Objectives: This study analyzed survival with respect to lymph node involvement to develop a new staging system for patients with esophageal cancer that accurately reflects prognosis.
Methods: The records of patients undergoing resection of primary esophageal cancer from 1989 to 1993 were reviewed. The data collected included patient age and sex, tumor histologic characteristics and location, the use of preoperative or postoperative radiation and chemotherapy, the type of resection, the depth of tumor invasion, the number and location of benign and malignant lymph nodes in the resected specimen, the disease status at last follow-up, and the first site of relapse. With an anatomically specific lymph node map, tumors designated in the current American Joint Committee on Cancer system as M1 because of extensive lymph node metastases were reclassified as N2, reserving the M1 category for visceral metastases. Survival was analyzed by the Kaplan-Meier method, and prognostic factors were assessed by log-rank and Cox regression analyses.
Results: There were 216 patients (159 men, 57 women) with a median age of 63.5 years. Adenocarcinoma of the distal esophagus or gastroesophageal junction was the most common tumor (127 patients, 59%) and Ivor Lewis esophagogastrectomy was the most frequently performed operation. Both lymph node location (N1 versus N2) and number (0 vs 1 to 3 vs 4 or more) significantly influenced survival.
Conclusions: A new staging system that adds an N2 M0 descriptor and reclassifies stage groupings reflects prognosis more accurately than does the current American Joint Committee on Cancer staging system. The number of positive lymph nodes is also an important stratification factor.