4074 Background: Currently, S-1 with cisplatin is the standard regimen for advanced gastric cancer (AGC) in Japan. However, the triple combination of S-1, cisplatin and paclitaxel yielded a stronger antitumor effect in our multicenter phase II study (Iwase H et al., ASCO 2008). Since the number of patients (pts) and follow-up period was limited, it was decided that additional safety and efficacy data should be generated. Thus, subjects were increased from an initial 42 pts to the final 63 pts in a multicenter setting. Methods: Paclitaxel (160 mg/m2/day) was administered by infusion for 3 hours on the first day. S-1 (70 mg/m2/day) was administered orally for 14 consecutive days from the first day. Cisplatin (60 mg/m2) was administered as a 24-h infusion on day 14 of every 28-day cycle. In pts who had a related episode of grade 3 or 4 toxicity, S-1 was reduced by 10mg/m2 and paclitaxel was reduced 25% for the subsequent cycle. The drug doses were never increased. Results: All 63 pts were assessed for safety and 60 pts were assessed for efficacy. The median age was 58 years. A total of 240 cycles were administered (median 3, range 1–10). Grade 3-4 toxicities included neutropenia in 33.3% (grade 4: 15.8%), thrombocytopenia in 14.3%, and anemia in 12.7%. In regards to the nonhematological toxicity, there was no grade 3-4. Grade 2 toxicities included nausea/vomiting in 28.5%, myalgia in 22.2%, neuropathy in 23.8%, rush in 9.5% and renal dysfunction in 7.9% of pts. All treatment related toxicities were resolved, and no toxic death was reported. Complete response (CR) was observed in 6 pts and partial response (PR) was observed in 32 pts (CR rate: 10%, PR rate: 53.3%). The overall response rate was 63.3% (38/60). The median survival time and progression free survival were 15.0 months and 7.5 months, respectively. The 1, 2 and 3-year cumulative survival rate were 65%, 15%, and 7%, respectively. Conclusions: These extended data confirm that this triple combination chemotherapy has a very desirable safety profile considering dose modifications and encouraging active efficacy data in AGC. Phase III studies are required to establish the optimum treatment regimens. No significant financial relationships to disclose.