Randomized trial of influenza vaccine with granulocyte-macrophage colony-stimulating factor or placebo in cancer patients.

PURPOSE: To determine whether granulocyte-macrophage colony-stimulating factor (GM-CSF) would improve response to influenza vaccination in cancer patients. PATIENTS AND METHODS: In a randomized, patient-blinded, placebo-controlled trial carried out in 1997 to 2000, 133 patients were stratified into five groups of treatment and disease. Single doses of standard split trivalent influenza vaccine and either placebo or 250 micro g of GM-CSF were administered at the same time. Hemagglutination inhibition assay titers were measured before and 4 weeks after vaccination. RESULTS: Standard analyses, which define response as at least a four-fold increase in titers, detect no effect of GM-CSF for any of the three influenza subtypes in the trivalent vaccines (P >or=.12). Analysis that includes the magnitude of the change in titers and combines responses of the subtypes suggests that the placebo group had the greater response (P =.051), thus indicating that GM-CSF does not improve response. Ancillary analyses show that response declines both with increasing age and with higher initial titers. The fraction of patients with at least a four-fold increase in titers was 0.36 (95% confidence interval, 0.29 to 0.42) CONCLUSION: A single 250- micro g dose of GM-CSF administered with the influenza vaccine does not improve response to vaccination. Response in cancer patients is low and declines as age and initial titer increase.