Analytical and Clinical Validation of an EPIC 900K v2 array-based methylation assay to predict recurrence in stage II melanoma

New York University School Of Medicine, New York NY

Investigators

Abstract

PROJECT SUMMARY In contrast to breast and colon cancers, where molecular pathology tests are routinely used to improve the accuracy of disease prognostication over routine histopathology, there are no validated molecular tests for locally advanced, node-negative melanoma (Stage II). Existing molecular biomarkers, particularly commercial tests developed to manage localized melanoma, have not robustly demonstrated clinical validity or utility as evidenced by the lack of recommendation by the National Comprehensive Cancer Network Melanoma Guidelines and other organizations. Recently, pembrolizumab, an anti-PD-1 immunotherapeutic agent, was approved as adjuvant therapy for surgically resected stages IIB and IIC melanoma based on data from a randomized, placebo- controlled clinical trial. Although recurrence-free survival (RFS) at the 18-month post-surgical follow up showed a statistically significant benefit associated with pembrolizumab, the RFS in the placebo group is fairly high. Given the potential toxicity of immunotherapy, and its variable efficacy and high cost, we need more accurate tools to appropriately select patients who have a high probability of tumor recurrence, and spare surgically cured patients the risks of overtreatment. Several publications, as well as our own preliminary data have identified strong associations between aberrantly methylated DNA loci and melanoma survival outcomes. Indeed, we identified a 13-probe signature using the EPICMethylation v1.0 array that was highly accurate in identifying melanoma recurrence and mortality among patients with stage II melanoma. The goal of the current proposal is to refine and validate (analytically and clinically) the recurrence prediction signature using a large prospective cohort of stage II melanoma patients whose tumors are analyzed with the updated MethylationEPIC v2.0 array. The UH2 phase will establish the minimum amount of tumor needed to reliably detect a melanoma signature (technical sensitivity), and will establish the precision and reproducibility of the assay. We will also examine the potential interfering effect of melanocytic nevi when they are associated with melanomas, and will conduct inter- laboratory harmonization studies prior to the UH3 studies. In the UH3 phase, we will use the analytically validated assay to: 1) analyze stage II melanoma tumors from a prospective cohort of patients with a minimum of three years of follow up data (among non-relapsed patients); 2) refine the recurrence prediction model; and 3) validate the model in a separate group of patients. The final deliverable will be a recurrence-prediction model that combines clinicopathological factors with a set of differentially methylated loci to more accurately classify patients into high- or low-risk for recurrence than a model based on clinicopathological factors alone. Since the Infinium MethylationEPIC BeadChip v2.0 array is currently in use in clinical laboratories, the results of this project could be quickly tested in other patient cohorts, incorporated into clinical utility trials, and broadly disseminated to the health care community at large.