Randomized phase III study of S-1 alone versus S-1 + cisplatin in the treatment for advanced gastric cancer (The SPIRITS trial) SPIRITS: S-1 plus cisplatin vs S-1 in RCT in the treatment for stomach cancer

Abstract

4514 Background: S-1 has been widely used against advanced gastric cancer (AGC) not only as monotherapy but also in combination with other cytotoxic compounds. Results of a phase I/II study combining S-1 + cisplatin (CDDP) were very encouraging with a high response rate (RR) of 76%, and the MST (Median Survival Time) of 383 days (Koizumi W et al, Br J Cancer, 2003). Based on these results, a phase III study comparing S-1 alone with S-1 + CDDP has been conducted to further evaluate the efficacy and safety for S-1 + CDDP as a standard treatment for AGC. Methods: This is a randomized, controlled, open-label, parallel, multicenter study. Patients (pts) are randomized to one of two treatment arms. Arm A: Pts receive oral S-1 (40 mg/m2) twice daily 28 days followed by 14 days rest. Arm B: Pts receive oral S-1 (40 mg/m2) twice daily 21 days followed by 14 days rest plus CDDP (60 mg/m2) iv on day 8. Eligibility criteria included unresectable/recurrent AGC, age 20–74, no prior chemotherapy for AGC. Primary endpoint was overall survival (OS). Main secondary endpoints included RR, time to treatment failure (TTF) and toxicity. Based on planned sample size of 284 pts, the trial was designed to have 90% power to detect an improvement in median OS from 8 to 12 months (2-sided log-rank test; significance level 0.05). Results: 305 pts (Arm A/B, 152/153) were randomized between Mar 2002 and Nov 2004. The eligible pts were 299 (Arm A/B, 150/149). Median age was 62.0/61.5 yrs. At a 2 yrs follow-up since last patient in, the MST for Arm A was 335.5 days (95%CI: 292.0 - 402.0) and for Arm B was 396.0 days (95%CI: 342.0 - 471.0). The OS for Arm B was superior to Arm A (log-rank p=0.0366, hazard ratio: 0.774, 95% CI: 0.608 - 0.985). RR was 31.1% for Arm A and 54.0% for Arm B. In Arm A vs Arm B, the most common grade 3/4 toxicities were: leucopenia, 2.0% vs 11.5%; neutropenia, 10.7% vs 39.9%; anemia (decreased Hb), 4.0% vs 25.7%; nausea, 1.3% vs 11.5%; anorexia, 6.0% vs 30.4%. No treatment related death was observed. Conclusions: The combination treatment of S-1 and CDDP met primary endpoint of OS, and was found to be effective and well tolerated in pts with AGC. Accordingly, this regimen can be regarded as one of first-line standard treatment for AGC. No significant financial relationships to disclose.