| 1Background and aimsEsophageal cancer (EC) is one of the most common malignancies in our country and worldwide. There was no obviously symptom in its early stage. Its worth pointing out that above of95percent of patients with first diagnose were patients with in middle-late stage. Its prognosis was worst; the survival rate of5-years was only fifteen percent or so, but the survival rate with patients in early stage was more than ninety-five percent. To our surprise, our study group founded that the survival time of some ESCC patients in middle-late stage were above10years, even more than20years after their surgical operations during the field research in high incidence area. These patients were defined as Ultra long-survival patients (ULS,≥10years).It is suggested that only clinical pathology stage of EC patients could not decide patients’ prognosis and survival conditions, maybe the inherent factors (gene and immunology) would play an important role on improving quality of life and prolonging survival time with EC patients. Clearly, it was very important to clarify the molecular mechanisms of survival difference and to determine the key survival-related gene and protein.Our recent GWAS studies on esophageal cancer from25000cases and normal control subjects have identified two susceptibility genes for esophageal cancer (Nat Genetic,2010), which demonstrate the inherited molecular basis. However, our GWAS study does not provides further information for the SNPs relevant to survival.The aim is to identify the key survival related genes and molecules and to investigate the clinical significance for these identified SNPs and molecular through the following3relevant directions:(1)To determine the key factors favorable for good prognosis by comparing clinic pathological changes in the patients with ULS and SS for ESCC;(2) To identify survival-related SNPs for esophageal cancer derived from our previous GWAS scanning data on the cases with positive (n=1077) and negative family history (n=1733) by genotype imputation and to explore the correlation with EC patients’clinical outcomes;(3) To investigate the percentage of survival-related immunological molecular, the phenotypic characteristics of T cells in the peripheral blood and the tumor microenvironment of EC patients, by using FCM and IHC and to analyze the association with clinical pathology characteristics and prognosis.2Study population2.1The total of30,718EC patients derived form45hospitals in the high-and low-incidence areas for EC in China. All of clinical pathology information of EC patients comes from henna key laboratory for esophageal cancer research. Finally there were30,718cases included7,180ULS cases and23,538SS cases were enrolled into the analysis according to the pathological diagnosis, treatment, lifetime follow-up results and questionnaires.2.2A total of6,047cases from Chinese Han population were enrolled in the study of GWAS.2,810samples, including1,077cases and1,733controls, were performed whole genome scanning.3,921cases, including2,406male who median survival time was3.05years and1,515females that were6.93years, were further enrolled for replication study. The peripheral venous blood samples (3-5ml) were collected from each patient and control and stored at-20℃.2.3A total of218cases and60controls from Chinese Han population were enrolled in the study. The peripheral venous blood samples (lOmL) were collected from each patient and control.218Patients were divided into48symptomatic patients (35male and13female),45patients who survival time below5years (28male and17female),34patients who survival time between5and10years (19male and15female) and91patients who survival time above10years (52male and39female), All the cases were confirmed as ESCC by histopathology. All participants gave their written consent to take part in the study. All of research programme submit the hospital ethics committee approval and approved.2.4The225surgically resected EC specimens were collected from multiple hospitals at the high-(n=168) and low-incidence areas (n=57) for ESCC in northern China.70cases were with positive ESCC family history (FH+) and155cases were with negative (FH-).3Study methods3.1Questionnaire, follow-up and Clinic pathology informationsQuestionnaire was undertaken for each case at interview. General information’s including gender, age, occupation; ethnicity, drinking or smoking, family history etc. were recorded. Clinic pathology data including diagnosis time, pathological diagnosis, histological type, differentiation, TNM stage, lymph node metastasis, type of treatment (surgery, chemotherapy or radiotherapy), general