T-cell lymphoid aggregates in bone marrow after rituximab therapy for B-cell follicular lymphoma: a marker of therapeutic efficacy?

International audienceRituximab, an anti-CD20 monoclonal antibody, is widely used in the treatment of B-cell lymphoma. Some reports have outlined histologic modifications in bone marrow specimens from patients treated with this antibody, notably the presence of CD3(+) lymphoid aggregates morphologically mimicking residual lymphoma. To gain insight into the significance of such infiltrates, serial BM trephines obtained in 39 patients with B-cell follicular lymphoma treated by rituximab and enrolled in the GOELAMS-GELA intergroup FL2000 protocol were reexamined. The 39 patients were 22 women and 17 men with a median age of 50 years (range, 29-75 years). All pretreatment bone marrow biopsies showed CD20(+) lymphomatous cells. A second biopsy was obtained between 30 and 100 days after the last rituximab injection: 19 (48%) were morphologically diagnosed as negative (no lymphoid infiltrates or only minor lymphoid aggregates) and 20 (51%) as positive because of persistent lymphoid nodules. After immunohistochemical analysis, 13 (33%) cases were reinterpreted as false-positive because of the complete absence of CD20(+) cells, with the lymphoid nodules consisting of CD3(+) and CD5(+) T cells. Most of them also expressed CD4(+), whereas only a few CD8(+) cells were present. Among these 13 false-positive cases, 12 were BCL2-IGH polymerase chain reaction-negative in the bone marrow aspirate at the time of biopsy. The 13th case turned out to be negative in the 18th-month bone marrow aspirate. In all of these cases, lymphoid aggregates had disappeared on bone marrow biopsies performed 18 months after treatment. After a mean follow-up of 4.5 years, 9 of 13 patients were in remission as compared with only 2 among the 7 patients with postrituximab persistent CD20(+) lymphomatous cells. There was no statistically significant difference between this false-positive group of patients and that with negative postrituximab bone marrow regarding sex, age, medullar involvement pattern before treatment, delay between rituximab treatment, and molecular status. Interestingly, we noted a more favorable outcome (70% versus 52% remission) for the false-positive cases, suggesting that these T-cell reactions could be the hallmark of specific antitumoral immunity after rituximab treatment and should be properly investigated