Multi-omics & AI-based imaging to identify biomarkers of molecular targets for breast cancer prevention - High mammographic breast density (MBD) presents two challenges: (1) it is associated with a 2-to 6-fold increase in breast cancer (BrCa) risk, and (2) it significantly decreases mammogram sensitivity. Since most women have high MBD, identifying modifiable factors contributing to MBD is critical for reducing BrCa risk, improving early detection, and increasing survival from screen-detected breast cancers. However, factors leading to high MBD remain poorly understood. We aim to identify modifiable factors influencing MBD that can be targeted for novel therapeutics to prevent BrCa. Gut bacteria, which can be modified by diet and oral agents, may be one such factor. Gut bacteria and their metabolites impact numerous processes linked to BrCa and MBD, including estrogen production, inflammation, and immune responses. Emerging studies suggest an association between the gut microbiome (GM) and BrCa and between the GM and host factors related to BrCa and MBD. However, the GM-MBD relationship remains unclear. Our preliminary data suggest that gut bacteria metabolites are associated with MBD and that there are significant differences in the GM between women with high vs. low MBD. However, the specific bacteria responsible for our observations are unknown. Moreover, no studies have examined the joint roles of the gut microbiome and systemic metabolome in MBD. Our study addresses this gap. We hypothesize that gut bacteria and their metabolites predict MBD. We propose identifying specific gut bacteria species, metabolites, and metabolic pathways associated with MBD to provide insight into ways to reduce MBD, decrease BrCa risk, increase mammogram sensitivity, and ultimately improve BrCa survival. We will use novel, automated AI-based image analysis software to quantify and classify MBD in a cohort of women with no history of cancer. We will use 16S rRNA gene amplification sequencing and metagenomic whole-genome shotgun sequencing to quantify species-level taxa in fecal samples. Linear Discriminant Analysis Effect Size (LEfSe) will identify taxa abundance differences between low vs high MBD groups (Aim 1). Metabolites will be quantified in banked serum samples using mass spectrometry and compared between MBD groups (Aim 2). Metabolites and metabolic enrichment analyses will also help further identify biologically active and meaningful bacteria species associated with MBD. We will identify gut bacteria predictive of MBD, facilitating the development of personalized BrCa surveillance and prevention care. We will also identify gut bacteria that can serve as intervention targets to reduce MBD. Our team will then be well- positioned to conduct human trials to assess how altering gut bacteria composition impacts MBD. We will also be well-positioned to pursue functional studies illuminating the mechanisms by which gut bacteria influence MBD, thus further identifying prevention targets. Responding to PAR-22-216, ours is a hypothesis-driven, correlative translational exploratory R21 study grounded in lab-based research that will use high-dimensional data and AI to identify (1) biomarkers predictive of MBD, and (2) potential targets and agents to reduce MBD.
