Comparison of neoadjuvant chemotherapy versus upfront surgery with or without chemotherapy for patients with clinical stage III esophageal squamous cell carcinoma.

Neoadjuvant chemotherapy (NAC) and chemoradiotherapy have been shown to extend postoperative survival, and preoperative therapy followed by esophagectomy has become the standard treatment worldwide for patients with esophageal squamous cell carcinoma (ESCC). The Japan Clinical Oncology Group 9907 study showed that NAC significantly extended survival in advanced ESCC, but the survival benefit for patients with clinical stage III disease remains to be elucidated. We compared the survival rates of NAC and upfront surgery in patients with clinical stage III ESCC. Consecutive patients histologically diagnosed as clinical stage III (excluding cT4) ESCC were eligible for this retrospective study. Between September 2002 and April 2007, upfront transthoracic esophagectomy was performed initially and, for patients with positive lymph node (LN) metastasis in a resected specimen, adjuvant chemotherapy using cisplatin and 5-fluororouracil every 3 weeks for two cycles was administered (Upfront surgery group). Since May 2007, a NAC regimen used as adjuvant chemotherapy followed by transthoracic esophagectomy has been administered as the standard treatment in our institution (NAC group). Patient characteristics, clinicopathological factors, treatment outcomes, post-treatment recurrence, and overall survival (OS) were compared between the NAC and upfront surgery groups. Fifty-one and 55 patients were included in the NAC and upfront surgery groups, respectively. The R0 resection rate was significantly lower in the NAC group than in the upfront surgery group (upfront surgery, 98%; NAC, 76%; P = 0.003). In the upfront surgery group, of 49 patients who underwent R0 resection and pathologically positive for LN metastasis, 22 (45%) received adjuvant chemotherapy. In the NAC group, 49 (96%) of 51 patients completed two cycles of NAC. In survival analysis, no significant difference in OS was observed between the NAC and upfront surgery groups (NAC: 5-year OS, 43.8%; upfront surgery: 5-year overall surgery, 57.5%; P = 0.167). Patients who underwent R0 resection showed significantly longer OS than did those who underwent R1, R2, or no resection (P = 0.001). In multivariate analysis using age, perioperative chemotherapy, depth of invasion, LN metastasis, surgical radicality, postoperative pneumonia, and anastomotic leakage as covariates, LN metastasis [cN2: hazard ratio (HR), 1.389; P = 0.309; cN3: HR, 16.019; P = 0.012] and surgical radicality (R1: HR, 3.949; P = 0.009; R2 or no resection: HR, 2.912; P = 0.022) were shown to be significant independent prognostic factors. In clinical stage III ESCC patients, no significant difference in OS was observed between NAC and upfront surgery. Although potential patient selection bias might be a factor in this retrospective analysis, the noncurative resection rate was higher after NAC than after upfront surgery. The survival benefit of more intensive NAC needs to be further evaluated.