An evaluation of prognostic factors and tumor staging of resected carcinoma of the esophagus

Objective: To evaluate prognostic factors and tumor staging in patients after esophagectomy for cancer, Summary Background Data: Several reports have questioned the appropriateness of the sixth edition of the International Union Against Cancer (UICC) TNM guidelines for staging esophageal cancer. Additional pathologic characteristics, besides the 3 basic facets of anatomic spread (tumor, node, metastases), might also have prognostic value. Methods: All patients who underwent resection of the esophagus for carcinoma between January 1995 and March 2003 were extracted from a prospective database. Univariate and multivariate analysis was performed to identify prognostic factors for survival. The goodness of fit and accuracy of 3 staging models (UICC-TNM, Korst classification, Rice classification) predicting survival were assessed. Results: A total of 292 patients (mean age, 63 years) underwent esophagectomy. The 5-year overall survival rate was 29% (median, 21 months). pT-, pN-, pM-stage, and radicality of the resection were independent prognostic factors. Subdivision of T1 tumors into mucosal and submucosal showed significant differences in 5-year survival between both groups: 90% versus 47%, respectively (P = 0.01). Subdivision of pN-stage into 3 groups based on the number of positive nodes (0, 1-2, and > 3 nodes positive) or the lymph node ratio (0, 0.01-0.2, and > 0.2) also refined staging (P = 0.001 and P < 0.001, respectively). The current subclassification of M1 (M1a and M1b) is not warranted (P = 0.41). The staging model of Rice was more accurate than the UICC-TNM classification in predicting survival. Conclusion: This study supports the view that the current (6th edition) UICC-TNM staging model for esophageal cancer needs to be revised.