Genetic determinants of mammographic density as a risk factor for breast cancer

Abstract

Background: Mammographic density (MD) for age and BMI is a strong risk factor (up to a 6 fold increase across extreme quantiles) for breast cancer. More than 60% of MD variation is estimated to be due to heritable (genetic) factors. Taking two different approaches, this work aimed to determine some of these genetic factors. The recently identified common genetic variants associated with small gradients in breast cancer risk and candidate common genetic variants identified using a genome wide association study (GWAS) of MD extremes were tested for an association with MD.

Methods: Germline DNA extracted from peripheral blood samples from 497 monozygotic (MZ) and 330 dizygotic (DZ) twin pairs, and 634 of their sisters from 903 families were genotyped for 22 independent variants (12 from associations with breast cancer and 10 from GWAS of MD extremes). Mammographic dense area, percent dense area and non-dense area were measured by three observers using a computer thresholding technique. Associations with MD measures adjusted for age, BMI and other determinants were estimated: (a) cross-sectionally using a multivariate normal model for pedigree analysis (P-values reported by Px), and (b) between-sibships and (c) within-sibships using orthogonal transformations of outcomes and exposures. A combined test of association (P-values reported by Pc) was derived using the independent estimates from (b) and (c). The distributions of P-values across variants were tested for a deviation from the uniform distribution (P-values reported by Pu).

Results: For the breast cancer associated common genetic variants tested, for dense area and percent dense area, the distributions of Pc-values deviated from the uniform distribution (both Pu<0.007), providing strong evidence that at least one genetic variant is associated with these MD measures. Consistent with their breast cancer associations, rs3817198 (LSP1) and rs13281615 (8q) were associated with dense area and percent dense area (all Px and Pc<0.05), and rs889312 (MAP3K1), rs2107425 (H19) and rs17468277 (CASP8) were marginally associated with dense area (some Px or Pc <0.05). For the candidate genetic variants from the GWAS of MD extremes the distributions of Pc-values deviated from the uniform distribution for dense area and percent dense area (Pu=0.07 and 0.009). One variant, rs10827227 (NRP1) showed strong evidence for an association with both dense area and percent dense area (Pc<0.009). For both approach, all associations were independent of menopausal status.

Conclusion: At least two common breast cancer susceptibility variants and one common variant identified through a GWAS for MD extremes were associated with MD measures that predict breast cancer. Together these variants explain about 1% of the variation in MD.