 0.001). The ROC curve analysis suggested that the AUC of MLR was the highest, and the MLR was the only marker to show statistical significance. Thus, we divided the patients into high-MLR (MLR ≥ 0.220) and low-MLR (MLR < 0.220) group.
The associations between MLR and clinicopathologic factors are presented in Table 2. The low-and high-MLR groups included 151 (67.1%) and 74 (32.9%) patients, respectively, with no statistically significant differences between the two groups in terms of age, body mass index (BMI), grade, depth of myometrial invasion (MMI), tumor size, LVSI status, and adjuvant treatment after surgery. Interestingly, significantly more patients in the high-MLR group experienced recurrence (20.3% vs. 1.9%, p < 0.001).
To evaluate the prognostic factors of recurrence, we used Cox's proportional hazards model (Table 3).
Univariate analysis revealed that DFS was significantly associated with grade, MMI, adjuvant radiotherapy (RT) and MLR. LVSI status and tumor size were not associated with DFS. Moreover, other inflammatory markers, such as NLR and PLR, were not associated with DFS either. In multivariate analysis, grade, MMI, adjuvant RT, and high MLR were independent prognostic factors for DFS.
According to Kaplan-Meier analysis, the 5-year DFS rates in the low-and high-MLR groups were 97.7% and 63.7% (p < 0.001), respectively, and the 5-year OS rates in these two groups were 97.5% and 96.7%, respectively (p = 0.397) (Figure 2). The two groups showed no statistically significant differences in terms of OS.
Discussion
The association between inflammation and cancer was first described by Virchow in 1863 [19]. Since then, numerous studies have highlighted the importance of inflammatory cells and cytokines, which are more likely to contribute to tumor growth, progression, and metastasis [20,21]. These findings indicate that a systemic inflammatory response is a basic feature of malignancy. Moreover, previous studies reported the association between the systemic inflammatory response and the prognosis of solid tumors, including gynecologic cancer [13][14][15][16][17][18].
The majority of endometrioid EC patients are diagnosed as showing stage I disease. Most of the stage I endometrioid EC patients were treated with surgery alone, but adjuvant treatment is recommended in patients with adverse risk factors. Pathologic factors that may influence the decision regarding adjuvant therapy include LVSI, grade, tumor size, and depth of invasion [4][5][6]. Patients without adverse risk factors are defined as showing low-risk EC. The low-risk group represents the largest group of patients with stage I EC and presents excellent survival outcomes [28]. However, 3%-10% of these patients experience relapse [3]. Thus, identification of novel indicators is essential to ensure prompt detection of probable recurrence. In the present study, the MLR was demonstrated as a surrogate marker for DFS in multivariate analysis. These results are in concordance with previous studies in which the MLR was suggested to be associated with survival in patients with endometrial cancer, ovarian cancer, colorectal cancer, and hepatocellular carcinoma [23][24][25][26]. Thus, the findings of this study indicate that the preoperative MLR was an independent predictor of recurrence in patients with stage I endometrioid EC, including low-risk EC, and the results provide a valuable clue for evaluation of the systemic inflammatory response to predict the recurrence of low-risk EC.
The precise mechanisms of the association between a high MLR and poor outcomes have not been clarified. The MLR is thought to reflect the balance between the unfavorable role of monocytes and the favorable prognostic effect of lymphocytes [27]. Monocytes are known to have pro-tumoral functions, such as differentiation into tumorassociated macrophages (TAMs), metastatic cell seeding, suppression of T cell function, angiogenesis, and extracellular matrix remodeling [28]. TAMs accelerate tumor progression and invasion by releasing growth factors and angiogenic factors [29]. Lymphocytes, in contrast, are usually known for their anti-tumor functions, which include induction of apoptosis and suppression of proliferation [8]. CD8 + T lymphocytes attack tumor cells via cytotoxicity, while CD4 + T lymphocytes exhibit potent anti-tumor immune response [30]. Thus, a low lymphocyte count and high monocyte count might be associated with cancer progression and a poor prognosis. An elevated MLR can be attributed to a relative increase in the monocyte count or relative decrease in lymphocyte count. Thus, the MLR may serve as a surrogate marker reflecting increased cancer aggressiveness. The NLR and PLR have also been suggested to be related to cancer patient prognosis. A higher NLR and PLR have been shown to be associated with poor prognosis in patients with EC [31,32]. However, in our study, higher NLR and higher PLR were not associated with poor survival.
Recent study by Crosbie et al. suggested C-reactive protein (CRP) as a prognostic biomarker in EC patients [33]. In this study, MLR was associated with adverse factors, but not overall, cancer-specific or recurrence-free survival in the multivariable analysis. A different conclusion might be reached based on alternative thresholds, since there are no clinically validated prognostic thresholds for MLR. The adverse risk factors in patients with stage I EC include high grade, deep MMI, LVSI, and tumor size [34,35]. Stratification of patients for adjuvant RT is based on these factors. In our study, adverse risk factors such as LVSI and large tumor size were not associated with survival outcomes while patients who received adjuvant RT had favorable outcomes. Thus, our findings highlight the potential benefit of RT in patients at increased risk of recurrence, especially LVSI and large tumor size.
The histologic grade of EC is an important factor associated with its prognosis. The majority of low-grade ECs tend to limit their spread to the surface of the endometrium, with a low likelihood of disease extension beyond the uterine corpus or the need for adjuvant therapy [36]. In our study, a higher grade was associated with an increased risk of recurrence. Interestingly, not only grade 3, but grade 2 EC was also associated with an increased risk of recurrence. All patients with grade 3 EC received adjuvant RT. However, patients with grade 2 EC received adjuvant RT only if they had additional risk factors such as deep MMI, LVSI, or a large tumor size. Thus, our findings indicate that not only grade 3, but grade 2 EC is also adverse risk factors in patients with stage I EC.
Our study had several limitations. First, it was a retrospective single-center study. Second, the number of enrolled patients was small. These results need to be confirmed in a large cohort. Third, since there was no defined MLR value, we had to set a cut-off value for our population.
In conclusion, we found that an elevated MLR was significantly associated with a lower DFS in stage I endometrioid EC patients. Our findings suggest that the MLR may be clinically reliable and useful as an independent prognostic marker for stage I endometrioid EC patients. Further prospective studies are needed to confirm our findings and to identify appropriate cut-off values.
Data Availability
The data that support the findings of this study are available on request from the corresponding author.
|
v2
|
2022-12-03T06:17:12.309Z
|
2022-11-07T00:00:00.000Z
|
254150418
|
s2ag/train
|
Laparoscopic atypical resection of subcardial gastric GIST guided by intraoperative endoscopy. A single-center study and a review of the literature.
BACKGROUND
Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. They are 1% of all gastrointestinal cancer and 60% of them affects the stomach. Up to 10% to 30% of GISTs are malignant. They occur in people over the age of 50 in both sexes. The most common symptoms of gastric GIST are bleeding, dyspepsia, vague abdominal pain or discomfort, and mass palpation. Some are asymptomatic and diagnosed incidentally. The first choice of treatment for primary localized gastric GISTs is surgery. The most suitable type of resection is not yet clear and it depends on size and location of tumor, especially for difficult localizations, such as subcardial, posterior wall and less curvature GISTs.
METHODS
We report a rare case of a patient with subcardial gastric GIST treated with laparoscopic atypical quadrangular resection guided by intraoperative endoscopy. Furthermore, we performed a review of the literature about this topic.
RESULTS
Despite the difficult localization an atypical resection of the gastric GIST was performed without breaking the lesion but preserving the lumen of the esofagogastrich junction.
CONCLUSIONS
An atypical quadrangular resection for subcardial gastric GISTs, located along the posterior wall and lesser curvature, can be a safe and reliable alternative technique. However, we believe that it should be performed by an experienced surgeon and endoscopist to decrease the risk of mass's break and the narrowing of the cardial region's lumen. In our literature's knowledge there aren't cases treated with this technique.
KEY WORDS
Gastric GIST, Gastrointestinal stromal tumors, Intraoperative endoscopy, Laparoscopic resection, Minimally invasive surgery.
|
v2
|
2022-12-07T18:12:46.923Z
|
2022-11-07T00:00:00.000Z
|
254331536
|
s2ag/train
|
Association of Serum Apelin levels with Peripheral Neuropathy in Type-2 Diabetes Mellitus
Objectives: To determine serum levels of Apelin in healthy individuals, type-2 diabetic patients and diabetic peripheral neuropathy patients (DPN) and to find the correlation with serum Apelin levels and serum fasting blood glucose, glycosylated hemoglobin.
Study Design: Cross-sectional analytical study.
Place and Duration of Study: Physiology Department and Centre for Research in Experimental and Analytical Medicine (CREAM) lab of the Army Medical College/National University of Medical Sciences (NUMS) Rawalpindi Pakistan, from Jan to Dec 2021.
Methodology: A total of 90 individuals comprising three Groups having 30 subjects in each were recruited. Group-I included thirty healthy subjects, and Group-II consisted of thirty newly diagnosed patients having T2DM, Group-III consisted of thirty T2DM patients with DPN. We used Michigan Neuropathy Screening Instrument (MNSI) to assess neuropathy. Serum Apelin and tumour necrosis factor-alpha (TNF-alpha) levels were recorded from the blood samples of all subjects by enzyme-linked immunosorbent assay (ELISA) using human serum Apelin and TNF-alpha ELISA kit catalogue no. E2014Hu (bio-assay technology) and E0082Hu (Bio-assay technology), respectively.
Results: The mean values of serum Apelin were higher in Group-III compared to Group-II and Group-I, and a statisticallysignificant difference was found (p value=0.001). The serum Apelin levels showed a strong negative correlation with Group-III serum FBG, HbA1c and TNF-alpha with r value -0.728, -0.79, and -0.95, respectively.
Conclusion: Serum Apelin has a beneficial role in DPN in T2DM, with reduced TNF-alpha levels as one of the possible mechanisms.Keywords: Apelin, Diabetic peripheral neuropathy, TNF-alpha, Type-2 diabetes mellitus.
|
v2
|
2022-12-24T16:02:51.776Z
|
2022-11-07T00:00:00.000Z
|
255018283
|
s2ag/train
|
The prevalence and risk factors of Spirocerca lupi in domestic dogs in the Mekong Delta of Vietnam
Spirocercosis is caused by Spirocerca spp., which is a chronic disease and might cause life-threatening due to forming cancer in oesophagus in canid carnivores. There are limited studies involving spirocercosis in domestic dogs. Thus, this study aims to investigate the prevalence and analyse risk factors involved in the S. lupi infection in Mekong Delta in Vietnam. In total, 400 fecal samples from domestic dogs were collected from May 2020 to May 2021. The overall prevalence of spirocercosis in domestic dogs in the Mekong Delta was 10.50% by copromicroscope and PCR methods. PCR targeted to the housekeeping gene cytochrome c oxidase I (cox-1) was applied to identify species of Spirocerca spp. and analyse the phylogenetic tree. Outdoor dogs had 5.48 times (CI 95% = 2.45-11.690, p < 0.001) higher risks of S. lupi infection compared to indoor dogs. Besides, seasons and age showed a correlation to the increase the risk of S. lupi infection, while neither dog breeds nor gender influenced the prevalence of this species. The cytochrome c oxidase I (cox-1) gene sequence of S. lupi in the Mekong Delta showed the high homologues to the S. lupi isolates in India, Israel, and the North of Vietnam and belonged to the S. lupi genotype 2.
|
v2
|
2022-11-08T14:06:21.700Z
|
2022-11-08T00:00:00.000Z
|
253386353
|
s2orc/train
|
Ceritinib (LDK378) prevents bone loss via suppressing Akt and NF-κB-induced osteoclast formation
Background Ceritinib is used for the treatment of patients with anaplastic lymphoma kinase (ALK)-rearranged non-small cell lung cancer (NSCLC), who are at the risk of developing bone metastasis. During bone metastasis, tumor cells release factors that induce osteoclast formation, resulting in osteolysis. However, the effect of ceritinib on osteoclast formation remains unclear. Methods Osteoclastogenesis was induced to assess the effect of ceritinib on osteoclast formation and osteoclast-specific gene expression. Western blotting was used to examine the molecular mechanisms underlying the effect of ceritinib on osteoclast differentiation. An in vivo ovariectomized mouse model was established to validate the effect of ceritinib in suppressing osteoclast formation and preventing bone loss. Results The differentiation of osteoclasts and the expression of osteoclast-specific genes were inhibited upon ceritinib stimulation. Ceritinib suppressed Akt and p65 phosphorylation during the receptor activator of nuclear factor kappa-B ligand (RANKL)-induced osteoclastogenesis. The administration of ceritinib to ovariectomized mice ameliorated trabecular bone loss by inhibiting osteoclast formation. Conclusions Ceritinib is beneficial in preventing bone loss by suppressing osteoclastic Akt and nuclear factor κB (NF-κB) signaling.
Introduction
Ceritinib (LDK378) is a second-generation anaplastic lymphoma kinase (ALK) inhibitor (1). It exhibits robust antitumor efficacy in patients with ALK-rearranged non-small cell lung cancer. Generally, it increases ALK selectivity approximately 20 times more potently than the first generation ALK-targeted compound, crizotinib (2). In crizotinib-resistant mutants, ceritinib can inhibit the activity of ALK mutants, including Leu1196Met, Gly1269Ala, Ile1171Thr, and Ser1206Tyr. Therefore, ceritinib was granted accelerated approval by the FDA in April 2014, for the treatment of ALK-positive NSCLC (3,4).
Approximately, 30-40% of patients with non-small cell lung cancer develop bone metastasis during the course of the disease (7). During non-small cell bone metastasis, osteolysis prevails over bone formation due to the activity of tumor-secreted factors, such as PTHrP, IL-11, IL-6, and TNF-a, which induce excessive osteoclast formation (8). In addition to bone metastasis, changes in bone mineral density also occur; moreover, according to clinical studies, the prevalence of osteoporotic vertebral fractures is 30-40% in lung cancer patients (9). Therefore, compounds, such as bisphosphonates, that can target osteoclast formation are preferred for the treatment of bone metastasis caused due to non-small cell lung cancer or to improve the bone mass in these patients (10)(11)(12).
Osteoporosis is highly prevalent worldwide with typical characteristics of low bone mineral density and high possibility of fracture. The prevalence of osteoporosis was calculated by several countries and indicated higher risks in women from 2 to 8 times one in men (13). Excessive formation and activation of osteoclasts leads to development of osteoporosis (14). Most of current clinical antiresorptive agents include estrogen; bisphosphonates (BPs); human monoclonal antibody against receptor activator of NF-kB ligand (RANKL) denosumab; selective estrogen receptor modulators (SERM) raloxifene; and strontium ranelate (SR). However, their efficacy and safety have not been established for long-term therapy completely (15). And there have been reported that BPs caused osteonecrosis of the jaw (ONJ); and alendronate causing atypical fractures (16), and estrogen replacement therapy brought concern of increased breast cancer risk (17,18). RANKL, the critical factor of osteoclast differentiation, downstream signaling pathways are activated including NF-kB, JNK, p38 MAPK, extracellular signal-related kinase and Akt. RANKL activates the IkB kinase (IKK) complex and thus phosphorylates NF-kB-associated IkBa to release RelA/P65 (one of the members of NF-kB family) to cytosol. Further P65 translocates to nucleus and activates downstream transcriptions (19,20). Phosphatidylinositol 3-kinase (PI3K) and further phospholipid dependent activation of the Akt (serine/threonine kinase) mediate to suppress the apoptotic function in various cells. Increased expression levels of phospho-Akt enhanced RANKLinduced osteoclastogenesis (21,22). Activation of Ras, Raf-1 and mitogen-activated protein kinase (MAPK) signaling cascades are essential in cell proliferation and differentiation (23, 24). And MAPK activates downstream of ERK1 (p44 MAPK) and ERK2 (p42 MAPK) to maintain osteoclast survival and polarity, and ruffled border formation (25)(26)(27).
