PD-035 Efficacy and safety of second-line irinotecan based chemotherapy in early relapse patients with gastric cancer after adjuvant chemotherapy: exploratory subgroup analysis of TRICS trial

Abstract

Introduction: Second-line chemotherapy (SLC) was reported to improve the survival of advanced gastric cancer (AGC). A randomized phase III study of second-line irinotecan plus cisplatin versus irinotecan alone in patients with AGC refractory to S-1 monotherapy, the TRICS trial, revealed that both irinotecan based chemotherapies were effective with favorable long-term survivals and generally well tolerated. Eligible patients included patients with recurrence within 6 months after the completion of adjuvant therapy with S-1 (early relapse cases), or patients with tumor progression after first-line S-1 for an advanced cancer (progressive cases), in the TRICS trial. However, it is unclear whether the efficacy and safety in early relapse cases were different from these in progressive cases or not. Methods: A total of 168 patients registered to the TRICS trial were classified as early relapse group (n = 111) or progressive group (n = 57), and 168 (irinotecan plus cisplatin, early relapse 56/83, 67%; irinotecan alone, 55/85, 65%) and 163 (55/81, 68%; 53/82, 65%) patients were analyzed for efficacy and safety, respectively. Median OS, PFS and RR for each disease type group were estimated. Differences in therapeutic efficacy between irinotecan plus cisplatin and irinotecan alone were also tested for each group. Results: In baseline characteristics, PS 0 were more frequently seen in early relapse group than in progressive group (86% vs 72%, p = 0.0328), and median age of early relapse group was younger than progressive group (66 vs 72, p = 0.0162). No differences of the median number of courses (6 vs 4 courses), treatment delays (53.7% vs 50.9%) and post-protocol therapy (73.8% vs 64.5%), were identified between each group. The median OS was 13.9 months in early relapse group and 10.0 months in progressive group (HR:0.746, p = 0.1019). No differences of the median OS were identified between irinotecan plus cisplatin and irinotecan alone regimen in either group (early relapse group: 14.0 vs 13.5, HR: 0.753, p = 0.1799) ( progressive group: 10.6 vs 10.0, HR: 0.876, p = 0.6493). Early relapse group showed better trends in PFS compared to the progressive group (4.7M vs 3.6M, HR: 0.650, p = 0.0149). No differences of the median PFS were identified between irinotecan plus cisplatin and irinotecan alone regimen in either group (early relapse group: 5.0 vs 4.5, HR: 0.819, p = 0.3220) (progressive group: 3.6 vs 3.2, HR: 0.868, p = 0.6359). Early relapse group was associated with significantly better RR compared to progressive group (21.3% vs 4.9%, p = 0.0177). Although there were no significant differences between irinotecan plus cisplatin and irinotecan alone regimen in either group (early relapse group: 19.6% vs 23.3%) ( progressive group: 11.1% vs 0%). There were no significant differences of both any grade and grade 3 or worse adverse events between each group. In early relapse group, compared to irinotecan alone regimen, irinotecan plus cisplatin showed significantly higher grade 3 or worse toxicities, only regarding anemia (20% vs. 0%). In progressive group, there were no significant differences of grade 3 or worse adverse events between each regimen. Conclusion: In second-line setting of AGC, irinotecan based chemotherapies were effective and generally well tolerated, especially in early relapse cases after adjuvant therapy with S-1.