Progressive Multifocal Leukoencephalopathy Associated with Rituximab Treatment of B-Cell Lymphoproliferative Disorders: A Report of 29 Cases from the RADAR Project.

Background- Progressive multifocal leukoencephalopathy (PML) is a demyelinating disorder of the brain caused by the reactivation of latent JC polyoma virus. Rituximab (Rituxan, Mabthera) is a B-cell depleting, monoclonal antibody which has been linked to reactivation of the hepatitis B virus. HIV infection, purine analog therapy, hematopoietic transplant and lymphoma have previously been associated with PML. We evaluated the characteristics of patients with B-cell lymphoproliferative disorders who developed PML after exposure to rituximab, and the quality of case reports available in the medical literature and at the FDA.

Methods- Data sources included 1 observation at Northwestern Memorial Hospital, 14 reports obtained from the FDA MedWatch database, 9 case reports from the medical literature, and 21 reports from the manufacturer, Genentech, that were submitted to the FDA. Reports that did not confirm the diagnosis of PML, were associated with HIV infection, or rituximab use for a rheumatologic indication were excluded.

Results- Of 47 unique case reports screened, 3 were excluded for rituximab use in rheumatologic disease, 2 for pre-existing HIV, and 13 for inadequate confirmation of the PML diagnosis. In the 29 remaining cases, survival of PML was reported in only one patient. Diagnosis was made by brain biopsy (n=6), autopsy (n=6), or MRI of brain AND positive JC virus by PCR (n=17). Five cases were seen in hematopoietic transplant recipients, 4 autologous and 1 allogeneic. Of the 24 patients who were not transplanted, 10 had purine analog exposure. Thirteen of the 14 patients that did not receive transplant or a purine analog received an alkylating agent, with 7 receiving standard R-CHOP. The median age of the 29 patients was 63.5 years (range 32–89), 15 were female and 14 male. Indications for rituximab were: large B-cell lymphoma (n=6), follicular lymphoma (n=6), CLL (n=5), Waldenstrom (n=2) and other non-Hodgkin lymphomas (n=10). Median time to diagnosis of PML from first and last rituximab doses was 10 and 5 months respectively. The median number of rituximab doses was 6. Quality comparison of source data is contained in table 1. Overall completeness ratio for literature and observed reports vs. FDA and manufacturer reports was 1.48.

Conclusions: We have found 29 cases of rituximab-associated PML, of which 14 were not associated with transplantation or purine analog exposure, suggesting an association of rituximab therapy independent of these other treatment related risk factors. FDA and manufacturer reports are inferior to those available in published case reports or through active surveillance. Further examination of the relationship of rituximab to PML, using an active surveillance strategy, is warranted.