Immune Correlates of GM-CSF and Melanoma Peptide Vaccination in a Randomized Trial for the Adjuvant Therapy of Resected High-Risk Melanoma (E4697).

E4697 was a multicenter intergroup randomized placebo-controlled phase III trial of adjuvant GM-CSF and/or a multiepitope melanoma peptide vaccine for patients with completely resected, high-risk stage III/IV melanoma. A total of 815 patients were enrolled from December 1999 to October 2006 into this six-arm study. GM-CSF was chosen to promote the numbers and functions of dendritic cells (DC). The melanoma antigen peptide vaccine (Tyrosinase, gp100, MART-1) in montanide was designed to promote melanoma-specific CD8 T-cell responses. Although the overall RFS and OS were not significantly improved with the vaccine or GM-CSF when compared with placebo, immunomodulatory effects were observed in peripheral blood and served as important correlates to this therapeutic study. Peripheral blood was examined to evaluate the impact of GM-CSF and/or the peptide vaccine on peripheral blood immunity and to investigate potential predictive or prognostic biomarkers. A total of 11.3% of unvaccinated patients and 27.1% of vaccinated patients developed peptide-specific CD8 T-cell responses. HLA-A2 patients who had any peptide-specific CD8 T-cell response at day +43 tended to have poorer OS in univariate analysis. Patients receiving GM-CSF had significant reduction in percentages of circulating myeloid dendritic cells (mDC) and plasmacytoid DC (pDC) at day +43. In a subset of patients who received GM-CSF, circulating myeloid-derived suppressor cells (MDSC), and anti-GM-CSF-neutralizing antibodies (Nabs) were also modulated. The majority of patients developed anti-GM-CSF Nabs, which correlated with improved RFS and OS. The assessment of cellular and humoral responses identified counterintuitive immune system changes correlating with clinical outcome. .