Rituximab for the Treatment of Adults with ITP: A Systematic Review.

Background: Rituximab, a monoclonal anti-CD20 antibody, is a promising therapeutic agent for adults with idiopathic immune thrombocytopenic purpura (ITP).

Objective: To determine the estimate of effect and to describe the safety of rituximab in adults with ITP.

Methods: Systematic review of the world literature and meta-analysis. We searched MEDLINE, EMBASE, the Cochrane Registry for Controlled trials, abstracts of the American Society of Hematology annual meetings from 1997 to 2004 and bibliographies of relevant articles. In duplicate and independently, studies were assessed for eligibility using the apriori inclusion criteria of idiopathic ITP and adults 16 years of age or older. All study designs and all languages were included. Agreement on study selection was good (k=0.87) and discrepancies were resolved by consensus in all cases. Data extraction was performed in duplicate. A random effects model was used to pool the estimates of the proportions for overall response, defined as a platelet count of 50 x10⁹/L or greater, and complete response, defined as a platelet count of 100 x10⁹/L or greater. Funnel plots were examined for evidence of publication bias.

Results: 39 studies enrolling a total of 365 patients were included. Studies were case reports, case series or single-arm cohorts; there were no comparative trials of either a randomized or non-randomized design. Small studies showing a large treatment effect were overrepresented. The median age of patients in all studies was 50 years (range 4 – 98) including one study of adults and children which did not report adult data separately. Prior to receiving rituximab, nearly all patients had failed treatment with steroids; 195/365 (53.4%) had failed splenectomy and 100/365 (27.4%) had failed treatment with danazol, immunosuppressives and/or cytotoxic agents. In most studies, rituximab was administered as a weekly infusion of 375mg/m² for 4 consecutive doses. Among 17 studies of 5 or more patients, the pooled estimate of the proportion for overall response was 61.2% (95% CI: 49.8%, 72.6%) and for complete response was 48.3% (95% CI: 36.1%, 60.4%). The median duration of response was 7 months (range 0.5 – 41) as reported in 31 of 39 studies. Infusion-related side effects occurred in 52 patients. Severe toxicities were reported in 15 patients and included bronchospasm, anaphylactoid reaction, serum sickness, interstitial pneumonitis, pulmonary embolism (PE), retinal artery thrombosis and infection. There were 8 deaths temporally related to rituximab including 3 fatal bleeds and 1 postoperative fatal PE.

Conclusions: Pooling results of the published literature demonstrates that rituximab achieves a platelet count response in 61.2% of adults with ITP, however without appropriate controls, improved outcome compared with standard therapy remains uncertain. In addition, heterogeneity of patient pre-treatment status, poor follow up data and publication bias limit the results. Significant morbidity and a high number of deaths complicated this treatment. Our results confirm the need for prospective randomized data before widespread use of this potentially toxic therapy can be recommended.