Background: Differentiation between exercise induced adaptive myocardial hypertrophy (athlete's heart) and hypertrophic cardiomyopathy (HCM) is currently based on echocardiographic and cardiac magnetic resonance (CMR) criteria, but these may be insufficient in patients with subtle phenotype expression. This study aimed to assess whether left ventricular (LV) fractal pattern could permit to differentiate athlete's heart from HCM.
Methods: We recruited retrospectively 61 elite marathon runners, 67 patients with HCM, and 33 healthy subjects. A CMR study was performed in all subjects and the LV trabeculae fractal dimension (FD) was measured in end-diastolic frames of each short-axis cine sequence. For group comparison, the ratio of maximal myocardial wall thickness (mMWT)/indexed LV end-diastolic volume (LVED) was determined.
Results: As compared with athletes, patients with HCM had significantly (p < 0.001) greater FD in the LV basal (1.30 ± 0.07 vs. 1.23 ± 0.05) and apical (1.38 ± 0.06 vs. 1.30 ± 0.07) regions and in the whole heart (1.34 ± 0.05 vs. 1.27 ± 0.05). FD increased with age, left atrial area and indexed left ventricular mass (p < 0.05 for all) and correlated negatively with LV and RV end-diastolic volumes (p < 0.05 each). The addition of whole heart FD to the ratio of maximal myocardial wall thickness/indexed LVEDV lead to an improvement in the ability to discriminate HCM with a net reclassification index (NRI) of 71%.
Conclusions: The FD regional distribution of the LV trabeculae differentiates patients with athlete's heart from patients with HCM. The addition of whole heart FD to the mMWT/indexed LVEDV ratio improves the predictive capacity of the model to differentiate both entities.
Keywords: Cardiac magnetic resonance; Fractal dimension; Hypertrophic cardiomyopathy; athlete's heart.
Copyright © 2021 Elsevier B.V. All rights reserved.