Group risk classification of non-Hodgkin's lymphoma.

Between 1980 and 1982, 162 patients with non-Hodgkin's lymphoma were treated with cyclophosphamide, adriamycin, vincristine and prednisone (CHOP) or CHOP plus bleomycin. At the moment when the patient were diagnosed all clinical characteristics were evaluated according to a multiple regression analysis model, which has the following three factors associated to bad prognosis: a quick clinical evolution (less than three months) bone marrow infiltration and high levels of lactate dehydrogenase (LDH). The Cox model of analysis also agreed that a quick clinical evolution and the high levels of LDH were bad prognosis factors. These two factors were associated with poor complete remission and short survival rates. A mathematical model was built based on the last two factors. Five groups of patients were observed with increasing risk of a poor response and a short survival rates, which allowed us to identified three prognostic groups with clear differences in both the duration of remission and survival. These groups were low, moderate and high-risk. Results analysis in this paper have important clinical implications for the design of the prospective clinical trials in patients with malignant lymphoma.