Is diastolic dysfunction a red flag sign in moderate aortic stenosis?

With rapidly ageing population, aortic stenosis (AS) has become one of the most common valvular heart diseases. Severe AS is clearly associated with increased mortality, and the current American College of Cardiology/American Heart Association guidelines designate a class I or IIa indication for aortic valve surgery in severe patients with AS with symptoms or with low left ventricular (LV) function. Historically, based on the lower risk of sudden death in moderate AS compared with severe AS, watchful observation was recommended to patients with moderate AS. In the last decade, prognostic findings on moderate AS have been gathered (table 1). A large echocardiographic national database study, including 3315 moderate AS with various LV systolic functions, suggested a poor survival rate (5year mortality: 56%). Recently, two academic institutional databases, including 1245 moderate AS, also showed poor prognosis during followup (median followup: 4.3 years; mortality: 45.3%). From a physiological viewpoint, narrowing of the aortic valve area leads to LV systolic/diastolic dysfunction due to LV pressure overloading. In patients with severe AS who particularly underwent transcatheter aortic valve replacement (TAVR), diastolic dysfunction (DD) has been described as an early marker of myocardial damage and an important prognostic information. Thaden et al reported an association between echocardiographic data and outcomes from a retrospective singlecentre study. Over a mean followup period of 7.3 years, increased left atrial pressure based on the American Society of Echocardiography/ European Association of Cardiovascular Imaging criteria remained an independent predictor of mortality after successful aortic valve replacement (AVR) (HR: 1.45; 95% CI 1.16 to 1.81). More recently, Ong et al clearly showed an association between DD grading at baseline and combined cardiovascular death/rehospitalisation at 1 year from the Placement of Aortic Transcatheter Valves (PARTNER) 2 and SAPIEN 3 registry analysed by core laboratories (n=1253). These previous studies described a consistent message of an increased risk of event with worsening grade and stage of DD in severe AS. What about the association of moderate AS with DD? In their Heart article, Stassen et al presented data from a large registry of patients with moderate AS and preserved LV systolic function from three academic institutions between October 2001 and December 2019. Moderate AS was defined as an aortic valve area of between 1.0 cm and 1.5 cm. Finally, 1247 patients with moderate AS were retrospectively included and divided into three groups (normal diastolic function, indeterminate diastolic function and DD) based on the guidelines. The aims of the study were (1) to evaluate the prevalence of DD and (2) to examine the prognostic implications of DD in moderate AS with preserved LV systolic function. The primary outcome was allcause mortality. The secondary outcome was a composite of allcause mortality and surgical or transcatheter AVR. The results showed DD was frequently present in patients with moderate AS (normal diastolic function: 32%; indeterminate diastolic function: 25%; DD: 43%). Over a median followup of 53 months, 39% of the patients died. The survival rate was 91% at 1 year and 65% at 5 years. The presence of DD was independently associated with allcause mortality (HR: 1.37; 95% CI 1.09 to 1.73) and the composite endpoint of allcause mortality and AVR (HR: 1.24; 95% CI 1.04 to 1.49) after adjustment for several clinical variables (figure 1). The strengths of the study are the large sample size and the welldesigned data collection. The main limitations of the study include the retrospective nature of the analysis and the lack of a validation cohort. Moreover, a major limitation in the assessment of DD is that there are many indeterminate cases using the guideline algorithm. Patients with indeterminate DD were not associated with higher eventfree survival compared with patients with normal diastolic function. Moreover, individual variables for assessment of LV diastolic function (E/e’, left atrial volume index and tricuspid regurgitation velocity) were not as highly associated with the outcomes. Because there are some limitations in the guideline algorithm, we need supplementary methods to determine the presence or absence of DD. After the introduction of strain imaging, it became clear that patients with normal left ventricular ejection fraction may have mildly reduced LV systolic function by global longitudinal strain. Therefore, LV strain imaging represents a supplementary test and is useful when echocardiographic indices of diastolic function are inconclusive. Tastet et al showed that the new staging system characterised the extraaortic valve cardiac damage using several parameters, including global longitudinal strain, in AS. Strain imaging may contribute to the additional value to the DD grading system in moderate AS with preserved LV systolic function. The lack of cardiac magnetic resonance data was another limitation because of not excluding cardiac amyloidosis in this cohort. Around 30% of the patients have diuretics and around 40% were of New York Heart Association class II–IV. Around half of this cohort seem to be patients with moderate AS with high prevalence of symptoms. Atrial fibrillation (AF) was relatively prevalent in this cohort (25%) and it is unclear how patients with AF were assessed. Variation in echocardiographic examinations of patients with AF among the three institutes might affect the results. During a median followup of 53 months, progression of AS severity can be expected in patients with an initial diagnosis of moderate AS. Although this is a crosssectional study evaluating the prognostic significance of different types of DD at the time of diagnosis of moderate AS, it is not clear whether progression to severe AS contributed to mortality during followup. The impact of AS progression on mortality needs to be clarified in future investigations.