Correlation of DNA methylation signatures with response to immune checkpoint inhibitors in metastatic melanoma.

Abstract

9561 Background: Immune checkpoint inhibitors (ICIs) have revolutionized the treatment of metastatic melanoma with objective response rates of 40-58%. However, reliable biomarkers to predict treatment response are lacking. Tumor tissue methylation profiles were recently proposed to have predictive value in various solid organ tumor entities. Methods: Patients with metastatic melanoma, American Joint Committee on Cancer (AJCC; 8th Edition) stage IV, receiving first-line ICI-based therapy, were retrospectively identified and formalin-fixed paraffin-embedded tumor tissue samples (FFPE) prior to ICI therapy were retrieved. We analyzed DNA methylation profiles of > 850.000 CpG sites in tumor specimens by Infinium MethylationEPIC microarrays. DNA methylation profiles were then correlated with radiological response (iRECIST). Results: 71 patients with metastatic melanoma (44 (62.0%) male, 27 (38.0%) female) were investigated; median progression free survival (PFS) was 8.5 months (range: 0 – 104.1 months) and median overall survival (OS) was 30.6 months (range: 0 – 104.1 months), respectively. 29 (40.8%) patients achieved an objective response to ICIs. Microarray analyses revealed a methylation signature including mainly hypomethylation, corresponding to the response to ICIs. Based on the 500 mostly differentially methylated genes, a total of 3 clusters were identified with the majority of responders being in cluster 2 (12/12) and 3 (12/22). The predictive performance of the methylation signature was high with 80% sensitivity, 81% specificity, and an AUC of 0.853. Conclusions: Our findings suggest that tumor DNA methylation profiling may be useful to predict response to ICIs in patients with metastatic melanoma.