A large-scale study of bone marrow involvement in patients with Hodgkin's lymphoma.

This study was designed to identify variables that can predict bone marrow involvement (BMI) in Hodgkin's lymphoma (HL), and to analyze the benefit of bilateral over unilateral bone marrow trephine biopsy (BMB). From 1982 to 2000, BMB had been performed at diagnosis in 1161 patients with HL who had been followed from the institutions participating in the Piemonte Hodgkin's Disease Registry. Six hundred and sixteen patients (53%) had received bilateral BMB, and the remaining 545 patients (47%) received unilateral BMB. The relationships between BMB results and other clinical features were retrospectively studied with both univariate and multivariate analyses. Ninety-two patients (8%) showed BMI: 51 of them were staged with bilateral and 41 with unilateral BMB. Among the 92 patients with BMI, a second extranodal involvement was present in only 25 patients (27%). In multivariate analysis, the 5 independent factors that predicted for BMI were B symptoms, infradiaphragmatic involvement, mixed cellularity (MC) and lymphocyte depleted (LD) histology, involvement of > or = 4 lymphatic areas, and liver involvement. The probability of BMI according to the presence of these variables was distributed as follows: 0.3%, 2.5%, 7.6%, and 27% in patients positive for 0, 1, 2, and > or = 3 factors, respectively. Among 51 patients staged with bilateral BMB, BMI was shown in both specimens in 33 cases (65%), whereas the positivity was limited to only 1 of the 2 specimens in the remaining 18 cases (35%). A score based on 5 variables can predict the probability of BMI, and BMB could be avoided in patients with a score of 0 and a probability of BMI of < 0.5%. When BMB is needed, the superiority of bilateral over unilateral biopsy is suggested.