Objective: We began a controlled clinical trial to assess efficacy and toxicity of surgery (S), surgery + radiotherapy (SRT), surgery + chemotherapy (SCT), and chemotherapy (CT) in the treatment of primary gastric diffuse large cell lymphoma in early stages: IE and II1.
Summary background data: Management of primary gastric lymphoma remains controversial. No controlled clinical trials have evaluated the different therapeutic schedules, and prognostic factors have not been identified in a uniform population.
Patients and methods: Five hundred eighty-nine patients were randomized to be treated with S (148 patients), SR (138 patients), SCT (153 patients), and CT (150 patients). Radiotherapy was delivered at doses of 40 Gy; chemotherapy was CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) at standard doses. International Prognostic Index (IPI) and modified IPI (MIPI) were assessed to determine outcome.
Results: Complete response rates were similar in the 4 arms. Actuarial curves at 10 years of event-free survival (EFS) were as follows: S: 28% (95% confidence interval [CI], 22% to 41%); SRT: 23% (95% CI, 16% to 29%); that were statistically significant when compared with SCT: 82% (95% CI, 73% to 89%); and CT: 92% (95% CI, 84% to 99%) (P < 0.001). Actuarial curves at 10 years showed that overall survivals (OS) were as follows: S: 54% (95% CI, 46% to 64%); SRT: 53% (95% CI, 45% to 68%); that were statistically significant to SCT: 91% (95% CI, 85% to 99%); CT: 96% (95% CI, 90% to 103%)(P < 0.001). Late toxicity was more frequent and severe in patients who undergoing surgery. IPI and MIPI were not useful in determining outcome and multivariate analysis failed to identify other prognostic factors.
Conclusion: In patients with primary gastric diffuse large cell lymphoma and aggressive histology, diffuse large cell lymphoma in early stage SCT achieved good results, but surgery was associated with some cases of lethal complications. Thus it appears that CT should be considered the treatment of choice in this patient setting. Current clinical classifications of risk are not useful in defining treatment.