Myostatin binds to and activates complex of ALK4/5 and strongly accelerates RANKL-mediated osteoclast formation. Meanwhile ALK inhibitory effect has been found in reducing osteoclast differentiation critically (28). Here we introduced an anti-cancer drug ceritinib (LDK378) with additional inhibiting effect on osteogenesis and potential in treating related osteoporosis. Ceritinib was reported as an inhibitor of anaplastic lymphoma kinase (ALK) (1).
A previous study reported that the first-generation ALKtargeted compound crizotinib could prevent prostate cancerassociated bone loss (29). In this study, we aimed to investigate the effects of ceritinib on osteoclast formation in vitro and in vivo. We demonstrated that ceritinib also has a beneficial effect in preventing bone loss through the inhibition of osteoclast formation. Our findings suggest that ceritinib can not only suppress ALK-rearranged non-small cell lung cancer but also prevent osteoclastic osteolysis.
Cell proliferation assay (CCK-8)
The cell counting kit-8 (CCK-8, Dojindo, Japan) was used to examine the proliferation of BMMs. Cells were seeded into 96well plates at a density of 3 × 10 3 cells per well and cultured at 37°C with 5% CO2 for 12 h to allow the cells to adhere. The original medium was replaced with 10% (v/v) CCK-8/complete high glucose DMEM or a-MEM, and the cells were incubated for an additional 2 h. The Infinite M200Pro microplate reader (Tecan Trading AG, Hombrechtikon, Switzerland) was used to measure the absorbance at 450 nm. Each experiment consisted of three individual replicates.
Osteoclast differentiation assay and bone resorption assay
Primary bone marrow-derived cells were extracted from the bone marrow of hindlimbs of 6 to 8-week-old C57BL/6J mice. After culturing in complete a-MEM (containing 10% FBS and 1% penicillin/streptomycin) in the presence of M-CSF (50 ng/ ml, R&D Systems, MN, USA) for 4 days, all attached cells were treated as osteoclast precursor cells (bone marrow-derived monocytes and macrophages, BMMs), which were maintained and used for the following experiments. BMMs were subsequently induced to differentiate, using M-CSF (50 ng/ml) and RANKL (100 ng/ml, R&D Systems, MN, USA), for 5 days.
M-CSF-dependent BMMs were seeded into 96-well plates at a density of 1 × 104 cells/well in complete a-MEM and stimulated with 100 ng/ml of RANKL, as well as with different doses of ceritinib (CAS No.: 1032900-25-6; Cat. No.: HY-15656; purity: 99.98%; MedChemExpress LLC, China). The medium containing RANKL and the drug treatments were changed every two days. After 5-6 days of incubation, when large multinucleated osteoclasts were observed in the RANKL-only positive control group, all cells were fixed with 4% paraformaldehyde (PFA) and stained to detect tartrateresistant acid phosphatase (TRAP). The TRAP-positive (violet) cells of more than or equal to 3 nuclei were defined as multinucleated cells (MNCs). The number, total area or area per cell were circled out manually, and quantified within a well as a whole, using the open-source software Fiji. Each set of experiments was repeated trice.
Then for bone resorption assay, M-CSF-dependent BMMs were also seeded in hydroxyapatite-coated Osteo Assay Surface Polystyrene Microplates (Corning Inc., NY, United States) at a density of 1.3 × 104 cells/well in complete a-MEM and stimulated with 100 ng/ml RANKL and respective agents for 7 days. Thereafter, the wells were incubated in 5% sodium hypochlorite solution to remove the cells. The bone resorption pits on the microplates were captured using a phase-contrast i n v e r t e d l i g h t m i c r o s c o p e ( I X 7 1 ; O l y m p u s , Hamburg, Germany).
Western blotting
BMMs were pre-stimulated with ceritinib for 2 h in serumfree medium and then stimulated with RANKL (100 ng/ml) for the indicated time points. Total proteins from whole cells were extracted with RIPA Lysis Buffer (medium-level intensity) (Beyotime, Shanghai, China) in the presence of PMSF and protease inhibitor cocktail; subsequently, these proteins were electrophoretically separated using ExpressPlus PAGE (GenScript Laboratories, Piscataway, NJ, USA) and transferred onto PVDF membranes using an e-blot device (GenScript Laboratories, Piscataway, NJ, USA). The membranes were blocked with 5% (w/v) skim milk in tris-buffered saline (TBS) containing 0.1% (v/v) tween-20 (TBST, pH 7.4), and incubated for 1 h at room temperature (RT). The membranes were washed three times for 15 min with TBST and incubated for 12 h at 4°C with the following primary antibodies: p-Akt, Cell Signaling Technology #4060; Akt, Cell Signaling Technology #9272; p-p65, Cell Signaling Technology #3033; p65, Cell Signaling Technology #8242; p-ERK, Cell Signaling Technology #9101; ERK, Cell Signaling Technology #9101; and Actb, Cell Signaling Technology #3700. Membranes were washed thrice with TBST and incubated with either an IRDye 800CW anti-mouse secondary antibody (LI-COR) or an anti-rabbit secondary antibody conjugated to fluorescence (LI-COR), at 1:10000 dilution. Finally, the membranes were washed thrice with TBST and visualized using the Odyssey near-infrared (NIR) fluorescence imaging system (LI-COR, NE, USA). The grayscale ratio of each phospho-protein to its respective total protein was calculated. The ratios were then standardized, using the first control lane as the baseline (1.00), as shown below each western blot lane in the figures.
Animal model
Female C57BL/6J (12-week-old) mice were purchased from Jihui (Shanghai, China) and maintained in an SPF-grade animal facility. Mice were anesthetized and subjected to ovariectomy (OVX) or sham operation. The mice were under observation for 1 week, and then 18 mice, in good condition, were randomly allocated into three groups (Sham, OVX, OVX plus ceritinib, n = 6 per group); mice in each group were intraperitoneally injected with PBS, PBS, and ceritinib (10 mg/ kg body weight), respectively. The treatments were administered twice per week for a total of 8 weeks. At the end of the treatment period, all mice were euthanized and weighed. Hindlimb tissues were fixed in 4% PFA for 48 h. The tibiae were analyzed with micro-CT first and then decalcified in 10% EDTA (pH = 7.4) for 21 days.
Finally, all these tissues were embedded in paraffin and sliced for histological examination, including hematoxylin and eosin (H&E) staining and TRAP staining. TRAP staining results were evaluated based on two parameters: N. Oc (number of osteoclasts)/BS (bone surface) and Oc. S (osteoclast surface)/ BS. All procedures were performed following the protocols approved by the Institutional Animal Care and Use Committee of Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine.
Micro-CT
Bone morphometry within the metaphyseal region of the proximal tibia was quantified with micro-CT (SkyScan 1176, Billerica, MA, USA) using an X-ray tube voltage of 50 kV and a current of 500 mA with a 0.5 mm aluminum filter. The resolution was set to 9 mm, and 1.8 mm of the bone sample was acquired for each CT scan. The 3D images of the scans were reconstructed and analyzed using the SkyScan NRecon program. Trabecular morphometry was characterized by measuring the bone volume fraction (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), and trabecular spacing (Tb.Sp). Cortical morphometry was characterized by measuring the bone area (BA), tissue area (TA), cortical area fraction (BA/TA), and cortical thickness (Ct.Th).
Statistical analyses
All data were analyzed with one-way ANOVA using GraphPad Prism 8 (GraphPad Software, San Diego, CA, USA). Once a significant effect was detected, a post hoc Student's t-test was performed for further analysis. Results are presented as mean ± standard deviation (SD); number ≥ 3; all the p-values are displayed in the figures results; p < 0.05, p < 0.01, and p < 0.001 were considered statistically significant (*), highly statistically significant (**), and extremely statistically significant (***). To reduce the risk of type I errors, we considered |beta| ≥.20 as significant regarding the predictions.
Ceritinib revealed no obvious cell toxicity in BMMs
We examined whether the inhibitory effect of ceritinib on osteoclast formation was due to the inhibition of cell proliferation. Cell viability of BMMs was examined by incubating the cells with 150, 300, 600, or 1200 nM ceritinib. Compared to that in the control group, no obvious cell toxicity was detected in BMMs upon ceritinib treatment, indicating that ceritinib at concentrations of 150, 300, 600, or 1200 nM did not inhibit BMM proliferation until 96 hours (Figure 1).
Ceritinib inhibited osteoclastogenesis inhibited osteoclast-specific gene expression in vitro
To investigate whether ceritinib has any effect on osteoclast formation, we cultured BMMs with different concentrations of ceritinib ( Figure 2A). Increased number of TRAP-positive and multinucleated osteoclasts was observed in the control group ( Figure 2B). However, fewer TRAP-positive osteoclasts were B C D A Frontiers in Endocrinology frontiersin.org observed in the presence of 150, 300, 600 or 1200 nM ceritinib ( Figure 2B). We noticed a dose-dependent inhibitory effect of ceritinib on osteoclast formation. There was a significant decline in the number ( Figure 2B) and total area ( Figure 2C) of TRAPpositive multinucleated cells (MNCs) upon ceritinib treatment. The osteoclasts were smaller in the presence of ceritinib ( Figure 2D). Taken together, these data suggest that ceritinib inhibits osteoclast formation in vitro. Next, we examined the bone resorption activity of osteoclasts, the bone resorption pit areas were unobvious in the 150, 300, 600 and 1200 nM ceritinib group compared to the positive control group (Figures 2E, F).
In agreement with the suppression of osteoclast formation and resorption activity in vitro, the expression of osteoclastspecific genes, such as cathepsin K (Ctsk) (Figure 2G), acid phosphatase 5 (Acp5) ( Figure 2H), and nuclear factor of activated T-cells, cytoplasmic 1 (Nfatc1) (Figure 2I), was significantly inhibited by ceritinib at concentrations above 150 nM. Together, these data confirmed the inhibitory effect of ceritinib on osteoclast-specific gene expression.
Ceritinib suppressed the phosphorylation of Akt and p65
Based on the observed inhibitory effect of ceritinib on osteoclast formation, we aimed to elucidate the possible underlying mechanisms. BMMs were treated with RANKL for 0, 5, 10, 20, 30, and 60 min in the presence or absence of ceritinib. As shown in Figure 3, RANKL activated Akt phosphorylation from 0 to 20 min. However, decreased Akt activation was observed in the ceritinibtreated group. In addition, phosphorylation of p65 was observed in the RANKL-treated group. However, ceritinib almost completely abolished the activation of p65. Interestingly, it did not alter the activation of the ERK signaling pathway (Figure 3). Collectively, ceritinib inhibited Akt and p65 phosphorylation during osteoclast formation.
Ceritinib prevented ovariectomy-induced bone loss in vivo
Finally, to validate the inhibitory effect of ceritinib on osteoclast formation and function, we used an ovariectomyinduced osteoporosis model. Reconstructed images of the 3D scans of the left tibia are presented. There was a distinct decrease in trabecular and cortical bone volume in the OVX group, confirming the successful establishment of an ovariectomy animal model (Figures 4A-J). Compared with that in the OVX group of trabecular bone, higher BV/TV ( Figure 4B) and Tb.N ( Figure 4C), lower Tb.Sp ( Figure 4D), and non-significant Tb.Th ( Figure 4E) were observed in the OVX plus ceritinib group, suggesting that ceritinib prevented bone loss. Cortical bone results suggest no significant changes of BA ( Figure 4G), TA ( Figure 4H), BA/TA ( Figure 4I), and Ct.Th ( Figure 4J) in the OVX plus ceritinib group.
Further, histomorphometry analysis confirmed the preventive effect of ceritinib on bone loss. H&E staining confirmed the increased bone mass in the ceritinib-treated group compared to that in the OVX group. (Figure 4K). To assess the condition of osteoclasts in vivo, TRAP staining was performed, which showed fewer osteoclasts and decreased osteoclast surface per bone surface in the OVX plus ceritinib group compared to those in the OVX group (Figures 4L-N). Collectively, these data support the conclusion that ceritinib inhibited osteoclast formation in the OVX model.
Discussion
Ceritinib is typically used to treat patients with ALKrearranged non-small cell lung cancer. Our study demonstrated that, in addition to inhibiting cancer development, ceritinib was also beneficial in preventing bone loss. Our data showed that ceritinib effectively inhibited osteoclast formation and osteoclastspecific gene expression. More importantly, the inhibition of osteoclast formation resulted in the prevention of trabecular bone loss in vivo. Our result also suggested that ceritinib presented non-significant cortical bone loss prevention unlike the trabecular bone. According previous studies, cortical bone can be differentially affected by hormones and medications compared with trabecular one (30). Bone loss was first found in trabecular bone after menopause, and cortical bone later. The other explanation was that the effect of ceritinib on cortical bone probably need more time to observe. Notably, advanced nonsmall cell lung cancer metastasizing to the bone leads to osteolytic bone invasion (31)(32)(33). However, ceritinib shows great potential in repressing such skeletal-related events, besides suppressing tumor growth and extending survival.
ALK is a receptor tyrosine kinase that activates the PI3K-Akt signaling pathway (6). A previous study reported that overexpression of EML4-ALK variant 3 in HEK293T cells led to enhanced phosphorylation of Akt, whereas increased phosphorylation of ERK1/2 was not prominent in COS-7 cells (34). On the contrary, another study showed that treatment of NCI-H2228, a human non-small cell lung carcinoma cell line, with CH5424802 (a selective ALK inhibitor) led to reduction in p-Akt expression (35). Remarkably, crizotinib-resistant H3122CR-1 cells showed a dramatic downregulation of ALK and p-ALK and upregulation of the Akt/mTOR/S6 kinase pathways. Activation of autophagy in these cell lines positively altered the Akt/mTOR signaling pathway (36). The relationship between ALK and Akt indicated that ALK-driven neuroblastomas gradually acquired resistance to ALK inhibitors, but this effect was attenuated when combined with a p53 activator. This shift towards apoptosis, and away from cellcycle arrest, is mediated by inhibition of the ALK-Akt-FOXO3a (38,39). This suggests that ceritinib, an ALK inhibitor, reduced the activity of ALK in order to negatively control the activity of Akt. In our study, we demonstrated that the underlying mechanism was due to the inhibition of phosphorylation of the Akt signaling pathways, which are canonical targets during osteoclastogenesis. Notably, Akt activates NF-kB signaling through phosphorylation of IkB (40,41). A previous study suggested that in uterine carcinosarcomas, the ALK-mediated Akt/NF-kB/Twist1 pathway participates during the initial stage and regulates morphological alterations towards the sarcomatous phenotype (42). In addition, the induction of Akt was found to activate NF-kB/p65-dependent transcription, probably through repression of IkBa expression (43). Therefore, logically, the inhibition of p65 phosphorylation might be influenced by Akt inhibition in our study.
To the best of our knowledge, no previous study has reported that ceritinib can inhibit osteoclast formation and thus prevent bone loss. Interestingly, the first generation ALK inhibitor crizotinib has been reported to inhibit osteoclast formation and prevent prostate cancer bone destruction (29). These data suggest that ALK inhibitors have a preventive effect on bone loss. However, to generalize the applications in other metabolic bone diseases, the effect of ceritinib on other cells, FIGURE 3 Molecular mechanisms of ceritinib in RANKL-mediated osteoclastogenesis. The status of different signaling pathways during RANKL-induced osteoclastogenesis: ceritinib inhibited the activation of phospho-Akt (Ser473) and phospho-p65. However, there was no significant change in ERK signaling. such as osteoblasts, should also be taken into consideration. Also, to provide the stronger evidence to realistic bone invasion of cancer therapy, the administration of ceritinib on the mouse cancer models with bone loss (e.g., breast cancer, prostate cancer) should be analyzed in the future.
