Endoscopic biopsy after chemoradiation therapy (CRT) for esophageal cancer has been used to determine response to treatment. We wanted to determine if endoscopic biopsy can accurately establish evidence of local pathologic complete response (pCR) in patients undergoing CRT.

We queried a prospectively maintained database for patients seen at Memorial Sloan-Kettering Cancer Center from 1996 to the present who underwent, (1) CRT for local-regionally advanced esophageal cancer, (2) post-CRT endoscopic biopsy, and (3) esophagectomy. Data points included pathology of post-CRT endoscopy and surgical specimens, tumor histology, and survival. Correlations were analyzed by the chi2 test and one-way analysis of variance. Survival comparisons were assessed using the Kaplan-Meier method and log-rank analysis.

One hundred fifty-six patients were identified. Over 80% of patients received cisplatin-based chemotherapy and 5040 cGy of radiation. One hundred eighteen patients had no tumor identified on endoscopic biopsy. A negative biopsy at endoscopy was a poor predictor of pCR (negative predictive value: 31%), with 69% having local disease at esophagectomy. A positive biopsy was predictive of residual disease (positive predictive value: 95%). Negative endoscopic biopsy better predicted a pCR for squamous cell carcinomas versus adenocarcinomas (P[r] < 0.001). Nodal status of surgical specimens was not correlated with post-treatment endoscopic findings. Survival was equivalent after surgery in patients with a negative endoscopic biopsy versus patients with positive pathology.

A negative endoscopic biopsy is not a useful predictor of a pCR after CRT, final nodal status, or overall survival.