【作者】 王强；

【导师】 赵群； 李勇；

【作者基本信息】 河北医科大学 ， 外科学（专业学位）， 2021， 博士

【摘要】 胃癌(gastric cancer,GC)是源于胃的恶性肿瘤,是全世界最常见的恶性肿瘤之一,也是癌症相关死亡主要原因。胃癌作为最常见的癌症之一,每年男性发病率为每100000人中有18例,女性发病率为每100000人中有9例。胃癌的形成机制复杂,主要涉及基因修复功能异常、细胞生长、死亡和凋亡基因的功能紊乱等,相关研究发现不同类型的胃癌存在不同发展机制、检测标记物、临床病理特征。为了明确胃癌的发展机制,为胃癌的治疗提供新的证据和方向,改善患者预后,研究胃癌分类的相关分子机制至关重要。TCGA分型和ACRG分型是近些年学者们提出的新的胃癌相关的分子分型方法,该分型方法的提出可以弥补传统病理分型的不足,为胃癌个体化治疗奠定基础。TCGA分型主要来源于欧洲人群,而ACRG分型来源于亚洲人群,主要来自日本和韩国,这两种分型在基因扩增、基因突变、临床特征、预后等各个方面均有显著的研究价值,对于临床的精准个体化治疗有重要意义。然而,TCGA分型和ACRG分型在河北省人群中的临床特征及其与临床参数和预后的关系尚不清楚。本研究旨在通过使用免疫组织化学(IHC)和二代测序(NGS)进行肿瘤分型,探讨两种分型方法和胃癌的临床病理特征、预后的关系,并通过大数据深入挖掘研究胃癌发生发展的分子机制,为胃癌研究提供新的方向和实验理论。第一部分 ACRG分型在河北省胃癌人群中的临床特征及预后意义目的:研究ACRG分型在GC组织中的表达情况,探讨胃癌分子分型、临床病理特征与预后的关系。进一步明确ACRG分型是否适用于河北人群,探讨ACRG分型对预后的影响。方法:1.采用免疫组化染色法检测65例患者GC组织中MDM2,P21,E-钙粘蛋白和波形蛋白表达水平。2.分析相关蛋白的表达水平与GC患者临床病理特征的相关性。结果:1.IHC染色显示MLH1,MDM2和P21蛋白主要位于细胞核中,而E-钙粘蛋白和波形蛋白则主要在肿瘤细胞的细胞质和膜中表达。2.不同年龄段(≤60岁和>60岁)胃癌病人的MLH1和MDM2表达水平有统计学差异(P<0.05)。3.E-钙粘蛋白和波形蛋白在不同位置的表达水平有统计学差异(P<0.05)。4.四种ACRG分子分型的GC患者在性别、年龄、肿瘤位置、Lauren分型或术后辅助治疗类型上均无明显差异(P>0.05)。5.MSS/TP53+胃癌在胃食管结合部(EGJ)(10.3%)显示出低频率,而大多数MSS/TP53-胃癌存在于远端胃(41.7%)。6.MSS/EMT GC患者倾向为弥漫型(53.8%),而MSS/TP53-GC患者为肠型(57.1%;MSS/TP53-GC患者倾向于肠型)。7.MSI与MSS/EMT预后明显不同(P<0.001),MSS/EMT与MSS/TP53-预后明显不同(P<0.001)。MSS/EMT预后最差,其次为MSS/TP53-、MSS/TP53+、MSI。8.ACRG分子分型(P=0.045)、Lauren分型(P=0.001)和辅助治疗(P=0.013)与GC的总生存率(OS)相关。小结:1.MLH1和MDM2蛋白与年龄显著相关。2.E-钙粘蛋白和波形蛋白与肿瘤的位置显著相关。3.MSS/EMT胃癌患者倾向为弥漫型,而MSS/TP53-胃癌患者则倾向为肠型。大多数MSS/TP53-胃癌存在于远端胃。4.在所有ACRG分型中,MSI的OS最好,复发率最低。第二部分 TCGA分型在河北省胃癌人群中的临床特征及预后意义目的:采用二代测序对胃癌进行综合分析,根据TCGA分型对65例胃癌患者进行分类。探讨胃癌分子分型、临床病理特征与预后的关系。进一步明确TCGA分型是否适用于河北人群,评价TCGA胃癌分子分型对预后的影响。方法:1.采用二代测序方法对65例患者GC组织的基因突变进行筛选和注释。2.应用相关统计学分析TCGA与GC患者OS之间的关系。结果:1.NGS测序结果显示在65例GC病例中,EBV+为6例(9.2%),MSI为15例(23.1%),GS为14例(21.5%),CIN为30例(46.2%)2.MSI在女性中更为常见(53.3%),而EBV+在男性中更为常见(83.3%)。3.65例胃癌中,TP53的突变率为80.0%,扩增百分比CCNE1为20.0%4.弥漫型、GS分子、MSS/EMT和接受XELOX辅助化疗患者的预后最差。5.Kaplan-Meier图显示多基因突变患者的预后较差。单变量和多变量分析表明TP53突变的患者(危险比=4.193,95%置信区间=1.260-13.945)的预后较差。小结:1.MSI和EBV+发生的概率有性别差异;2.GS和MSI在不同的年龄段发生概率有差异;3.CIN胃癌在胃食管连接处更为常见;4.弥漫型、GS分子、MSS/EMT和接受XELOX辅助化疗的患者的预后最差;5.突变率最高的基因是TP53,扩增百分比最高的是CCNE1;6.多基因突变患者的预后较差;7.TP53突变是GC的独立预后指标。第三部分 ACRG亚型MSI和EMT存在预后差异的分子机制探索目的:利用来自NCBI的基因表达数据库(GEO)、癌症基因组的大型公共数据集数据库(TCGA)研究ACRG分型的生物信息分析,并探讨其在胃癌中的临床意义,对MSI和EMT突变频率进行分析,探讨MSI和EMT相关的差异表达基因,并进行GO富集分析和KEGG分析,对MSI和EMT相关基因和通路进行分类。从基因层面探讨MSI和EMT型胃癌的影响因素。方法:1.利用TCGA数据库和来源于NCBI的GEO数据库中的数据,分析比较TCGA数据库与GSE62254数据集中MSI和EMT表达水平的差异,探讨GC患者预后。2.对MSI和EMT相关的差异表达基因进行GO富集分析和KEGG分析。3.单变量Cox风险回归分析来探讨差异基因与胃癌患者预后的关系。结果:1.对GSE62254胃癌数据集和TCGA数据库的综合分析显示,EMT阳性与不良总体存活显著相关,且MSI的总体生存显著优于EMT。2.MSI的突变频率明显高于EMT。3.8个基因与胃癌患者EMT和MSI分型显著相关,分别是THBS4、CPE、RSPO3、APOD、CRISPLD1、GHR、SFRP4、NAP1L3。4.THBS4、CPE、RSPO3、APOD、CRISPLD1、GHR、SFRP4、NAP1L3在EMT患者中表达量显著高于MSI,且有统计学差异。小结:1.MSI的突变频率明显高于EMT;2.MSI患者的总体生存显著优于EMT;3.通过生物信息学分析发现THBS4、CPE、RSPO3、APOD、CRISPLD1、GHR、SFRP4、NAP1L3与胃癌患者的分类及预后显著相关;4.THBS4、CPE、RSPO3、APOD、CRISPLD1、GHR、SFRP4、NAP1L3在EMT患者中表达量显著高于MSI,且有统计学差异。更多还原

