简介 肝内胆管细胞癌 (iCCA) 是一种预后不良的原发性肝恶性肿瘤。目前的预后方法对于有手术切除病变的患者最为准确。然而，相当比例的 iCCA 患者不适合手术。我们旨在开发一种基于临床变量的通用分期系统来确定所有 iCCA 患者的预后。 方法 拆分队列包括 2000 年至 2011 年见过的 436 名 iCCA 患者。对于外部验证，将纳入 2000 年至 2014 年见过的 249 名 iCCA 患者。进行生存分析以确定预后预测因子。全因死亡是主要终点。 结果 东部合作肿瘤学组织 (ECOG) 状态、肿瘤数量、肿瘤大小、转移、白蛋白和 CA 19-9 被纳入 4 分期算法中。Kaplan Meier 估计了 1 年生存率，I、II、III 和 IV 分期分别为 87.1% (95%CI 76.1-99.7)、72.7% (95%CI 63.4-83.4)、48.0% (95%CI 41.2-56.0) 和 16% (95%CI 11-23.5)。 单因素分析表明，与 I 期（参考）相比，II 期（HR:1.71；95%CI 1.0-2.8）、III 期 (HR:3.32；95%CI 2.07-5.31) 和 IV 期 (HR:7.44；95%CI 4.61-12.01) 的死亡风险有显著差异。Concordance indices 表明，新分期系统对于预测拆分队列中的死亡率优于 TNM 分期，p 值 <0.0001。然而在验证队列中，两个分期系统之间的差异并不明显。 讨论 提出的独立验证的分期系统使用非组织病理学数据成功地将患者分为四个分期。与 TNM 分期相比，该分期系统具有更好的预后准确性，可以帮助医生和患者对 iCCA 进行治疗。 版权所有 ©2023 The American College of Gastroenterology。

IntroductionIntrahepatic cholangiocarcinoma (iCCA) is a primary liver malignancy with poor prognosis. Current prognostic methods are most accurate for patients with surgically-resectable disease. However, a significant proportion of patients with iCCA are not surgical candidates. We aimed to develop a generalizable staging system based on clinical variables to determine prognosis of all iCCA patients.MethodsThe derivation cohort included 436 patients with iCCA seen between 2000 and 2011. For external validation, 249 patients with iCCA seen from 2000 to 2014 were enrolled. Survival analysis was performed to identify prognostic predictors. All-cause mortality was the primary end-point.ResultsEastern Cooperative Oncology Group (ECOG) status, tumor number, tumor size, metastasis, albumin, and CA 19-9 were incorporated into a 4-stage algorithm. Kaplan Meier estimates for 1-year survival were 87.1% (95%CI 76.1-99.7), 72.7% (95%CI 63.4-83.4), 48.0% (95%CI 41.2-56.0) and 16% (95%CI 11-23.5), respectively for stages I, II, III, and IV. Univariate analysis yielded significant differences in risk of death for stages II (HR:1.71; 95%CI 1.0-2.8), III (HR:3.32; 95%CI 2.07-5.31), and IV (HR:7.44; 95%CI 4.61-12.01) compared to stage I (reference). Concordance indices showed the new staging system was superior to the TNM staging for predicting mortality in the derivation cohort, p-value <0.0001. In the validation cohort however, the difference between the 2 staging systems was not significant.DiscussionThe proposed independently-validated staging system uses non-histopathologic data to successfully stratify patients into four stages. This staging system has better prognostic accuracy compared to the TNM staging and can assist physicians and patients in treatment of iCCA.Copyright © 2023 by The American College of Gastroenterology.