Marina Alessandra Pereira ,Marcus Fernando Kodama Pertille Ramos ,André Roncon Dias ,Leonardo Cardili ,Rafael Dyer Rodrigues de Moraes ,Renan Ribeiro E Ribeiro ,Venancio Avancini Ferreira Alves,Bruno Zilberstein,Evandro Sobroza de Mello,Ulysses Ribeiro Jr
1Department of Gastroenterology,Instituto do Cancer do Estado de S?o Paulo,Hospital das Clinicas,Faculdade de Medicina,Universidade de S?o Paulo,(ICESP-HCFMUSP) S?o Paulo,SP 01246-000,Brazil;2Department of Pathology,Instituto do Cancer do Estado de S?o Paulo,Hospital das Clinicas,Faculdade de Medicina,Universidade de S?o Paulo,(ICESP-HCFMUSP) S?o Paulo,SP 01246-000,Brazil
Abstract Objective:Remnant gastric cancer (RGC) is usually associated with a worse prognosis.As they are less common and very heterogeneous tumors,new prognostic and reliable determinants are required to predict patients’ clinical course for RGC.This study aimed to investigate the tumor-infiltrating lymphocytes (TILs) and programmed cell death ligand 1 (PD-L1) status as prognostic biomarkers in a cohort of patients with RGC to develop an immunerelated score.Methods:Patients with gastric cancer (GC) who underwent curative intent gastrectomy were retrospectively investigated.RGC resections with histological diagnosis of gastric adenocarcinoma were enrolled in the study.The risk score based on immune parameters was developed using binary logistic regression analysis.RGCs were divided into high-risk (HR),intermediate-risk (IR),and low-risk (LR) groups based on their immune score.The markers(CD3+,CD4+/CD8+T cells and PD-L1) were selected for their potential prognostic,therapeutic value,and evaluated by immunohistochemistry (IHC).Results:A total of 42 patients with RGC were enrolled in the study.The score based on immune parameters exhibited an accuracy of 79% [the area under the receiver operating characteristic curve (AUC)=0.79,95%confidence interval (95% CI),0.63-0.94,P=0.002],and the population was divided into 3 prognostic groups: 10(23.8%) patients were classified as LR,15 (35.7%) as IR,and 17 (40.5%) as HR groups.There were no differences in clinicopathological and surgical characteristics between the three groups.In survival analysis,HR and IR groups had worse disease-free survival and overall survival rates compared to the LR group.In the multivariate analysis,lymph node metastasis and the immune score risk groups were independent factors related to worse survival.Conclusions:A scoring system with immune-related markers was able to distinguish prognostic groups of RGC associated with survival.Accordingly,tumor-infiltrating immune lymphocytes and PD-L1 status may serve as a potential prognostic biomarker for patients with RGC.
Keywords: Stomach neoplasms;gastric remnant;tumor-infiltrating lymphocytes;immune microenvironment
Gastric cancer (GC) remains an important cause of cancer mortality worldwide (1).Among the various risk factors that contribute to the onset of the disease,it is known that previous gastrectomy is related to the occurrence of tumors in the gastric remnant (2,3).
Remnant gastric cancer (RGC) can occur either after a previous resection for benign or malignant lesions,and its incidence is reported to comprise 1%-8% of all GC (4-6).However,as RGC is a less common and very heterogeneous tumor,doubts about its prognosis remain unclear (4,5).
Following surgical treatment with curative intention,the main prognostic factor in GC is the pathological stage(7,8).Despite that,the prognostic value of the tumor-nodemetastasis (TNM) staging system in RGC is questionable,especially regarding the N stage (9).Due to the previous reconstruction method and type of previous lesion (benign or malign),changes can occur in the perigastric lymphatic flow due to the lymphadenectomy.Further,the number of lymph nodes dissected may not be optimal to correctly predict the prognosis (4-6,10).Thus,new prognostic determinants are essential to more reliably predict the survival of these patients.
