Purpose: Chemoradiotherapy (CRT) is an alternative to surgery for resectable locally advanced esophageal carcinoma (RLA-EC). We investigated the heterogeneity of the treatment benefits across subgroups of patients, defined according to the radiation scheme.
Patients and methods: Between February 1993 and December 2000, 451 patients were enrolled. The following two schemes were allowed: protracted radiotherapy (P-RT), which scheduled 46 Gy over 4.5 weeks or split-course radiotherapy (SC-RT) with two 1-week courses of 15 Gy. Two courses of cisplatin and fluorouracil were delivered concomitantly. In case of exclusive CRT, a further course of 20 Gy over 2 weeks in the P-RT group and one 1-week course of 15 Gy in the SC-RT group were delivered with three courses of chemotherapy. SC-RT and P-RT were administered to 285 patients (64%) and 161 patients (36%), respectively.
Results: For P-RT versus SC-RT, the response rate to induction CRT was 67% v 68%, respectively (P = .09), and 2-year local relapse-free survival rate was 76.7% v 56.8%, respectively (P = .002). Shorter tumor length and P-RT were associated with better local control in multivariate analysis (P = .002 for both). After a median follow-up time of 47.4 months, 2-year overall survival rate was 37.1% for P-RT compared with 30.5% for SC-RT (P = .25). Independent prognostic factors on survival were tumor diameter (P = .02), weight loss of 10% or less (P = .05), and response to induction CRT (P = .002).
Conclusion: Patients with RLA-EC treated with P-RT had better local control than patients treated with SC-RT. Response to induction CRT is a determinant prognostic factor on survival.