Purpose: Randomized controlled trials (RCTs) of prophylactic granulocyte colony-stimulating factors (G-CSF) have demonstrated a significant reduction in febrile neutropenia (FN) after systemic chemotherapy. Several RCTs have been published recently that investigate the impact of G-CSF on mortality and relative dose-intensity (RDI).

Methods: A comprehensive systematic review and meta-analysis of all reported RCTs comparing primary prophylactic G-CSF with placebo or untreated controls in adult solid tumor and malignant lymphoma patients was undertaken without language restrictions, using electronic databases, conference proceedings, and hand-searching techniques. Two reviewers extracted data independently. Summary estimates of relative risk (RR) with 95% CIs were estimated based on the method of Mantel-Haenszel and DerSimonian and Laird.

Results: Seventeen RCTs were identified including 3,493 patients. For infection-related mortality, RR reduction with G-CSF compared with controls was 45% (RR = 0.55; 95% CI, 0.33 to 0.90; P = .018); for early mortality (all-cause mortality during chemotherapy period), it was 40% (RR = 0.60; 95% CI, 0.43 to 0.83; P = .002); and for FN, it was 46% (RR = 0.54; 95% CI, 0.43 to 0.67; P < .001). Average RDI was significantly higher in patients who received G-CSF compared with control patients (P < .001). Bone or musculoskeletal pain was reported in 10.4% of controls and 19.6% of G-CSF patients (RR = 4.03; 95% CI, 2.15 to 7.52; P < .001). Significant reductions in FN with G-CSF were observed in studies allowing secondary G-CSF prophylaxis in controls and in the three trials with concurrent prophylactic antibiotics in both treatment arms.

Conclusion: Prophylactic G-CSF reduces the risk of FN and early deaths, including infection-related mortality, while increasing RDI and musculoskeletal pain. There are insufficient data to assess the impact of G-CSF on disease-free and overall survival.