Mechanical dispersion is a superior echocardiographic feature to predict exercise capacity in preclinical and overt heart failure with preserved ejection fraction.

Hortegal RA; Hossri C; Giolo L;   R; Gun C; Assef J; Moriya HT; Franchini KG; Feres F; Meneghelo R;

The international journal of cardiovascular imaging [Int J Cardiovasc Imaging] 2023 Mar 31.
Date of Electronic Publication: 2023 Mar 31.

Background: Heart Failure with Preserved Ejection Fraction (HFpEF) is a syndrome characterized by different degrees of exercise intolerance, which leads to poor quality of life and prognosis. Recently, the European score (HFA-PEFF) was proposed to standardize the diagnosis of HFpEF. Even though Global Longitudinal Strain (GLS) is a component of HFA-PEFF, the role of other strain parameters, such as Mechanical Dispersion (MD), has yet to be studied. In this study, we aimed to compare MD and other features from the HFA-PEFF according to their association with exercise capacity in an outpatient population of subjects at risk or suspected HFpEF.
Methods: This is a single-center cross-sectional study performed in an outpatient population of 144 subjects with a median age of 57 years, 58% females, referred to the Echocardiography and Cardiopulmonary Exercise Test to investigate HFpEF.
Results: MD had a higher correlation to Peak VO2 (r=-0.43) when compared to GLS (r=-0.26), MD presented a significant correlation to Ventilatory Anaerobic Threshold (VAT) (r=-0.20; p = 0.04), while GLS showed no correlation (r=-0.14; p = 0.15). Neither MD nor GLS showed a correlation with the time to recover VO2 after exercise (T1/2). In Receiver Operator Characteristic (ROC) analysis, MD presented superior performance to GLS to predict Peak VO2 (AUC: 0.77 vs. 0.62), VAT (AUC: 0.61 vs. 0.57), and T1/2 (AUC: 0.64 vs. 0.57). Adding MD to HFA-PEFF improved the model performance (AUC from 0.77 to 0.81).
Conclusion: MD presented a higher association with Peak VO2 when compared to GLS and most features from the HFA-PEFF. Adding MD to the HFA-PEFF improved the model performance.