目的 基于中国胆囊癌研究小组（CRGGC）胆囊癌专病队列数据，验证美国癌症联合委员会（AJCC）第8版分期系统在胆囊神经内分泌癌（GB-NEC）预后评估中的适用性，评估不同手术方案及化疗方案的疗效差异。方法 回顾性分析CRGGC胆囊癌专病队列数据中2010—2017年全国49家医院收治的81例GB-NEC病人的临床资料。对纳入病例的临床病理特征、手术及化疗情况进行分析，并采用Kaplan-Meier法和Cox回归模型进行生存分析。结果 GB-NEC病人的男女比为1∶1.89。神经内分泌标记物突触素（Syn）、嗜铬粒蛋白A（CgA）、CD56阳性率分别为89.2%、73.8%和85.7%，Ki67强阳性（>50%）率达79.7%。行根治性手术病人占比为76.3%。中位总体生存期（mOS）为11.0（7，27）个月，中位无进展生存期（mPFS）为7.0（3，20）个月，5年生存率为13.2%。单因素生存分析结果显示，AJCC分期越高，病人预后越差（P<0.001）。TNM Ⅲ期病人中，根治性手术组mOS显著优于非根治性手术组（21.0个月 vs. 9.0个月，P=0.041）。进展期（Ⅲ期+Ⅳ期）病人中，采用依托泊苷+顺铂（EP）/伊立替康+顺铂（IP）化疗组mOS显著优于非EP/IP化疗组（27.0个月 vs. 8.0个月，P=0.001）。多因素生存分析结果显示，手术方式（非根治 vs. 根治，HR=2.710，P=0.001）、N分期（N1+N2 vs. N0，HR=2.054，P=0.007）、化疗方案（非EP/IP方案 vs. EP/IP方案，HR=3.576，P=0.001）是影响GB-NEC预后的独立因素。结论 第8版AJCC胆囊癌TNM分期系统与GB-NEC病人预后密切相关，可作为临床预后评估的初步参考。根治性手术可延长Ⅲ期病人总体生存时间，可作为首选的临床治疗手段；EP/IP方案可显著延长进展期病人生存时间，是潜在优选化疗方案；无淋巴结转移、根治性手术、EP/IP化疗方案是GB-NEC病人预后的独立影响因素。

Objective To validate the applicability of the AJCC 8th edition staging system in predicting prognosis for gallbladder neuroendocrine carcinoma (GB-NEC) and to evaluate the comparative efficacy of different surgical and chemotherapy regimens using data from the Chinese Research Group of Gallbladder Cancer (CRGGC) specialized cohort. Methods Clinicopathological data of 81 patients diagnosed with GB-NEC between 2010 and 2017 across 49 hospitals nationwide were retrospectively analyzed from the CRGGC database. Patient demographics, pathological characteristics, surgical procedures, and chemotherapy regimens were examined. Survival analysis was conducted using the Kaplan-Meier method and Cox proportional hazards regression models. Results The cohort exhibited a male-to-female ratio of 1:1.89. Immunohistochemically, the neuroendocrine markers synaptophysin (Syn), chromogranin A (CgA), and CD56 demonstrated positive rates of 89.2%, 73.8%, and 85.7%, respectively, with Ki67 strong positivity (>50%) observed in 79.7% of cases. Radical resection was achieved in 76.3% of patients. Median overall survival (mOS) was 11.0 (7, 27) months, median progression-free survival (mPFS) was 7.0 (3, 20) months, and the 5-year survival rate was only 13.2%. Univariate analysis revealed that higher AJCC stage correlated with significantly worse prognosis (P<0.001). In patients with TNM stage Ⅲ disease, those undergoing radical surgery had a significantly longer mOS than those receiving non-radical surgery (21.0 months vs. 9.0 months, P=0.041). For patients with advanced disease (stage Ⅲ and Ⅳ), treatment with etoposide plus cisplatin (EP) or irinotecan plus cisplatin (IP) resulted in a significantly superior mOS compared to non-EP/IP regimens (27.0 months vs. 8.0 months, P=0.001). Multivariate analysis identified surgical approach (non-radical vs. radical, HR=2.710, P=0.001), nodal status (N1+N2 vs. N0, HR=2.054, P=0.007), and chemotherapy regimen (non-EP/IP vs. EP/IP, HR=3.576, P=0.001) as independent prognostic factors for GB-NEC. Conclusion The 8th edition AJCC TNM staging system for gallbladder cancer closely correlates with the prognosis of GB-NEC and offers a preliminary reference for clinical prognostic assessment. Radical resection significantly prolongs overall survival in patients with TNM stage Ⅲ disease and should be considered the preferred treatment modality. The EP/IP regimen substantially improves patient survival and represents a potentially optimal chemotherapy strategy. Absence of lymph node metastasis, radical surgical resection, and administration of EP/IP chemotherapy constitute independent predictors of prognosis in GB-NEC.