Summary
Background Although its efficacy is unproven, 5-fluorouracil plus cisplatin (FP) is used to prevent postoperative relapse in gastric cancer. We investigated the safety and feasibility of S-1 plus cisplatin (SP) vs. FP for stage IIIB-IV (M0) gastric cancer. Methods Following curative resection, 41 stage IIIB-IV (M0) gastric cancer patients were assigned to SP (eight 14-day cycles of S-1 [40 mg/m2 twice daily] plus cisplatin [60 mg/m2 day 1] administered every 3 weeks) or FP (six 3-day cycles of FU [1 g/m2 per day] plus cisplatin [80 mg/m2 day 1] every 4 weeks). Doses were reduced based on predefined criteria. Results Patient characteristics were balanced between the two arms. In total, 124 cycles of SP (N = 20, median = 7, range 1–8) and 113 cycles of FP (N = 21, median 6, range 1–6) were administered. The median relative dose intensity per patient was 75% (49.99–100%) for S-1, 100% (75–100%) for cisplatin in SP, and 100% (64–100%) for 5-FU, 100% (60–100%) for cisplatin in FP. The relative dose intensity of FP was stable, while that of SP decreased during treatment. After median follow-up of 7.9 months (3.8–14.55), the median RFS was not reached. Relapse occurred in two (10%) patients on SP and five (23.8%) in the FP arm (P = 0.24). The incidence of grade 3–4 granulocytopenia was 36.8% with SP and 14.3% with FP. Grade 3–4 non-hematologic toxicities included fatigue (5.2% with SP vs. 4.8% with FP), vomiting (10.5% with SP vs. 0% with FP), and infection (5.2% with SP vs. 0% FP). Conclusion S-1 plus cisplatin was feasible and tolerable as adjuvant treatment for stage IIIB-IV (M0) gastric cancer. However, because of decreased relative dose intensity during treatment, further study is warranted to determine optimal dosage and combination.