基于唾液酸化相关长链非编码RNA构建卵巢癌预后模型的研究  

Establishment of a prognostic model for ovarian cancer based on sialylation-related long chain non-coding RNA

作　　者：胡超扬 曹晓华 刘格良 李渊 齐荣暄 张豪 于琦 孙焱 HU Chaoyang;CAO Xiaohua;LIU Geliang;LI Yuan;QI Rongxuan;ZHANG Hao;YU Qi;SUN Yan(College of Management of Shanxi Medical University,Taiyuan,Shanxi,030600;Shanxi Key Laboratory of Big Data for Clinical Decision Research,Taiyuan,Shanxi,030600)

机构地区：[1]山西医科大学管理学院,山西太原030600 [2]山西省临床决策研究大数据重点实验室,山西太原030600

出　　处：《实用临床医药杂志》2023年第16期29-36,42,共9页Journal of Clinical Medicine in Practice

基　　金：山西省研究生教育教学改革基金资助项目(2022YJJG112);山西省回国留学人员科研基金资助项目(HGKY2019057);山西省研究生教育创新基金资助项目(2022Y426);山西省重点研发计划基金资助项目(201903D311011);山西医科大学2022年度校级研究生精品示范课程建设基金资助项目(山医大研[2022]11号-No.3);山西省研究生实践创新基金资助项目(2023SJ167)。

摘　　要：目的基于唾液酸化相关长链非编码RNA(lncRNA)构建卵巢癌(OC)预后模型,分析患者预后免疫反应与抗癌药物敏感性。方法从癌症基因图谱(TCGA)数据库获取OC基因表达数据和临床数据;采用相关性分析筛选唾液酸化相关lncRNA;采用Lasso和Cox回归分析筛选卵巢癌生存相关唾液酸化lncRNA(OCSS lncRNA)并构建预后模型;通过生存分析、受试者工作特征(ROC)曲线等评估模型的效能;应用单因素和多因素Cox回归分析筛选OC的独立预后因素,并绘制列线图;采用CIBERSORT算法、肿瘤免疫功能障碍和排斥(TIDE)评分评估OC患者免疫细胞浸润与免疫治疗获益情况;采用药物敏感性分析获取潜在治疗药物。结果构建了7个OCSS lncRNA组成的OC预后模型;生存分析显示,总数据集、训练组和验证组中,高风险者的总体生存率(OS)均低于低风险者,差异有统计学意义(P<0.05),总数据集中高风险者的无进展生存期OS低于低风险者,差异有统计学意义(P<0.05)。OC预后模型1、3、5年的ROC曲线显示,模型效能较好且优于其他临床特征;单因素和多因素Cox回归证实,年龄和风险评分是独立预后因素(P<0.05);列线图结合校正曲线表明,预测的患者OS与实际OS基本一致。免疫细胞浸润分析显示,γδT细胞和M1巨噬细胞的含量在高、低风险组中的差异有统计学意义(P<0.05);免疫治疗敏感性分析显示,高、低风险组的TIDE评分差异有统计学意义(P<0.05)。药物敏感性分析筛选出9种对高风险OC患者具有较高治疗价值的药物(P<0.01)。结论7个OCSS lncRNA构建的OC预后模型有助于预测OC患者预后,并为OC患者的临床应用、免疫与药物治疗提供参考依据。Objective To establish a prognostic model for ovarian cancer(OC)based on sialylation-related long non-coding RNA(lncRNA),and to analyze prognostic immune response of patients and sensitivity to anticancer drugs.Methods The Cancer Genome Atlas(TCGA)database was utilized to acquire gene expression data and clinical data for OC;correlation analysis was used to screen for sialylation-related lncRNA;the Lasso and Cox regression analyses were used to screen for ovarian cancer survival-related salivation-related lncRNA(OCSS lncRNA),and a prognostic model was established;the efficiency of the evaluation model was assessed by survival analysis and receiver operating characteristic(ROC)curve;single-factor and multi-factor Cox regression analyses were applied to screen for independent prognostic factors for OC,and a Nomogram was drawn;the CIBERSORT algorithm and Tumor Immune Dysfunction and Exclusion(TIDE)score were used to evaluate immune cell infiltration and immunotherapy benefits in OC patients;the drug sensitivity analysis was used to obtain potential therapeutic drugs.Results A prognostic model for OC comprising 7 OCSS lncRNAs was established;the survival analysis revealed that the overall survival rate(OS)of high-risk patients was significantly lower than that of low-risk patients in the total dataset,training set and validation set(P<0.05),and the progression-free survival OS of high-risk patients was also significantly lower than that of low-risk patients in the total dataset(P<0.05).The ROC curves of the OC prognosis model at 1 year,3,and 5 years demonstrated that the efficiency of model was excellent and better than other clinical characteristics;the single-factor and multi-factor Cox regression confirmed that age and risk score were independent prognostic factors(P<0.05);the Nomogram combined with calibration curve demonstrated that the predicted OS of patients was basically consistent with the actual OS.Immune cell infiltration analysis showed significant differences in the contents ofγδT cells and M1 macroph

关 键 词：卵巢癌 唾液酸化长链非编码RNA 转录组 生存预后 免疫浸润 药物敏感性 

分 类 号：R737.31[医药卫生—肿瘤] Q[医药卫生—临床医学]