Left ventricular deformation predicts major adverse cardiac events following acute myocardial infarction independently of afterload and ventricular-arterial coupling.

Abstract

Background: Load dependence on left ventricular (LV) strain is under constant debate with its interference with prognostic implications remaining unclear. Consequently, we sought to investigate their interaction and prognostic value following acute myocardial infarction (AMI) using state-of-the-art cardiac magnetic resonance (CMR) imaging.

Methods: In total, 1235 patients (n = 795 ST-elevation [STEMI] and 440 non-STEMI) underwent CMR in median 3 days following AMI. Infarct characteristics were described by CMR using tissue characterisation (infarct size, microvascular obstruction, area at risk) and deformation imaging including LV global longitudinal and circumferential strain (GLS/GCS). Non-invasive haemodynamic indices included effective arterial elastance Ea (end-systolic pressure (ESP)/stroke volume) and the non-geometric LV end-systolic afterload index NGI [(ESP × LV end-systolic volume (ESV))/LV mass] for estimation of LV afterload. LV contractility was assessed using end-systolic elastance Ees (ESP/LV ESV). Ventriculo-arterial coupling was described as Ea/Ees. Major adverse cardiac events (MACE) were recorded within the first year.

Results: All haemodynamic indices were impaired in patients with MACE during follow-up compared to patients without (p < 0.001-0.005). Ventriculo-arterial coupling showed the highest correlation to infarct properties (infarct size r = 0.51, p < 0.001) and deformation imaging (GLS r = 0.54, GCS r = 0.72, p < 0.001). GLS and GCS were associated with MACE independently of all haemodynamic indices (p < 0.001 for all except of GCS-Ea/Ees p = 0.024).

Conclusions: Non-invasive haemodynamic indices are associated with outcome following AMI with ventriculo-arterial coupling showing the most prominent association to infarct properties and outcome. GCS showed higher correlation to haemodynamic indices compared to GLS whilst both are independent predictors for MACE.