Pathological complete response in patients with esophageal cancer after the trimodality approach: The association with baseline variables and survival—The University of Texas MD Anderson Cancer Center experience

BACKGROUND: Reports are limited regarding clinical and pretreatment features that might predict a pathological complete response (pathCR) after treatment in patients with esophageal cancer (EC). This might allow patient selection for different strategies. This study examines the association of a pathCR with pretreatment variables, overall survival (OS), recurrence-free survival (RFS), and patterns of recurrence in a large cohort from a single institution. METHODS: The baseline clinical features of 911 consecutive patients with EC who were treated with trimodality therapy from January 2000 to November 2013 were analyzed. A pathCR was defined as a surgical specimen with no residual carcinoma (primary or nodes). Logistic regressions were used to identify independent baseline features associated with a pathCR. We applied log-rank testing and Cox models to determine the association between a pathCR and the time-to-event outcomes (OS and RFS). RESULTS: Of 911 patients, 218 (23.9%) achieved a pathCR. The pathCR rate was 23.1% for adenocarcinoma and 32.2% for squamous cell carcinoma. A lower pathCR rate was observed for 1) older patients (>60 years), 2) patients with poorly differentiated tumors, 3) patients with signet ring cells (SRCs), and 4) patients with a higher T stage. Patients with a pathCR had longer OS and RFS than those without a pathCR (P =.0021 and P =.0011, respectively). Recurrences occurred more in non-pathCR patients. Distant metastases were the most common type of recurrence. PathCR patients developed brain metastases at a marginally higher rate than non-pathCR patients (P =.051). CONCLUSIONS: In this large cohort study, a pathCR is confirmed to be associated with better OS and RFS. The presence of a poorly differentiated tumor or SRCs reduces the likelihood of a pathCR. Future research should focus on molecular classifiers. Cancer 2017;123:4106–4113.