Impact of patient selection, disease progression, and adverse events on esophageal cancer outcomes after trimodality therapy.

Abstract

Neoadjuvant chemoradiation followed by surgery (NeoCRT) has been advocated as standard therapy for resectable esophageal cancer. Our objective was to compare oncologic outcomes between NeoCRT and upfront surgical resection (SURG). We conducted a single-institution, retrospective review of all potentially resectable esophageal cancer patients treated with NeoCRT or SURG. From 2003 to 2010, 151 patients had NeoCRT (n = 48; 31.8%) or SURG (n = 103; 68.1%). Histology was mostly adenocarcinoma (77.5%) or squamous carcinoma (19.2%). Mean radiation dose was 44 ± 0.1 Gy, and 80.8% received platinum-based doublet chemotherapy. There were more women in the SURG group (23.3% vs 4.2%; p < 0.01) and more cardiovascular comorbidity in the NeoCRT group (39.6% vs 21.4%; p = 0.027). There was no difference in age, histology, R0 resection rate, and treatment-related mortality (NeoCRT = 4.2%; SURG = 3.9%; p = 0.15). Failure to undergo resection after NeoCRT (n = 11; 22.9%) was mainly due to disease progression (n = 6) or treatment-related mortality (n = 4). Resection could not be performed in 4 SURG patients (3.9%; p < 0.001; unresectable = 2; occult metastases = 2). NeoCRT did not improve median survival (NeoCRT = 29 ± 6; SURG = 26 ± 3 months; p = 0.376) or recurrence-free interval (NeoCRT = 25.8 ± 5; SURG = 19.4 ± 2 months; p = 0.19). Complete pathologic response (n = 8; 21.6%) was not associated with improved survival. If we exclude from analysis NeoCRT patients who did not undergo surgery, survival was significantly improved after NeoCRT (NeoCRT = 41 ± 15; SURG = 24 ± 8 months; p = 0.0082). Patient selection and early assessment of treatment response may be key factors in identifying the best candidates for trimodality therapy.