Abstract PS3-01-09: Associations of genetically predicted metabolites with mammographic breast density and breast cancer risk in premenopausal women

Abstract Background: Genome-wide association studies (GWAS) have identified genetic loci associated with mammographic breast density (MBD) and breast cancer risk, but the mechanisms underlying these associations remain unclear. Metabolite quantitative trait loci (mQTL) analyses can map genetic loci associated with metabolites, potentially offering insight into mechanisms involved in risk, but there is limited data on how genetically predicted metabolites are associated with MBD and breast cancer risk in premenopausal women. To address this, we performed mQTL analyses in non-Hispanic White (NHW) and non-Hispanic Black (NHB) women and examined their overlap with established GWAS loci for MBD and breast cancer. Methods: Our analysis included 494 NHW and 163 NHB women who had screening mammograms at Washington University in St. Louis. Plasma metabolomic profiling (Metabolon®) identified 1,074 metabolites, and genotyping (Infinium microarray, Illumina) covered 2,028,571 loci. Metabolite data were batch-normalized, metabolites with 10% missingness were excluded, and remaining missing values were imputed. We conducted mQTL mapping using linear regression in PLINK 2.0, stratified by race and adjusted for age, body mass index, alcohol use, oral contraceptive use, and first five genotype principal components. Significant mQTLs (FDR-p0.05) were compared to GWAS loci for breast cancer, two fine mapping studies, and GWAS of MBD. Results: In NHW women, 30 mQTLs were associated with 54 metabolites overlapping breast cancer GWAS loci. Key pathways included xenobiotics (anti-inflammatory/immunosuppressant drugs, chemicals, tobacco metabolites), lipid (fatty acid, corticosteroid, bile acid metabolism), and amino acid (tryptophan, methionine/cysteine, tyrosine metabolism). In NHB women, 2 mQTLs were associated with 3 metabolites overlapping breast cancer GWAS loci: rs16866849 (diltiazem, ß =0.2, FDR-p=1.6×10-6, cardiovascular drug), and rs9397068 with two metabolites (2-hydroxyibuprofen, ß=5.3, FDR-p=9.6×10-3, analgesics; and 2,2'-methylenebis(6-tert-butyl-p-cresol), ß=14.6, FDR-p=9.6×10-3, chemical). Additionally, in NHW women, 5 mQTLs were associated with 5 metabolites overlapping GWAS loci of MBD. For dense area, these included rs150249911 (levulinate (4-oxovalerate), ß=6.5, FDR-p=2.3×10-7, food component/plant), rs150249911 (N-acetyltyrosine, ß=5.0, FDR-p=3.3×10-4, tyrosine metabolism), rs2042239 (5-acetylamino-6-formylamino-3-methyluracil, ß=1.3, FDR-p=3.3×10-2, xanthine metabolism), and rs833472 (3-hydroxystachydrine, ß=11.0, FDR-p=2.5×10-3, food component/plant). For percent mammographic density, rs61941038 (3-hydroxystachydrine, ß=10.4, FDR-p=7.4×10-3, food component/plant) was associated. For non-dense area, rs78395856 (taurochenodeoxycholic acid 3-sulfate, ß=3.7, FDR-p=3.3×10-2, bile acid metabolism) was associated. Conclusions: Our novel study identified genetic loci associated with metabolites that overlap with established MBD and breast cancer risk loci with differences in NHW and NHB women. These findings suggest biological mechanisms of genetic risk and provide a basis to prioritize metabolites and pathways for targeted prevention. Citation Format: G. Pourali, L. Lyu, X. Guo, A. Toriola. Associations of genetically predicted metabolites with mammographic breast density and breast cancer risk in premenopausal women abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-01-09.