T w o hundred and eighteen untreated patients with non-Hodgkin's lymphomas, classified according to the scheme of Rappaport et al., were investigated for bone marrow involvement. No pattern of pretreatment laboratory abnormalities predicted which patients would have a positive bone marrow for lymphoma. Open marrow biopsy demonstrated lymphoma after needle biopsy was negative, a n d both biopsy techniques were clearly superior to bone marrow aspiration in identifying marrow involvement. Bone marrow involvement correlated with advanced stage, cellular composition of the lymphoma, and, in the nodular lymphomas, splenomegaly and constitutional symptoms. Patients with histiocytic lymphomas uncommonly had initial marrow involvement whereas patients with mixed and lymphocytic types were frequently affected. Nodular or diffuse patterns did not influence the incidence of marrow involvement, b u t patients with nodular lymphomas a n d positive marrows survived significantly longer than those with diffuse lymphomas.
INCE RAPPAPORT ET AL. PROPOSED THE S classification of all malignant lymphomas
into nodular or diffuse patterns each composed of histiocytes and/or lymphocytes,7 little clinical data have been gathered to judge the usefulness of this classification. I n particular, no information has been accumulated about the frequency of bone marrow involvement in relation to each histologic type, the clinical setting in which such involvement occurs, and its prognostic significance. Furthermore, in many centers, a bone marrow aspirate may be the only technique used to assess bone marrow during staging. This paper compares our experience with marrow aspiration as well as needle and open bone marrow biopsies in staging patients with non-Hodgkin's lymphomas and analyzes the clinical features of patients with proven marrow involvement.