Determinants of the efficacy of prophylactic granulocyte transfusions: a meta-analysis

Abstract

Background: Cytokines that improve granulocyte collection yields have recently become available, and may lead to a new series of trials of granulocyte transfusion (GTX) therapy. We conducted a meta-analysis of studies of prophylactic GTX in order to identify the determinants of efficacy of this intervention, and to assist in the design of future trials of GTX for the treatment of patients with overwhelming infection.

Methods: Randomized controlled trials of the efficacy of prophylactic GTX published in English in 1970-1995 were retrieved, and eight studies were eligible for analysis. Summary relative odds (RR) of bacterial or fungal infection, death, or death from infection in transfused patients vs. controls were computed for patient subsets defined on the basis of dose of granulocytes transfused, assessment of leukocyte compatibility, duration of neutropenia, and infection rate of controls. The random-effects method was used for all analyses.

Results: Assessment of leukocyte compatibility prior to the transfusion, dose of granulocytes transfused, and duration of neutropenia in enrolled patients could account, respectively, for the variation in findings across published reports in terms of all three, two, and one of the outcome measures studied. Transfusion of adequate doses of compatible leukocytes significantly reduced the relative risk (RR) of infection, death, and death from infection in transfused patients vs. controls (RR = 0.075, RR = 0.224, and RR = 0.168, respectively; P < 0.05).

Conclusion: Two necessary elements in the design of future trials of therapeutic GTX should be the transfusion of high doses of granulocytes and the provision of leukocytes that are crossmatch-compatible with the recipient's serum.