Currently,as an alternative to better understand the heterogeneity of GC and to predict outcomes,studies have focused on the investigation of several biomarkers (11-14).In this sense,the importance of the tumor microenvironment (TME) and the immune system for tumor progression has gained increasing attention (11-13).The presence of tumor-infiltrating lymphocytes (TILs) is closely related to prognosis in several studies,where high infiltration of CD3+and CD8+T lymphocytes have been associated with improved survival.Besides,TILs density,expressed as CD4+/CD8+T cells tissue ratio,in addition to programmed cell death ligand 1 (PD-L1) expression,has been proposed as potential prognostic biomarkers,demonstrating a critical role in T cell-mediated host immunity (12,15).However,to date,there are no data available on the role of inflammatory infiltrate in RGC.
Thus,this study aimed to evaluate the TME in patients with RGC to establish a prognostic scoring system with immune-related markers.Clinicopathological characteristics and outcomes were also evaluated according to the risk groups determined based on the score for immunological status.
Patients with GC who underwent curative intent gastrectomy between 2009 and 2019 in Instituto do Cancer do Estado de S?o Paulo,Hospital das Clinicas,Faculdade de Medicina,Universidade de S?o Paulo were retrospectively investigated.Study inclusion criteria were: 1) RGC patients who underwent complete total gastrectomy;2)histological diagnosis of gastric adenocarcinoma;and 3)formalin-fixed paraffin-embedded tissue (FFPE) blocks available for analysis.Previously gastric resection could be due to benign disease (peptic ulcer) or gastric adenocarcinoma. Wedge resections,palliative and emergency surgeries were excluded from the study.
Clinical data were obtained from a prospectively collected medical database,including age,sex,preoperative body mass index (BMI),albumin level,hemoglobin level,neutrophil-lymphocyte ratio (NLR),and physical status based on the American Society of Anesthesiologists (ASA)classification.Pathologic TNM staging was defined according to the 8th edition of the TNM of the International Union Against Cancer (UICC) (8).
The extent of surgery and LN dissection was established by the attending surgeon to achieve a microscopically margin-negative resection (R0).Surgery followed the recommendations of the Japanese Gastric Cancer Association and the guidelines of the Brazilian Consensus on Gastric Cancer (7,16).All patients were operated in a high-volume center by specialized surgeons.
Perioperative or postoperative platinum-based chemotherapy (CT) was administered according to clinical indication (T2-T4 and/or N+).
The postoperative follow-up was performed quarterly in the first 2 years,and every 6 months in the following years.Follow-up tests for recurrence detection were performed based on the presence of symptoms.The study was approved by the Hospital Ethics Committee and by National Research Ethics Committee (CAAE:25516719.3.0000.0065),and registered online (https://plataformabrasil.saude.gov.br).Informed consent was waived by the local Ethics Committee given the retrospective nature of the study.
Tissue microarrays
For tissue microarrays (TMA) construction,all hematoxylin and eosin (HE)-stained slides were reviewed,and the most representative tumor areas of each case were selected.Tissue cores (2 mm in diameter) were taken from the corresponding FFPE tissue blocks (donor block) and transferred to the recipient block using a precision mechanized TMA system.Three to six tumor samples(cores) and two non-neoplastic areas were selected per case.Then,TMA blocks were sectioned at 4 μm thick and used for immunohistochemistry (IHC) staining.The final result of each marker was defined according to the mean expression obtained by the sum of all spots.
IHC
The following primary antibodies were used to IHC reactions: anti-CD3+(polyclonal),anti-CD4+(clone SP35),anti-CD8+(clone C8/144B),anti-CD163 (clone 10D6),and anti-PD-L1 (clone SP142),anti-programmed cell death protein 1 (PD-1;clone PDCD1/922).
