Abstract:Highlights•Using fully-automated LVGLS was reproducible and feasible in ≥moderate-severe AR.•Patients with LVGLS<15% had reduced survival as compared to those with LVGLS ≥15%.•LVGLS determined survival in head-to-head comparison with LVEF, LVESVi, and LVESDi.•Combined LVGLS<15% plus LVESVi≥45 ml/m2 discriminated mortality in asymptomatic AR.BackgroundWhether automated left ventricular global longitudinal strain (LVGLS) is associated with outcomes in asymptomatic aortic regurgitation (AR) is unknown.ObjectiveTo explore the impact of automated LVGLS on survival and compare it with conventional LV parameters in chronic asymptomatic AR.MethodsLVGLS (presented as absolute value) was measured using fully-automated two-dimensional strain software in consecutive patients with isolated chronic ≥moderate-severe AR between 2004 and 2020; incremental value of LVGLS was assessed. Limited correction of endocardial border tracking was performed if needed.ResultsOf 550 asymptomatic patients (age 60±17 years; 86% men), average LVGLS was 17±3% (1st and 2nd tertile, 15.8% and 18.5%). In 16% cases, tracking border was partially corrected; average time for analysis was 25±5 seconds. At a median (interquartile range) of 4.8(1.5-9.9) years, 87 patients had died (19 died after aortic-valve surgery [AVS]). Separate multivariable models adjusted for age, sex, Charlson index, AR severity and time-dependent AVS demonstrated that LVEF(hazard ratio[HR] per 10%: 0.9), LV end-systolic volume index(LVESVi; HR per 5ml/m2: 1.08) and LVGLS(HR per unit: 0.87) were independently associated with death(all p≤0.018); however, LVGLS remained statistically significant (HR: 0.86-0.9; P≤0.007) as compared head-to-head to LVEF, LVESVi and LV end-systolic dimension index(LVESDi). The association of LVGLS and mortality was consistent across all subgroups (P for interaction all≥0.08). Spline curves showed that continuous risk of death rose at LVGLS <15%. Those with LVGLS <15% had 2.6-fold risk of death [95% CI 1.54-4.23] while those with LVGLS <15% plus LVESVi≥45ml/m2 had 3.96-fold risk[95% CI 1.94-8.03].ConclusionIn this large cohort of asymptomatic patients with ≥moderate-severe AR, automated LVGLS was feasible, efficient, and independently associated with death on head-to-head comparison with conventional LVEF, LVESDi, and LVESVi. The automated-LVGLS threshold of <15% alone or combined with LVESVi≥45ml/m2, was significantly associated with increased mortality risk and may be considered in early surgery decision-making.Graphical abstractDownload : Download high-res image (222KB)Download : Download full-size image

Automated Global Longitudinal Strain Exhibits a Robust Association with Death in Asymptomatic Chronic Aortic Regurgitation