The assessment of aortic stenosis: looking beyond the valve

This editorial refers to ‘Reappraising myocardial fibrosis in severe aortic stenosis: an invasive and non-invasive study in 133 patients’†, by T.A. Treibel et al., on page 699. The high prevalence of aortic stenosis (AS) and the significant mortality that is associated with the symptomatic stage have stimulated research in this field that led to a rapid evolution of diagnostic and therapeutic modalities.1 While the new therapeutic techniques, in particular percutaneous interventions, rapidly made their way into daily practice and have remarkable success, the development of diagnostic modalities translates slowly into clinical routine. Current ESC guidelines on the management of valvular heart disease focus on the echocardiographic [transthoracic echocardiogram (TTE)] assessment of the valve regarding mean pressure gradient (MPG), aortic valve area (AVA), and calcification, and name ejection fraction and hypertrophy as only extravalvular imaging finding.2 Even though this may be sufficient for the majority of patients with AS, in particular for symptomatic patients, the management of asymptomatic patients, and patients with discrepant findings or confounding diseases remains controversial.3 Hitherto, many advanced imaging techniques have been used to characterize AS but are not yet incorporated into guideline recommendations. Recent studies have shown that small alterations of the diagnostic algorithms lead to non-negligible changes in the classification of AS severity. For example, up to 52% of the patient population with severe AS according to TTE-based AVA, normal transvalvular flow, and discordant MPG were reclassified as true moderate AS only by measuring the left ventricular outflow tract (LVOT) dimension with computed tomography (CT) instead of TTE and the evaluation of a fused AVA.4 Likewise, correcting AVA for post-stenotic pressure recovery that occurs downstream of the aortic valve, using the energy loss index (ELI) as a simple TTE-based approximation, led to a reclassification of 28% of patients from severe to moderate AS in a study by Stähli et al.5 The assessment of ELI has also proven to provide independent and additional prognostic information to that derived from conventional measures of AS severity.6 In addition to the TTE-based assessment of energy loss, the CMR-based quantification of turbulent kinetic energy has shown that bicuspid aortic valves and dilated ascending aortic geometries are associated with higher energy losses that are not assessable by current TTE measures.7 On the other side of the valve, the myocardial texture, assessed by cardiac magnetic resonance (CMR) has been the subject of intensive research in patients with AS. Manifest myocardial fibrosis as determined by late gadolinium enhancement (LGE) has been shown to hold prognostic information in asymptomatic patients with AS8 and to be an independent predictor for post-operative clinical outcome in patients undergoing aortic valve replacement.9 While LGE-CMR undoubtedly provides important clinical information, it also has some limitations. First of all, the evaluation of LGE requires normal reference myocardium and provides predominantly qualitative information, which is difficult to quantify, in particular in non-ischaemic heart disease. In diseases with more diffuse than focal myocardial fibrosis, such as AS, an accurate assessment of collagen content is hardly possible with LGE.10 Those limitations can be overcome with newer CMR techniques for tissue characterization, such as T1 mapping. Since the first description of a modified look locker inversion recovery (MOLLI) sequence for myocardial T1 mapping by Messrogli et al.11 in 2004, the method has been rapidly evolving and is currently on the verge of moving from clinical research to clinical practice. It allows generation of maps of the native myocardial relaxation times and, when used in combination with post-contrast agent measurements, the extent of extracellular volume (ECV) can be quantified and displayed. At that time, myocardial tissue characterization may be the fastest flourishing field in CMR research, and elevated native T1 times as well as expansion of ECV have been found in many commonly encountered diseases including myocardial infarction, dilated cardiomyopathy, myocarditis, or systemic diseases. T1 mapping techniques have also proven to identify ischaemic and infarcted myocardium under adenosine stress12 and to distinguish hypertrophic cardiomyopathy (HCM) from athletes’ hearts,13 which indicates potential for direct clinical utility. In patients with AS, previous studies have shown that ECV mapping can reveal myocardial involvement by elevated ECV values in both degenerative and congenital AS.14 , 15 These trials already taught us that elevated ECV values are on the one hand strongly correlated with histological collagen volume fraction (CVF) and on the other hand associated with functional parameters such as echocardiographic indexes of diastolic dysfunction. In the trial by Treibel and colleagues, presented in this issue of the journal, 133 patients with severe AS were investigated using biopsy, CMR, TTE, and clinical/laboratory testing, providing the largest combined histology–multimodality imaging study in AS.16 In addition, 10 myocardial samples from autopsies of subjects who died of non-cardiovascular causes were studied. The imaging protocol included TTE for the assessment of valve area/velocities and diastolic function, and CMR for investigation of structure, function, and tissue characterization, with LGE for focal fibrosis and ECV for diffuse fibrosis. Importantly, the authors found that neither histological CVF nor the CMR parameters ECV and LGE capture the AS-induced fibrosis in its totality. The manuscript describes three patterns of fibrosis in severe AS: endocardial fibrosis, microscars, and diffuse interstitial fibrosis. While biopsies best determined the transmural gradient of fibrosis and microscars, LGE was significantly correlated with CVF independently of coronary artery disease or other confounders. ECV, in contrast, was the only imaging marker of fibrosis that correlated with the functional limitations of the patients, while not being correlated with CVF as reported in previous studies. To account for the different aspects of fibrosis, captured with each method, the authors propose a combination of LGE and ECV, which allows for an improved phenotyping of AS patients according to their myocardial response to AS. Unexpectedly, the authors found that AS severity did not associate with CVF, endocardial thickness, LGE, ECV, NT-proBNP levels, or the degree of LV remodelling, which may indicate that these parameters independently describe different effects of AS on the cardiovascular system. The authors should be commended for their effort in conducting the largest histologically validated multimodal imaging study in AS so far. With this new perspective on myocardial fibrosis and its invasive and in particular non-invasive investigation in AS patients added to the literature that has grown during the last decade, is CMR ready to change daily clinical practice for AS patients? Unfortunately not (yet). This manuscript is the latest one in a long line of studies that showed impressive ways to assess AS with CMR, indicating potential for improved risk stratification and therapeutic decision-making. However, although the diagnostic armamentarium to characterize almost every aspect of AS by CMR is larger than ever before, hardly any of those techniques are used in clinical routine to guide the treatment of AS patients. While most of the advanced methods are built on a sound theoretical base, have successfully passed feasibility studies, and showed exciting results in first clinical trials, the evidence, needed to include these methods in the guidelines, is still missing. The tremendous gap between the information that we would be able to assess and that we actually use to stratify AS patients (Take home figure) in view of the ongoing controversies and inconsistencies in AS classification is an obligation to investigate further which of the innovative CMR methods can really enhance treatment and outcome for AS patients. To achieve this goal, it needs the effort to set up a large longitudinal multicentre trial that combines all advanced CMR methods from myocardial tissue characterization and strain analysis to flow investigation in order to identify the parameters that are best suited to complement the assessment of AS severity. Schematic representation of the multiple valvular and extravalvular structures that are affected by aortic stenosis and accessible with up-to-date imaging technology. The imaging parameters that are currently named in the ESC guidelines on the management of valvular heart disease to influence the indications for aortic valve replacement are listed on the left. The right side displays a selection of cardiac magnetic resonance (CMR)-based parameters that have been shown to provide additional information in patients with aortic stenosis. Conflict of interest: none declared.