胃癌进展相关胃菌群的特征分析  

Analysis characteristics of microbiota associated with gastric cancer progression

作　　者：孙伟丽 刘居新 王莉莉[3] 张秋月 王鹏 徐晓娜[3] 董全江 SUN Weili;LIU Juxin;WANG Lili;ZHANG Qiuyue;WANG Peng;XU Xiaona;DONG Quanjiang(Department of Central Laboratory,the Affiliated Qingdao Municipal Hospital of Dalian Medical University,Qingdao 266071;Department of Clinical Laboratory,the Affiliated Qingdao Municipal Hospital of Qingdao University Medical College;Department of Central Laboratory,the Affiliated Qingdao Municipal Hospital of Qingdao University Medical College,China)

机构地区：[1]大连医科大学附属青岛市市立医院中心实验室,山东青岛266071 [2]青岛大学医学院附属青岛市市立医院检验科 [3]青岛大学医学院附属青岛市市立医院中心实验室

出　　处：《胃肠病学和肝病学杂志》2021年第10期1101-1106,共6页Chinese Journal of Gastroenterology and Hepatology

基　　金：国家自然科学基金面上项目(31870777)。

摘　　要：目的分析晚期胃癌(advanced gastric cancer,AGC)患者和早期胃癌(early gastric cancer,EGC)患者菌群的结构、组成,进一步确定胃癌进展相关的菌群特征。方法本研究共纳入60例胃癌患者,黏膜标本取自于癌性病灶切缘>5 cm的正常胃黏膜组织。提取组织基因组DNA,采用高通量测序分析胃黏膜菌群,以EGC组菌群为对照,利用线性判别效应(linear discriminant analysis effect size,LEfSe)分析AGC组和EGC组菌群组成差异;采用受试者工作特性曲线下面积-随机森林(optimizing the area under the receiver operating characteristic curve of the random forest,AUC-RF)算法建立随机森林(random forest,RF)模型,对胃癌进展相关菌群特征进行分析。结果EGC组菌群Shannon多样性指数为2.454,明显低于AGC组(2.673),差异有统计学意义(t=2.577,P=0.013),而Chao 1指数EGC组略高于AGC组,但差异无统计学意义(t=0.619,P=0.538)。主成分分析(principal coordinate analysis,PCoA)结果发现两组患者胃菌群出现明显集中,并可以相互区分(P=0.04)。LEfSe分析结果显示,LDA值≥2.0的细菌共12个属,分别为Thermus、Burkholderia、Pseudomonas、Brevibacillus、Stenotrophomonas、Salinivibrio、Uruburuella、Schlegelella、Bradyrhizobium、Nocardioides、Bacillus和Ochrobactrum。采用AUC-RF算法得到一个最优的24个细菌属组合能够最大限度地区分AGC组和EGC组(AUC=0.916,95%CI:0.841~0.990,灵敏性=0.800,特异性=0.767)。结论在胃癌进展过程中,胃菌群在组成及多样性方面发生明显改变,并发现了24个细菌属组合能够很好地区分AGC组和EGC组菌群,与胃癌进展密切相关。本研究从菌群角度为辅助临床判断胃癌进程,为探索胃癌进展相关的菌群变化提供新的理论依据和新路径。Objective To analysis the structure and composition of microbiota in patients with advanced gastric cancer(AGC)and patients with early gastric cancer(EGC),and further determine the characteristics of microbiota related to gastric cancer progression.Methods A total of 60 patients with gastric cancer were enrolled in the study,and the biopsy was taken at least 5 cm away from the cancerous lesion.Genomic DNA was extracted from gastric mucosa.Gastric mucosal microbiota were analyzed by high-throughput sequencing.Linear discriminant analysis effect size(LEfSe)was used to analyze the difference of microbiota composition between AGC patients and EGC patients.We used optimizing the area under the receiver operating characteristic curve of the random forest(AUC-RF)algorithm to establish random forest(RF)model that was used to identify the characteristics of bacteria associated with gastric cancer progression.Results Shannon index of EGC group was 2.454,while AGC group was 2.673,there was significantly higher in AGC group than in EGC group(t=2.577,P=0.013).In contrast,there was no significant difference in the Chao 1 index between EGC group and AGC group(t=0.619,P=0.538).Principal coordinate analysis(PCoA)showed apparent separation of the two stages of gastric cancer on the plot(P=0.04).At the genus level,LEfSe analyses identified 12 genera(Thermus,Burkholderia,Pseudomonas,Brevibacillus,Stenotrophomonas,Salinivibrio,Uruburuella,Schlegelella,Bradyrhizobium,Nocardioides,Bacillus and Ochrobactrum)with LDA scores≥2.0.A RF classification model was built with the AUC-RF algorithm that the results showed a minimal set of 24 bacterial genera that maximally differentiated AGC from EGC(AUC=0.916,95%CI:0.841-0.990,sensitivity=0.800,specificity=0.767).Conclusion In the process of gastric cancer progression,gastric microbiota changed significantly in terms of composition and diversity.RF model identified an optimal 24 bacterial signature that distinguished EGC from AGC with high accuracy,which may be associated to gastric cancer prog

关 键 词：晚期胃癌 早期胃癌 胃菌群 随机森林 线性判别效应 

分 类 号：R735.2[医药卫生—肿瘤]