基于SEER数据库构建转移性胆囊癌的预后列线图

Constructing a prognostic nomogram for metastatic gallbladder carcinoma based upon SEER data

目的构建可有效预测转移性胆囊癌(metastatic gallbladder carcinoma,MGBC)病人预后的列线图模型。方法提取2004-2017年美国国立癌症研究所的监测、流行病学和结果(SEER)数据库所登记的709例MGBC病人临床病理资料。采用简单随机法,按7∶3的比例分为发现集(n=496)与验证集(n=213)。使用Cox比例风险回归模型筛选影响胆囊癌肿瘤特异性生存期(gallbladder carcinoma cancer-specific survival,GCSS)的影响因素。发现集数据中根据GCSS的独立影响因素构建列线图模型,然后利用验证集完成外部验证。利用一致性指数、绘制校准曲线及受试者工作特征曲线评估模型的预测精度。利用Kaplan-Meier生存分析和log-rank检验分析高、低风险组间的生存差异。结果多因素分析提示年龄、肿瘤分化程度、是否手术、是否放疗、是否化疗是独立的危险因素。发现集一致性指数为0.704,验证集一致性指数为0.732。根据列线图模型进行风险分层,Kaplan-Meier生存分析显示:发现集和验证集的低风险组的预后均明显优于高风险组。结论建立的列线图模型可准确、直观地预测MGBC病人的生存概率。

Objective To construct and validate an effective prognostic nomogram for metastatic gallbladder carcinoma(MGBC).Methods This study retrospectively analyzed 709 MGBC patients from 2004 to 2017 in the database of Surveillance,Epidemiology and End Results(SEER).The data were randomized into the sets of discovery(n=496)and validation(n=213).The independent prognostic factors affecting the gallbladder carcinoma cancer-specific survival(GCSS)of patients were obtained from univariate and multivariate Cox regression analyses.A nomogram was constructed from the discovery sets.The comprehensive survival probability scores of each patient were assessed by the independent prognostic factors and the results validated in validation sets.The predictive accuracy of the nomogram was determined by concordance index(C-index),calibration curve and receiver operating characteristic(ROC).Kaplan-Meier survival analysis and log-rank test were performed for examining the difference in survival between the high/low-risk groups.Results Multivariate analysis of the discovery sets revealed that independent factors for GCSS were age,degree of tumor differentiation,whether or not surgery,radiotherapy or chemotherapy.In the discovery sets,C-index of the nomogram for predicting survival was 0.702.In the validation sets,C-index of nomogram discrimination was 0.732.Risk stratification was performed with the nomogram.Kaplan-Meier survival analysis indicated that for both discovery and validation sets,low-risk group had a significantly better prognosis than high-risk group.Conclusion The above nomogram can accurately and intuitively predict the survival probability of MGBC patients.