Primary tumor classification according to methylation pattern is prognostic in patients with early stage ER-negative breast cancer

Abstract

Breast cancer patients with similar clinical stage may experience different disease outcomes. Aberrant DNA methylation of primary breast tumors can have impact on the clinical outcome. This study aimed to assess clinical utility of tumor-specific methylated sequences (MINT17, 31) and tumor-related gene (RARβ2) methylation classification in primary breast tumors. Absolute quantitative assessment of methylated alleles (AQAMA) was used to determine the methylation index (MI) of MINT17, MINT31, and RARß2 in 242 primary tumors of early stage breast cancer patients. Patients were classified into three methylation groups: meth-N, with normal methylation levels of all biomarkers; meth-L, with one biomarker hypermethylation; and meth-H, with hypermethylation of >1 biomarker. Disease outcome of methylation groups was compared during follow-up. MI of all biomarkers was successfully obtained in 237 tumors of which 79 (33%) were classified as meth-N, 86 (36%) as meth-L, and 72 (30%) as meth-H. Meth-H status was a risk factor for distant recurrence (DR) (log-rank P=0.007) and shorter disease-free survival (DFS) (log-rank P=0.039). Methylation classification had strongest prognostic value for patients with ER-negative tumors. In multivariate analysis (n=222), ER-negative meth-H patients had a 4.1-fold increased risk of DR (95% CI 1.80-9.59; meth-N HR 1.0, P=0.001), a 4.2-fold increased risk of overall recurrence (OR) (95% CI 1.88-9.47; meth-N HR 1.0, P=0.001), and a 3.1-fold shorter DFS (95% CI 1.57-5.98; meth-N HR 1.0, P=0.003). Methylation classification of primary breast cancer is an independent prognostic factor for disease outcome in patients with ER-negative tumors. The study's findings will have to be confirmed in an independent dataset.