Briefly,slides were dewaxed,deparaffinized in xylene,and rehydrated in graded alcohols.Antigen retrieval was performed by immersing the slides in citrate buffer(pH=6.0) and pressure cooking.Endogenous peroxidases were blocked with 0.3% hydrogen peroxidase.Slides were incubated at 4 °C with the primary antibody.Subsequently,the avidin-biotin free short polymer-based peroxidase amplification system was used,and diaminobenzydine was applied for the development of reaction products.All slides were counterstained with hematoxylin.
IHC and pathological evaluation
The expression of CD3+,CD4+,and CD8+T cells was evaluated in tumor-infiltrating leukocytes and determined according to the percentage of positively stained cells for each case.For analysis,TILs density reported as CD4/CD8 tissue ratio (15) and CD3+were classified into two groups according to the cutoff value related to survival determined by the receiver operating characteristic (ROC) curve,and defined as low-and high-CD levels.
PD-L1 staining was assessed using a combined positive score (CPS) method,as previously reported (11).CPS was defined based on the number of tumor cells and immune cells staining for PD-L1,divided by the total number of viable tumor cells,multiplied by 100.PD-L1 was defined as positive for CPS≥1.
The positivity of PD-1 and CD163 (M2 macrophage marker) was also assessed on the basis of the percentage of staining cells at 0-100%,and cases were classified into lowand high-density groups based on the median value.
Cases were classified as microsatellite instability (MSI) if at least one of the markers (MLH1,MSH2,MSH6 and PMS2) showed a complete loss of nuclear staining within the tumor,whereas those that maintained expression of all markers were defined as microsatellite stable (MSS status).All IHC analysis was performed by pathologists,blinded to clinical and outcome data.
Prognostic scoring system with immune-related markers to predict disease-free survival (DFS)
We used binary regression analysis to assess the abilities of immune-related markers to predict DFS in RGC patients.The score was built with recurrence or death considered as the main outcome.The odds ratios (ORs) with a 95%confidence interval (95% CI) were calculated and the points were assigned to each category (CD3,CD4/CD8 ratio,and PD-L1).Then,the total score was calculated for each patient (based on regression coefficients).The area under the ROC curve (AUC) was used to evaluate the metric performance of the score,and defined cutoff values.Subsequently,patients were divided into three risk groups:low-,moderate-and high-risk groups,based on the risk score of IHC parameters.The predictive capacity of the three immune-risk groups was tested by Kaplan-Meier survival analysis.
Statistical analysis
The differences between the groups were analyzed using the Chi-square or Fisher’s exact test for nominal variables andt-test,analysis of variable (ANOVA),or Kruskal-Wallis test for continuous variables.
The ROC curves with AUC values were used to evaluate the ability of prognosis classification for TILs to CD3+and CD4/CD8+T cells ratio.The optimal cutoff values for both were determined by maximizing Youden’s index(sensitivity+specificity -1).For the score,the optimal cutoff value was also determined by the point where the Youden’s index is maximum,to maximize both sensitivity and specificity.And,in order to define the group at the greatest risk,a strict criterion level was adopted to determine the point that minimizes the weighted number needed to misdiagnose (cut-off value associated to high specificity and low sensitivity).
Survival was estimated by the Kaplan-Meier method,and the log-rank test was used to identify statistical differences across groups.DFS was calculated from the date of surgery until the date of recurrence or death.Overall survival (OS)was measured from the date of surgery to the date of death from any cause.Patients alive were censored at the date of the last follow-up.The Cox proportional hazards model was used to identify risk factors related to survival.Significant variables in the univariate analysis were included in the multivariate model to verify those independently associated with survival outcomes.The results were reported as hazard ratios with 95% CI.A P value of ＜0.05 was considered statistically significant.Statistical analyses were performed using SPSS software(Version 20.0;IBM Corp.,New York,USA).