In conclusion, our data demonstrate that ceritinib can inhibit osteoclast formation by suppressing Akt and p65 phosphorylation, thereby preventing bone loss in vivo. Taken together, the use of ceritinib in patients with non-small cell lung cancer might improve their bone quality.
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
This study was reviewed and approved by Institutional Animal Care and Use Committee of Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine.
|
v2
|
2022-11-08T14:08:05.181Z
|
2022-11-08T00:00:00.000Z
|
253386061
|
s2orc/train
|
Immunogenic cell death-related gene landscape predicts the overall survival and immune infiltration status of ovarian cancer
Background: Ovarian cancer (OC) is the most troubling malignant tumor of the female reproductive system. It has a low early diagnosis rate and a high tumor recurrence rate after treatment. Immunogenic cell death (ICD) is a unique form of regulated cell death that can activate the adaptive immune system through the release of DAMPs and cytokines in immunocompromised hosts and establish long-term immunologic memory. Therefore, this study aims to explore the prognostic value and underlying mechanisms of ICD-related genes in OC on the basis of characteristics. Methods: The gene expression profiles and related clinical information of OC were downloaded from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) database. ICD-related genes were collected from the Genecards database. ICD-related prognostic genes were obtained by intersecting ICD-related genes with the OC prognostic-related genes that were analyzed in the TCGA database. Functional enrichment, genetic mutation, and immune infiltration correlation analyses were further performed to identify underlying mechanisms. Subsequently, we developed a TCGA cohort-based prognostic risk model that included a nine-gene signature through univariate and multivariate Cox regression and LASSO regression analyses. Meanwhile, external validation was performed on two sets of GEO cohorts and the TCGA training cohort for three other common tumors in women. In addition, a nomogram was established by integrating clinicopathological features and ICD-related gene signature to predict survival probability. Finally, functional enrichment and immune infiltration analyses were performed on the two risk subgroups. Results: By utilizing nine genes (ERBB2, RB1, CCR7, CD38, IFNB1, ANXA2, CXCL9, SLC9A1, and SLAMF7), we constructed an ICD-related prognostic signature. Subsequently, patients were subdivided into high- and low-risk subgroups in accordance with the median value of the risk score. In multivariate Cox regression analyses, risk score was an independent prognostic factor (hazard ratio = 2.783; p < 0.01). In the TCGA training cohort and the two GEO validation cohorts, patients with high-risk scores had worse prognosis than those with low-risk scores (p < 0.05). The time-dependent receiver operating characteristic curve further validated the prognostic power of the gene signature. Finally, gene set enrichment analysis indicated that multiple oncological pathways were significantly enriched in the high-risk subgroup. By contrast, the low-risk subgroup was strongly related to the immune-related signaling pathways. Immune infiltration analysis further illustrated that most immune cells showed higher levels of infiltration in the low-risk subgroup than in the high-risk subgroup. Conclusion: We constructed a novel ICD-related gene model for forecasting the prognosis and immune infiltration status of patients with OC. In the future, new ICD-related genes may provide novel potential targets for the therapeutic intervention of OC.
Introduction
Among all malignant diseases of the female reproductive system, ovarian cancer (OC) is one of the most troublesome. The most recent cancer statistics show that OC is expected to account for 19,880 new cases and 12,810 fatalities in the United States in 2022 (Siegel et al., 2022). OC has an incidence rate that ranks second among all gynecological tumor diseases (17.3%), but its mortality rate jumps to the first place (39.0%) mainly due to the following reasons: First, the ovaries are located deep in the pelvis, and the early stage of OC has almost no symptoms. Therefore, most patients are already in the advanced stage when they are diagnosed with OC (Wu et al., 2022). Second, the effectiveness of initial treatment in patients with OC has been limited due to widespread drug resistance, with 70%-80% of patients experiencing relapse within 2 years (Blagden et al., 2018). Despite advances in contemporary medical technology, the 5year survival rate for OC remains lower than 50% (Yang et al., 2022). Therefore, new biomarkers are urgently needed to predict and improve the prognosis of patients with OC.
Immunogenic cell death (ICD) is a unique form of regulatory cell death that can participate in immunity, as well activate fitness in immunocompetent hosts through the release of damageassociated molecular patterns (DAMPs) and cytokine immunity and establish long-term immune memory, which is critical for eradicating pathogens and balancing antitumor immunity to affect the tumor immune cycle (Galluzzi et al., 2020). Two anticancer drugs based on ICD have been developed. One is belantamab mafodotin, which was approved by the FDA in 2020 for the treatment of adult patients with relapsed or refractory multiple myeloma; this drug induces ICD in vitro and may contribute to T cell-mediated antitumor responses (Tzogani et al., 2021). The other is lurbinectedin, which has been approved by the FDA for the treatment of small cell lung cancer (Markham, 2020). Multiple ongoing clinical trials have shown that after ICDinducing chemotherapy, tumors tend to transition from "cold" tumors that respond poorly to immunotherapy to "hot" tumors that respond well to immune checkpoint inhibitors (Jia et al., 2020). The study of these ICD-based therapies undoubtedly provides a new direction for the study of the immunotherapy of "cold" tumors, such as OC. However, the role of ICD-related genes in OC prognosis is still largely unknown.
In this study, we downloaded the gene expression profiles and related clinical information of OC from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) database and collected ICD-related genes from the Genecards database. OC prognosis-related genes were screened from the TCGA cohort and intersected with the ICD-related gene set to obtain ICDrelated prognostic genes. Functional enrichment, genetic mutation, and immune infiltration analyses were further carried out to identify underlying mechanisms. Subsequently, through Cox and LASSO regression analyses, we developed a TCGA cohortbased prognostic risk model that included a nine-gene signature. At the same time, we performed external validation with the GEO cohort. Subsequently, we constructed a nomogram by integrating clinicopathological data and prognostic gene signatures to predict patient survival. Finally, we analyzed the functional and immunological differences between high-and low-risk subgroups depending on the above-mentioned risk groups.
Data collection and preprocessing
The RNA expression matrix, related clinical information, and RNA expression data from the clinical and follow-up information on OC, cervical cancer, endometrial cancer, and breast cancer were downloaded from TCGA database (https:// portal.gdc.cancer.gov/projects/TCGA-OV/). RNAseq data were converted from fragments per kilobase per million format into the transcripts per million (TPM) format and subsequently log2 (TPM + 1) transformed to shrink the numeric range of the data for further analysis. At the same time, samples with incomplete survival data were ruled out. By using the R "survival" package, the expression data were grouped by the median, and overall survival (OS)-related prognostic molecules were screened out through COX regression analysis. The obtained data were considered significant when p < 0.01 was satisfied. The Genecards Database (https://www.genecards.org/) is a comprehensive database that integrates genomic, transcriptomic, proteomics, genetic, clinical, and functional information and other resources and is freely available to users (Fishilevich et al., 2017). It was used to retrieve and download ICDrelated genes by applying the keyword "immunogenic cell death." Then, the relevance score provided by the database was utilized to screen out molecules for further research. In this work, the relevance score was computed by factoring in the importance of the different resources associating the gene with the disease (Safran et al., 2021). The median relevance score was set as the threshold to screen out the related genes with strong correlation. OC prognosis-related molecules and ICD-related genes were intersected to obtain the OC ICD-related prognostic gene set, which was used to construct a prognostic model. Finally, two datasets were downloaded from the GEO database (GSE26712 https://www.ncbi.nlm.nih.gov/geo/ query/acc.cgi?acc=GSE26712, GSE32062 https://www.ncbi.nlm. nih. gov/geo/query/acc.cgi?acc = GSE32062) for external validation.
Gene expression analysis
Normalized mRNA expression data that had been uniformly processed through the Toil (Vivian et al., 2017) processes of the TCGA-OV cohort and GTEx were downloaded from the UCSC Xena browser (https:// xenabrowser.net/datapages/), which was used to compare the expression of prognostic-relevant gene sets between tumor samples and normal samples. Before analysis and comparison, duplicate samples were removed, and the RNAseq data in TPM format were log2 transformed. Finally, the R "ggplot2" package was used for visualization.
Functional enrichment analysis
The STRING database (Szklarczyk et al., 2021) (https:// string-db.org/) aims to integrate all known and predicted physical interactions and functional associations between proteins. We used this database to obtain the protein-protein interaction network (PPI) of ICD-related genes. We further organized the PPI network lattice map by using the R "igraph" package. Next, the R "clusterProfiler" package was used to perform Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis to obtain a deepened understanding of the functional roles of the ICD-genes.
Genetic mutation and immune correlation analysis
The cBioPortal database (Gao et al., 2013) (http://www. cbioportal.org/) is a web-based repository for exploring, visualizing, and analyzing multidimensional cancer genomics data. We used this website to examine the genetic mutation status of 22 ICD-related genes in the TCGA cohort. Subsequently, we applied the ssGSEA algorithm built into the GSVA package to calculate the degree of infiltration of 24 types of immune cells (Bindea et al., 2013), including activated DC (aDC); B cells, CD8 T cells, cytotoxic cells, DC, eosinophils, immature DC (iDC), macrophages, mast cells, neutrophils, NK CD56bright cells, NK CD56dim cells, NK cells, plasmacytoid DC (pDC), T cells, T helper cells, T central memory (Tcm), T effector memory (Tem) cells, T follicular helper (Tfh) cells, T gamma delta (Tgd) cells, Th1 cells, Th17 cells, Th2 cells, and Treg cells.
Construction and validation of a prognostic ICD-Related gene signature
First, a univariate Cox regression model was utilized to evaluate the interaction between ICD-related genes and the OS of OC in the TCGA cohort. ICD-related genes with p < 0.05 were identified as factors with potential prognostic value. Subsequently, Lasso regression analysis based on the R "glmnet" package was applied to further screen variables. Ten-fold cross-validation was used to identify the optimal value of λ. Subsequently, genes screened by Lasso regression analysis were input into the multivariate Cox regression model, and the risk score of the ICD-related prognostic model was determined in accordance with the multivariate regression coefficient. Next, the samples were divided into the high-risk and low-risk subgroups on the basis of the median of the risk score as a cutoff. The risk score distribution, survival status, and heatmaps of characteristic ICD-related gene expression were also plotted. The OS Kaplan-Meier curve between the two subgroups was drawn by using the R "survival" package, and the log-rank test was performed. The receiver operating characteristic (ROC) curves of the 3-and 5-year survival rates between the two subgroups were analyzed by utilizing the R "timeROC" package to evaluate the predictive ability of the prediction model with the ICD-related gene signature. Finally, we also analyzed and plotted the progression-free interval (PFI) and disease-specific survival (DSS) Kaplan-Meier curves to estimate the Frontiers in Genetics frontiersin.org robustness of the prognostic model that we constructed. At the same time, external validation was performed on the two GEO cohorts GSE26712 and GSE32062. Finally, sample validation was performed with three other common female malignancies.
Establishment of the prognostic nomogram
The associations of relevant clinicopathological variables (age, stage, and residual tumor) and ICD-related risk score Frontiers in Genetics frontiersin.org 04 with OS were measured by using univariate and multivariate Cox proportional hazards regression models. Time-dependent ROC curves were further analyzed and plotted to evaluate the prognostic value of this nomogram. Furthermore, we drew calibration curves by using the R "rms" package and the R "survival" package to assess the agreement between actual and nomogram-predicted survival probabilities.
Functional and immunological analysis of ICD-Related prognostic signature
We performed gene set enrichment analysis (GSEA) on differentially expressed genes between high-risk and low-risk subgroups by using the "clusterProfiler" R package (Subramanian et al., 2005), which used the genome "c2. cp.kegg.v6.2. symbols.gmt" as the reference, to further investigate the underlying functional mechanisms of ICDrelated prognostic features. A gene was considered to be significantly enriched when it satisfied false discovery rate <0.25 and P. adjust <0.05. Subsequently, we compared the difference in immune-infiltrating cell scores between the high-and low-risk subgroups to further clarify the correlation of prognostic models with immune status.
Statistical analysis
All statistical analyses in this work were performed by using R software (v3.6.3, https://www.r-project.org/) and corresponding software packages. We used the Wilcoxon rank sum test to compare differences between groups. Spearman's rank correlation coefficients were calculated to determine the correlation between variables. Kaplan-Meier analysis using the log-rank test was used to assess survival between different subgroups. p < 0.05 was considered statistically significant unless stated otherwise.
Results
Before presenting the results, we provide the overall flow chart of this work in Figure 1 to help readers understand our work. The baseline clinical characteristics of the patients with OC in this study are summarized in Table 1. We enrolled 374 patients with OC from TCGA as the derivation cohort and 445 patients with OC from GEO as the validation cohort, which included 185 patients in the GSE26712 cohort and 260 patients in the GSE32062 cohort.
Biological function analysis of ICD-Related prognostic genes
We first applied the STRING database to analyze the PPI network of the 22 ICD-related prognostic genes. We set the parameter to "Homo sapiens" with medium confidence (0.400). We obtained enrichment p < 1.0e −16 ( Figure 3A). We further drew PPI network dot plots showing high and low expression levels and combined scores ( Figure 3B). Subsequently, we performed GO and KEGG enrichment analysis on the genes. Under the conditions of P. adj <0.05 and q value <0.2, we identified 216 BPs, 15 CCs, 5 MFs, and 18 KEGGs (Supplementary Table S3). GO enrichment analysis revealed that ICD-related prognostic genes were significantly enriched in a variety of immune-related biological processes. KEGG pathway analysis demonstrated that the enriched pathways of these cancer-related genes were mainly necroptosis, JAK−STAT signaling pathway, and natural killer cell-mediated cytotoxicity. We visualized three representatives of each item, and the results are presented in Figure 3C. Frontiers in Genetics frontiersin.org
Mutation and immunological analysis of ICD-Related prognostic genes
The extent of inherited mutations partly explains the role of genes in disease progression. We searched the cBioPortal website for genetic mutations in the 22 genes ( Figure 4A). We discovered that among the genes, RB1 had the highest mutation frequency, accounting for 11% of mutations, and its main mutation type was deep deletion. Five genes had the mutation frequency of 4%. They included ERBB2, SLAMF7, CEACAM1, ELN, and TAP1, and their mutation form was mainly amplification.
We applied the ssGSEA method to calculate the correlation between each factor and the 24 types of immune cells ( Figure 4B) to further evaluate the relationship between the 22 ICD-related prognostic genes and immune cell infiltration. Our results demonstrated that EPHA2, ERBB2, ICOSLG, and SLC9A1 were weakly associated with immune cells. In addition, except for SLC6A4, most of the other genes were positively correlated with immune cells. These findings echoed the results of the functional enrichment analysis discussed above.
Construction of ICD-Related prognostic signature in the TCGA cohort
First, we conducted univariate Cox regression analysis on the above 22 ICD-related genes, and our results revealed that 15 genes were significantly associated with the prognosis of patients with OC ( Figure 5A). Next, we included these prognostic genes in further LASSO Cox regression analysis. Nine characteristic ICD-related prognostic genes were identified on the basis of the base penalty parameter (λ) (Figures 5B,C). These genes included ERBB2, RB1, CCR7, CD38, IFNB1, ANXA2, CXCL9, SLC9A1, and SLAMF7.
