Gene methylation in random FNA samples as biomarkers for breast cancer.
518 Background: Current methods to determine breast cancer risk are insufficiently sensitive to select women most likely to benefit from preventive strategies. We hypothesized that candidate gene promoter hypermethylation may provide an individualized risk profile. We performed a prospective study to determine whether DNA cumulative methylation index (CMI) varies by menstrual phase or menopausal status, and to correlate CMI with established risk factors. Methods: We obtained random fine needle aspiration (rFNA) samples from healthy women age 35-60 and determined their menopausal and menstrual status, lifetime Gail risk, mammographic breast density, and cytologic atypia assessed as the Masood score. We evaluated CMI of 11 candidate genes in rFNA cells using the Quantitative Multiplex Methylation-Specific PCR (QM-MSP) technique. We used Wilcoxon test and ANOVA model to compare CMI across menopausal and menstrual (follicular, mid-cycle, luteal) categories, respectively. We used linear regression model to adjust for age and BMI. Methylation scores were log-transformed in the analysis. Results: We enrolled 390 women at the Avon Breast Centers at Johns Hopkins and Northwestern, the majority through the Love/Avon Army of Women, and 380 completed study procedures. Median age 50 (36-60), mean BMI 28 (18.7-50.8), 52% were postmenopausal. Mean life-time Gail risk 14.6 (5.6-54.1), mean percent mammographic density 19.6 (2.5-72.8), and mean Masood score (N=354) 13.6 (7-18). QM-MSP analysis was completed on 229 samples. We did not observe differences in CMI among menopausal (P=0.4895) or menstrual categories (P=0.2333). There was no association between CMI and life-time Gail risk (P=0.706) or breast density (P=0.4116). We observed a significant correlation between CMI and Masood score (P=0.0167). Conclusions: CMI correlates with degree of cytologic atypia and is potentially a robust indicator of breast cancer risk since it does not vary with menstrual or menopausal status. Next, we will select genes that best reflect changes in the clinical parameters to create a gene methylation signature that will be validated in other studies and correlated with breast cancer risk.