Among the 623 patients who underwent gastrectomy with lymphadenectomy with curative intent at Instituto do Cancer do Estado de S?o Paulo,Hospital das Clinicas,Faculdade de Medicina,Universidade de S?o Paulo,surgical procedures due to RGC were performed in 60 patients.Of these,42 patients met the inclusion criteria and were enrolled in the study.
The mean age of all patients was 68.5 years [standard devivation (SD)=7.6,range,51.6-83.1 years],and 83.3%were male.Patients were previously operated on due to peptic ulcer and adenocarcinoma in 35 and 7 cases,respectively.Roux-en-Y reconstruction was performed in 5 cases,and gastrojejunostomy (Billroth II) reconstruction in 37 cases.
The mean staining for CD3+,CD4+,CD8+,and CD4+/CD8+ratio assessed by IHC was: 40% (SD=17.3,median=45.9);20.1% (SD=11.0,median=22.1);24.3%(SD=12.7,median=30);and 1.1 (SD=0.8,median=0.9);respectively.Positivity for PD-L1 was observed in 13(30.9%) cases (Figure 1).The mean positive staining for CD163+and PD-1 was 30.3% (SD=15.3,median=26.7)and 14.7% (SD=25.4,median=0.65),respectively.
Considering the IHC markers,an association was observed between PD-L1-positive tumors and an increase in TILs CD8+(PD-L1-negative=20.3%±12.7%vs.PDL1-positive=33.7%±7.2%,P＜0.001),and with a lower CD4+/CD8+ratio (PD-L1-negative=1.3±0.8vs.PD-L1-positive=0.6±0.3,P=0.017).Infiltration by CD163+macrophages was higher in PD-L1-positive RGC (PD-L1-negative=25.8%±10.4%vs.PD-L1-positive=41.2%±19.6%,P=0.023).The CD3+T cells was also higher in PD-L1 positive cases,although without reaching statistical significance (PD-L1-negative=37.8%±10.2%vs.PD-L1-positive=44.8%±4.9%,P=0.089).Also,a positive correlation was observed between CD8+T cell and CD163+(r=0.396,P=0.010).
Prognostic scoring system with immune-related markers
Figure 1 Immunohistochemical findings.Gastric adenocarcinoma showing positive staining for CD3+T cells (A),CD8+T cells (B),CD4+T cells on TILs (C) and positivity staining for PD-L1 (D).20× magnification.
Based on the cutoff value determined by ROC curve,CD3+was dichotomized in low-CD3+(＜48%) and high-CD3+(≥48%) levels (AUC=0.67,95% CI: 0.49-0.85,P=0.068);and CD4+/CD8+ratio in low-CD4+/CD8+(＜0.7) and high-CD4+/CD8+(≥0.7) levels (AUC=0.62,95% CI: 0.43-0.81,P=0.097) (Supplementary Figure S1).
For score development,CD3+and CD4+/CD8+ratio levels,and CPS status were included in the model(Supplementary Table S1).After the binary logistic regression analysis,the following points (pts) were assigned for each category: high-CD4+/CD8+=1 pt;CPS-negative=3 pts.;and low-CD3+=8 pts.;as shown inFigure 2,for a maximum score value of 12.TILs high-CD3+,low-CD4+/CD8+,and CPS-positive were equal to zero.Subsequently,the score based on immune parameters was established for each patient.
The final score based on the three variables exhibited an accuracy of 79% (AUC=0.79,95% CI: 0.63-0.94,P=0.002)(Figure 2) and the population was divided into 3 prognostic groups: low-risk (LR,score 0-3),intermediate-risk (IR,score 4-9),and high-risk (HR,score 11-12).Accordingly,10 (23.8%) patients were classified LR,15 (35.7%) as IR,and 17 (40.5%) as HR groups.
Clinical and surgical characteristics of the three risk groups are summarized inTable 1.No differences were observed regarding sex,age,comorbidities,ASA classification,laboratory tests,primary surgery diagnosis or type of reconstruction.