【Abstract】 Gastric cancer(GC)is a malignant tumor originating from the stomach.It is one of the most common malignant tumors in the world,and also the leading cause of cancer-related death.As one of the most commonly diagnosed cancers,GC has an annual incidence of 18/100000 in men and 9/100000 in women.The underlying mechanism of GC is complex,mainly involving gene repair dysfunction,cell growth,death and apoptosis gene dysfunction and so on.Related studies have found that different types of GC have different mechanisms of development,markers for detection of GC and clinicopathological characteristics.It is of vital importance to investigate the molecular mechanism of GC classification,so as to clarify the development mechanism of GC,thereby providing new evidence and direction for the treatment of GC,and improving the prognosis of patients.TCGA classification and ACRG classification are new molecular classification methods related to GC put forward by scholars in recent years,which can overcome the shortcomings of traditional pathological classification and lay a foundation for individualized treatment of GC.TCGA classification mainly arises from European population,whereas ACRG classification is mainly based on Asian population,predominantly from Japan and South Korea.These two genotypes have significant values in scientific research with respect to gene amplification,gene mutation,clinical characteristics,and prognosis,and are of great significance for accurate diagnosis and individualized treatment in clinical practice.However,the clinical characteristics of TCGA classification and ACRG classification in GC population in Hebei Province and their relationship with clinical parameters and prognosis remain poorly understood.The purpose of this study was to explore the relationship of the two classification methods with clinicopathological characteristics and prognosis of GC after using immunohistochemistry(IHC)and next-generation sequencing(NGS)for tumor classification,to analyze their relationship with clinical prognosis by big data mining,and to gain insight into the molecular mechanism of the occurrence and development of GC,thereby providing a new direction and experimental theory for GC research.Part One Clinical characteristics and prognostic significance of ACRG classification in GC population in Hebei ProvinceObjective: To study the expression of ACRG classification in GC tissues and to explore the relationship between molecular classification,clinicopathological characteristics and prognosis of GC,to further clarify whether ACRG classification is appropriate for GC population of Hebei Province,and to explore the effect of ACRG classification on prognosis.Methods:1.Immunohistochemical staining was used to detect the expression level of MDM2,P21,E-cadherin and vimentin in 65 GC patients.2.The correlation of the expression level of related proteins with the clinicopathological characteristics of patients with GC was analyzed.Results:1.IHC staining showed that MLH1,MDM2 and P21 proteins were mainly located in the nucleus,while E-cadherin and vimentin were mainly expressed in the cytoplasm and membrane of tumor cells.2.Significant difference was found in the expression levels of MLH1 and MDM2 proteins in GC tissues of patients of different age groups(≤60 vs.