Résumé : Purpose: The currently used prognostic models for patients with nonmetastatic clear cell renal cell carcinoma (ccRCC) are based on clinicopathologic features and might be improved by adding molecular markers. Epigenetic alterations occur frequently in ccRCC and are promising biomarkers. The aim of this study is to identify prognostic promoter methylation markers for ccRCC.Experimental Design: We integrated data generated by massive parallel sequencing of methyl-binding domain enriched DNA and microarray-based RNA expression profiling of 5-aza-2'-deoxycytidine-treated ccRCC cell lines to comprehensively characterize the ccRCC methylome. A selection of the identified methylation markers was evaluated in two independent series of primary ccRCC (n = 150 and n = 185) by methylation-specific PCR. Kaplan-Meier curves and log-rank tests were used to estimate cause-specific survival. HRs and corresponding 95% confidence intervals (CI) were assessed using Cox proportional hazard models. To assess the predictive capacity and fit of models combining several methylation markers, HarrellC statistic and the Akaike Information Criterion were used.Results: We identified four methylation markers, that is, GREM1, NEURL, LAD1, and NEFH, that individually predicted prognosis of patients with ccRCC. The four markers combined were associated with poorer survival in two independent patient series (HR, 3.64; 95% CI, 1.02-13.00 and HR, 7.54; 95% CI, 2.68-21.19). These findings were confirmed in a third series of ccRCC cases from The Cancer Genome Atlas (HR, 3.60; 95% CI, 2.02-6.40).Conclusions: A four-gene promoter methylation marker panel consisting of GREM1, NEURL, LAD1, and NEFH predicts outcome of patients with ccRCC and might be used to improve current prognostic models. Clin Cancer Res; 23(8); 2006-18. ©2016 AACR.