Predictors of survival in patients with persistent nodal metastases after preoperative chemotherapy for esophageal cancer.
Academic Article
Overview
abstract
OBJECTIVE: In patients with esophageal cancer, a complete pathologic response after preoperative therapy is universally regarded as a favorable prognostic factor. However, less is known about factors predictive of outcome in patients with persistent nodal disease. The purpose of this study is to determine which variables affect survival in this patient population. METHODS: We reviewed a prospectively maintained esophageal cancer database. Patients with positive lymph nodes after preoperative therapy and surgery were selected. Predictors of survival were examined univariately using the log-rank test. Factors identified at P < .20 by univariate analysis were selected for inclusion in a multivariate model. RESULTS: Ninety-six patients with 1 or more positive nodes received preoperative therapy. Pathologic T classification was 0 to 2 in 25 (26%) patients and 3 to 4 in 71 (74%) patients. In 29 (30%) patients, nonregional nodal disease was present (M1). Final pathologic stages were IIB in 18 (19%), III in 49 (51%), and IV in 29 (30%). Postoperatively, 44 (46%) patients received additional chemotherapy. On univariate analysis, pathologic stage, pathologic T classification, and number of positive nodes significantly affected overall survival. On multivariate analysis, clinical stage (hazard ratio [HR], 2.25; P = .05), pathologic T classification (HR, 3.06; P = .006), and number of positive nodes (HR 1.03 per node, P = .09) were significant predictors of overall survival. CONCLUSION: Long-term survival can be achieved in patients with esophageal cancer who have persistent nodal disease after neoadjuvant therapy and surgical resection. Clinical stage, pathologic T classification, and number of positive nodes best predict survival. Nonregional nodal disease does not adversely affect outcome. Postoperative chemotherapy conferred no additional survival benefit in this patient population.
