Reappraising myocardial fibrosis in severe aortic stenosis: an invasive and non-invasive study in 133 patients

Aims To investigate myocardial fibrosis (MF) in a large series of severe aortic stenosis (AS) patients using invasive biopsy and non-invasive imaging. Methods and results One hundred thirty-three patients with severe, symptomatic AS accepted for surgical aortic valve replacement underwent cardiovascular magnetic resonance (CMR) with late gadolinium enhancement (LGE) and extracellular volume fraction (ECV) quantification. Intra-operative left ventricular (LV) biopsies were performed by needle or scalpel, yielding tissue with (n = 53) and without endocardium (n=80), and compared with 10 controls. Myocardial fibrosis occurred in three patterns: (i) thickened endocardium with a fibrotic layer; (ii) microscopic scars, with a subendomyo-cardial predominance; and (iii) diffuse interstitial fibrosis. Collagen volume fraction (CVF) was elevated (P<0.001) compared with controls, and higher (P<0.001) in endocardium-containing samples with a decreasing CVF gradient from the subendocardium (P = 0.001). Late gadolinium enhancement correlated with CVF (P<0.001) but not ECV. Both LGE and ECV correlated independently (P<0.001) with N-terminal pro-brain natriuretic peptide and high-sensitivity-troponin T. High ECV was also associated with worse LV remodelling, left ventricular ejection fraction and functional capacity. Combining high ECV and LGE better identified patients with more adverse LV remodelling, blood biomarkers and histological parameters, and worse functional capacity than each parameter alone. Conclusion Myocardial fibrosis in severe AS is complex, but three main patterns exist: endocardial fibrosis, microscars (mainly in the subendomyocardium), and diffuse interstitial fibrosis. Neither histological CVF nor the CMR parameters ECV and LGE capture fibrosis in its totality. A combined, multi-parametric approach with ECV and LGE allows best stratification of AS patients according to the response of the myocardial collagen matrix.