Background
Prespecified final 5-year follow-up analysis of the ASCOT (JCOG1202) phase 3 trial: multicenter, open-label, randomized trial in Japan. N=440 with resected biliary tract cancer (extrahepatic/intrahepatic cholangiocarcinoma, gallbladder, ampullary) randomized to adjuvant S-1 vs observation. The primary analysis (Lancet 2023) showed improved 3-year OS (77.1% vs 67.6%). S-1 is an oral fluoropyrimidine (tegafur/gimeracil/oteracil) widely used in East Asia.
Results
Interventions and follow up: Arm A: S-1 40 mg/m² PO twice daily, 4 weeks on / 2 weeks off, ×4 cycles (n=218)
Arm B: Observation (n=222)
Primary endpoint: OS (intention-to-treat)
mFollow up: 61.5 mo
Results: 5-year OS (primary): 64% vs 52%; HR 0.72 (95% CI 0.55–0.95), one-sided P=.01
Median OS: 8.2 yr vs 5.9 yr
Median RFS: 6.1 yr vs 3.5 yr; unstratified HR 0.80 (95% CI 0.62–1.03), stratified HR 0.76 (95% CI 0.59–0.98)
Subgroups: treatment effect generally consistent across subgroups
Arm B: Observation (n=222)
Primary endpoint: OS (intention-to-treat)
mFollow up: 61.5 mo
Results: 5-year OS (primary): 64% vs 52%; HR 0.72 (95% CI 0.55–0.95), one-sided P=.01
Median OS: 8.2 yr vs 5.9 yr
Median RFS: 6.1 yr vs 3.5 yr; unstratified HR 0.80 (95% CI 0.62–1.03), stratified HR 0.76 (95% CI 0.59–0.98)
Subgroups: treatment effect generally consistent across subgroups
Adverse events
Grade ≥3 (any): NR in the follow-up abstract; in the primary report the most common grade ≥3 events with S-1 were biliary tract infection, neutropenia, and skin toxicity, and no new safety signals were reported with longer follow-up
Treatment-related deaths: NR
Treatment-related deaths: NR
Conclusions
With 5 years of follow-up, adjuvant S-1 continues to improve OS over observation after resection of biliary tract cancer (5-year OS 64% vs 52%; HR 0.72), with median OS extended by more than 2 years. The RFS benefit is modest and reaches significance only in the stratified analysis. These mature data confirm adjuvant S-1 as a standard of care for resected BTC in Japan and as a fluoropyrimidine alternative to capecitabine.
Key Limitations
Open-label, Japan-only trial; S-1 is not approved in the US or Europe, and its pharmacokinetics and tolerability differ in Western populations (higher GI toxicity), so results are not directly transferable. The comparator was observation, not capecitabine (BILCAP), so the two adjuvant standards have never been compared head-to-head. The RFS benefit is borderline (unstratified HR 0.80, CI crossing 1), and the OS benefit is larger than the RFS effect, raising questions about post-relapse therapy imbalances. Only four cycles (24 weeks) of S-1 were given, shorter than the 6 months of capecitabine in BILCAP. Detailed long-term toxicity was not reported in the abstract.
Clinical Context
Adjuvant capecitabine for 6 months (BILCAP) is the NCCN- and ESMO-endorsed standard after resection of BTC in Western practice; adjuvant S-1 (ASCOT) is the standard in Japan. ASCOT was the first phase 3 trial in resected BTC to meet its primary OS endpoint (BILCAP was positive only in the per-protocol analysis). The 5-year data strengthen the case for adjuvant fluoropyrimidine in all resected BTC; gemcitabine-based adjuvant regimens (PRODIGE 12, BCAT) remain negative. Ongoing trials are testing gemcitabine-cisplatin plus/minus immunotherapy in the adjuvant/perioperative setting.