>60years of age)(P < 0.05).3.The expression levels of E-cadherin and vimentin in different locations were statistically different(P < 0.05).4.There was no significant difference in sex,age,tumor location,Lauren classification or postoperative adjuvant therapy among the four ACRG subtypes of GC patients(P > 0.05).5.MSS/TP53+esophagogastric junction(EGJ)tumor(10.3%)showed low frequency,while most MSS/TP53-subtype were present in the distal stomach(41.7%).6.Patients with MSS/EMT subtype mostly had diffuse GC(53.8%),while patients with MSS/TP53 subtype tended to develop intestinal GC(57.1%;patients with MSD/TP53-had intestinal GC).7.The prognosis was significantly different between MSI subtype and MSS/EMT subtype(P < 0.001),as well as between MSS/EMT subtype and MSS/TP53-subtype(P < 0.001).The prognosis of MSS/EMT subtype was the worst,followed by MSS/TP53-,MSS/TP53+ and MSI.8.ARCG molecular subtypes(P = 0.045),Lauren classification(P =0.001)and adjuvant therapy(P = 0.013)were associated with overall survival(OS)in GC.Conclusions:1.MLH1 and MDM2 proteins were significantly correlated with age.2.E-cadherin and vimentin were significantly correlated with the location of the tumor.3.Patients with MSS/EMT subtype tended to have diffuse GC,while those with MSS/TP53-subtype tended to have intestinal GC.Most of the MSS/TP53-subtype was present in the distal stomach.4.Among all ACRG subtypes,MSI subtype has the best OS and the lowest recurrence ratePart Two Clinical characteristics and prognostic significance of TCGA classification in GC population in Hebei ProvinceObjective: To make a comprehensive analysis of GC by NGS and classify 65 patients with GC according to TCGA classification,and to explore the relationship between molecular classification,clinicopathological characteristics and prognosis of GC,further determining whether TCGA classification is appropriate for GC population in Hebei Province,thereby evaluating the effect of TCGA molecular classification of GC on prognosis.Methods:1.The gene mutations in the GC tissues of 65 patients were screened and annotated by the NGS method.2.The relationship between TCGA and OS in patients with GC was analyzed using correlation statistics.Results:1.NGS showed that among 65 GC cases,there were 6(9.2%)with EBV+subtype,15(23.1%)with MSI subtype,14(21.5%)with GS subtype,and 30(46.2%)with CIN subtype.2.MSI subtype was more common in women(53.3%),whereas EBV+subtype was more common in men(83.3%).3.In 65 cases of GC,the mutation rate of TP53 was 80.0%,and percentage of CCNE1 amplification was 20.0%.4.Patients with diffuse type,molecular subtype of GC,MSS/EMT subtype and XELOX adjuvant chemotherapy had the worst prognosis.5.Kaplan Meier graphs showed that patients with multiple gene mutations had poor prognosis.Univariate and multivariate analysis