Fibroblast activation protein in aortic stenosis: integrating in vivo imaging with a bioinformatics approach

Abstract Introduction Aortic stenosis has no medical therapy to attenuate disease progression. The underlying pathogenesis is unknown with a lack of delineation of responsible cellular subtypes and pathways driving disease initiation and progression. High-resolution transcriptomic profiling of explanted tissue can enable the identification of disease driving pathways and novel therapeutic targets. Positron emission tomography (PET) may be able to label and to track such pathways in vivo. Methods Bulk- and single nucleus RNA-sequencing was performed on diseased aortic valves from deceased patients with a spectrum of aortic stenosis severity and normal aortic valves from patients undergoing transplant for ischaemic cardiomyopathy. In a linked prospective observational cohort, patients with aortic stenosis and matched controls underwent PET and cardiac computed tomography (CT) at baseline and ≥1 year later. Results Bulk RNA expression (control, n=12; mild/moderate aortic stenosis, n=9; severe aortic stenosis, n=18) revealed a disease-specific gene signature for valvular interstitial cells. A specific cluster of these cells expressing chrondo- and osteo-genic profiles were associated with disease severity, with specific upregulation of fibroblast activation protein (FAP) (Figure 1). Eighty-six patients (11 with aortic sclerosis, and 25 each with mild, moderate or severe aortic stenosis; 72±11 years, 32% female) and 9 control subjects (71±8 years, 33% female) underwent PET with the novel positron emitting radiotracer gallium-68 fibroblast activation protein inhibitor (68GaFAPI, which binds specifically to FAP. 68GaFAPI uptake was seen in all patients with aortic sclerosis or stenosis with no uptake seen in the control population. 68GaFAPI uptake increased with disease severity (ANOVA p0.001) and correlated with peak aortic jet velocity on echocardiography (r=0.532, p0.001) and CT calcium score (r=0.577, p0.001) (Figure 1). After a median of 16 months, baseline 68GaFAPI uptake was associated with an increase in peak aortic jet velocity (r=0.397, p=0.01) and the CT calcium score (r=0.473, p0.001). Explanted aortic valves were collected from patients in the imaging study who underwent aortic valve replacement. In vivo 68GaFAPI uptake co-localised with FAP on immunohistochemistry, seen around developing but not established calcific nodules. 68GaFAPI high regions were associated with increased proportions of FAP mRNA by single cell RNA sequencing compared with 68GaFAPI low regions within the same patient (Figure 2). Conclusions In this integrated translational and clinical imaging study, we have identified a disease driving cell population in aortic stenosis which can be labelled and tracked over time using 68GaFAPI. FAP is an important disease biomarker in the search for a medical therapy for this condition, with the potential for 68GaFAPI to assess response to novel therapies and identify patients most likely to benefit from them.