Integrating Molecular Pathology, Radiology, and Genetics to Improve Breast Cancer Risk Prediction
Abstract
Each year, over a million women in the U.S. undergo breast biopsies (i.e., tissue sampling) to determine whether a suspicious lesion found on a mammogram or on physical examination represents breast cancer (BC). Although 75% of breast biopsies show benign breast disease (BBD), which is reassuring, numerous studies have established that women diagnosed with BBD experience a 1.5- to 4-fold increased risk of developing BC in the future compared with women not diagnosed with BBD. Classification of BBD by pathologists as non-proliferative, proliferative without atypia or atypical hyperplasia identifies groups of patients at progressively increased BC risk; however, pathology alone is insufficient to provide accurate risk information for individual women. We propose to leverage our recent discovery of genetic, radiologic, and molecular risk factors for BC to develop improved approaches for individual risk prediction, enabling most women to receive reassurances of relative safety, while identifying high-risk women who might benefit from intensive screening or therapy that would prevent their cancers entirely or detect them at an early, curable stage. Further, our unique study will compare the molecular pathology of biopsies to the radiological lesions from which they were derived, thereby improving our ability to assess future risks posed by specific mammographic findings (e.g., microcalcifications) of radiologically similar appearing lesions that are not biopsied to rule out ductal carcinoma in situ or invasive BC but may represent markers of BC risk throughout the breasts. Many factors impact BC risk, but few are strong individually; thus, a multi-modality approach that combines factors with additive impact is needed to achieve a breakthrough in BC risk prediction. We have identified several candidate markers that address this requirement. First, research has identified variation in genes, referred to as single nucleotide polymorphisms (SNPs), which individually are associated with small BC risks but which when combined into a polygenic risk score (PRS) are highly predictive. We analyzed a PRS developed, tested, and validated among over 200,000 women, in a subset of five studies, demonstrating that women with BBD and the highest 10% of PRSs experienced an 11-fold increase in BC risk compared with women without BBD in the lowest PRSs. Importantly, risks associated with PRS and BBD were additive and similar across studies. Also, we have demonstrated that mammographic density and certain diagnoses of BBD contribute additively to BC risk prediction, and more recently we and others have determined that women whose mammograms show multiple microcalcifications are at increased risk. In our data, BBD severity and microcalcifications are independent BC risk factors. Further, several novel histopathologic features that are not routinely assessed microscopically by pathologists and molecular markers in breast tissues are also linked to BC risk. While multi-modal risk prediction approaches have been proposed, we aim to integrate them into a single clinical approach to achieve a breakthrough and link risk to specific carcinogenic mechanisms (hormones, inflammation, growth factors, etc.) that may point to prevention strategies for future testing in clinical trials. To achieve our major goals within the 4-year life cycle of this grant, we will identify women diagnosed with BBD who progressed to BC and women matched on age and follow-up time who remained cancer-free at Mayo Clinic, Rochester, and University of North Carolina, Chapel Hill. We will collect germline DNA for genetic testing, radiological images for analysis of mammographic density and additional features such as artificial intelligence (AI) and perform microscopic, AI, and molecular analyses of biopsies. We will define features from these approaches that contribute additively to predict BC risk and re-evaluate the genetic and radiological markers among screened women at Mayo wit
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 04, 2024
- Source ID
- HT94252310235
Entities
People
- Melissa Troester
Organizations
- United States Army
- University of North Carolina at Chapel Hill