论文部分内容阅读
目的:探讨肝内胆管癌患者根治性切除术后辅助化疗的临床价值及获益人群。方法:采用回顾性队列研究方法。回顾性收集2010年1月至2018年12月于国内10家三甲医院接受意向性根治性切除术的685例肝内胆管癌患者的临床和病理学资料。男性355例，女性330例；年龄［n M（IQR）］59（14）岁（范围：22~83岁）。应用倾向性评分匹配法对辅助化疗组及未辅助化疗组患者的组间差异进行均衡，使用Log-rank检验比较两组患者预后，以中位无复发生存时间（14个月）为目标变量构建贝叶斯网络无复发生存预测模型，依据多态Birnbaum重要度计算对相关预后影响因素重要度进行排序，进一步建立生存概率预测表，分析辅助化疗的获益人群。n 结果:685例患者中，接受辅助化疗214例，未接受辅助化疗471例。经倾向性评分匹配后，共筛选出124对患者，辅助化疗组患者的总体生存时间及无复发生存时间均优于未辅助化疗组（总体生存时间：32.2个月比18.0个月，n P=0.003；无复发生存时间：18.0个月比10.0个月，n P=0.001）。贝叶斯网络无复发生存预测模型的曲线下面积为0.7124。预后影响因素重要度排序结果依次为脉管侵犯、神经浸润、N分期、T分期、肝包膜侵犯、辅助化疗、肿瘤位置、年龄、肿瘤分化程度、性别、甲胎蛋白水平和术前黄疸。基于重要度大于0.1的变量（脉管侵犯、神经浸润、N分期、T分期）与辅助化疗建立生存概率预测表，结果显示，所有患者均可从辅助化疗获益（无复发生存时间≥14个月概率增幅2.21%~7.68%），早期患者行辅助化疗后无复发生存时间≥14个月的概率提升更为明显。n 结论:肝内胆管癌患者根治术后行辅助化疗可明显延长患者的总体生存时间及无复发生存时间，早期患者从辅助化疗中的获益可能更大。“,”Objectives:To investigate the clinical value of adjuvant chemotherapy(ACT) in patients with intrahepatic cholangiocarcinoma(ICC) who underwent radical resection and to explore the optimal population that can benefit from ACT.Methods:A retrospective cohort study method was adopted. The clinical and pathological data of 685 patients with ICC who underwent curative intent resection in 10 Chinese hepatobiliary surgery centers from January 2010 to December 2018 were collected;There were 355 males and 330 females. The age(n M(IQR)) was 58(14) years (range: 22 to 83 years). Propensity score matching(PSM) was applied to balance the differences between the adjuvant and non-adjuvant chemotherapy groups. Log-rank test was used to compare the prognosis of the two groups of patients. A Bayesian network recurrence-free survival(RFS) prediction model was constructed using the median RFS time (14 months) as the target variable, and the importance of the relevant prognostic factors was ranked according to the multistate Birnbaum importance calculation. A survival prognostic prediction table was established to analyze the population benefiting from adjuvant chemotherapy.n Results:Among 685 patients,214 received ACT and 471 did not receive ACT. A total of 124 pairs of patients were included after PSM, and patients in the ACT group had better overall survival (OS) and RFS than those in the non-ACT group(OS: 32.2 months n vs. 18.0 months,n P=0.003;RFS:18.0 months n vs. 10.0 months,n P=0.001). The area under the curve of the Bayesian network RFS prediction model was 0.7124. The results of the prognostic factors in order of importance were microvascular invasion,perineural invasion,N stage,T stage,hepatic envelope invasion,adjuvant chemotherapy,tumor location,age,pathological differentiation,sex,alpha-fetoprotein and preoperative jaundice. A survival prediction table based on the variables with importance greater than 0.1 (microvascular invasion,perineural invasion,N stage,T staging) and ACT showed that all patients benefited from ACT (increase inthe probability of RFS≥14 months from 2.21% to 7.68%), with a more significant increase in the probability of RFS≥14 months after ACT in early-stage patients.n Conclusion:ACT after radical resection in patients with ICC significantly prolongs the OS and RFS of patients, and the benefit of ACT is greater in early patients.