On this basis, we calculated the risk score of our new prognostic model through multivariate Cox regression analysis Frontiers in Genetics frontiersin.org 08 by using the following formula: risk score = 0.0521608 × ERBB2 expression +0.20999961 × RB1 expression + (−0.0850947) × CCR7 expression + (−0.1610273) × CD38 expression + (−0.2437315) × IFNB1 expression +0.09886907 × ANXA2 expression + (−0.0133657) × CXCL9 expression +0.22922048 × SLC9A1 expression + (−0.0492765) × SLAMF7 expression + (−2.427119). Finally, we explicitly separated the patients in the TCGA cohort into the high-and low-risk subgroups on the basis of their intermediate survival scores. Risk factor maps, including risk score distributions, the survival status of patients, risk subgroups, and the heatmap of risk gene expression in the two cohorts were also drawn. As shown in Figure 6A, the high-risk subgroup had significantly more deaths than the low-risk subgroup. Similarly, Kaplan-Meier curve analysis illustrated that patients in the highrisk subgroup had a significantly poorer OS than those in the low-risk subgroup (p < 0.001, Figure 6B). Time-dependent ROC analysis was utilized to further assess OS predictive power ( Figure 6D), yielding the AUC values of 0.599 and 0.692 for the prediction of OS at 3 and 5 years, respectively. We also predicted survival differences in PFI and DSS between the two subgroups to demonstrate the robustness of our predictive model. Our results showed that the high-risk subgroup had worse PFI and DSS outcomes than the low-risk subgroup with p = 0.007 and p < 0.001, respectively ( Figures 6C,E).
Validation of the ICD-Related prognostic signature in the GEO cohort
We assessed the reproducibility of the prognostic signature in two independent GEO cohorts. (GSE26712 and GSE32062) to validate our findings in the TCGA cohort. Patients in both cohorts were divided into the high-and low-risk subgroups in accordance with the median risk score derived from the risk score formula above, and survival and time-dependent ROC curve analyses were performed on OS. Kaplan-Meier analysis showed that in both cohorts, the high-risk group had Frontiers in Genetics frontiersin.org 10 3.6 Validation of the ICD-Related prognostic signature in other common female malignancies We also performed survival analysis and time-dependent ROC curve analysis by using the data of three common female malignant tumors (cervical, endometrial, and breast cancers) in TCGA to further verify the universality of the prognostic signature that we constructed. Kaplan-Meier analysis showed that OS, PFI, and DSS in the high-risk group were significantly worse than those in the low-risk group. (p = 0.014, 0.005, and 0.024) (Supplementary Figures S1B,C,E). In addition, the results of time-dependent ROC curve analysis revealed that our constructed prognostic signature can well predict the survival time of cervical cancer (Supplementary Figure S1D). The AUC at 5 years, which had the value of 0.656, was the highest. However, the prediction signature that we constructed did not demonstrate obvious prognostic ability for the data of endometrial cancer (Supplementary Figure S2). Finally, for breast cancer (Supplementary Figure S3), the OS of the high-risk group was significantly worse (p = 0.032) than that of the low-risk group. However, no significant difference was found for PFI and DSS. Meanwhile, the results of the time-dependent ROC curve showed that the predictive ability of the ICD-related gene signature for breast cancer was not ideal.
Establishment of a reliable nomogram for the prediction of OC prognosis
Age, FIGO stage, residual tumor, and risk score were included in the univariate and multivariate Cox regression models. The univariate cox regression results showed that age, FIGO stage, residual tumor, and risk score were all significantly correlated with OS ( Figure 8A). Multivariate Cox regression analysis revealed that age, residual tumor of 1-10 mm, residual tumor >20 mm, and risk score did appear to be independent prognostic factors for the OS of patients with Frontiers in Genetics frontiersin.org OC ( Figure 8B). Next, we incorporated these clinicopathological parameters and the risk score to construct a nomogram to evaluate survival outcome ( Figure 8C), with high total points indicating the worsened prognosis of the patient. Timedependent ROC curves were further plotted to confirm that the nomogram was highly powerful in predicting patient survival Frontiers in Genetics frontiersin.org 12 outcomes ( Figure 8D). In addition, the calibration curve plotting the predicted probability of survival was generated. As shown in Figures 8E,F, the calibration curve and C index (Concordance = 0.659 [se = 0.023]) indicated that the prediction results of the nomogram had good fit.
Functional enrichment analysis of the ICD-Related prognostic signature and interaction with immune cell infiltration
We performed GSEA to characterize the biological differences between high-and low-risk subgroups as shown in Figures 9A,B. Cytokine-cytokine receptor interaction, chemokine signaling pathway, NK cell-mediated cytotoxicity, oxidative phosphorylation, systemic lupus erythematosus, JAK-STAT signaling pathway, cell adhesion molecule cams, T cell receptor signaling pathway, Toll-like receptor (TLR) signaling pathway, proteasome, and primary immunodeficiency were enriched in the low-risk subgroups. At the same time, tumorigenic pathways, such as hedgehog signaling pathway, ECM receptor interaction, WNT signaling pathway, calcium signaling pathway, pathways in cancer, MAPK signaling pathway, and TGF beta signaling pathway, were significantly activated in the high-risk subgroup, suggesting that the high-risk subgroup had a detrimental effect on survival outcomes.
We quantified the differences in the scores of 24 immuneinfiltrating cells between the high-and low-risk subgroups to further explore the correlation between our constructed genetic signature risk score and immune status. Wilcoxon rank sum test showed that only NK cells and Tcm were enriched in the highrisk subgroup. However, most immune-infiltrating cells were significantly enriched in the low-risk subgroup. These cells included pDC, T cells, T helper cells, TFH, Th1 cells, Th2 cells, Treg cells, aDC, B cells, CD8 T cells, cytotoxic cells, DC, macrophages, and NK CD56dim cells ( Figure 9C,D).
Discussion
OC, as the main cause of death by gynecological malignancies, is a perennial object of interest of researchers. New treatment options, such as targeted therapy, biological therapy, and immunotherapy, have been introduced, but their results continue to be unsatisfactory. In recent years, immune checkpoint blockers have been recognized as the most promising method for cancer treatment. However, in a variety of immunologically "cold" tumor types, including OC, their therapeutic efficacy is largely limited by factors, such as the lack of tumor antigens, the activation of T cells, priming, and infiltration (Bonaventura et al., 2019). Recently, the advent of ICD has heralded a new dawn in the diagnosis and treatment of Frontiers in Genetics frontiersin.org OC. Studies have clearly pointed out that ICD-based cancer vaccines can be immunogenic against "cold" tumors while increasing sensitivity to immunotherapy (Jin and Wang, 2021). Therefore, understanding the intimate relationship between differentially expressed genes in OC and ICD will allow us to explore the therapeutic potential of ICD-based treatments on a deep level. We performed the first comprehensive identification and investigation of the ICD-related prognostic signature of OC. Our work can expand the ideas for the improvement of OC prognostic prediction and the guidance of individualized treatment. In this study, we mined 22 ICD-related genes with distinct prognostic implications in OC. We applied unpaired samples to further analyze the expression of these 22 genes in OC tissue in TCGA and the corresponding normal tissue in GTEx. We found that only five genes (MITF, CCR7, JAK1, ELN, and SLC6A4) functioned as tumor suppressors, whereas the remaining 17 genes functioned as oncogenes. Next, we performed further functional analysis on these genes. As a result, we discovered that the vast majority were involved in immune-related biological processes. This finding was consistent with the characteristics of ICD. Our KEGG enrichment pathway analysis results showed that ICD-related genes were involved in the JAK-STAT signaling pathway and natural killer cell-mediated cytotoxicity. Previous studies have demonstrated that the ICD-related gene IFN can induce a variety of cell phenotypes by activating the JAK-STAT signaling pathway (Medrano et al., 2017). Minute et al. cocultured tumor cells and cytotoxic immune cells (such as T lymphocytes and NK cells) and observed the presence of ICD markers. They found that cytotoxic immune cells could induce the release of DAMPs, further triggering the antitumor immune response (Minute et al., 2020). Consistent with our enrichment analysis results, this phenomenon suggested that cytotoxicity is a type of ICD.
The immune system is our primary defense mechanism against exogenous and endogenous threats (Lakins et al., 2018). It mainly includes innate immune cells (such as macrophages, dendritic cells, and natural killer cells) and adaptive immune cells (such as T and B cells) (Chu et al., 2022). Numerous studies have demonstrated that genetic mutations in genes can affect tumor immune status (Xu et al., 2014). Mutations in U2AF1 have been shown to activate innate immune pathways in myeloid malignancies (Smith et al., 2019). Furthermore, p53 mutations can support immune dysfunction by altering the tumor microenvironment, disrupting innate immunity by modulating the TLR signaling pathway, and promoting immune privilege and the ability to survive by disrupting cell-mediated immunity (Agupitan et al., 2020). Therefore, we analyzed the genetic mutation status of ICDrelated genes and their association with immune-infiltrating cells. We found that only RB1 mutations were dominated by deep deletion mutations. By contrast, the mutations of other genes were dominated by amplification. The results of immune infiltration analysis showed that most genes were positively correlated with immune-infiltrating cells. This situation indicated that the prognostic model we constructed may have a certain predictive ability in immunotherapy.
Next, we constructed the nine-gene signature of the prognostic risk model by performing univariate/multivariate Cox regression and LASSO regression analyses. This signature included ERBB2, RB1, CCR7, CD38, IFNB1, ANXA2, CXCL9, SLC9A1, and SLAMF7. ERBB2, commonly referred to as HER2, is a oncogene located on chromosome 17 that encodes a member of the epidermal growth factor receptor family of receptor tyrosine kinases . ERBB2 amplification and mutation have been identified in many cancer types (Abrahao-Machado and Scapulatempo-Neto, 2016;Lin et al., 2021). They promote cancer cell growth and invasion and portend poor prognosis (Uchida et al., 2021). Currently developed nanoparticles targeting ERBB2 can enhance ICD effects at tumor sites (Zheng et al., 2020). CCR7 is a lymphocyte-specific G protein-coupled receptor (Birkenbach et al., 1993) that executes a unique antagonistic role in tumorigenesis by transferring tumor cells to the T cell region of lymph nodes (Zlotnik et al., 2011). It is essential for initiating adaptive immune responses. Interestingly, however, we found that some studies point to the opposite role of CCR7; specifically, CCR7 is induced in some cancer cells and contributes to metastasis formation (Gerken et al., 2022). CD38 is a messenger for intracellular calcium mobilization (Schmid et al., 2011), which participates in and regulates immune cell differentiation, activation, and tolerance (Malavasi et al., 2008). Several studies have shown that the expression of CD38 in tumors can induce proliferation and inhibit apoptosis (March et al., 2007) and participate in processes, such as tumor cell energy metabolism (Liao et al., 2020), and immune tolerance and resistance (Chen et al., 2018). IFNB1 is a cytokine of the wellknown signaling protein type I interferon family and is involved in cell differentiation and antitumor defense (Daman and Josefowicz, 2021). It has powerful antiproliferative, proapoptotic, antiangiogenesis, and immunomodulatory functions (Ambjorn et al., 2013). The ANXA2 gene encodes a member of the calcium-dependent phospholipid-binding protein family, which may play a role in regulating cell growth and signal transduction pathways (Bharadwaj et al., 2013). A growing body of evidence shows that the dysregulation of ANXA2 expression is associated with tumorigenesis and immunity in a variety of cancers, such as glioma , oral squamous cell carcinoma (Ma and Wang, 2021), pancreatic cancer (Karabulut et al., 2020), colorectal cancer (Hu et al., 2020), breast cancer (Sharma and Jain, 2020), thyroid cancer (Qin et al., 2020), and gastric cancer (Han et al., 2017). A recent experiment explored the effect of OC cell-derived exosomal ANXA2 on peritoneal implantation and tumor metastasis and its underlying mechanism (Gao et al., 2021). CXCL9, a member of the CXC chemokine subfamily, encodes secreted proteins that participate in immune regulation and inflammatory processes (Balkwill, 2004). Research on its role in tumors remains controversial. It can either act as an oncogenic factor to promote tumor progression or as a tumor suppressor to exert antitumor effects (Xiu and Luo, 2021). Our analysis showed that it may function as the former. SLC9A1 is a member of the solute carrier family 9. It encodes a protein that acts as a plasma membrane transporter with a crucial role in regulating pH homeostasis, cell migration, and cell volume (Cardone et al., 2019). SLAMF7 is a member of the signaling lymphocyte activating molecule family of receptors, which could be involved in adaptive immune responses. (Cannons et al., 2011). Studies have indicated that its expression has prognostic significance in cancers dominated by T cell exhaustion (O'Connell et al., 2021). In addition, a recent study found that SLAMF7 can mediate ICD in colorectal cancer cells (Roh et al., 2021). These genes can promote or inhibit tumor immune pathways through various mechanisms. However, whether they play an important role in the prognosis of patients with OC by affecting ICD remains unclear.
In recent years, on the basis of the characteristics of ICD, researchers have explored immunotherapy based on ICD in an attempt to overcome the limitations of conventional tumor treatment. However, the potential relationship between ICD and OC prognosis has yet to be elucidated. Our study divided OC samples from the TCGA database into high-and low-risk subgroups on the basis of the median risk score and assessed their prognostic value for OS, PFI, DSS, and 3-year and 5-year survival ROC curves. Patients in the high-risk subgroup had a significantly worse prognosis than those in the low-risk subgroup. The risk model we constructed had good predictive value for the 3-and 5-year survival rates of patients. In addition, we performed external validation by using two sets of GEO cohorts, further illustrating the power of the prognostic model. We also used the data on cervical, endometrial, and breast cancers in the TCGA database to validate the usefulness of our prognostic signature for multiple cancers, and our results showed that the prognostic signature that we constructed has a certain universality. Subsequently, we constructed a nomogram incorporating relevant clinicopathological factors to predict survival probability. After completing these works, we analyzed the differentially expressed genes between high-and low-risk subgroups then performed GSEA. Interestingly, we found that various cancer-related pathways were enriched in the high-risk subgroup. This situation further explained the poor prognosis of the high-risk subgroup. Enriched pathways in the low-risk subgroup were mainly closely related to the immune response. Studies have also illustrated that most T-cell markers, including CD8 + T cells, Th1, and Tem cells, were closely associated with good prognosis (Bindea et al., 2013). Therefore, we further analyzed the differences in the abundance of immune-infiltrating cells between the two groups. In coincidence with previous research findings, our results showed that most immune-infiltrating cells were significantly enriched in the low-risk subgroup. This situation suggested that the risk score formed by our nine-gene signature was inversely associated with immune cell infiltration and that our model may predict immune responses in tumors. Similarly, Wang et al. recently constructed an ICD-related classification signature to predict prognosis and immunotherapy response in head and neck squamous cell carcinoma . In their prognostic signature, the high-risk cohort score also corresponded to poor OS. In addition, patients with high-risk scores were inversely associated with CD8 T cells. This association was also confirmed in our study. Therefore, we can reasonably speculate that the identification of ICD-related biomarkers may be beneficial for a variety of malignant tumors and that these biomarkers can help identify patients with tumors who can benefit from immunotherapy .
In summary, we provided additional insights into the association between ICD-related genes and OC prognosis. Our newly constructed ICD signature demonstrated certain sensitivity and specificity as a prognostic predictor of OC. The prognostic nomogram based on the ICD-related signature also showed an excellent ability to forecast the OS of patients with OC. However, our research continues to have some deficiencies. The ICD-related prognostic model that we established was based only on public databases for bioinformatics analysis. In the future, we will perform experimental studies and validate the comprehensive roles of ICD-related genes in the progression of OC.
Data availability statement
The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material.
Author contributions
WZ conceived the study, wrote the manuscript, and completed the figures and tables. TL and LJ refined the method. JC and QL completed the validation of the data. WJ conceived the study and organized and edited the text.
