目的：分析肝脏手术切缘对肝内胆管癌（intrahepatic cholangiocarcinoma,ICC）病人预后的影响,并探索ICC切缘推荐距离。方法：收集2011年1月至2017年1月北京肿瘤医院肝胆外科ICC病人的临床病理资料。用单因素和多因素Cox分析影响ICC预后的危险因素,并用Kaplan-Meier方法制作无复发生存（recurrent free survival,RFS）期和总生存（overall survival,OS）期曲线。采用限制性立方样条（restricted cubic spline,RCS）分别探索R0切缘距离与OS期和RFS期的关系,并探索手术切缘推荐距离。结果：本研究共71例病人,随访2~107个月,中位RFS期8个月,中位OS期17个月。1、3、5年的RFS率分别为35%、20%、10%,OS率分别为68%、38%、23%。Cox回归分析发现,美国癌症联合委员会（American Joint Committee on Cancer,AJCC）分期、手术切缘、甲胎蛋白（alpha-fetoprotein,AFP）、糖类抗原19-9（carbohydrate antigen 19-9,CA19-9）是ICC病人预后的独立危险因素。共R0切除59例（83%）,R1切除12例（17%）。比较R0与R1切除病人的术前临床病理因素,发现肿瘤单发、AJCC分期或日本肝癌研究组（Liver Cancer Study Group of Japan,LCSGJ）分期处于Ⅰ期、Ⅱ期时,R0切除的可能性更大。本研究原始数据为病理切缘,考虑到病理标本制作存在缩水,将病理切缘换算为手术切缘。RCS和Kaplan-Meier法分析发现,宽切缘组（0.5 cm<X₁<2.8 cm）与窄切缘组（0.1 cm<X₂≤0.5 cm）相比,RFS期更长。结论：R1切除是ICC病人预后的独立危险因素。对于肿瘤单发、AJCC分期或LCSGJ分期处于Ⅰ期、Ⅱ期的ICC病人,行肝切除术时,R0切除的可能性更大。至少实现0.5 cm以上手术切缘,以改善预后。

Objective To analyze the effect of surgical margin on the prognosis of patients with intrahepatic cholangiocarcinoma (ICC), and to explore the distance of margin of ICC hepatectomy to be recommended. Methods The clinical and pathological data of patients with ICC were collected from January 2011 to January 2017 in the Cancer Hospital in Beijing. Single-factor and multi-factor Cox analysis were used to explore risk factors that affected the prognosis of ICC. Kaplan-Meier method was used for recurrent free survival (RFS) and overall survival (OS) curves. Restricted cubic splines (RCS) were used to study the relationship between R0 margin distance and OS or RFS, and the recommended distance of surgical margin. Results There were 71 patients with median RFS 8 months and OS 17 months after follow-up from 2 to 107 months in this study. One, 3 and 5 year RFS were 35%, 20% and 10%, and OS 68%, 38%, and 23%, respectively. It was shown by Cox regression analysis that the American Joint Committee on Cancer (AJCC) staging, surgical margin, alpha-fetoprotein, carbohydrate antigen 19-9 were independent risk factors of the prognosis of ICC patients. R0 resection was in 59 (83%) cases and R1 resection in 12 (17%) cases. R0 resection was found to be more likely in the patients when tumor was single, in stage Ⅰ or stage Ⅱ of AJCC or of Liver Cancer Study Group of Japan (LCSGJ) after comparison the preoperative clinicopathological factors between R0 resection and R1 resection. Original data was pathological margin and for the shrinkage during preparation of pathological specimens, the pathological margin converted to surgical margin. Using restricted cubic spline and Kaplan-Meier method it was shown that wider margin group (>0.5 cm-<2.8 cm) had longer RFS than more narrow margin group (>0.1 cm-≤0.5 cm). Conclusions R1 resection would be an independent risk factor for the prognosis of ICC patients. When the tumor is single, AJCC staging or LCSGJ staging in stage Ⅰ or stage Ⅱ, it could be more likely to have R0 resection and surgical margin at least 0.5 cm or more could improve the prognosis.