Clinicopathological features of primary gastric lymphoma.

Background and Objectives: Surgery has been the mainstay of the treatment of primary gastric lymphoma, but the value of surgical treatment needs reevaluation. Methods: Thirty-two patients with primary non-Hodgkin B-cell lymphoma of the stomach were examined retrospectively to evaluate prognostic factors and their impact on survival. All patients had undergone abdominal exploration for radical surgery between 1979 and 1992. The prognostic factors in view of survival after treatment were determined with both univariate and multivariate analyses. Results: The resectability rate was 66% (21/32) and radical resections had been performed on 53% (17/32). The overall median survival was 65 months and the overall 5-year survival was 56%. The 5-year survival rates related to a modified Ann Arbor classification as follows: I 1E, 86%; I 2E, 100%; II 1E, 44%; II 2E, 37%; IIIE, 20%; and IVE, 0%. Univariate analysis using Kaplan-Meier estimates showed that radical surgery, Ann Arbor stage, patient's age, and lymph node involvement were significant prognostic factors. According to Cox proportional regression analysis, only Ann Arbor stage, radical surgery, and age were significant independent variables. Conclusions: According to our experience, surgery is still needed for the treatment of primary gastric lymphomas, but the benefits of primary chemotherapy or adjuvant chemotherapy using cytotoxic drugs must be determined in large prospective controlled trials. (C) 1999 Wiley-Liss, Inc.