Outcomes for Esophageal Squamous Cell Carcinoma Treated with Curative Intent in a Western Cohort: Should Multimodal Therapy Be the Gold Standard?

Abstract

The standard of care for treatment of oesophageal squamous cell carcinoma (SCC) continues to evolve. Neoadjuvant chemoradiotherapy (neoCRT) provides a significant survival benefit compared to surgery alone but it is unclear whether definitive chemoradiation (dCRT) is superior. Retrospective analysis of outcomes from patients treated in a national high-volume centre (2000-2014) where both neoCRT and dCRT are used with curative intent. Propensity score match analysis was used to match patients undergoing dCRT with those undergoing surgery ± neoCRT. A total of 668 patients were treated for SCC in this time period, 361 (54.0%) of whom were treated with curative intent. In patients treated with curative intent, 179 (49.6%) had dCRT, and of these 32 (18%) did not complete the treatment regimen. One hundred and seven patients (29.6%) underwent surgery only, and 75 patients (20.8%) had multimodal therapy. The proportion of patients treated with curative intent increased over this time period. The five-year disease-specific and overall survival rate of patients treated with multimodal therapy was 62 and 50%, respectively, compared with 25 and 20% for patients the dCRT group and 44 and 38%, respectively, for the surgery only cohort (p < 0.001). Patients with a complete pathological response had a 90% five-year disease-specific survival and 76% overall survival rate. Multimodal treatment rather than dCRT was a significant predictor of overall survival (OR 1.7 95% CI 1.3-2.4, p = 0.002). In 106 patients matched, those undergoing dCRT had a significantly poorer overall survival versus those receiving surgery as a component of their care (20.47 ± 3.74 months versus 30.65 ± 10.07 months, p = 0.002). This study provides evidence, consistent with CROSS data, that multimodal therapy for SCC can provide excellent outcomes with respect to overall survival, pathologic complete response rates, R0 resections and treatment-related mortality. A large RCT with specific arms for multimodal, dCRT and surgery alone is required.