Abstract
Introduction
Muscular architecture of the heart is three dimensionally complex and contractility parameters vary widely. Cardiac magnetic resonance (CMR) feature tracking is a largely available and facile method to assess myocardial strain at different layers of the myocardium.
Purpose
Assessing and compare the myocardial longitudinal (GLS) and circumferential strain (GCS) at three distinct layers of the myocardium in patients with heart failure (HF).
Methods
59 patients with a clinical diagnosis of HF who were post-hoc subdivided according to the measured EF and echo assessment of diastolic impairment into 3 groups, following ESC guidelines, were included: (1) patients with HF with preserved ejection fraction (HFpEF) where EF >50% and diastolic dysfunction (E/e' ratio) is present and plasma levels of natriuretic peptides are elevated, (2) patients with HF with mid-range ejection fraction (HFmrEF), where EF = 40–49% and similar additional criterias are present, (3) patients with HF with reduced ejection fraction (HFrEF) where EF <40%. Exclusion criteria: valvulopathy, arrhythmias, insufficient acquisition and artefacts.
Results
Strain values are the highest in the Endo− and progressively lower in the Myo− and Epi− layers with a gradient present in all groups but decreasing in HFmEF and further in HFrEF. GLS decrease with the severity of the disease in all 3 layers Normal > HFpEF > HFmrEF > HFrEF (Endo−: −23.0±3.5 vs −20.0±3.3 vs −16.4±2.2 vs −11.0±3.2, p<0.001, Myo−: −20.7±2.4 vs −17.5.0±2.6 vs −14.5±2.1 vs −9.6±2.7, p<0.001, Epi−: −15.7±1.9 vs −12.2±2.1 vs −10.6±2.3 vs −7.7±2.3, p<0.001) (Figure A), GCS is not different between the Normal and HFpEF (Endo−: −34.5±6.2 vs −33.9±5.7, p=0.51; Myo−: −21.9±3.8 vs −21.3±2.2, p=0.39; Epi−: −11.4±2.0 vs −10.9±2.3, p=0.54) but markedly lower in systolic HF groups Normal > HFmrEF > HFrEF (Endo−: −34.5±6.2 vs −20.0±4.2 vs −12.3±4.2, p<0.001; Myo−: −21.9±3.8 vs −13.0±3.4 vs −8.0±2.7, p<0.001; Epi−: −11.4±2.0 vs −7.9±2.3 vs −4.5±1.9) (Figure B). ROC analysis renders Endo− GCS (AUC=0.89) and respectively Endo− GLS (AUC=0.74) as optimal to detect contractile impairment in HF with Youden's thresholds of −20.2 for Endo− GLS and, respectively, −28.1 for Endo− GCS. Endo− GCS is not different between control and HFpEF and GLS impairment is present only inconstantly in HFpEF.
Conclusions
Feature tracking CMR successfully assess layer-specific myocardial strain and emerges as a powerful tool in functional stratification of patients with HF. Strain amplitude varies consistently throughout the myocardium and its quantification warrants careful standardization. Sub-endocardial strain values of strain are comparatively the highest and show most predictive power to detect contractile impairment. Underlying systolic impairment is present only in a subgroup of patients with HFpEF and only GLS and not the GCS is for this purpose a useful diagnostic tool.