肝内胆管细胞癌根治性切除术后临床预后影响因素的研究

目的探讨肝内胆管细胞癌（ICC）行根治性切除（R0切除）术后的独立预后影响因素。 方法回顾性分析中山大学附属第一医院自2002年1月至2018年8月行R0切除并经病理确诊的237例ICC患者的临床病理及随访资料，采用Kaplan-Meier法构建生存曲线；单因素及多因素COX回归分析筛选影响患者生存预后的独立因素。一致性指数（C-index）、时间依赖性受试者工作特征曲线（ROC曲线）、曲线下面积（AUC）及校准曲线用于评价COX回归模型的预测效能。 结果237例ICC患者的中位生存期为19.93个月（95%CI=14.31~26.21），术后1、2、3年生存率分别为62.3%、45.0%及35.6%。单因素及多因素回归分析显示谷氨酰转移酶（GGT） >53 U/L（P=0.001）、CA125>17.4 U/ml（P<0.001）、淋巴结转移（P=0.039）、肿瘤分化不良（P<0.001）是ICC患者行R0切除术后预后不良的独立危险因素。按照删除概率P>0.1的标准，纳入癌胚抗原（CEA）（P=0.092）后的预后模型C-index为0.74（95%CI=0.68~0.78），不同随访时间点对应的AUC值均在0.8左右，模型具有良好的预测效能。 结论GGT、CEA、CA125、淋巴结转移情况及肿瘤分化程度可作为ICC患者R0切除术后临床预后的预测指标。

Prognostic factors of intrahepatic cholangiocarcinoma patients undergoing curative resection

ObjectiveTo identify the independent prognostic factors of intrahepatic cholangiocarcinoma (ICC) patients after receiving curative (R0) resection. MethodsFrom January 2002 to August 2018, the clinicopathological and prognosis data of two hundred and thirty-seven patients with the pathological diagnosis of ICC in the First Affiliated Hospital of Sun Yat-sen University who underwent curative resection were analyzed retrospectively. The survival curve was constructed by Kaplan-Meier method. Univariate and multivariate COX regression analysis were used to screen independent factors affecting survival and prognosis of patients. C-index, ROC curve, AUC curve and calibration curve were used to evaluate the predictive effectiveness of COX regression model. ResultsThe median survival time of ICC patients was 19.93 months (95%CI=14.31-26.21), and the 1-, 2-, and 3-year survival rates were 62.3%, 45.0%, and 35.6%, respectively. Univariate and multivariate regression analysis demonstrated that gamma-glutamyl transpeptidase (GGT) >53 U/L (P=0.001), CA125>17.4 U/ml (P<0.001), lymph node metastasis (P=0.039) and poor tumor differentiation (P<0.001) were the adverse factors of prognosis after R0 resection in ICC patients. According to the criteria for deleting probability: P>0.1, the C-index of the prognostic model containing CEA (P=0.092) was 0.74 (95%CI=0.68-0.78). The AUC values corresponding to different follow-up time points were all around 0.8, and the model had good prediction efficiency. ConclusionGGT, CEA, CA125, lymph node metastasis and poor tumor differentiation are prognostic predictors of ICC patients after receiving R0 resection.