Comparative Analysis of Methylation of Genes Associated with Breast Cancer and Prostate Cancer, and the Diagnostic Potential of the Method on its Basis

Abstract

Objective: to compare the methylation rates of nine genes in prostate cancer and breast cancer and to identify a single panel of genes that is promising for its use as an additional marker in the diagnosis of these types of cancer. Material and methods. The methylation analysis of promoter areas of 9 genes ( RARβ , HIN1 , DAPK , RASSF1A , GSTP1 , CCND2 , p16 , APC and hMLH1 ) was performed by means of methylspecific polymerase chain reaction with electrophoretic detection. Results. We have identified and compared the methylation rates of the 9 genes in the biopsy material of 68 patients with prostate cancer, 39 patients with benign prostatic hyperplasia / prostatic intraepithelial neoplasia (BPH/PIN), surgical material from 104 female patients with breast cancer and 38 patients with benign tumors of the mammary gland (fibroadenoma, nodal mastopathy). Conclusion. It has been found that methylation of five genes ( RARβ , HIN1 , CCND2 , APC and GSTP1 ) out of the nine analyzed genes is statistically significantly more common in samples of malignant tumor tissue compared to that in samples with benign pathology both among patients with prostate diseases and among patients with breast pathology. The methylation rate of each of these five genes is significantly higher in the tumor tissue of patients with prostate cancer compared to those with breast cancer. Methylation of two or more genes out of five in prostate cancer was detected in 89.7 % of cases, in BPH / PIN - in 23.1 %, which is 3.9 times as little, OR - 29.1 (95 % CI [9.9-85, 6]), p < 0.001. The diagnostic sensitivity of the method was 89.7 % with specificity of 76.9 % and accuracy of 85.0 %. Methylation of one or more genes out of five in the group of patients with breast cancer was detected in 80.8 % cases, with benign tumors of the mammary gland in 13.2 % cases, which is 6.1 times as little, OR - 27.7 (95 % CI [9.6-80.0]), p < 0.001. Accordingly, the sensitivity of the method was equal to 80.8 %, specificity - 86.8 %, and accuracy - 82.4 %.