Funding
This study was funded by Natural Science Foundation of Heilongjiang Province (YQ 2020H035), National Cancer Center Climbing Fund (NCC201908B07) and the Nn10 program of Harbin Medical University Cancer Hospital, China.
Frontiers in Genetics frontiersin.org
|
v2
|
2022-11-08T15:09:37.784Z
|
2022-11-08T00:00:00.000Z
|
253387297
|
s2orc/train
|
NAVIGATOR: an Italian regional imaging biobank to promote precision medicine for oncologic patients
NAVIGATOR is an Italian regional project boosting precision medicine in oncology with the aim of making it more predictive, preventive, and personalised by advancing translational research based on quantitative imaging and integrative omics analyses. The project’s goal is to develop an open imaging biobank for the collection and preservation of a large amount of standardised imaging multimodal datasets, including computed tomography, magnetic resonance imaging, and positron emission tomography data, together with the corresponding patient-related and omics-related relevant information extracted from regional healthcare services using an adapted privacy-preserving model. The project is based on an open-source imaging biobank and an open-science oriented virtual research environment (VRE). Available integrative omics and multi-imaging data of three use cases (prostate cancer, rectal cancer, and gastric cancer) will be collected. All data confined in NAVIGATOR (i.e., standard and novel imaging biomarkers, non-imaging data, health agency data) will be used to create a digital patient model, to support the reliable prediction of the disease phenotype and risk stratification. The VRE that relies on a well-established infrastructure, called D4Science.org, will further provide a multiset infrastructure for processing the integrative omics data, extracting specific radiomic signatures, and for identification and testing of novel imaging biomarkers through big data analytics and artificial intelligence.
Background
Biobanks are born as repositories of biological materials and related data, which are collected for a broad spectrum of further and future analysis (e.g., genetic, genomic, epigenetic, miRNA, proteomics, and transcriptomics) [1]. Only in the last few years, they have also begun to include imaging data, and real imaging biobanks have begun to exist.
The Imaging Biobanks Working Group of the Research Committee was established by the European Society of Radiology in 2014 [2]. They defined imaging biobanks as 'organized databases of medical images, and associated imaging biomarkers shared among multiple researchers, linked to other biorepositories' . From this definition, it is evident that an imaging biobank is not simply a system of archiving and transmitting images as are the picture archiving and communication systems (PACS) used in hospitals. An imaging biobank not only allows the storage and retrieval of medical images and associated metadata, but the added value is that these data are linked to the imaging biomarkers, extracted through the radiomics analysis of the imaging data (typically radiology imaging data) and to clinical, molecular, biological, and genomic data of a typical tissue biobank. The big potential of imaging biobanks is the ability to provide integrative omics data for studying advanced imaging techniques on imaging pools with sufficient sample size. This availability of data is necessary for the researchers to find an association between phenotype and genotype [3], to design and validate new imaging biomarkers, and to understand their biological significance, which may be a crucial point in precision medicine [4]. Once large high-quality and well-curated data sets are available within the biobank, they can be used for data mining.
In this article, we describe an establishment of Italian regional imaging biobank project, NAVIGATOR, its infrastructure, ethical and sustainability challenges, research and clinical consortium structure, the future development, and prospective.
A regional imaging biobank
NAVIGATOR is an Italian regional project that is designed to develop an imaging biobank open to various stakeholder communities and devoted initially to the oncological domain of solid tumours. The main deliverables and milestones of the project are summarised in Table 1.
In Europe, several projects aim to build an imaging biobank, such as the euCanSHare project [5], the PRIMAGE project [6], or the Lifebrain project [7], but most of them are under development as part of SC1-DT-TDS-05-2020 H2020 call ' AI for Health Imaging' [8]. The creation of an H2020 call entirely dedicated to the development of large cancer imaging repositories demonstrates that Europe has underlined the importance of access to multicentre cancer imaging datasets for the artificial intelligence (AI) communities and industries.
Despite this great international interest, in Italy, there are currently no examples of imaging biobanks; therefore, the creation of an oncologic imaging biobank in Tuscany is a challenge that could help the development of precision medicine in oncology in Italy. Although NAVIGA-TOR differs from PACS in several aspects, it will follow the work and future developments of the regional PACS initiative, to promptly deliver strategies and solutions for the smooth integration of the biobank with this system. This will further enable the positioning of NAVIGATOR as an additional service of the regional healthcare system. NAVIGATOR will deliver proper image acquisition guidelines and ethical, security, and privacy operation policies (i.e., governance) that may be needed to establish the biobank as a regional service. Then, it will be possible to connect NAVIGATOR to the other existing biobank infrastructures such as the European Biobanking and BioMolecular resources Research Infrastructure-European Research Infrastructure Consortium [9].
Nowadays, some imaging biomarkers are routinely used to support decision-making in cancer clinical management [10][11][12][13][14][15]. However, the integration of imaging biomarkers into clinical practice is still restricted to a limited number of imaging biomarkers due to several shortcomings: harmonisation of data acquisition and analysis, lack of international standards, and availability of good quality validated data sets. To overcome these criticalities, the development of imaging biobanks is an emerging field proposed as the radiological counterpart of the more common biobank of biological materials. The NAVIGATOR project inserts into this context aiming to the realisation of a regional imaging biobank.
Since the beginning of this last decade, radiomics has represented the first real attempt to transform the old qualitative or semiquantitative clinical imaging into the complete data-driven and quantitative perspective, trying to embrace the emerging paradigm of the science of big data and AI [16][17][18][19][20][21]. Specifically, since the NAVIGATOR project was created as an imaging biobank, the main focus will be the analysis of the information extracted from these images, using radiomics and conventional machine learning on the one hand and images and deep learning directly on the other. From a methodological point of view, we will focus on reproducibility, accountability, explainability, and interpretability of our machine learning models to obtain valuable and safe results. An example of the typical analysis workflow (radiomic feature extraction associated with conventional machine learning and deep learning processing) performed on a cohort of prostate cancer patients participating in the NAVIGATOR project can be found in the paper by Bertelli et al. [22]. The framework offered by NAVIGATOR has the key ingredients to get improvements in AI and radiomics applied to medical imaging: researchers with diverse backgrounds (clinicians, radiologists, biomedical engineers, information scientists, physicists, mathematicians) will join their efforts to collect large and curated data (following acquisition standards) and to design, develop, and perform processing pipelines based on deep learning and radiomics to provide the community with data and innovative tools for the automatic analysis of radiological data and metadata.
Ethical and legal challenges
Among the members of the European Union (EU), only a few legal systems have adopted a special law on research biobanks, establishing organisational requirements for the infrastructure and the management of biological samples, as well as the protection of the rights of the subjects involved, which must be reconciled with the needs of science. In the Italian legal system, which is characterised as a hybrid system [23], the discipline of the establishment of biobanks is placed in the face of the spaces left empty by the law. This discipline is often the result of the interpretations of individuals and soft law instruments operating on principles that are often formed in the practice of those individuals operating in the sector. Ultimately, it all depends on the absence of a law aimed at governing the phenomenon in an orderly manner. In the absence of a specific binding regulatory framework, the approach becomes how to properly use the existing regulation with other sectors. We appeal to the rules governing the protection of personal data and the provisions governing clinical trials, always making use of the classic categories of the Italian civil code. These schemes have allowed operators to trace a legal horizon; however, they do not always provide an adequate regime. Each model borrowed, albeit for different reasons, could be deficient and inappropriate. Even the EU Clinical Trials Regulation [24] is not a model that follows specific challenges and needs of research biobanks, primarily because the establishment of biobanks does not imply any clinical trials; moreover, the provisions on experimentation do not contemplate the hypothesis of a 'sharing of the material' differently from what happens in research [25]. The legislator will soon no longer be able to exempt itself from dealing with research biobanks as a research model in which different skills -ethics, information technology, medicine, law, and aspects of the administration -are called into question.
In the lack of specific regulation for the biobank phenomenon, the experience of others becomes precious. In this framework, the experience of the project NAVIGA-TOR could be analysed and taken into consideration.
Partnership
For the successful design and implementation of NAVI-GATOR, a very high level of interdisciplinary consortium was required, with expertise ranging from protocols for imaging acquisition and reporting to AI model development. The NAVIGATOR consortium relies on a strong regional network of hospitals, university hospitals, and research institutions, whose multidisciplinary expertise guarantees coverage of all the requirements.
On the clinical side, the NAVIGATOR consortium has the possibility of reuniting the three territorial areas of the public regional health system. This represents a strength of the project, which can therefore include a wide range of cases and therefore greater reliability and effectiveness in objective research. The University of Pisa (UNIPI), the Azienda USL Toscana Centro (AUSL TC), the Azienda Ospedaliera Universitaria Senese (AOUS), and the Azienda Ospedaliera Universitaria Careggi (AOUC) constitute the clinical team.
On the technical side, the knowledge and expertise essential for the successful development of NAVIGATOR are given by the research centres' partners which include the Institute of Information Science and Technologies of the National Research Partner Council (ISTI-CNR) and 'Nello Carrara' Institute of Applied Physics of the Partner National Research Council (IFAC-CNR). The Azienda Regionale di Sanità (ARS) will contribute with clinical records and data science expertise. The international actors partnered in NAVIGATOR will support the coordination with other emerging initiatives at the EU level, granting the international bases and timely outreach of project outcomes.
Ethical and legal aspects of a biobank: moving in the new world data sharing economy
Many ethical and legal issues, which focus on the balance between research activity and the rights and freedoms of individuals, are related to the institution of a research biobank. There is a lack of systematic international or national rules that address these issues. Therefore, this paragraph aims to present a reflection on some ethical-legal aspects addressed in the creation of the NAVIGATOR biobank. This means that the relationship between General Data Protection Regulation (EU) 2016/679 (GDPR) and images must first be understood. According to the Recital n. 51 to the GDPR, the processing of images should not systematically be the processing of special categories of personal data. Clinical images are covered by the definition of biometric data only when processed through a specific technical means that allows the unique identification or authentication of a natural person. In the case of a biobank for medical imaging, the images of clinical examinations do not qualify as personal data unless they can be associated with the personal data of the patient, even if they are pseudonymised. Biobanks for medical imaging will anonymise or pseudonymise the personal data of the patient. For this purpose, a list of personal data of the patient that can be stored and associated with the clinical images for a biobank has been realised at the starting point of the project. The list supports the selection of old clinical images and personal data associated with patients that were acquired before the establishment of the biobank for medical imaging when former patients have not given specific informed consent for biobanking activities. The list of personal data of the patient that can be stored and associated with the clinical images has an impact on several matters for the future biobank: the joint controller agreement for managing health data in the project consortium, the secondary use of associated medical imaging and personal data acquired before the constitution of the biobank, and the consent for the biobanking purpose.
Joint controller agreement for managing personal health data under GDPR
The issue of sharing personal data in large research consortiums or biobank infrastructures commonly arises and similar cases in which data processing takes place in an intragroup context [26]. According to the aims of the GDPR, organisations are obliged to demonstrate that their processing activities are compliant with the Data Protection Principles (Rec. 85; Art. 5 (2) and Rec. 74; Art. 24 GDPR). An arrangement between joint controllers can help organisations to demonstrate compliance with all the principles of the regulation: principles of lawfulness, fairness and transparency, purpose limitation, data minimisation, accuracy, storage limitation, integrity, and confidentiality. Under the NAVIGATOR project, more controllers jointly determine the purposes and means of the processing of personal data, and they are thus joint controllers (Rec. 79; Art. 4 (7) and Art. 26). The essence of the arrangement shall be made available to the data subject, and special attention will be posted on NAVIGA-TOR service infrastructure to optimise the value of imaging data in connection with other clinical data (clinical health records). The service infrastructure in the NAVI-GATOR project is a platform to manage health personal data. The Guidelines 07/2020 on the concepts of controller and processor in the GDPR, adopted on September 02, 2020, by the European Data Protection Board, remind us that research projects by several research institutes can decide to participate in a specific joint research project and to use to that end the existing platform of one of the institutes involved in the project. Each institute feeds personal data it holds into the platform for joint research and uses the data provided by others through the platform for carrying out the research. In this case, all institutes qualify as joint controllers for the personal data processing that is done by storing and disclosing information from this platform since they have decided together the purpose of the processing and the means to be used (the existing platform). Each of the institutes however is a separate controller for any other processing that may be carried out outside the platform for their respective purposes. A joint controller agreement among the NAVIGATOR partners will be arranged with the elements established by Art. 26 GDPR and will reflect the respective roles and relationships of the joint controllers vis-à-vis the data subjects as well as considering the service infrastructure.
Consent procedures for collecting medical images
The emergence of biobanks as a vital research tool in the medical sciences has led to a widespread debate in the literature about how to best handle the consent procedures governing the enrolment of participants in research and the subsequent use of participant samples and data in other studies. Informed consent is an extremely important tool to implement various international, EU, and national mandatory laws (e.g., on the protection of personal data, use of biological material, clinical trials) and, above all, to make biomedical activities consistent with fundamental ethical principles, such as dignity and self-determination (see, e.g., Charter of Fundamental Rights of the European Union, European Convention on Human Rights, Convention on Human Rights and Biomedicine -Oviedo, 4 April 1997, and its Additional Protocols).
To comply with these principles and laws, information must be provided in a way that meets the requirements of the different legal sources in a text that addresses the different topics. Therefore, not only the requirements of a specific legal framework, even as extremely important as Regulation (EU) 2016/679, should be considered but the entire legal and ethical framework underlying informed consent. In any case, Regulation (EU) 2016/679 on the protection of individuals regarding the processing of personal data and on the free movement of such data (hereinafter GDPR) lays down special rules for consent and rights in the context of research activities. When the broad consent model is applied for biobanking activities, general consent is gathered at the time of enrolment. Subsequently, samples stored in the biobank can be used for new studies that fall within the scope of the consent without reobtaining consent from participants. Medical researchers defend the broad model by arguing that it is the best way to make large-scale biobank research feasible.
The GDPR considers the situation in which it is not possible to fully identify the purposes of personal data processing for scientific research at the time of data collection. The derogation of the principle of the 'granularity' in research is allowed by the 'Recital' no. 33 of the GDPR; the GDPR and other European sources extend the effectiveness of consent. The principle of the limitation of purpose prescribes that 'the processing of personal data for purposes other than those for which the personal data were initially collected should be allowed only where the processing is compatible with the purposes for which the personal data were initially collected' (Recital 50), but nevertheless, 'further processing for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes shall, in accordance with Article 89 (1), not be considered to be incompatible with the initial purposes' (Article 5 (1) (b) GDPR). For purposes of this type, a sort of presumed consent is given. The same approach was chosen by the Council of Europe in its Recommendation on the Protection of Health-Related Data of 2019 which replaced the above-mentioned Recommendation of 1997 (see Article 4 (1) (b)). This recommendation also seems to consider that it may be difficult to provide detailed information to the data subject about the use of health-related data at the time of collection (see Article 11 (2)). Starting from the aforementioned framework, dealing with the problem of future research depends on the current state of knowledge. The NAVIGATOR consent form will be realised according to the Recital n. 33, and so, data subjects shall 'have the opportunity to give their consent only to certain areas of research or parts of research projects to the extent allowed by the intended purpose' if -and only if -the research project is 'in keeping with recognised ethical standards for scientific research' . Despite this, such consent should still be in line with the applicable ethical standards for scientific research. Taking into consideration the last affirmation and the meaning of scientific research in the EU framework, we can identify the possible problems in dealing with future research. In relation to biobanking activities and building a research database, the future research has lawful broad consent for the use and reuse of personal data only if further activity could qualify as a 'genuine research project' developed in the framework of the activities of public or private research organisations. This is true if the future research is set up by relevant sector-related methodological and ethical standards in conformity with good practice and with the proper oversight of an ethics committee.
