Abstract
INTRODUCTION: Left ventricular (LV) longitudinal function is a prognostic marker of hospitalization and mortality in LV dysfunction. Recently, left atrial (LA) reservoir and conduit strain have also been presented as independent prognostic markers. However, the atria and ventricles are coupled in the fibrous atrioventricular plane (LA-LV coupling). The degree to which the LA strain is affected, or even determined, by the LV longitudinal function in LV dysfunction has been explored by echocardiography, but not by cardiac magnetic resonance imaging (CMR). Therefore, we aimed to quantify the association between LV longitudinal ventricular function and LA strain using CMR feature-tracking. METHODS: Three hundred and forty-two patients with LV dysfunction (including heart failure with reduced ejection fraction (HFrEF), candidates for cardiac resynchronization therapy (CRT) implantation, and ischaemic heart disease (IHD)), and 19 healthy controls (HC) who had undergone CMR were retrospectively included. LV global longitudinal strain (LV-GLS), LV atrioventricular plane displacement (AVPD), and LA-GLS (i.e. reservoir strain) were analysed in long-axis views using CMR feature-tracking. RESULTS: LA-GLS was lower in the LV dysfunction group when compared to HC (12 ± 8% vs 19 ± 7, P < .001), mirroring reductions in LV-GLS (-10 ± 5% vs -19 ± 3, P < .001), and LV-AVPD (9 ± 3 vs 15 ± 2 mm, P < .001). The coefficient of determination (r2) between LV-GLS and LA-GLS was .40 (95% CI 0.32-0.48) for the whole cohort, and 0.39 (95% CI 0.31-0.47) between LV-AVPD and LA-GLS. CONCLUSION: In a large cohort comprising both patients with LV dysfunction and HC, LA reservoir function quantified as LA-GLS was to a large extent determined by LV longitudinal function. LA function may not be an independent marker of global cardiac function for certain patient groups where diminished LA function can be a reflection of LV dysfunction.