Regarding pathological characteristics (Table 2),the groups were similar for aspects evaluated,including number of lymph nodes,grade of histological differentiation,lymphatic/vascular/perineural invasion,pT and pN stage.Intestinal type and less advanced tumors were more common in the IR group,but without reaching statistical significance.There was also no difference in the presence of MSI,PD-1 positivity and CD163+infiltration between the three risk groups.
Figure 2 ROC for diagnostic accuracy of the score system with immune-related markers,and tables with points assigned to each category along with risk groups based on immune score.The AUC was 0.79 (95% CI,0.63-0.94,P=0.002),demonstrating a good diagnostic accuracy for immune score to predict survival in patients with RGC.ROC,receiver operating characteristic;AUC,area under the ROC curve;95% CI,95% confidence interval;RGC,remnant gastric cancer.
The median length of hospital stay was 12 d,and there were no differences between the groups regarding the occurrence of postoperative complications (LR=30%vs.IR=13.3%vs.HR=17.6%,P=0.693).CT was administered for 3 (30%),4 (26.7%) and 6 (35.3%) of patients in LR,IR and HR groups,respectively (P=0.912).
Survival analysis
After a mean follow-up of 32 months (median 30.1 months),18 patients had disease recurrence,and 26 died.The DFS and OS rate for the entire cohort was 34.7% and 36.5%,respectively.The median DFS was 16.2 months and the median OS was 27 months.
Survival according to the three risk groups are demonstrated inFigure 3.HR and IR groups had worse DFS rates compared to the LR patients (P=0.002 and P=0.046,respectively).Furthermore,survival was also significantly poorer in the HR group compared to the IR group (P=0.034).The median DFS for IR and HR groups was 20.4 and 8.8 months,respectively.
Considering the OS,the LR group had better survival than IR and HR patients (P=0.043 and P=0.008,respectively).The median OS was 14.5 and 4.9 months for IR and HR groups,respectively (P=0.102).
In the multivariate analysis,lymph node metastasis and the score groups IR and HR (reference: LR) were independent factors related to worse DFS and OS(Tables 3,4).
RGC is generally associated with a worse prognosis and frequently diagnosed at an advanced stage (4-6,9).Since they are poorly addressed by studies,new prognostic determinants in conjunction with the TNM stage are required to predict patients’ clinical course more precisely for RGC.
Therefore,in the present study,we investigated the TILs and PD-L1 status as prognostic biomarkers in a cohort of patients with RGC to develop an immune-related score.The markers (CD3+,CD4+/CD8+T cells,and PDL1) were selected for their potential prognostic and therapeutic value.As a result,we proposed an immune score that allowed the stratification of patients into three risk groups independently associated with survival.
Currently,cancer-related inflammation,including TME and immunological profile,has been investigated in several cancers to understand the tumor progression,as well as resistance or tumor response to chemotherapy and/or immunotherapy (12,15,17).Evidence has shown a positive correlation between the presence of lymphocytes in tumor tissues and increased patient survival,where several types of TILs are associated with better disease outcomes for GC(12,15,17).At the same time,intense TILs in GC have been associated with PD-L1 expression,playing a central role in the functional suppression of the T cell-dependent immune response and the subsequent immune evasion of cancer cells (14,18).
Most reports demonstrate that a high infiltrate of CD3+and CD8+T cells in GC is significantly associated with lower frequencies of lymph node metastasis and less advanced TNM stage (12,15).Some results showed that high CD3+T cell is associated with a lower TNM stage,but when evaluating the TILs for subpopulation CD4+andCD8+T cells alone,there is no relationship between high and low TILs density and tumor stage (19,20).On the other hand,TILs reported as CD4+/CD8+tissue ratio and PD-L1 expression showed a significant correlation with lymphovascular invasion and TNM stage in GC patients (15).