indicated that patients with TP53 mutations(risk ratio = 4.193,95% CI = 1.260-13.945)had poor prognosis.Conclusions:1.There were gender differences in the occurrence probability of MSI and EBV+subtypes.2.The occurrence probability of GS subtype and MSI subtype was different in different age groups.3.CIN subtype of GC were more common at EGJ.4.Patients with diffuse type,GS molecular subtype,MSS/EMT subtype and XELOX adjuvant chemotherapy had the worst prognosis.5.TP53 had the highest mutation rate,and CCNE1 had the highest percentage of amplification.6.Patients with multiple gene mutations had poor prognosis.7.TP53 mutation was an independent prognostic indicator of GC.Part Three Study on the molecular mechanism underlying prognostic difference between MSI and EMT subtypes of ACRGObjective: To explore the clinical significance of ACRG classification in GC after the bioinformatics analysis by using the gene expression database(GEO)from NCBI and the large public dataset database(TCGA)of cancer genome,to explore the differentially expressed genes related to MSI and EMT by analyzing the mutation frequency of MSI and EMT,and to classify MSI and EMT related genes and pathways by GO enrichment analysis and KEGG analysis,so as to explore the influencing factors of MSI and EMT subtypes of GC at the genetic level.Methods:1.The difference of expression level between GSE62254 and TCGA data set was analyzed and compared using GEO data from NCBI and data from TCGA,to explore the prognosis of GC patients.2.The differentially expressed genes related to MSI and EMT were analyzed by GO enrichment analysis and KEGG analysis.3.Univariate Cox risk regression analysis was used to explore the relationship between differential genes and prognosis of patients with GCResults:1.GSE62254 and TCGA GC data set and comprehensive analysis showed that there was a significant correlation between EMT-positive tumor and poor OS,and the OS of MSI subtype was significantly better than that of EMT.2.The mutation frequency of MSI was significantly higher than that of EMT.3.Eight genes were significantly correlated with EMT and MSI classification ation in GC patients,including THBS4,CPE,RSPO3,APOD,CRISPLD1,GHR,SFRP4 and NAP1L3.4.APOD was poorly expressed in MSI but highly expressed in EMT.The expression of CPE was low in MSI.CRISPLD1 was highly expressed in EMT.The expression of GHR was low in MSI.The expression of NAP1L3 was low in MSI and high in EMT.The expression of RSPO3 was low in MSI and high in EMT.SFRP4 was highly expressed in both MSI and EMT,and was higher in EMT than in MSI.THBS4 was highly expressed in EMT.Conclusions:1.In this study,the OS of MSI was significantly better than that of EMT.3.Bioinformatics analysis showed that THBS4,CPE,RSPO3,APOD,CRISPLD1,GHR,SFRP4,and NAP1L3 were significantly correlated with OS of GC patients.4.APOD,CPE,GHR,NAP1L3 and RSPO3 were poorly expressed in MSI,but SFRP4 was highly expressed.APOD,CRISPLD1,NAP1L3,RSPO3,SFRP4 and THBS4 were highly expressed in MSI.更多还原