Neoadjuvant vs definitive concurrent chemoradiotherapy in locally advanced esophageal squamous cell carcinoma patients

• Main Authors: Chih-Yi Chen, Chia-Chin Li, Chun-Ru Chien
• Format: Article
• Language: English
• Published: BMC 2018-07-01
• Series: World Journal of Surgical Oncology
• Subjects: Esophageal squamous cell carcinoma; Concurrent chemoradiotherapy; Esophagectomy
• Online Access: http://link.springer.com/article/10.1186/s12957-018-1444-0

Abstract Background The optimal treatment for locally advanced esophageal squamous cell carcinoma remains unclear. We compared the clinical outcomes of neoadjuvant concurrent chemoradiotherapy (CCRT) followed by esophagectomy [the surgery group] and CCRT without surgery [the CCRT group] in patients with squamous cell carcinoma from an Asian population. Methods Eligible patients diagnosed from 2008 to 2015 were identified through the Taiwan Cancer Registry. To balance observable potential confounders, we constructed a 1:1 propensity score-matched cohort [surgery vs CCRT]. We compared the hazard ratios between the surgery and CCRT groups for death using a robust variance estimator. We also evaluated the outcomes of patients for freedom from local regional recurrence (FFLRR) and esophageal cancer-specific survival (ECSS). Extensive supplementary analyses were performed to examine the robustness of our findings. Results Our study population included 298 patients balanced with respect to the observed covariables. The hazard ratio of death was 0.56 [95% confidence interval 0.42~0.75] when surgery was compared to CCRT. The results remained significant in the FFLRR and ECSS outcomes. In the supplementary analyses, our results also remained significant when additional covariables were taken into consideration or when the definition of the index date was changed. Conclusions When compared to definitive CCRT, neoadjuvant CCRT followed by esophagectomy was associated with improved overall survival for locally advanced esophageal squamous cell carcinoma. However, given the nonrandomized nature of the study and the sensitivity to potentially unmeasured confounders, our results should be interpreted cautiously.