Randomized phase III study of adjuvant chemotherapy with S-1 versus capecitabine in patients with stage III colorectal cancer: Updated results of Japan Clinical Oncology Group study (JCOG0910)

Abstract

Background: We demonstrated that the second planned interim analysis (median follow-up, 23.7 months) of the JCOG0910 failed to show non-inferiority of adjuvant S- 1 to capecitabine in disease-free survival (DFS: relapse, second malignancy, death are events) at ASCO2015 (abstract # 3512), and the results were opened by the recommendation of JCOG Data and Safety Monitoring Committee. We updated the follow-up data to confirm the conclusion at the interim analysis. Methods: Key eligibility criteria were: stage III, colorectal adenocarcinoma except for lower rectal cancer, R0 with D2/3 lymph node dissection. Patients were randomized to 8 courses of capecitabine (1,250 mg/m2 twice daily, days 1-14, every 3 weeks) or 4 courses of S-1 (40 mg/m2 twice daily, days 1-28, every 6 weeks). Primary endpoint was DFS. Planned sample size was 1,550 in order to provide 80% power with a noninferiority margin at a hazard ratio (HR) of 1.24 and 1-sided a=0.05. This trial is registered with UMIN-CTR, #UMIN000003272. Results: 1,564 patients were randomized to capecitabine (n=782) or S-1 (n=782). At the end of the follow-up period of 3 years, 69% of required events (368/535) were observed, with a median follow-up for all randomized patients of 4.13 years, 3-year DFS was 81.7% (95% CI, 78.8 - 84.2%) in capecitabine and 78.3% (75.2 - 81.0%) in S-1. The HR of DFS was 1.22 (95% CI, 1.00-1.50) and the non-inferiority of S-1 was not demonstrated (P for non-inferiority = 0.448). Three-year relapse-free survival (RFS: relapse, death are events) was 84.6% in capecitabine and 81.5% in S-1. The HR of RFS was also 1.21 (95% CI, 0.96-1.53). Three-year overall survival (OS) was 96.3% in capecitabine and 95.4% in S-1. The HR of OS was 1.18 (95% CI, 0.83-1.68). In the subgroup analyses, no significant interactions were identified between the major baseline characteristics. Conclusions: These updated results confirmed that S-1 could not be demonstrated to be non-inferior to capecitabine in DFS. Adjuvant capecitabine remains the standard treatment and S-1 is not recommended.