Table 1 Clinical and surgical characteristics of RGC according to the risk groups based on immune status
The prognostic role of tumor-infiltrating immune cells even in GC is widely discussed,being subject to variations that also include the combination of markers evaluated.Investigating three immunological parameters,it was found that patients with lower TILs ratio (CD4+/CD8+＜0.2) and PD-L1-positive expression (CPS≥1) had a significantly better OS than those with higher TILs and PD-L1-negative (15).Similarly,a study performed in advanced GC patients by TMA demonstrated that high density of CD3+and CD8+TILs was associated with improved survival.In contrast,PD-L1 positivity and density of CD4+TILs had no impact on prognosis (13).
Likewise,in our cohort the low-risk immune group was associated with better survival,which represents the RGC with high CD3+,CD8+TILs predominance,and PD-L1-positive.At the same time,the HR group (low-CD3+,high-CD4+/CD8+and PD-L1-negative) had a significantly worse survival than both the LR and IR groups.But interestingly,we found no association between the immune-score groups and the presence of lymph node metastasis.Similar findings were also described by Kimet al.,who did not found any association between density of CD8+TILs and clinicopathological characteristics,and GC with high-density CD8+or FoxP3+TILs showed significantly higher OS rates than those with low-density(20).Thus,the impact of TILs on survival,even if there is no relationship with prognostic characteristics of the tumor,seems to reinforce the critical role of T cellmediated host immunity in repressing tumor progression.
Among the mechanisms presented that justify the role of TILs in contributing to better prognosis,it is believed that tumor clones with potential for metastatic spread,sincethey generally have a greater amount of aberrantly expressed proteins (including those that contribute to metastasis) that can act as tumor-associated antigens,making these clones more prone to destruction by immune reactionsin situ.Furthermore,the presence of TILs reflects a functional immune system,which can exert its antitumor activity on metastatic cells foci that have spread to lymph nodes (12,21).An association between TILs and PD-L1 status with response to therapy has also been demonstrated.Patients with an enriched immune response and PD-L1-positive GC who received chemotherapy had no significant benefit in survival compared to those treated with surgery alone (15,17,18).Unfortunately,due to the small number of cases,we were not able to assess the risk groups in terms of response to CT.
Table 2 Pathological characteristics of RGC according to the risk groups based on immune status
Figure 3 Disease-free survival (A) and overall survival (B) of patients with RGC according to the risk group based on the score system with immune-related markers: high-risk,intermediate-risk,and low-risk group.RGC,remnant gastric cancer.
Table 3 Univariate and multivariate analysis for DFS
Table 4 Univariate and multivariate analysis for OS
Remarkably,although in MSI GC,the TILs represent the main immune-active cellular components of the TME that contribute to better prognosis,in the RGC,there was no association between immune risk groups and the presence of MSI.Indeed,the frequency of MSI was 23.8%in the RGC cases,which is higher than the rate commonly reported in primary GC (14).It suggested that MSI may be an important factor in the carcinogenesis of metachronous multiple GC,which includes RGC patients who had undergone a previous gastrectomy for GC (22).Furthermore,our previous study was demonstrated that the presence of MSI and PD-L1 expression is quite high in RGC tumors compared to primary GC (23).Watanabeet al.(24) also detected a high frequency of PD-L1 expression in RGC,but a low frequency of MSI(1.1%) (24).
To add information about the immune status,in addition to TILs and PD-L1,we also provided the IHC results for PD-1 and M2 macrophage marker CD163 in RGC.The binding of the receptor PD-1 to its ligand PD-L1 results in inhibition of T-cell effector function and decreases cytotoxic activity,and a strong correlation has already been described among the densities of CD3,CD4,CD8,and PD-1 in GC (25).Meanwhile,M2 macrophages are associated with promoting tumor growth,angiogenesis,and suppressing adaptive immunity.Some previous analyses demonstrate that high infiltration of CD163+macrophages was a negative prognostic factor for survival in GC,and significant associations were also observed between CD163+infiltration and PD-L1 expression (26,27),which was also observed in our study.However,no difference was observed between both PD-1 and CD163+and the risk groups.Possible reasons may include the difficulty in interpreting results from small cohort of cases.