physical examination (height, weight, blood pressure) were also retrieved from hospital. Menstrual and reproductive history (including age at menarche, menopausal status and, if relevant, age at menopause, number of births and age at first giving birth); the questionnaire, home interview and/or telephone were performed for survival follow-up. All statistical analyses were performed using IBM SPSS ver.19.0. The chi-square test was also applied for comparisons among different groups (a=0.05was considered as the test standard). Cox multivariate analysis assessed the survival for ESCC patients.3.2Associations of Genetic Variants and Survival of Patients with EC Genomic DNA was extracted by standard procedures using Flexi Gene DNA kits (QIAGEN, Germany) and the concentration was measured accurately, and normalized to50ng/μl with1.8-2.0OD in scanning study and15ng-20ng/μl with1.7-2.0OD in replication study, respectively.In scanning study, genotyping was performed using Illumina Human610-Quad BeadChips according to the Infinium HD protocol from Illumina. In replication study, genotyping was performed using Sequenom iPLEX Chip according to Sequenom MassArray System.Sample rates and SNPs rates, minor allele frequency (MAF), Hardy-Weinberg equilibrium rates (HWE) were calculated, respectively. Quality controlled data were analyzed and outputted with Plinkl.03software. P value, odds ratio (OR) and95%confidence interval (95%CI) were calculated with Cochran-Armitage trend test. PLINK was used to examine genome-wide association of ESCC risk. The chi-square test, Kaplan-Meier survival analysis and Cox model was used to analyze genome-wide association of overall-survival. a=0.05was considered as the test standard.3.3The detection of PD-1, TIM-3and CD27in peripheral blood lymphocytes of ESCC by using FACSThe phenotype moleculars (PD-1, CD27, TIM-3) of T cells in the peripheral blood of278EC patients and45health controls by FACS. The serum concentration of IL-2, IL-10, TNF-a and IFN-y in patients was determined by enzyme linked immunosorbent assay (ELISA). The data were analyzed by SPSS19.0software. The chi-square test, ANOVA were applied for comparisons among different groups (a=0.05was considered as the test standard).3.4The distribution of PD-1, TIM-3and CD27in tumor tissue microenvironment lymphocytes of ESCCIHC was performed to detect the expression of PD-1, CD27and TIM-3of T cells in150ESCC tumor tissues and75health controls. All statistical analyses were performed using IBM SPSS ver.19.0. The chi-square test was also applied for comparisons among different groups (a=0.05was considered as the test standard). Cox multivariate analysis assessed the association of the expression of PD-1, CD27and TIM-3in tumor tissues with the survival for ESCC patients.4Results4.1Survival analysis on ULS and SS patients with ESCCThere were30,718cases included7,180ULS cases and23,538SS cases were enrolled into the analysis. Among7,180ULS cases,4,053(57%) were males with the average age of55+9.6years old;3,127(43%) were females with the average age of56±8.4years old; among23,538SS cases,15,554(66%) were males with the average age of61±9.3years old,7,984(34%) were females with the average age of62±9.4years old.The ULS patients were primarily peoples with males (4,053/7,180), diagnosis aged above fifty-six years old (5,195/7,180), middle and advanced stage (2,994/3,313), moderate-differentiated (1,208/2,067), fewer lymph node metastasis (2,168/3,212), T3stage (1,387/1,842), surgical treatment males (2,955/3,705).The ratio of the patients with middle and advanced stage was lower in the ULS patients than in SS patients (91%vs.96%,x2=69.09, P<0.05); similarly, the ratio of the patients with well-differentiated in ULS group was2-folds higher than in SS group (28%vs.10%, x2=614.16, P<0.05). the survival time of patients with different differentiated have significantly difference (x2=9.67; P=0.00), the best patients is that with well-differentiated. T3stage was lower in ULS patients than that in SS patients (75%vs.86%,x2=16.55, P<.05). In addition, the patients with ULS had a higher frequency of positive family history (40%vs.29%,x2=68.98, P<0.05) and surgical treatment record (80%vs.60%,x2=390.61, P<0.05). The ratio of male patients who diagnosis age bellowed45years in ULS group is two folds than in those with SS (8.7%vs.3.5%); the female is four folds (6.7%vs.1.7%).Cox multivariate analysis assessed sex, age and tumor differentiated degree are associated with improved long-term survival for ULS patients. Risks of long-term survival time were significantly higher in female than in male patients (RR=1.68;95%CI:1.19-2.34,). A