Molecular Imaging to Identify Predictors of Aortic Valve Disease Activity

Structural imaging modalities, such as echocardiography (echo) and computed tomography (CT), measure the structural consequences of aortic valvular calcification and have proven invaluable for grading AS severity. However, these techniques cannot capture the dynamic microcalcification processes that drive the underlying disease and have proven insufficient for predicting AS progression. In contrast, molecular imaging with 18F-NaF positron emission tomography/computed tomography (18F-NaF PET/CT) is the gold-standard method for noninvasively measuring the central pathophysiological process driving the progression of AS. This imaging approach uses 18F-sodium fluoride, a tracer that binds preferentially to hydroxyapatite in regions of developing microcalcification. This enables quantification of microcalcification activity (which precedes visible macrocalcification) and provides insights into the underlying disease activity that drives AS progression. Notably, optimized protocols for 18F-NaF PET/CT yield a reproducible biomarker to track disease activity and predict progression. Furthermore, recent advances in PET technology (such as total-body PET scanners) have improved sensitivity by 10- to 40-fold, further enhancing the quantification of the microcalcification rate underlying AS progression.

Overall, the assessment of microcalcification activity shifts the focus from static anatomical assessments to measurement of the dynamic pathological process driving AS progression. However, while 18F-NaF PET/CT is unsuitable for routine clinical use (because it is costly and currently limited to research applications), it can be effectively leveraged to train predictive models and deploy them as clinical calculators to estimate the rate of microcalcification from readily available biomarker data. The resulting clinical models and calculators could greatly enhance clinical care by identifying patients at risk of rapid progression, refining risk stratification and prognostic accuracy, and opening avenues for personalized monitoring and treatment in calcific AS.

The overarching goal of this Clinical Imaging Study is to develop novel predictive models and clinical calculators that capture AS microcalcification activity and forecast the risk of progressive aortic valve stenosis among participants with mild-to-moderate AS. The investigators will achieve this by performing advanced multimodality imaging, including 18F-NaF PET/CT, to quantify microcalcification activity in deeply phenotyped patients using clinical, imaging, and blood plasma biomarkers. Computational approaches, including AI tools, will identify optimal combinations of imaging and non-imaging predictors of microcalcification activity. The investigators will then develop accessible clinical calculators and test the hypothesis that these models and calculators enhance identification of rapid progressors through cross-validation and external validation in the Population Project cohort (which undergoes identical biomarker phenotyping).