Use cases
The NAVIGATOR clinical team (University of Pisa, Azienda USL Toscana Centro, Azienda Ospedaliera Universitaria Senese, Azienda Ospedaliera Universitaria Careggi) will guarantee the clinical case study for the biobank. Three major abdominal neoplasms have been chosen: prostate cancer, rectal cancer, and gastric cancer.
Prostate cancer is the most prevalent male malignancy [27]. Treatment recommendations are currently based on three levels of risk stratification (low, intermediate, and high) according to prostate-specific antigen, Gleason score, and T category, with active surveillance offered to low-risk men [28]. However, some low-risk patients may harbour diseases that are more aggressive; thus, risk stratification needs to be improved. Magnetic resonance imaging (MRI) plays a crucial role in imaging prostate cancer [29].
Colorectal cancer is the third most common cancer and the third leading cause of cancer deaths in both males and females [30]. The interest for this neoplasm is related to the identification of the locally advanced cancer (T3 stage) with a complete response to neoadjuvant chemoradiotherapy that could be excluded from surgery [31]. The identification of such patients prior to treatment and the response to treatment are based only on MRI imaging [32].
Gastric cancer is the fourth most common neoplasm in the world and the third leading cause of cancer death [33]. The treatment is complex and multimodal today, and the entire decision-making process is largely driven by imaging [34]. The task of imaging is to be able to discern between early and advanced forms of gastric cancer, with the possibility of endoscopic versus surgical resections and/or neo-adjuvant therapy, as well as recognise forms liable to multimodal treatment. The prognostic value of microsatellite instability in intestinal-type noncardia cancer and the different clinical impact of gastric cancers are interesting goals to predict in the pretreatment phase [35]. These neoplasms are divided into mesenchymal type and epithelial type, based on molecular characteristics [35].
The reason for the choice of these three neoplasms was slightly different: for prostate cancer, to improve risk stratification; for rectal cancer, to identify with certainty the complete response to neoadjuvant treatment in the case of stage T3; and for gastric cancer, to identify those forms liable to multimodal treatment and in particular the ones which respond to neoadjuvant therapy and, thus, may benefit from aggressive multimodal treatment. The radiomics-based biomarkers may be an additional tool to better elucidate these questions.
Platform architecture
The ultimate objective of NAVIGATOR is to deliver the technological platform whose high-level architecture, depicted in Fig. 1, will provide the following: • An innovative open imaging biobank was heterogeneous, anonymised (or pseudonymised) imaging, and related patient data will be collected from multiple clinical sources and shared among multiple researchers, by relying on a flexible data model complaint to state-of-the-art clinical data and metadata standards. • A virtual research environment (VRE) that will offer web access to a digital laboratory where clinicians, scientists, and clinical stakeholders can experiment, discover, test, and validate novel or known biomarkers over the biobank, via a set of available data analytics tools based on radiomics and AI algorithms (Fig. 1).
The biobank
The NAVIGATOR imaging biobank will integrate heterogeneous, pseudonymised patient-related data from multiple clinical sources, ranging from PACS archives, relative medical reports, and biomarkers to public clinical-health records collected from the database of the Agenzia Regionale di Sanità (Toscana, Italy). The anonymisation tools integrated into the clinical PACS will serve the deletion of sensitive information (i.e., patient's name and date of birth), which will be maintained by the clinical centres according to the definition of GDPR Recital 26. The underlying patient centric, agnostic data model will be shaped up by well-established metadata standards -such as Digital Imaging and Communications in Medicine (DICOM), Minimum Information About BIobank data Sharing (MIABIS), Observational Medical Outcomes Partnership (OMOP)-common Data Model (DM) -and devised to flexibly adapt to any form of patientrelated data that may be identified and integrated into the biobank in the future. This will ensure the biobank scalability and ability to integrate with tumour types beyond the ones targeted by the project. An initial stage of the project will identify all the clinical and molecular variables to be stored and linked to the imaging data. These variables will be suitably stored in a database implementing an electronic clinical reporting form, while the imaging data will be stored in a dedicated repository of DICOM files, as better detailed in the following.
The NAVIGATOR VRE
The NAVIGATOR VRE will offer web access to a digital laboratory' where clinicians, scientists, and clinical stakeholders can upload data into the biobank and collaborate, experiment, discover, test, and validate novel or known biomarkers, thanks to available data analytics tools based on radiomics and AI libraries. VRE will supply the users with the following: • User dashboard: users can access a local space where they can keep and share data via an online, accesscontrolled file system (dropbox like); they can integrate, test, and share their methods with others and exchange messages via social tools and mailing lists. • Experiment handling: users can integrate methods to perform experimental actions in this domain, such as image normalisation, segmentation, feature extraction, classification, and regression; can select existing methods, configure, and experiment with the biobank data or manually uploaded images; and can share their methods and data for the sake of science reproducibility. • Biobank management: users can upload studies and related images, as well as anonymised patient and diagnostic data, into the biobank. The VRE will implement a framework envisioned by many in the domain [36,37], enabling scientific research communities to mine the rich content of multimedia health data and to carry out various types of proofs and validation of imaging biomarkers in a controlled and collaboration-based pipeline environment. Following openscience principles, NAVIGATOR will sustain cooperative research to ensure reproducibility and verifiability of biomarker models, thus fostering standardisation while reducing variability.
Platform implementation
The NAVIGATOR VRE will be delivered as an extension and customisation of the D4Sci ence. org infrastructure, a well-established production-ready technology today supporting the operation of various European H2020 and Horizon Europe Research Infrastructures (e.g., SoBig-Data.eu [38], AriadnePlus [https:// ariad ne-infra struc ture. eu/], BlueCloud [https:// blue-cloud. org/], AGINFRA [https:// plus. aginf ra. eu/]). This infrastructure is in turn powered by the gCube software toolkit 1 . VREs, science gateways, virtual laboratories, and other similar terms [39] are used to indicate web-based systems emerged to provide researchers with integrated and user-friendly (transparent) access to data, services, and computing resources of interest for a given investigation that is usually spread across many and diverse data and computing infrastructures. VREs hide to scientists, often without any information technology background, the complexity of sophisticated computing stacks and provide them with intuitive user interfaces they can use to perform experiments, enact collaboration among colleagues, and control access to their algorithms and data, for the sake of investigations.
Many frameworks can be used to build such systems. Shahand et al. [40] have identified eleven frameworks explicitly exploited to develop science gateways including Apache Airavata [https:// airav ata. apache. org/], Catania SG Gateway [41], Globus [42], HUBzero(+Pegasus) [43], ICAT Job Portal [44], WS-PGRADE/gUSE [45]. The D4Science infrastructure is singles out for its innovative approach offering components and tools to operate a data service infrastructure capable of supporting the dynamic creation of custom VREs by integrating data, algorithms, and services from specific disciplines. Its VREs, available at D4sci ence. org via dedicated portals, are today serving different disciplinary domains [46][47][48][49], which (i) include services that operate on the same cloud storage and computation resources, located at Institute of Information Science and Technologies of the National Research Partner Council and partly at GARR Consortium, 2 and (ii) can rely on state-of-the-art big data analysis tools built on such cloud resources. Overall, D4Science is currently serving more than thousands of users (more than 7,000 in September 2018). In the period January-September 2018, the users served by this infrastructure and its VREs performed: a total of 50,127 sessions, with an average of about 5,569 sessions per month; a total of 4,288 social interactions, with an average of about 476 interactions per month; and a total of 150 million of analytics tasks, with an average of about 16 million tasks per month. The experiences made while exploiting gCube to operate the D4Sci ence. org infrastructure demonstrate not only the production readiness and flexibility of customisation of the software but also that the principles governing VREs delivery and system openness are key in the modern science settings [50].
By realising the NAVIGATOR infrastructure on top of D4Science, we will, therefore, (i) ensure the ability of the biobank and VREs to be sustained over time, (ii) deliver all the flexibility required to extend VRE functionality and biobank features in the future, and (iii) operate a production system by adopting an economy of scale, scalable, and performing approach.
As shown in Fig. 2, D4Science operates a three-tier architecture whose components (orange boxes) can be customised to different use cases and applications and be delivered as web applications, via dedicated VREs, to specific groups of users. More specifically, D4Science will be instantiated to deliver the NAVIGATOR biobank and applications as explained in following paragraphs.
Biobank
The storage framework of D4Science offers a distributed file system accessible to users (humans and services) and a database overlay, supporting several data management systems as Fig. 2), in combination with a structured repository of DICOM files for the storage of the related images (i.e., file-based store in Fig.2). The image upload phase ensures that DICOM tags are extracted from the images and inserted into the database to enable query and data filtering based on these.
VRE
It is instantiated to provide the three sets of functionalities described in the architecture section (red boxes).
The web applications will be developed by the project to address the specific requirements of the NAVIGA-TOR community, ensuring different user typologies are granted access to the application and data with dedicated access rights. The Collaborative Framework offers tools to exchange messages and experiences as well as a workspace (personal file system) where data can be uploaded and shared via fine-grained access control with other users. The analytics framework offers Jupyter notebooks and RStudio environments that users can adopt out of the box to operate over data extracted from the biobank (or otherwise uploaded by the user in the workspace); in scope, the method engine allows scientists to integrate their custom data analytics methods (Java, Python, R, C++, etc.) and web services to share them with other users (individuals or groups) as a service.
Precision and personalised medicine
Medicine is experiencing a major change in the last decades, moving from reactive to proactive approaches by providing personalised medical solutions. As far as oncology is concerned, precision medicine highly relies on increasingly detailed characterisation of disease states using the multiple omics platforms for better individualise diagnostics, prognostics, and therapeutics, thus setting a cornerstone of integrative omics medicine, the last frontier of medical sciences. Precision medicine holds promise for better personalisation of oncological care; meantime, the personalisation strongly relies on modelling individual case characteristics and their variability, as opposed to the classical approach, named 'one size fits all' , based on generalised protocols derived from (average) evaluations of entire populations. To achieve a personalised approach presupposes (A) the integration of heterogeneous patients' data with a holistic approach, to account for tumour genotypic and phenotypic heterogeneity, and (B) an omics perspective, i.e., the production, via analytics methods, of a large amount of data mineable to describe and interpret tumour biology and evolution [51][52][53][54].
The NAVIGATOR biobank will allow for the collection and preservation of a large amount of high-quality, standardised imaging data and related omics data in a privacy-preserving model, including computed tomography, magnetic resonance imaging, and positron emission tomography data for various tumour settings, patients' clinical data from regional healthcare services, and other omics data. These imaging data will be used for the extraction of imaging biomarkers based on image analysis tools for radiomics and the identification and testing of novel imaging biomarkers through big data analytics and AI.
Digital patient model
Knowledge extraction in big data analysis is an evergreen challenge, and many techniques have been developed and tested in many application domains. Of course, in any domain, the dataset's dimension and the number of data types considered are critical parameters: a highly heterogeneous set of data could be very difficult to interpret, not depending on the specific task. In oncology, current research is going towards merging radiomics into holomics (i.e., secured access, sharing, and integration of all health data) for precision/personalised medicine: radiomics algorithms now include genomics and immunomics data to improve patient stratification [55]. For example, radiogenomics and radioimmunomics, alone or in combination with other data, improve the accuracy of the prediction of prognosis, treatment response, and outcomes (overall survival or toxicity) [56][57][58][59].
In NAVIGATOR, an initial large set of imaging and integrative omics data will be collected for three cancer types. A key issue for data collection is to have proper patient stratification. Even if unsupervised techniques have been used largely in genomics for patient stratification, such as clustering [60], deep neural networks [61,62], and topological data analysis [63], they suffer from interpretability and could not be considered trustworthy by medical doctors. All the data collected in NAVIGA-TOR (i.e., standard and novel imaging biomarkers, nonimaging data, and health agency data) will be integrated with prior medical knowledge (for patient stratification) and used to create a digital patient model, to support the reliable prediction of the disease phenotype and patients' risk stratification.
Merging insights from clinical-, imaging-, and molecular-based data in the digital patient model will ensure a more comprehensive risk stratification in oncology with high accuracy, paving the way to true precision oncology. Of course, specific issues arising from the use of AI algorithms and data Analysis will be addressed, such as reproducibility, bias assessment, and the monitoring of the prediction performances.
Data integration
In NAVIGATOR, data integration is a fundamental activity as it provides the necessary technological means to connect different and heterogeneous data sources. The purpose of the biobank of NAVIGATOR is to set up an infrastructure that can store and give access to images and biological markers for three different cancer types. Besides these kinds of datasets, the NAVIGA-TOR biobank also considers exploiting an additional data source, provided by Agenzia Regionale di Sanità, containing medical administrative records. These records contain information about the interaction of patients with the local health service, including their hospitalisation, prescribed drugs, and others. This data is different from the typical focused data on a specific pathology but can provide an overview of the patient's historical health life.
The NAVIGATOR project plans to evaluate the use of this data source in several aspects, including the following: • To enable the extraction of biomarkers considering also the administrative medical information of the patient • To help in assessing the accuracy of biomarkers by comparing their relevance with patients affected with different administrative records • To determine the correspondence of the medical administrative history of a patient to the values of specific biomarkers
Ethical and legal aspects
There are many questions to be addressed in the creation of a biobank, including the scale of collaboration, the need for data sharing, written rules, and regulations to cover areas of responsibility from data ownership to research dissemination [26]. The formal management structure arising from NAVIGATOR project will be realised through the following: • A joint controllers agreement for health personal data management, reflecting the compliance with principles of lawfulness, fairness and transparency, purpose limitation, data minimisation, accuracy, storage limitation, integrity, and confidentiality • The realisation of a dedicated broad consent defining the meaning of future research (primary use) and a systematic approach for the collecting of medical imaging and clinical data • The adoption of a protocol for the reuse of archived personal data that will be subject to an initial and continuous assessment by the competent ethical committees and by other authorities • The correct conservation and management of images and connected data • The guarantee of the provision of all useful information to the person from whom the material was origi- nally taken and the protection of the integrity of the images and personal data itself • The respect of the duty to inform donors, families, institutions, and public and private entities about who (research groups or laboratories) collaborate with the infrastructure or about the results of the activities carried out.
Conclusion
The general goal of the NAVIGATOR project is to advance imaging-based translational research for precision medicine in oncology, through quantitative imaging and integrative omics analyses. To this aim, NAVIGA-TOR will deliver an infrastructure that will offer the collection and preservation into an imaging biobank with a large amount of imaging and related omics data. To process the integrative omics data, a dedicated VRE tools will be available for the extraction of imaging biomarkers and the identification and testing of novel imaging biomarkers. The biobank will contain an initial large set of imaging and omics data (more than 1,500 patient cases), collected during the project course, for three cancer types that cover the most clinically relevant and impacting cases. For each considered cancer type will be created a digital patient model which schematises the cancer phenotyping, stratified risks, and responsivity to therapy. Finally, at the end of the project, the main issues related to the compliance with legal and ethical issues for the creation of an imaging biobank will be addressed. The achieving of this goal is essential given the lack of systematic international or national rules that address the whole of these issues.
|
v2
|
2022-11-09T06:16:57.469Z
|
2022-11-08T00:00:00.000Z
|
253396167
|
s2ag/train
|
Folic acid engineered sulforaphane loaded microbeads for targeting breast cancer.
