Rituximab for Patients with Chronic Lymphocytic Leukemia: A Systematic Review

Chronic lymphocytic leukemia (CLL) accounts for 25% of all leukemias and remains incurable with conventional chemotherapy (CX). Rituximab (R) may be an effective treatment option for CLL patients with the potential to improve overall survival (OS) when given in combination with CX but there is also a risk of more side effects such as infections. This review aims to summarize the evidence for this new treatment option by evaluating the effects on OS, progression-free survival (PFS) and side effects.

MEDLINE and CENTRAL were systematically searched for randomized controlled trials up to April 2010. Trials of patients with CLL comparing CX including R and treatment with CX alone (CX identical in both groups) were included. Clinical trials with previously untreated and pre-treated patients were explored in the main meta-analyses. Trial selection, quality assessment and data extraction were done independently by two review authors. Dichotomous data were analyzed as relative effect measures (i.e. relative risk, RR) with 95% confidence intervals (CI). Time-to-event outcomes were analyzed with hazard ratios (HR) and 95% CI in a random effects model.

A total of 992 records were screened. Two eligible trials with 921 untreated (GCLLSG CLL 8 trial (CLL8) and CALGB 9712 trial (CALGB 9712)) and two eligible trials with 604 pre-treated patients (REACH trial (REACH) and NCRI CLL 201 trial) that were fitting the inclusion criteria were identified. The NCLRI CLL 201 trial provided response data only. CALGB 9721 did not report HRs or P-values on OS or PFS and the survival curves for PFS and OS were of low quality, so the provided data were not included in the meta-analysis. Both, CLL8 and REACH examined patients receiving fludarabine (F) and cyclophosphamide (C) with or without R and were meta-analyzed with regard to PFS and OS. PFS (1342 patients) was significantly longer for FCR (HR 0.65, 95% CI [0.48, 0.88]). Analysis of OS (1368 patients) also showed a significantly longer survival for FCR (HR 0.73 (95% CI [0.58, 0.93]). CLL8 provided HRs for the different disease stages and showed significantly improved OS after FCR for Binet B patients only [Binet A: HR 0,19, 95% CI [0.23, 1.613]; Binet B: HR 0.45, 95% CI [0.295, 0.689]; Binet C: HR 1.4, 95% CI [0.843, 2.620]). REACH had not been significant regarding OS (HR 0.83, 95% CI [0.59, 1.17]). With regard to severe hematologic toxicity, meta-analysis of CLL8, REACH and CALGB 9721 showed a significantly higher risk of neutropenia (RR 1.46, CI 95% [1.03, 2.08]) for regimens including R, but there was no statistical difference for thrombocytopenia (RR 1.06, CI 95% [0.60, 1.87]), anemia (RR 0.89 [0.63, 1.26]) or the incidence of severe infections (RR 1.08 CI 95% [0.86, 1.35]). REACH reported a higher rate of secondary malignancies in the FCR-arm (FCR (7%), FC (5%)).

This systematic review demonstrates significantly longer OS for CLL patients that received FCR compared to FC (HR 0.65, 95% CI [0.48, 0.88]) and confirms better PFS for patients receiving FCR (HR 0.73 (95% CI [0.58, 0.93]). Adverse events (particularly neutropenia) occurred more often when patients were treated with CX plus R but did not result in an increased infection rate. However, data of CLL8 were only available from the abstract and need to be subsequently confirmed.

Hallek:Roche: Consultancy, Honoraria, Research Funding.