Developing CVD prediction models for women with early stage breast cancer - ABSTRACT In a Renewal Application of the Pathways Heart Study (R01CA214057), we extend our research by developing, testing, and validating two cardiovascular disease (CVD) risk prediction models, leveraging available clinical and “precision cardio-oncology” data in stage I-III breast cancer (BC) patients. BC survivors are at high risk of developing and dying from CVD following BC diagnosis compared to women without a BC history due to cardiotoxic BC treatments and shared risk factors between BC and CVD. Clinical risk prediction tools are needed to identify BC survivors at high risk of serious, life-threatening CVD so that appropriate surveillance and management can prevent and mitigate sequelae of CVD. Risk prediction tools are commonly utilized in clinical breast oncology and cardiology, but to date, no established CVD risk prediction tools for BC patients are available that account for CVD risk factors and BC treatment history. Recently, genomic and blood-based biomarker data have been shown to substantially increase the predictive ability of risk prediction models that rely solely on clinically available data. The Pathways Heart Study (R01CA214057) is an established cohort of 14,942 women diagnosed with stage I-IV BC at Kaiser Permanente Northern California (KPNC) from 2005-2013 and followed through 2021, which expanded upon the original Pathways Study cohort of 4,504 BC patients (U01CA195565). In this Renewal Application, we enrich the cohort by adding 25,684+ new stage I-III BC patients and 5 years of follow-up on existing participants, creating a cohort of 40,626+ women diagnosed with stage I-III BC from 2005-2024 and followed through 2026. The expanded cohort adds contemporary BC treatment exposure data and increases the power to build prediction models with clinical utility. We leverage existing genetic data from the Pathways Study and cardiometabolic biomarker data from the Pathways Breast White Adipose Tissue Inflammation study (R01CA241409), collect new data on high- sensitivity cardiac troponin T (hs-cTnT), extract new CVD risk factors from KPNC clinical records, and utilize two external validation cohorts. Specific Aims: Aim 1) To develop and evaluate a prediction model to quantify risk of serious CVD events (heart failure/cardiomyopathy, ischemic heart disease, stroke, clinically significant arrhythmia, and CVD death) in women with a diagnosis of stage I-III BC using readily available KPNC clinical data; Aim 2) To develop a “precision cardio-oncology” risk prediction model for serious CVD using data from women with stage I-III BC in the KPNC Pathways Study with available genotype and biomarker data collected near the time of diagnosis; and Aim 3) To validate and compare the performance of models developed in Aims 1 and 2 using two external cohorts. If successful, our models will accurately identify early-stage BC patients at high risk of developing serious life-threatening CVD who need increased surveillance and management. As warranted, the next steps in our research will be to develop and test an online tool for clinical use, and to continue to develop and validate the models as new BC treatments become available.