Abstract
PURPOSE: We conducted a double-blind, placebo-controlled trial to evaluate the effect of granulocyte-macrophage colony-stimulating factor (GM-CSF) and peptide vaccination (PV) on relapse-free survival (RFS) and overall survival (OS) in patients with resected high-risk melanoma. PATIENTS AND METHODS: Patients with completely resected stage IV or high-risk stage III melanoma were grouped by human leukocyte antigen (HLA) -A2 status. HLA-A2-positive patients were randomly assigned to receive GM-CSF, PV, both, or placebo; HLA-A2-negative patients, GM-CSF or placebo. Treatment lasted for 1 year or until recurrence. Efficacy analyses were conducted in the intent-to-treat population. RESULTS: A total of 815 patients were enrolled. There were no significant improvements in OS (stratified log-rank P = .528; hazard ratio, 0.94; 95% repeated CI, 0.77 to 1.15) or RFS (P = .131; hazard ratio, 0.88; 95% CI, 0.74 to 1.04) in the patients assigned to GM-CSF (n = 408) versus those assigned to placebo (n = 407). The median OS times with GM-CSF versus placebo treatments were 69.6 months (95% CI, 53.4 to 83.5 months) versus 59.3 months (95% CI, 44.4 to 77.3 months); the 5-year OS probability rates were 52.3% (95% CI, 47.3% to 57.1%) versus 49.4% (95% CI, 44.3% to 54.3%), respectively. The median RFS times with GM-CSF versus placebo were 11.4 months (95% CI, 9.4 to 14.8 months) versus 8.8 months (95% CI, 7.5 to 11.2 months); the 5-year RFS probability rates were 31.2% (95% CI, 26.7% to 35.9%) versus 27.0% (95% CI, 22.7% to 31.5%), respectively. Exploratory analyses showed a trend toward improved OS in GM-CSF-treated patients with resected visceral metastases. When survival in HLA-A2-positive patients who received PV versus placebo was compared, RFS and OS were not significantly different. Treatment-related grade 3 or greater adverse events were similar between GM-CSF and placebo groups. CONCLUSION: Neither adjuvant GM-CSF nor PV significantly improved RFS or OS in patients with high-risk resected melanoma. Exploratory analyses suggest that GM-CSF may be beneficial in patients with resected visceral metastases; this observation requires prospective validation.