Implementing Integrated Genomic Risk Assessments for Breast Cancer: Lessons Learned from the eMERGE Study

ObjectiveTo develop and implement a pipeline for integrated breast cancer risk assessment using the BOADICEA model within the eMERGE study, incorporating polygenic risk scores (PRS), monogenic variants, family history, and clinical factors. Materials and MethodsA pipeline was deployed across ten eMERGE clinical sites, integrating data from REDCap surveys, PRS reports, monogenic reports, and pedigrees via CanRisk Application Programming Interface (API). The process included design, customization, technical implementation, testing, and refinement. ResultsThe pipeline successfully generated integrated risk scores for >10,000 females. Of these, 3.6% were classified as high-risk ([≥]25% lifetime risk), and 0.9% harbored rare pathogenic variants in BRCA1, BRCA2, PALB2, or PTEN. High PRS only scores were identified in 5.6% of participants. Among those with high PRS, 34% also had high-risk based on integrated scores. API and User Interface (UI) results were highly concordant, with an average difference of 0.13%. DiscussionKey challenges included integrating diverse data sources, handling missing data, and standardizing pedigree formats. Risk classification discrepancies highlighted the need for refined communication strategies. ConclusionThis study demonstrates the feasibility of PRS-integrated breast cancer risk assessment in clinical settings but underscores challenges in data integration and risk communication. Future work should enhance recalibration for diverse populations and streamline workflows for risk interpretation and update.