Supervised clustering of peripheral immune cells associated with clinical response to checkpoint inhibitor therapy in patients with advanced melanoma.

Background and purpose: Immune therapy with checkpoint inhibitors (CPIs) is a highly successful therapy in many cancers including metastatic melanoma. Still, many patients do not respond well to therapy and there are no blood-borne biomarkers available to assess the clinical outcome. Materials and methods: To investigate cellular changes after CPI therapy, we carried out flow cytometry-based immune monitoring in a cohort of 90 metastatic melanoma patients before and after CPI therapy using the FlowSOM algorithm. To evaluate associations to the clinical outcome with therapy, we divided the patients based on progression-free survival. Results: We found significant associations with CPI therapy in both peripheral blood mononuclear cell and T-cell subsets, but with the most pronounced effects in the latter. Particularly CD4+ effector memory T-cell subsets were associated with response with a positive correlation between CD27+HLA-DR+CD4+ effector memory T cells in a univariate (odds ratio: 1.07 [95% confidence interval 1.02-1.12]) and multivariate regression model (odds ratio: 1.08 [95% confidence interval 1.03-1.14]). We also found a trend towards stronger accumulation of CD57+CD8+ T cells in non-responding patients. Conclusion: Our results show significant associations between immune monitoring and clinical outcome of therapy that could be evaluated as biomarkers in a clinical setting.