FDG PET in evaluation of bone marrow involvement in patients with lymphoma – Correlation with subtypes and bone marrow biopsy patterns

1682 Objectives: Detection of bone marrow (BM) involvement in lymphoma is important for staging. BM biopsy suffers from the problems of sampling error and low sensitivity. BM involvement also varies according to histologic subtypes. The aim of this study was to determine the role of PET imaging in the evaluation of BM involvement in patients with lymphoma and correlate the results with the histologic subtypes. Methods: PET images and BM biopsy findings of 102 patients with newly diagnosed lymphoma at initial staging were reviewed. BM involvement with PET was considered to be present if PET showed focal or diffuse FDG uptake prior to therapy and the findings resolved after therapy. BM biopsy findings were analyzed with respect to cellularity and patterns of involvement. The two diagnostic modalities (PET vs. biopsy) were also compared according histologic subtypes. The distribution of various subtypes was Hodgkin’s disease ( HD, n=23), diffuse B cell lymphoma ( DLBCL, n=38), follicular lymphoma ( FL, n=22), Marginal Zone lymphoma ( MZL, n=8), mantle cell lymphoma ( n=7), T cell ( n=4). Results: PET showed active uptake in the marrow in 27 of 102 patients (26 %). Twenty of 27 patients (75%) had focal and 7 (25 %) showed diffuse uptake. Biopsy showed marrow involvement in 21 patients (20 %). Including both PET and biopsy finding, marrow involvement was found in 39 patients, almost a two fold increase in the number of cases detected. Of these 9 had marrow involvement detected by both PET and biopsy, 12 had marrow involvement only on biopsy and 18 patients (46%) had marrow involvement detected only by PET. The 18 patients with + PET but negative biopsy were predominantly DLBCL (n=10) and HD (n=6). The predominant histology among the 12 patients who had marrow involvement only on biopsy but not on PET was FL (n=7), followed by marginal zone lymphoma (n=3). In 10 of these 12 (83%) patients the involvement by lymphoma was less than 10 % and predominantly in the paratrabecular region of the marrow. Conclusions: These findings suggest that PET imaging has a higher detection rate of marrow involvement in lymphoma than biopsy. PET was superior to biopsy in detecting BM involvement in Patients with DLBCL and HD. Biopsy was more useful in patients with FL and MZL. PET may be negative when the involvement is paratrabecular and of a lower degree.