P2457Left atrial strain parameters predict clinical outcomes in patients with HFmrEF and HFpEF

Abstract Background Heart failure with preserved ejection fraction (HFpEF) and with mid-range ejection fraction (HFmrEF) are conditions sharing similar diagnostic criteria and symptoms. Left ventricular (LV) diastolic dysfunction has often been considered central to the pathophysiology of both, as determined by current conventional echocardiographic measures of elevated filling pressure and/or an increased in left atrial (LA) volume. However, the independent prognostic importance of LA function, beyond clinical predictors and LV features, in those groups remains unclear. Aims To evaluate the prognostic value of LA function measured by speckle tracking echocardiography in patients with new diagnosis of HFpEF and HFmrEF. Methods A total of 171 consecutive patients diagnosed with HFpEF with an Ejection Fraction (EF) ≥50%, or HFmrEF with an EF of 40–49% were prospectively enrolled and followed up. Follow-up was completed in 134 patients with HFpEF (n=100) and HFmrEF (n=34) for a median period of 57 months. Baseline characteristic and comorbidities were recorded, as well as conventional resting echocardiographic parameters of systolic and diastolic function. Left atrial phasic function was analysed using 2D-Specle Tracking. Peak-atrial longitudinal strain (PALS) and peak-atrial contraction strain (PACS) were measured, reflecting LA reservoir and pump function respectively. We looked at the relationship of traditional parameters of systolic and diastolic function and LA strain, with the following cardiovascular (CV) events: acute coronary syndromes, new atrial fibrillation, cerebrovascular events, HF-related admissions and cardiac deaths. Results During a median follow up period of 57 months (range 11.10), of 134 patients, 42 (31%) had at least one CV event, 10 (8%) out of whom had at least one HF-related hospitalization and 5 (4%) had cardiac death. 92 patients (69%) remained event-free. In univariate analysis, lower PALS and PACS were associated with increased risk for the composite endpoint of all CV-events (PALS OR=0.92 p<0.01 and PACS OR=0.92 p<0.05), but only PALS predicted HF admissions (PALS OR=0.86 p<0.05). In multivariate analysis after adjusting for LV EF, E/e' and LAVi, PALS remained significant at predicting CV-events (OR=0.91 p<0.05) and HF admissions (OR=0.85 p<0.05), but PACS did not. Amongst all echocardiographic parameters measured, PALS had the greatest Area Under the Curve at Receiver Operating Characteristics analysis (AUC=0.78 p<0.05) and a PALS value of <23% best predicted future HF-related admissions (sensitivity 0.9, specificity 0.6). Conclusion LA dysfunction in HFpEF and HFmrEF is associated with a higher risk of CV events and HF hospitalizations. The association of LA strain with HF hospitalization remained significant after adjustment for systolic and diastolic parameters.