Granulocyte colony-stimulating factors for febrile neutropenia prophylaxis: systematic review and mixed method treatment comparison

Background This study assesses the efficacy of three granulocyte colony-stimulating factors (G-CSFs; pegfilgrastim, filgrastim and lenograstim) in preventing febrile neutropenia (FN).

Methods A systematic review was undertaken. Head-to-head studies were combined using direct meta-analyses. In addition, an indirect Bayesian mixed treatment comparison (MTC) was undertaken to facilitate comparison between G-CSFs where there were no direct trials, and to allow data from all trials to be synthesised into a coherent set of results.

Results The review identified the following studies comparing G-CSF prophylaxis to no primary G-CSF prophylaxis: 5 studies of pegfilgrastim, 9 studies of filgrastim and 5 studies of lenograstim. In addition, 5 studies were identified comparing pegfilgrastim to filgrastim. The two synthesis methods (meta-analysis and MTC) demonstrated that all three G-CSFs significantly reduced FN rate. Pegfilgrastim reduced FN rate to a greater extent than filgrastim (significantly in the head-to-head meta-analysis and in the MTC of all studies, and not quite significantly when the MTC was restricted to RCTs only). In the absence of direct trials, the MTC gave an 80-86% probability that pegfilgrastim is superior to lenograstim in preventing FN, and a 71-72% probability that lenograstim is superior to filgrastim.

Conclusions Prophylaxis with G-CSFs significantly reduces FN rate. A head-to-head meta-analysis shows pegfilgrastim to be significantly superior to filgrastim in preventing FN events, while an MTC demonstrates that pegfilgrastim is likely to be superior to lenograstim.