Background: Adjuvant CAP is the standard of care for resected CCA according to the BILCAP trial. However, the prognosis of patients with resected CCA is still poor. As GemCis is the standard first-line therapy for unresectable/metastatic BTC, we investigated the role of adjuvant GemCis in resected BTC. Because BTC is heterogeneous disease according to the primary tumor site, this study included only pts with resected LN+ extrahepatic CCA. Methods: STAMP is a multicenter, open-label, randomized phase 2 study. Pts with ≥ 19 years, ECOG PS 0/1, adenocarcinoma of perihilar or distal bile duct, at least one regional LN metastasis (N1 or greater), complete macroscopic (R0 or R1) resection within 12 weeks before randomization were eligible. Distant metastasis or R2 disease, previous chemotherapy or radiotherapy, or a serum CA 19-9 level ≥ 100 U/mL were ineligible. Pts were randomized 1:1 to GemCis (Gem 1,000 mg/m2 IV, and Cis 25 mg/m2 IV on day 1 and 8, every 3 weeks) or CAP (1,250 mg/m2 orally twice daily on days 1-14, every 3 weeks) for 8 cycles. Tumor response was performed every 12 weeks for the first 2 years, followed by every 24 weeks for the next 3 years. Primary endpoint was disease-free survival (DFS). Secondary endpoints were overall survival (OS) and safety. This study was designed to improved 2-year DFS rates from 22% (CAP) to 40% (GemCis). Considering follow-up loss rates of 10% with a 1-sided type I error of 0.1 and a type II error of 0.2, a total of 100 patients (50 in each arm) were required. Results: Between JUL 2017 and NOV 2020, a total of 101 pts (50 for GemCis group and 51 for CAP group) were included in the ITT population. Perihilar and distal bile duct were primary tumor sites in 45 pts (44.6%) and 56 pts (55.4%), respectively and 32 pts (31.7%) had R1 resection. Pts characteristics were well balanced between two arms. With median follow-up duration of 28.7 mo (IQR 17.2-39.4), the 2-year DFS rates were 38.5% (1-sided 90% CI, 29.5-47.4%) in GemCis group and 25.1% (17.4-33.5%) in CAP group. The median DFS were 14.3 mo (10.7-16.5 mo) in GemCis group and 11.1 mo (8.4-12.7 mo) in CAP group (HR=0.96 [0.71-1.30], p=0.86). The median OS were 35.7 mo (29.5 mo-not estimated [NE]) in GemCis group and 35.7 mo (30.9 mo-NE) in CAP group (HR=1.08 [0.72-1.64], p=0.81). Grade 3-4 adverse events (AEs) occurred in 42 pts (84.0%) and 8 (16.0%) in GemCis and CAP groups, respectively. The most common AE of grade 3-4 was neutropenia (n = 36, 72.0%) in GemCis group and hand-foot skin reaction (n = 4, 8.0%) in CAP group. Conclusions: In this study including prognostically homogeneous pts population, GemCis was feasible as adjuvant therapy, but failed to improve survival outcomes compared to CAP. CAP should remain standard adjuvant therapy for resected BTC. Clinical trial information: NCT03079427.