Non-targeted cancer therapy poses a huge risk to the cancer patients' life due to high toxicity offered by chemotherapy. Breast carcinoma is one of such deleterious disease, demanding a highly effectual treatment option which could reduce the toxicity and extend survival rate. Since, folate receptors extensively display themselves on the cancer cell surface, targeting them would help to ameliorate the progression and metastasis. Considering this, we envisaged and developed sulforaphane loaded folate engineered microbeads to target breast cancer cells over-expressing folate receptors. The surface engineered microbeads were optimized and developed using emulsion gelation technique, among which the best developed preparation demonstrated the particle size of 1302 ± 3.98 µm, % EE of 84.1 ± 3.32% and in vitro drug release of 98.1 ± 4.42%@24h. The spherical sized microbead showed controlled release with improved haem-compatibility in comparison to the bare drug. Free radical scavenging activity by ABTS assay showed strong anti-oxidant activity (IC50-20.62 µg/ml) of the targeted microbeads with profound cancer cell suppressing effect ((IC50-17.48 ± 3.5 µM) as observed in MCF-7 cells by MTT assay. Finally, in comparison to lone SFN, the targeted therapy showed enhanced uptake by the intestinal villi indicating a suitable oral targeted therapy against breast carcinoma.
|
v2
|
2022-11-09T06:16:57.521Z
|
2022-11-08T00:00:00.000Z
|
253396162
|
s2ag/train
|
Docetaxel and fluorouracil as first-line therapy for gastric cancer with bone marrow metastasis and disseminated intravascular coagulation.
Gastric cancer with bone marrow metastasis and disseminated intravascular coagulation constitutes a highly aggressive gastric cancer subtype which presents a peculiar biological behavior and very poor prognosis. Retrospective studies have shown chemotherapy could prolong survival, but a prospective trial is still unavailable. This study is the first prospective clinical trial to evaluate the safety and efficacy of chemotherapy for advanced gastric cancer patients with bone marrow metastasis.
|
v2
|
2022-11-09T06:16:57.553Z
|
2022-11-08T00:00:00.000Z
|
253396367
|
s2ag/train
|
The genetic polymorphisms in the SP4 gene and the risk of gastric cancer.
Aim: Gastric cancer (GC) is the leading cause of cancer death, and is associated with host genetic factors. This study aimed to determine the impact of SP4 polymorphisms on GC. Materials & methods: Four hundred and eighty-nine GC patients and 481 healthy subjects were recruited. The association between single nucleotide polymorphisms and GC risk was investigated by logistic regression analysis. Results: It was observed that rs39302 and rs7811417 were related to a decreased GC risk. Stratified analyses showed that rs39302 decreased GC susceptibility at ages ≤60 years, in men, GC patients who had previously smoked and drank. rs7811417 had a risk-decreasing impact on the patients aged ≤60 years, in men, GC patients who were nonsmoking and nondrinking. rs35929923 decreased the GC risk of patients in grade III-IV and the lymph node metastasis subgroup. Conclusion: SP4 gene polymorphisms are associated with GC risk.
|
v2
|
2022-11-09T06:16:57.612Z
|
2022-11-08T00:00:00.000Z
|
253396623
|
s2ag/train
|
Disparity in checkpoint inhibitor utilization among commercially insured adult patients with metastatic lung cancer.
BACKGROUND
There is a lack of evidence from nationwide sample on the disparity of initiating immune checkpoint inhibitors (ICIs) after metastatic lung cancer diagnosis.
METHODS
We identified metastatic lung cancer patients diagnosed between 2015 and 2020 from a large nationwide commercial claims database. We analyzed the time from metastatic lung cancer diagnosis to ICI therapy using Cox proportional hazard models. Independent variables included county-level measures (quintiles of percentage of racialized population, quintiles of percentage of population below poverty, urbanity, and density of medical oncologists) and patient characteristics (age, sex, Charlson comorbidity index, Medicare Advantage, and year of diagnosis). All tests were two-sided.
RESULTS
A total of 17,022 patients were included. Counties with a larger proportion of racialized population appeared to be more urban, have a greater percentage of its residents in poverty, and have a higher density of medical oncologists. In Cox analysis, the adjusted hazard ratio of the second, third, fourth, and highest quintile of percentage of racialized population were 0.89 (95% confidence interval [CI]: 0.82-0.98), 0.85 (95% CI: 0.78-0.93), 0.78 (95% CI: 0.71-0.86), and 0.71 (95% CI: 0.62-0.81), respectively, compared to counties in the lowest quintile. The slower ICI therapy initiation was driven by counties with the highest percentage of Hispanic population and other non-Black racialized groups.
CONCLUSIONS
Commercially insured patients with metastatic lung cancer who lived in counties with greater percentage of racialized population had slower initiation of ICI therapy after lung cancer diagnosis, despite greater density of oncologists in their neighborhood.
|
v2
|
2022-11-09T06:16:57.739Z
|
2022-11-08T00:00:00.000Z
|
253396278
|
s2ag/train
|
The prognostic value of ADAMTS8 and its role as a tumor suppressor in breast cancer.
A disintegrin-like and metalloprotease with therombospondin type1 motif 8 (ADAMTS8) plays an important role in many malignancies. However, the clinical and biological significance of ADAMTS8 in breast cancer remain unknown. In this study, the clinical data from 1066 breast cancer patients were analyzed by The Cancer Genome Atlas (TCGA) database, and were analyzed using the correlation between ADAMTS8 expression and the clinicopathological features and prognoses. The CCK-8 assay, clone formation assay, flow cytometry and Transwell assay were used to characterize the effects of ADAMTS8 on proliferation, migration and invasion of breast cancer cells. Gene set enrichment analysis (GSEA) and western blotting were used to identify the potential molecular mechanism on how ADAMTS8 exert its biological function. ADAMTS8 overexpression correlated longer overall survival (OS) and progression-free survival (PFS). ADAMTS8 was considered as an independent prognostic factor for OS. ADAMTS8 overexpression inhibited breast cancer cell proliferation, migration and invasion in vitro, and induced G2/M cell cycle arrest. ADAMTS8 was also involved in cell cycle regulation and was associated with the EGFR/Akt signaling pathway. ADAMTS8 knockdown showed the reverse effect. Together, the results showed that ADAMTS8 functioned as a tumor suppressor gene (TGS) and could be a prognostic biomarker for breast cancer.
|
v2
|
2022-11-09T06:16:57.786Z
|
2022-11-08T00:00:00.000Z
|
253396916
|
s2orc/train
|
Neurocognition in adults with intracranial tumors: does location really matter?
Objective As preservation of cognitive functioning increasingly becomes important in the light of ameliorated survival after intracranial tumor treatments, identification of eloquent brain areas would enable optimization of these treatments. Methods This cohort study enrolled adult intracranial tumor patients who received neuropsychological assessments pre-irradiation, estimating processing speed, verbal fluency and memory. Anatomical magnetic resonance imaging scans were used for multivariate voxel-wise lesion-symptom predictions of the test scores (corrected for age, gender, educational level, histological subtype, surgery, and tumor volume). Potential effects of histological and molecular subtype and corresponding WHO grades on the risk of cognitive impairment were investigated using Chi square tests. P-values were adjusted for multiple comparisons (p < .001 and p < .05 for voxel- and cluster-level, resp.). Results A cohort of 179 intracranial tumor patients was included [aged 19–85 years, median age (SD) = 58.46 (14.62), 50% females]. In this cohort, test-specific impairment was detected in 20–30% of patients. Higher WHO grade was associated with lower processing speed, cognitive flexibility and delayed memory in gliomas, while no acute surgery-effects were found. No grading, nor surgery effects were found in meningiomas. The voxel-wise analyses showed that tumor locations in left temporal areas and right temporo-parietal areas were related to verbal memory and processing speed, respectively. Interpretation Patients with intracranial tumors affecting the left temporal areas and right temporo-parietal areas might specifically be vulnerable for lower verbal memory and processing speed. These specific patients at-risk might benefit from early-stage interventions. Furthermore, based on future validation studies, imaging-informed surgical and radiotherapy planning could further be improved. Supplementary Information The online version contains supplementary material available at 10.1007/s11060-022-04181-7.
Introduction
Treatments for intracranial tumors have tremendously evolved throughout the last decades [1,2]. Although survival rates have been rising gradually, many survivors experience medical and psychological sequelae in their daily life [3]. One topic that has increasingly received attention, is the risk for neurocognitive decline in this population [4,5]. However, the prevalence rates of this problem is diagnosis-specific and has been reported very inconsistently to date, and individual risk profiling consequently remains an important goal for the scientific neurooncological community [6]. A few steps in this direction have been taken, including investigations of individual risk factors such as fatigue and emotional difficulties [7], cognitive reserve and education level [7,8], genetic subtypes [9], molecular tumor profiling [10], as well as detailed treatment characteristics (e.g. cranial radiation dosimetry, chemotherapeutic agents, neurosurgical strategy including fiber tract analysis and neuromonitoring) [4,5,11].
Although the number of studies on potential individual risk factors for cognitive decline are growing, the predictive value of neuroimaging features remains inconclusive [12,13]. Multiple imaging studies have evidenced radiation-induced neurological damage which can be related to neurocognitive decline in multiple domains [14]. However, it remains uncertain to which extent the tumor itself (its focal as well as compressing or infiltrative effect) plays a role in baseline cognitive performance. Imaging-based predictions of cognitive functioning before the start of (radiation) therapy have only received limited attention [15]. Hence, the question arises to which extent treatment should be adapted to the tumor location, sparing functionally crucial areas for cognitive outcomes. To address regional sensitivity of tumoral damage and its functional impact in detail, lesion-behavior analyses have been accumulating [16]. Initially, individual cases and clinical observations helped to address focused region-based lesion-behavior investigations, for instance suggesting the Broca area to be an important functional hub for speech production [17]. However, since the 20th century, the integration of clinical neuroradiological knowledge and neurological observations has gradually provided more insights into structure-function relationships. Furthermore, the neuroscientific community is shifting towards pre-intervention cerebral network analyses [18]. To investigate functional brain hubs, we can nowadays profit from more advanced imaging analysis techniques. For instance, analyses such as voxel-based approaches have been proposed and optimized [19,20].
In addition, neurocognitive test assessments consisting of multiple assessments can provide a more comprehensive overview of the complex individual cognitive profile of patients compared to single neuropsychological tasks.
The combination of neuropsychological assessment and advanced neuroimaging techniques, can result in improved detection and individual neuropsychological risk profiling. In neuro-oncology specifically, existing voxel-based lesionsymptom mapping studies have mainly focused on glioma patients so far [21][22][23][24], with lower cognitive scores in case of lesions in middle temporal gyrus [23,[25][26][27], with possibly left hemisphere dominance for language-dependent tasks [23,26] and right dominance for visual attention or processing speed [22]. However, baseline performance of the complete population of intracranial tumors has not yet received attention in this field to date. Furthermore, the existing findings remain very inconsistent, with relatively small cohorts and univariate imaging statistics [21,22,24,28]. Therefore, an investigation of a large neuro-oncological cohort covering all intracranial tumor types was performed in this study [29].
Participants
Adult intracranial tumor patients aged between 18 and 80 years old who were scheduled to receive cranial radiotherapy, received cognitive testing (2-6 weeks post-surgery) as part of their standard care between April 2019 and March 2022. Diagnoses consisted of adult intracranial tumors (i.e. meningiomas, gliomas, vestibular schwannomas, pituitary adenomas, craniopharyngiomas and others). Exclusion criteria consisted of being unable to perform the cognitive tests (e.g. due to malaise, substantial hearing or vision loss, or chronic fatigue), an MRI-scan of insufficient quality at baseline.
Materials
Each patient underwent a baseline neuronavigation MRI procedure for radiotherapy planning. 3 T MR scans (1 mm slice thickness) were acquired in standard sagittal T2-weighted FLAIR scans and a gadolinium (Gadovist © 1.0 mmol/ml 0.1 mL/kg bodyweight) contrast-enhanced axial T1-weighted sequence (T1w). A baseline pre-radiation cognitive test assessment was additionally acquired. This test battery included the Hopkins Verbal Learning Test (HVLT), the Controlled Oral Word-Association Test (COWA) and the Trail Making Test, taking about 30 min in total. Each cognitive test score (HVLT learning phase, HVLT delayed recall, COWA phonemic fluency, TMT A, TMT B) was normalized to a z-score using test-specific international normative data [30][31][32]. Based on these tests, we estimated immediate and delayed recall, phonemic fluency, processing speed and cognitive flexibility, respectively. In addition, education level was requested and graded according to the Dutch Verhage scale (1964) [33] (i.e. level 1 = less than 6 years of primary education, up to level 7 presenting a university degree). Data including demographic characteristics (gender and age) and treatment-related characteristics (histological tumor subtype, type and location of surgery) were recorded as potential covariates of interest.
Image processing
Pre-radiation gross tumor volumes (GTVs) (i.e. complete tumor volume or post-surgical residual tumor tissue + resection cavity) were delineated by an experienced radiation neuro-oncologist based on the anatomical MRI scans, which were co-registered with a CT scan for radiotherapy treatment planning. These delineations were performed according to (inter)national guidelines and double-checked by a second rater [34,35]. For statistical analyses, CT scans were then (reversely) linearly registered (rigid transformation) to the post-contrast T1w MRI scan. The same transformation was applied to the GTVs (with nearest neighbor interpolation). Second, after skull stripping, the brains on pre-contrast T1w MRI scans were non-linearly registered (rigid, affine and deformable transformation) to a population-based brain T1w MRI template (i.e. ICBM-MNI). Again, this transformation was applied to the GTVs (with nearest neighbor interpolation). All registrations were performed using Advanced Normalization Tools [36]. Once all GTVs were in template space, voxel-wise statistics were performed.
Statistical analyses
First, frequencies of impairment per test are reported. Impairment was defined as ≥ 2 standard deviations below the normative mean for each test separately [37]. Frequencies of impairment on each task were compared between the different histological subtypes (a), as well as between WHO grades (b) and surgery subgroups (no surgery, biopsy, resection) (c), using likelihood chi-square tests (G 2 ) (for which (b) and (c) were only performed within the glioma and meningioma subgroups). Second, support vector regression voxel-based lesion-symptom mapping was implemented. This model predicted each (normalized) cognitive test score based on lesion location in each voxel, after regressing out lesion volume of both the behavioral data and lesion data. Covariates in the model predicting test scores included age at assessment, gender and education. The models were repeated to correct for potential tumor-and surgery-specific effects by including histological tumor subtype and type of surgery (no, biopsy, resection) as additional covariates. Permutation testing (with 1000 permutations) was performed with the level of significance set at p < 0.001 at voxel-level, and at p < 0.05 at cluster-level. Only voxels occurring in at least 10 patients were included to reach sufficient lesion affection and remove spurious voxels from analyses. These analyses were conducted using the MATLAB-based multivariate lesion symptom mapping toolbox [38].
Results
In total, 179 patients were included in this study (for demographic information, see Table 1). The majority of patients were diagnosed with gliomas (n = 126), followed by meningiomas (n = 28), vestibular schwannomas (n = 9), pituitary adenomas (n = 9), craniopharyngiomas (n = 4) and others (n = 4). A heatmap of voxel-wise lesion prevalence is presented in Fig. 1. Tumor-specific heatmaps (and low-vs. high-grade glioma heatmaps) are available in Supplementary Figs. 1-5. Lesions were most often detected in (or surrounding) the temporal and frontal lobes, more specifically involving the left inferior temporal gyrus, right superior temporal gyrus and right anterior cingulate cortex (Fig. 1). These locations mainly match with the gliomas and meningioma subgroups, given that these cover the majority of patients.
Regarding WHO grade in glioma patients, significantly lower scores were found in the WHO 4 subgroup for processing speed (TMT A) ( Regarding the different neurosurgerical procedures, no group differences (between "no surgery" versus "biopsy" versus "resection") in impairment frequencies were found within the glioma, nor within the meningioma subgroup (Suppl . Tables 2 and 3).
