STUDY PURPOSE: To assess the relative efficacy of granulocyte-colony stimulating factor (G-CSF) administered as a primary prophylaxis to patients with cancer receiving myelosuppressive chemotherapy
TYPE OF STUDY: Meta-analysis and systematic review
DATABASES USED: PubMed, EMBASE, Science Citation Index, Cochrane Database of Systematic Reviews, Cochrane central register of controlled clinical trials, Database of Abstracts of Reviews of Effects, Health Technology Assessment Database, and the NHS Economic Evaluation Database plus manual searches of original publications (The search included publications from January 1990 to September 2013.)
KEYWORDS: Febrile neutropenia, filgrastim, G-CSF, lipegfilgrastim, meta-analysis, and pegfilgrastim
INCLUSION CRITERIA: Randomized, controlled trials that compared primary prophylaxis (PP) with filgrastim, pegfilgrastim, lenograstim, or lipegfilgrastim with a placebo, no G-CSF PP, or PP with a different G-CSF in adult patients receiving myelosuppressive chemotherapy for solid tumors or non-Hodgkin lymphoma
EXCLUSION CRITERIA: If patients had received G-CSF for established febrile neutropenia (FN), or different doses of the same G-CSF in each treatment arm, and if patients had leukemia or multiple myeloma, or bone marrow or peripheral-blood stem cell transplantation; studies also excluded if they were economic analyses, evaluated investigational or unapproved drugs, or were published in languages other than English
TOTAL REFERENCES RETRIEVED: 4,790
EVALUATION METHOD AND COMMENTS ON LITERATURE USED: Two independent reviewers evaluated publications identified in the search and compared them to the search criteria for relevance. Disagreements were resolved by consensus (this process was not detailed). The data of interest were then extracted – these included protocol designs and patient, disease, and treatment characteristics. Statistic analyses included using an OR for febrile neutropenia incidence. For treatment effects, direct and indirect comparisons were made using mixed treatment comparisons (MTC) and ORs, and pairwise meta-analyses were calculated for conventional random effects. In addition, Bayesian statistics were applied for treatment effects as a posterior distribution. Finally, a full metaregression analysis and treatment effect with adjustment for relative dose intensity (RDI) was intended to be conducted; however, there weren’t enough data by group because of too much heterogeneity between studies for the RDI adjustment.