Prognostic impact of echocardiographic predictors of subclinical systolic dysfunction in patients with moderate aortic stenosis

Abstract Introduction Moderate aortic stenosis (AS) marks a critical point in AS severity spectrum, paralleled by significant cardiac structural changes. Concentric left ventricular (LV) hypertrophy, in response to the increased afterload, enables the preservation of a normal ejection fraction (EF) despite significant underlying myocardial dysfunction. Reductions in LVEF may serve as early indicators of the LV's diminished capacity to withstand this increased afterload, potentially leading to LV fibrosis and irreversible myocardial damage. Research has shown that LVEF may not recover in some patients following aortic valve replacement (AVR). Consequently, there is a growing debate concerning higher LVEF thresholds for triggering AVR in patients with severe AS, aiming to intervene before clinical systolic dysfunction ensues and irreversible damage occurs. Meanwhile, a combination of objective indicators may help to identify those patients AS at high risk before a decline in LVEF occurs. Aim To explore the role of baseline left ventricle global longitudinal strain (GLS) and strain rate (SR) as markers of subclinical LV dysfunction, as predictors of LVEF depression, and their impact on survival of patients with moderate AS. Methods Patients with moderate AS and LVEF ≥ 50% at least in 2 echocardiograms were retrospectively identified. Prosthetic and bicuspid valves were excluded. Baseline LV GLS and SR were used as covariates in cox regression models to predict the cumulative incidence of LVEF depression over 5 years and overall survival after multivariable adjustment including time-dependent AVR. Results 574 patients were included (age 76 ± 9 years; 51% female; median follow-up time of 8.97 years). The average baseline aortic peak velocity was 3.4±0.8 m/s, mean pressure gradient was 25±9 mmHg, aortic valve area was 1.1± 0.3 cm2, and LVEF 60±5%. The average GLS was -17±7 % while the peak systolic SR was 1.2 ± 0.5/s. Both baseline GLS (HR= 0.88 for each -1%; 95% CI= 079-0.99; p= 0.04) and SR (HR= 0.89 for each +0.1/s; 95% CI= 0.80-0.99; p=0.03) were associated with 5-year incidence of LVEF depression (<50%). Risk discrimination improved by incorporating SR on top of GLS (increase in area under curve from 0.78 to 0.84; 95% CI= 0.01-0.10; p= 0.04). Furthermore, a nonlinear relation was found between GLS (optimal boundary <-15.8%; p=0.35) and SR (optimal boundary >0.96/s; p=0.34) and overall survival. Conclusions This study identifies baseline GLS and SR as predictors of LVEF depression in moderate AS underscoring their importance in detecting subclinical LV dysfunction. The combination of GLS and SR may enhance the prediction of cardiac function decline, suggesting their potential utility in guiding early intervention strategies. Although their direct relationship with survival needs further exploration, these findings advocate for incorporating GLS and SR in AS management to inform timely AVR decisions and improve patient outcomes.