DOI: 10.1093/europace/euag105.323 ISSN: 1099-5129

Left atrial structural and functional assessment by cardiac magnetic resonance in atrial fibrillation-mediated cardiomyopathy

T Efraimidis, E Nyktari, O Kariki, K Pamporis, A Saplaouras, P Mililis, S Koskina, S Kotoulas, T Chatziantoniou, S Dragasis, K Letsas, M Efremidis

Abstract

Introduction

Atrial fibrillation–mediated cardiomyopathy (AF-CM) is a reversible cause of left ventricular (LV) dysfunction. Its atrial phenotype remains incompletely characterized, especially in comparison to non-reversible heart failure with reduced ejection fraction (HFrEF). Cardiac magnetic resonance (CMR) can be used for precise assessment of atrial remodeling.

Purpose

To compare LA structure and function between patients with AF-CM, HFrEF, and controls (normal LVEF) using CMR.

Methods

Forty-seven patients (Controls = 11, AF-CM = 22, HFrEF = 14) underwent CMR to measure LA volume index (LAVI), sphericity, EF, longitudinal strain (total=eR, conduit=eCD, contraction=eCT) and LGE burden. Group differences were analyzed using Kruskal–Wallis with Dunn-Holm correction; monotonic trends (Control→AF-CM→HFrEF) were tested by Jonckheere–Terpstra.

Results

Groups were comparable in age (Control 56.1 ± 13.3 y, AF-CM 62.6 ± 9.8 y, HFrEF 62.6 ± 12.2 y; p = 0.26) and sex (18 %, 23 %, and 14 % female; p = 0.82). Persistent AF was higher in AF-CM and HFrEF (9 % vs 41 % vs 36 %; p = 0.18).

Regarding LA structural parameters, LAVI was significantly larger in HFrEF (62.9 ± 17.0 mL/m²) than in AF-CM (44.7 ± 14.2 mL/m²) or controls (46.9 ± 17.5 mL/m²; p = 0.003) while LGE burden did not display noteworthy differences either for IIR thresholds of 1.2 (3.75 ± 2.71 % vs 3.99 ± 4.00 % vs 4.06 ± 5.11 %; p = 0.84) or 1.32 (0.82 ± 1.02 % vs 1.33 ± 1.78 % vs 1.58 ± 3.64 %; p = 0.93) . LA sphericity values were also similar among groups (79.8 ± 3.0 % vs 80.7 ± 3.9 % vs 81.2 ± 3.5 %; p = 0.52).

Concerning the LA functional parameters, LA-EF declined progressively from controls to HFrEF (52.4 ± 13.6 % vs 46.4 ± 11.9 % vs 38.4 ± 13.6 %; p = 0.038), with significant difference between controls and HFrEF (p = 0.036) but not between AF-CM and HFrEF (p= 0.17). Total LA longitudinal strain (eR) decreased stepwise (18.7 ± 4.4 % vs 17.5 ± 3.8 % vs 12.7 ± 3.4 %; p = 0.001), as did conduit (10.9 ± 3.0 % vs 9.3 ± 2.9 % vs 7.4 ± 1.9 %; p = 0.011) and contraction strain (7.8 ± 2.5 % vs 8.2 ± 2.9 % vs 5.2 ± 3.9 %; p = 0.032). Post-hoc analysis showed that AF-CM displayed significant or borderline higher strain values compared to HFrEF ( eR p=0.003, eCD p=0.08, eCT p=0.04) with non-significant differences to controls.

Overall comparing AF-CM to HFrEF we found significant differences to structural (LAVI) and functional parameters (LA longitudinal strain) while AF-CM values were only slightly reduced compared to Controls in a non-significant way (p>0.10). Furthermore, trend analysis demonstrated ordered worsening across control → AF-CM → HFrEF for LA-EF (p = 0.004), total (p = 0.0006) and conduit strain (p = 0.0010).

Conclusion

AF-CM LA remodeling is characterized by preserved atrial size and mild functional impairment compared to controls and markedly better metrics than HFrEF, placing AF-CM in the spectrum between normal atrial mechanics and advanced heart failure.

More from our Archive