Randomized phase II study of preoperative concurrent chemoradiotherapy with or without induction chemotherapy with S-1 and oxaliplatin in patients with resectable esophageal cancer.

4093 Background: The value of adding induction chemotherapy (ICT) to preoperative chemoradiotherapy followed by surgery has not been delineated well. Methods: Patients with stage II, III or IVA (by AJCC 6th ed.) esophageal cancer were randomly allocated to either 2 cycles of ICT (oxaliplatin 130 mg/m2 on day 1 and S-1 at 40 mg/m2 bid on days 1-14, every 3 weeks), followed by concurrent chemoradiotherapy (CCRT) (46 Gy, 2 Gy/day with oxaliplatin 130 mg/m2 on day 1 and 21 and S-1 30 mg/m2 bid, 5 days/week during radiotherapy) and surgery (arm A, n=48), or the same chemoradiotherapy followed by surgery without ICT (arm B, n=49). Primary outcome was to compare pathologic complete response (pCR). Results: Thirty six and 35 patients underwent surgery with or without ICT, respectively. pCR rate among those who underwent surgery was significantly lower in arm A (30.6% vs. 54.3%, p=0.043). However, no difference in progression-free survival (PFS) and overall survival (OS) was observed with a median follow-up of 19.5 mo (95% CI, 19.1-22.4). Two-year PFS rate was 63.8% in arm A and 55.2% in arm B (p=0.626) and 2-year OS rate was 70.1% and 62.6%, respectively (p=0.515). While 47 (arm A) and 48 (arm B) patients received at least 44 Gy of radiotherapy, relative dose intensity (RDI) for oxaliplatin during CCRT was significantly lower in arm A vs. arm B (92.7% ± 19.6% vs. 99.7 ± 1.8%, p=0.017). RDI for S1 did not significantly differ (94.1% ± 17.3% vs. 98.5% ± 5.9%, p=0.095). G3/4 thrombocytopenia was significantly common in arm A (37.5% vs. 4.1%, p <0.001), which contributed to lower RDI of oxaliplatin. Three patients in arm A, compared to none in arm B, failed to survive for 90 days after surgery. Conclusions: Adding this ICT to preoperative chemoradiotherapy seems to cause lower pCR rate and higher toxicity during CCRT.