? DESCRIPTION (provided by applicant): The goal of this proposal is to establish a novel image-based breast cancer risk factor based on functional behavior of breast fibroglandular tissue that can improve risk stratification of women with mammographically dense breasts. Over 40% of screening-eligible women have mammographically dense breasts, which refers to a high amount of fibroglandular tissue relative to fat on a mammogram. These women are at elevated risk of developing breast cancer, based on a large body of evidence showing density as one of the strongest risk factors, and are increasingly encouraged to discuss their density-related risk with providers. Yet, the discriminatory accuracy of breast density is weak, limiting its usefulness in clinical risk prediction. We postulate that mammographic density does have importance in risk assessment, with the critical caveat that not all dense tissue confers equivalent risk. Because mammography creates an anatomic depiction of the breast, it is inherently unable to distinguish the known functional heterogeneity of breast fibroglandular tissue among women that likely has important implications in regards to breast cancer risk. We propose that functional behavior of fibroglandular tissue or level of background parenchymal uptake (BPU) that can be assessed on molecular breast imaging (MBI) exams allows for a non-invasive means for determining the subset of women with dense tissue who are at greatest risk of developing breast cancer. In Aim 1, we will establish a retrospective cohort study of 3300 women who had screening MBI performed at Mayo Clinic Rochester between the years 2004 - 2015. MBI exams, mammograms, and medical history information will be obtained; assessments of BPU and mammographic density from existing imaging will be performed; and incident breast cancers in the cohort will be identified via the Mayo Clinic tumor registry and mailed follow-up. In Aim 2, we will use this well-characterized cohort to examine the association of BPU (assessed both categorically and quantitatively) and breast cancer risk. Cox proportional-hazards regression will be employed in a multivariable model with adjustments for covariates, including mammographic density. An exploratory analysis of the interaction between BPU and mammographic density will also be performed using a Gradient Boosting Machine learning algorithm. We hypothesize that high BPU will be associated with increased cancer risk relative to low BPU, beyond the association observed between density and risk. We will also explore whether BPU stratifies the risk associated with dense breasts. This work is the first evaluation of BPU on MBI as a breast cancer risk factor and the first investigation into whether functional behavior of fibroglandular tissue can discriminate risk of mammographically dense tissue. For increasing numbers of women who are now undergoing MBI for supplemental breast screening, this additional risk information will be valuable in guiding clinical decisions regarding tailored screening and risk-reduction options.