trend of decreasing survival with increasing age (RR for old patients versus younger,0.82;95%CI,0.46-0.94) and with declining differentiated degree (RR for moderately, poorly differentiated versus higher differentiated,0.39,0.38;95%CI:0.19-0.82,0.19-0.74). Tumor invasion, lymph node metastasis, treatment ways and family history were not the independent impact facors for the survival conditions and prognosis with ULS patients.4.2Associations of Genetic Variants and Survival of Patients with EC We conducted a genome-wide association study on3,921ESCC patients using40SNPs (p≤1x10-7). The results demonstrated that rs1、rs2、rs3SNPs were associated with overall survival in patients with ESCC:The survival time of patients with the rsl AA、GG or GA, the rs2AA、GA of GG, the rs3CC、CT of TT genotype were significantly difference (x2=12.258,8.399,6.671; P=0.002,0.015,0.036);After excluding confounding factors such as sex, age, smoking, drinking, clinical staging, Cox multivariate survival analysis revealed that the genetic variants of rs1、 rs2、rs3were all not the significant independent impact factors for the survival in ESCC patients(P=0.43,0.52,0.21).After investigating the patterns of the recombination and linkage disequilibrium (LD) around the risk-associated SNPs and the gene(s), one gene at each locus was implicated:rsl within CAMTA1at1p36.31-p36.23chromosome、rs2within cyclin L2at1p36.33chromosome and rs3within TENM4at11q14.1chromosome.4.3The distribution of PD-1, TIM-3and CD27in peripheral blood lymphocytes of ECThe percentage of PD-1, TIM-3of T cells detected by FACS were significantly higher, CD27was lower, in peripheral blood lymphocytes of EC compared with normal group (p=3.55E-28,1.06E-37,1.57E-5).The expression of PD-1was associated with sex%age and tumor stage (P<0.05), patients in middle-and late stage, male, whose age aboved56years increased the expression of PD-1. The expression of TIM-3was associated with gender and age (P <0.05), male patients, the older age, the higher level. The decrease expression of CD27with increased age (P<0.05).The expression of IL-10, TNF-a and IFN-y in different groups were significantly difference(p=4.44E-5,1.97E-5,0.003), not the IL-2(p=0.323).4.4The distribution of PD-1, TIM-3and CD27in tumor tissue microenvironment lymphocytes of ESCCThe expressions of TIM-3, PD-land CD27in SS, ULS and control groups were all significantly difference(p=0.000,1.54E-12,0.001). further two comparison demonstrated that the expression of PD-1was significantly difference between SS and ULS group(p=4.95E-4), not the CD27and TIM-3(p=0.611,0.242).The presence of CD27, TIM-3and PD-1in tumor tissues were not associated with patients sex, age, family history, incident area (p>0.05), the presence of PD-1 was related with tumor invasion and lymph node metastasis (p<0.05), not the The presence of CD27and TIM-3. The percentage of PD-1was higher in stage III-IV tumors than I-II tumors (x2=9.442; P=0.009) and in tumors with lymph node metastasis than in that with no metastasis (x2=8.599; P=0.014).Kaplan-Meier survival analysis showed that the percentage of CD27and TIM-3were not associated with survival (x2=1.283,0.112; P=0.52,0.945), but the survival of the negative expression of PD-1was higher than those of positive expression (x2=14.155, P=0.001). Multivariate analysis demonstrated that CD27, TIM-3and PD-1was not independent prognostic factor for survival (P=0.172,0.275,0.049).5Conclusions5.1The ULS patients were primarily peoples with male, who diagnosis aged above fifty-six years old, middle and advanced stage, moderate-differentiated, fewer lymph node metastasis, T3stage, surgical treatment males. Cox multivariate analysis assessed that gender, age and differentiation, menopause state for the female patients, were the independent prognostic factors for survival.5.2Kaplan-Meier survival analysis demonstrated that genetic variants of rs1、rs2and rs3were associated with survival time.one gene at each locus was implicated:rsl within CAMTA1at1p36.31-p36.23chromosome、rs2within cyclin L2at1p36.33chromosome and rs3within TENM4at11q14.1chromosome.5.3The expressions of PD-1, TM-3in peripheral blood lymphocytes and tumor microenvironment of EC patients were higher than that of control group, CD27was lower. Moreover, the expression of PD-1was higher in SS group than that in ULS, not the TIM-3and CD27.5.4IHC study suggested that the expression of PD-1in tumor tissue micro-environment was related with tumor invasion and lymph node metastasis, the worst survival with the best positively expression. The percentage of CD27, TIM-3was not associated with patients’ survival. |
PDF Full Text Request