Prediction of outcome in "resectable" esophageal carcinoma.

A retrospective review was performed of 51 patients with esophageal carcinoma, deemed "resectable" by preoperative workup (e.g., CT scan, barium swallow), who presented to Hahnemann University Hospital between 1980 and 1991. This represented 21.8% of the total number of patients (234) with esophageal cancer who presented during that time period. At exploration, only 21 of the 51 patients (41%, or 9% overall) were truly resectable; 59% had more extensive disease than was appreciated preoperatively and that precluded resection for cure. Of the 21 patients resected for cure, 24% were alive at two years and only 5% were alive at 3 years. Neither age, gender, tumor type nor location in the esophagus significantly affected overall survival. Furthermore, none of these parameters, taken as independent variables, were able to predict true resectability at the time of operation. We conclude that preoperative assessment of resectability, even in those patients who appear to be good candidates for cure, remains imprecise at best. Given an operative mortality rate of 6-8% (in most series) and an overall 3- to 5-year survival rate of less than 10% (even in patients thought to have had curative resections), we reinforce the fact that meticulous patient selection and multimodality management strategies remain the keys to making any impact on this disease.