Novel Characterization of Sex Specific Biologic Signatures in Valvular Heart Disease

Overview

This project aims to validate sex-specific biologic signatures associated with aortic valve disease developed in a large multicenter CMR registry, using unsupervised phenomapping. The aim to use standard and advanced CMR techniques (MRF, DTI, chemical exchange transfer, and radiomics analysis) is to determine advanced CMR predictors of reverse remodeling following aortic valve surgery and develop sex-specific thresholds for risk. Infrastructure developed by this study will enable development of an innovative, scalable, sex-specific precision medicine cardiovascular imaging pipeline to determine overall risk and treatment response.

Chronic valvular heart disease leads to significant left ventricular (LV) remodeling. Current national guidelines for surgical/procedural referral for valvular heart disease do not consider sex differences in presence of symptoms, LV remodeling and dysfunction. However, our prior research in patients with chronic aortic regurgitation, validated by our recent multicenter study, demonstrated that despite higher left ventricular function, and less ventricular dilation, females experienced more heart failure symptoms, fewer referrals for surgical intervention, and higher prevalence of adverse outcomes compared to males. Similarly, published HVTI data have demonstrated increased adverse outcomes in women referred for surgical mitral valve intervention. Cardiac magnetic resonance (CMR) provides an exciting opportunity to characterize sex differences in LV remodeling.In combination with conventional CMR measures, novel CMR techniques such as Magnetic Resonance Fingerprinting (MRF), Diffusion Tensor imaging (DTI) and radiomics analysis provide tissue level specificity with potential to enhance phenomapping. Limitations in understanding sex-specific remodeling patterns stem from heterogeneity of presentation, which confound traditional analytic methods. Phenomapping, a method of machine learning, clusters imaging features and patients into distinct phenotypic groups. Unsupervised phenomapping enables unbiased grouping of patients by both clinical characteristics as well as complex imaging features. In recent studies, this unbiased phenomapping approach demonstrates superior risk stratification of cardiac disease compared to traditional approaches that can be used to guide individualized treatment The aim to use advanced CMR techniques (MRF, DTI, chemical exchange transfer, and radiomics analysis) is to determine advanced CMR predictors of reverse remodeling following procedural valve intervention and develop sex-specific thresholds for risk. Results from this study would enable the development of sex-specific precision medicine pathway, augmented by advanced imaging features, to better predict overall risk and treatment response, and thus enable novel patient selection criteria. Study hypothesis: Radiomics, MRF, chemical exchange transfer, and DTI will elucidate distinct sex-specific biologic signatures, in addition to standard CMR imaging features, and are associated with adverse outcomes, and reverse remodeling following surgical/procedural valve intervention.