The current study has some limitations.First,this is a single-center retrospective study,and only patients who underwent gastrectomy with curative intent were included.Thus,possible selection bias was unavoidable,and the number of cases is more limited.Some analyses were underpowered due to the insufficient number of patients.We also included patients with previous resection for malignant disease importantly,all tumors developed in the gastric remnant after a period of more than 5 years after previous gastrectomy.Therefore,it is a new primary cancer,not a recurrence of the disease.All analyses were performed in TMA samples,and differences about the technical methodology,antibodies,clones and cutoff value used may exist when comparing the results with other series (12).Regarding survival analysis,we perform univariate analyses of all covariates first and take only those which were significant in the univariate analysis into the multivariate analysis.Although there may be questions that sometimes a variable considered plausible of interference does not enter the multivariable model,we also evaluated survival by including variables that did not exhibit significance in the model (Supplementary Table S2).
There is no current consensus on TILs scoring in GC.Some authors grouped their cases for TILs in high and low by using the median,mean or 75th percentile as cutoffs(12,15,28).In our study,we performed an evaluation system that includes an analysis of the percentage of TIL,and used the ROC curve to determine the best cut-off value.Therefore,multicentric studies or additional studies with a larger number of patients are required to confirm our findings and ensure external validation of results.
Despite that,we sought to ensure tumor representativeness as recommended by tissue microarrays guidelines (29),and the results were based on the integrated evaluation of the IHC markers.Our patient series comprises a well-characterized western cohort of RGC treated at a referral center.A systematic review and meta-analysis showed that a small number of studies are from the West,and there is a lack of evidence about the biologic differences among GC from different geographic regions,including in immune signature between East Asian and Western patients (12).Furthermore,our study represents one of the largest single institution western series of RGC including only patients treated with curative intent in a recent 10-year period.Some western studies,despite having more cases,comprise multicenter studies that integrate patients treated over a period of more than 20 years.This may have an influence on the outcomes due to heterogeneity in the treatment adopted,both in terms of differences between Centers and differences in treatment over the years,since improvements in adjuvant treatments and perioperative care have been incorporated in more recent periods (30-34).
To date,there is no tissue-based marker that is recommended as a prognostic factor in RGC,and most of the analyses performed in GC series are restricted to a single marker,besides not including RGC.Ultimately,as the TNM stage system cannot reflect the full information of TME,TILs information may be useful as a complementary tool in the diagnosis of RGC.
So,we developed an immune-related score that could not only help predict the prognosis of RGC,but also reveal intratumoral immune response,which may expand the development of new treatment strategies combining chemotherapy,targeted therapy,and immunotherapy for this group of patients as well.In light of current advances in cancer immunotherapies,a comprehensive understanding of tumor-immune interactions in patients with RGC in the context of host immune response is imperative.
A scoring system with immune-related markers was able to distinguish prognostic groups of RGC associated with survival.Accordingly,tumor-infiltrating immune lymphocytes and PD-L1 status may serve as a potential prognostic biomarker,which may reflect the antitumor immune response.
None.
Conflicts of Interest: The authors have no conflicts of interest to declare.
Figure S1 ROC for the diagnostic accuracy of TILs.(A) ROC for TILs CD3+: The AUC was 0.67 (95% CI,0.49-0.85,P=0.068),with an estimated value of 48% for optimal cutoff;(B) ROC for TILs CD4+/CD8+ratio: the AUC was 0.62 (95% CI,0.43-0.81,P=0.097) with an estimated value of 0.7 for optimal cutoff.ROC,receiver operating characteristic;TIL,tumor-infiltrating lymphocyte;AUC,area under the ROC curve;95% CI,95% confidence interval.
Table S1 Multivariate model for risk of relapse/death with immune related markers
Table S2 Multivariate analysis for DFS and OS＊