With regard to verbal memory, immediate and delayed recall (HVLT A, B) were significantly associated with lesions affecting the left temporal lobe (i.e. involving the superior gyrus and temporal pole) (see Fig. 2). More specifically, the significant lesion cluster predicting delayed recall was most closely located to the left hippocampus, while the significant cluster associated with immediate recall was located more laterally affecting the parahippocampal area and insular region. For processing speed (TMT A), a significant cluster was encountered surrounding the right temporo-parietal junction, extending to the dentate gyrus of the hippocampus and fornix (see Fig. 3). A smaller significant area associated with cognitive flexibility (TMT B) was largely overlapping with the cluster associated with processing speed. However, this area did not reach significance at cluster-level.
Regarding verbal fluency, a few voxels in the left superior temporal gyrus were significantly associated with fluency scores. However, these associations were not significant at cluster-level. Finally, all significant clusters were reproduced when tumor subtype and resection subtype were additionally included as covariates.
Discussion
This cohort study of patients with intracranial tumors showed cognitive impairments in about 20-30% of cases. More specifically, tumor histology was significantly associated with impairment on the processing speed task. Grading analyses further showed an increased risk of impairment for high-grade compared to low-grade gliomas in processing speed, cognitive flexibility and delayed memory, while no acute surgery-effects were found. Based on our multivariate voxel-wise analyses, tumor location was only significantly predictive of verbal memory and processing speed, involving the left superior temporal gyrus, temporal pole and (para-)hippocampus, and right temporo-parietal junction, hippocampal and fornix areas, respectively.
These findings are in line with evidence from functional MRI studies, showing activity in similar areas during memory [39] and visual processing or flexibility tasks [40,41]. Although previous VLSM studies also mostly reported significant findings in temporal gyri [23,[25][26][27], the existing results remained inconsistent [21,22,24,28]. This heterogeneity could partly be attributed to the use of different test materials across studies, but also to the different predominant lesion locations. More specifically, in our cohort, the glioma subtype appeared to be mainly located Fig. 3 Voxel-wise p-map for predictions of processing speed and cognitive flexibility performance Note. This p-map shows green to red voxels where lesions were significantly associated with processing speed scores (as estimated with TMT-A) and cognitive flexibility scores (as estimated with TMT-B), in panels A and B, respectively.
Red indicates the clusters that were significant at cluster level. In panel A, these clusters were located surrounding the right temporoparietal junction, extending to the dentate gyrus of the hippocampus and fornix. In panel B, no voxels were significant at cluster level. Images are shown according to radiological convention 1 3 in left temporal and frontal areas. This predominancy is in concordance with the recent findings of Habets et al. (2019) [21]. In contrast to most VLSM studies, the current study allows to perform a comparison at a larger scale (between tumoral location, histological and grading effects) by covering the entire adult intracranial tumor population. The fact that cognitive impairment was only histology-related for processing speed, with glioma (and schwannoma, albeit based on a small cohort n = 9) patients being potentially at higher risk, confirms that these specific tumoral cells interact with brain areas associated with processing speed, while compression by meningiomas seem to have a smaller effect. In particular, we found associations between lesions affecting the right temporo-parietal area, hippocampal and fornix and lower processing speed scores. Not only might the tumoral cells interact differently, also the predominant locations (and related functional outcomes) of the different tumor types are different (as shown in the tumor-specific heatmaps). More specifically, significant clusters related to processing speed and cognitive flexibility are mainly overlapping with supratentorial locations of gliomas (not with the heatmaps of other subtypes), whereas the areas related to immediate and delayed verbal memory recall overlap with both glioma and meningioma locations (again not the other subtypes). This can suggest that mainly glioma and meningioma patients with tumors located in (or surrounding and hence affecting) the encountered cluster areas are specifically at risk for such specific cognitive problems in daily life. Moreover, cognitive tests such as HVLT and TMT can be particularly important to implement in clinical routine for these populations.
Although we did not encounter surgery-related associations with impairment frequencies, we need to keep in mind that the distribution regarding surgery was skewed, with the majority that underwent surgery (41 biopsy and 103 had resection). Furthermore, cognitive assessments took place only shortly after surgery (app. 2-6 weeks), so we can only conclude that there were no acute symptoms. Longer followup is required to assess if and how specific surgical techniques (biopsy and resection) might lead to certain cognitive sequelae at a later timepoint.
Besides the histological subtypes, also the aggressiveness of the tumor can differently affect healthy brain tissue. For instance, glioblastoma is known to invade and grow faster, which results in edema, a remarkable mass effect and intracranial pressure, leading to substantial damage to the healthy tissue [42]. More specifically, the brain network could reorganize more efficiently in case of slower tumoral growth patterns (e.g. low-grade gliomas) compared to the more aggressive tumor types (e.g. glioblastomas). Hence, in this study we additionally investigated the impact of WHO grade, which showed high-grade glioma patients indeed to be specifically at increased risk for impairment in processing speed, cognitive flexibility, and delayed memory. Although tumor locations of these two subgroups differed in our study, with high-grade gliomas occurring more often in the right hemisphere, and low-grade gliomas more often in the left hemisphere (see Appendix), the lesion-outcome locations do not exactly match with these grade-specific heatmaps. Hence, tumor grade appears to be an additional risk factor on top of the encountered task-specific lesion location, for gliomas specifically. Detailed analyses of tumoral molecular markers have previously also been associated with brain atrophy (e.g. IDH status [43], 1p/19q co-deletion and TERT promoter mutation [44,45]) and cognitive performance (e.g. IDH-1 expression, CD3, ATRX, BDNF, EAAT1, GAT-3, SRF, NLGN3, CK2Beta and P-STAT5b, NLGN3 and CK2Beta [45]). Still, interactions between such detailed molecular features and lesion location-behavior relationships need further investigation in the future. In contrast to gliomas, we did not find grading effects for meningioma patients.
Although tumor-related factors can play an important role in neurocognitive outcomes of a patient, the final outcome of the patient is more complicated than tumor-related only. Neurodegenerative, connectome and metabolic changes are not only affected by the tumor or surgery, but also by patient-related factors such as age, cognitive reserve or education level, gender, and genetic factors [46], as well as additional treatments (including anti-epileptic drugs [47] and corticosteroid treatment and possible medical complications (including epileptic seizures [48]). Each of these components is intrinsically related to the tumor type, which can complicate the correct risk stratification for neurocognitive decline. Hence, heterogeneity of findings across VLSM studies can partly be explained by patient-related differences in the investigated samples, as well as in the statistical approach, either including covariates or not. In this study, covariates age, gender, education, tumor volume, histology/grade were all included in the full models. More large-scale studies applying models that can sufficiently explain the existing large variability in neuropsychological performance of neuro-oncological patients would be recommended. Furthermore, VLSM analyses focus on brain areapredicted outcomes from a localism perspective, while the neuroscientific field is moving towards connectomics [18]. Applying connectomic and network-based lesion symptom mapping imaging techniques (including resting state fMRI and diffusion-weighted MRI to estimate functional and structural brain networks, respectively) combined with daily life measures for neuro-oncological patients might accelerate building our knowledge on which brain connections to spare during surgery, irradiation, and to stimulate during interventions. Even more, given that brain reorganization, consequently functional hubs (and thus regional vulnerability) could depend on the subtype and aggressiveness of the tumor, network analyses could provide more insight into the 1 3 differential and dynamic structure-function relationships in future studies.
Patients with gliomas or meningiomas affecting the abovementioned temporal and temporo-parietal areas could possibly benefit from early (i.e. from diagnosis onwards) onset neurorehabilitation interventions [49], with computerized interventions, transcortical magnetic stimulation [50], and psychopharmacology [51], of which each need further investigation for effectiveness. Interventions such as stimulation might also implement information from VLSM studies, with a potential focus on temporo-parietal areas for specific cognitive decline in verbal memory and attention.
Alongside interventional trials, at this point, prevention remains key. Both prevention and intervention which can spare neuropsychological functioning is important, as it is not only important for daily quality of life of the patient, but also for patients to understand the treatment, informed consent of the treatment and for treatment adherence [52,53].
Some limitations and strengths need to be considered when interpreting the results. First, we need to mention that the majority of patients in this cohort were diagnosed with a glioma, and the remaining tumor types covered relatively small subgroups. Hence, the final results can be mainly driven by the glioma subgroup and its predominant locations. However, we note that there is a lack of research for the non-glioma populations, which is mainly due to their low prevalence rates. Therefore, we aimed to investigate all intracranial tumor types and to cover the entire brain tissue for brain-behavior analyses. Hence, the included lesions were defined as "affected brain tissue" more in general, which could include tumoral tissue, cavity as well as compression by the tumor. To account for this heterogeneity in lesions, analyses were maximally corrected for tumor histology subtype as well as surgery type. Second, as delineations were provided by different radiation oncologists in clinical care, inter-rater variability cannot be excluded. Still, each of these clinicians were trained according to a standardized protocol, and delineations were double-checked by a second colleague. Third, the scans were acquired on two different MRI scanners, which results in inter-scanner variability. Hence, the images were intensity-normalized before the registrations to the common template. Fourth, neuropsychological assessments were acquired by assessors trained by a neuropsychologist, but not necessarily blinded to the patient information. Fifth, specific test materials were selected according to the recommendations of the European Particle Network. We cannot exclude the possibility that our imaging-related findings are specific to the applied tests (i.e. learning of word lists in HVLT and sequential ordering of numbers in TMT-A), rather than generalizable to more general cognitive domains. For instance, one earlier study in patients with meningiomas showed frontal rather than temporal brain involvement in the computerized cognitive flexibility tasks that were used [54]. Relatedly, cognitive impairment was defined as deviating scores that exceed the cut-off of two standard deviations below the norm. By selecting this relatively stringent cut-off, we cannot exclude the possibility of daily life impact in patients who did not exceed this cut-off for the measured cognitive tests. Previous research often used less stringent cut-offs, which could have affected our results which could thus mainly focus on the most affected patients. Finally, some patients had a subtotal resection. In other words, some delineated volumes pre-RT consisted of both tumoral tissue and a resection cavity. In this study, the gross tumor volume, consisting of residual tumor (if any) and cavity, was used as volume of interest, as each voxel is non-healthy brain tissue (i.e. tumoral or resected tissue) and can therefore be involved in cognitive decline.
Regarding the strengths of this study, multiple tumor types were investigated in this study, covering all brain areas that are potentially important for cognitive outcomes. Most earlier studies only focused on glioma tumors only, while this study shows different heatmaps for each histological subtype, and functional brain clusters that could be important predictors for cognitive outcomes. Not only gliomas can affect these specific areas, but also meningiomas, and in case of larger tumors even other subtypes can lead to compression of these specific brain areas as well.
From a statistical point of view, a state-of-the art procedure of a support vector multivariate analysis was selected, as multiple voxels are considered at once, which reduces the number of applied tests, and inter-voxel relationships are considered. Furthermore, the criterion of 'sufficient lesion affection' was fulfilled since voxels were only included with the minimum of 10 patients having a lesion in that voxel [19]. In addition, tumor volume was incorporated as factor of interest for both the lesion (or voxel) locations as well as for the cognitive scores, as it was regressed out of both, as recommended for VLSM studies most recently [38]. Permutation testing was chosen with stringent voxel-and clusterlevel significance levels to sufficiently correct for multiple testing [20]. This approach solves the issue of the unmet assumption of uneven distributions of voxel-values.
Conclusion
In this cohort study, variable test-specific cognitive impairment was observed in about 20-30% of neuro-oncological patients, of which processing speed was specifically histology-related and mainly impaired in glioma and vestibular schwannoma patients. In the lesion-symptom mapping analyses, tumors affecting the left temporal areas and right temporo-parietal areas were related to verbal memory and processing speed, respectively. Both gliomas and meningiomas can occur within or compressing these specific areas, so they might benefit from early interventions, if the lesions specifically involve damage or compression of temporoparietal areas. Even more, processing speed or flexibility might be crucial to assess in glioma patients, while verbal memory should be assessed in both glioma and meningioma subgroups, with high-grade glioma patients being most at risk. Future multi-diagnosis multivariate VLSM studies are required to confirm these findings.
|
v2
|
2022-11-09T06:16:57.904Z
|
2022-11-08T00:00:00.000Z
|
253396130
|
s2ag/train
|
Predicting Molecular Subtype and Survival of Rhabdomyosarcoma Patients using Deep Learning of H&E Images: A Report from the Children's Oncology Group.
PURPOSE
Rhabdomyosarcoma (RMS) is an aggressive soft-tissue sarcoma which primarily occurs in children and young adults. We previously reported specific genomic alterations in RMS which strongly correlated with survival; however, predicting these mutations or high-risk disease at diagnosis remains a significant challenge. In this study, we utilized convolutional neural networks (CNNs) to learn histologic features associated with driver mutations and outcome using H&E images of RMS.
PATIENTS AND METHODS
Digital whole slide H&E images were collected from clinically annotated diagnostic tumor samples from n=321 RMS patients enrolled in Children's Oncology Group (COG) trials (1998-2017). Patches were extracted and fed into deep learning CNNs to learn features associated with mutations and relative event-free survival risk. The performance of the trained models was evaluated against independent test sample data (n=136) or holdout test data.
RESULTS
The trained CNN could accurately classify alveolar RMS, a high-risk subtype associated with PAX3/7-FOXO1 fusion genes, with an ROC of 0.85 on an independent test dataset. CNN models trained on mutationally-annotated samples identified tumors with RAS pathway with a ROC of 0.67, and high-risk mutations in MYOD1 or TP53 with a ROC of 0.97 and 0.63, respectively. Remarkably, CNN models were superior in predicting event-free and overall survival compared to current molecular-clinical risk stratification.
CONCLUSION
This study demonstrates that high-risk features, including those associated with certain mutations, can be readily identified at diagnosis using deep learning. CNNs are a powerful tool for diagnostic and prognostic prediction of rhabdomyosarcoma which will be tested in prospective COG clinical trials.
|
v2
|
2022-11-09T06:16:57.938Z
|
2022-11-08T00:00:00.000Z
|
253397375
|
s2ag/train
|
Combining prostate health index and multiparametric magnetic resonance imaging may better predict extraprostatic extension after radical prostatectomy.
BACKGROUND
In patients undergoing radical prostatectomy (RP) for prostate cancer (PCa), preoperative prediction of extraprostatic extension (EPE) can facilitate patient selection for nerve-sparing procedures. Since both multi-parametric magnetic resonance imaging (mpMRI) and prostate health index (PHI) have shown promise for the diagnosis and prognostication of PCa, we investigated whether a combination of mpMRI and PHI evaluations can improve the prediction of EPE after RP.
METHODS
Patients diagnosed with PCa and treated with RP were prospectively enrolled between February 2017 and July 2019. Preoperative blood samples were analyzed for PHI (defined as (p2PSA/fPSA) × √tPSA), and mpMRI examinations were performed and interpreted by a single experienced uroradiologist retrospectively. The area under the receiver operating characteristic curve (ROC) was used to determine the performance of mpMRI, PHI, and their combination in predicting EPE after RP.
RESULTS
A total of 163 patients were included for analysis. The pathological T stage was T3a or more in 59.5%. Overall staging accuracy of mpMRI for EPE was 72.4% (sensitivity and specificity: 73.2% and 71.2%, respectively). The area under the ROC of the combination of mpMRI and PHI in predicting EPE (0.785) was higher than those of mpMRI alone (0.717, p = 0.0007) and PHI alone (0.722, p = 0.0236). mpMRI showed false-negative non-EPE results in 26 patients (16%), and a PHI threshold of >40 could avoid undiagnosed EPE before RP in 21 of these 26 patients.
CONCLUSION
The combination of PHI and mpMRI may better predict the EPE preoperatively, facilitating preoperative counseling and tailoring the need for nerve-sparing RP.
|
v2
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.