# Left Ventricular Twist and Circumferential Strain from MRI Tagging Predict Early Cardiovascular Disease in Duchenne Muscular Dystrophy

## TL;DR

This study shows that MRI-based measurements of heart function can detect early signs of heart disease in boys with Duchenne Muscular Dystrophy before traditional indicators show problems.

## Contribution

The study introduces new MRI-derived biomarkers for early detection of heart disease in DMD patients before LVEF or LGE abnormalities occur.

## Key findings

- Peak LV twist was significantly impaired in LGE(−) DMD patients compared to healthy volunteers.
- LV biomarkers like twist and Ecc outperformed traditional metrics in distinguishing DMD patients from healthy controls.
- Age correlated with worsening heart function in LGE(+) DMD patients but not in LGE(−) patients.

## Abstract

Background/Objectives: Duchenne Muscular Dystrophy (DMD) is a prevalent fatal genetic disorder, and heart failure is the leading cause of mortality. Peak left ventricular (LV) circumferential strain (Ecc), twist, and circumferential-longitudinal shear angle (θCL) are promising biomarkers for the improved and early diagnosis of incipient heart failure. Our goals were as follows: 1) to characterize a spectrum of functional and rotational LV biomarkers in boys with DMD compared with healthy age-matched controls; and 2) to identify LV biomarkers of early cardiomyopathy in the absence of abnormal LVEF or LGE. Methods: Boys with DMD (N = 43) and age-matched healthy volunteers (N = 16) were prospectively enrolled and underwent a 3T CMR exam after obtaining informed consent. Breath-held MRI tagging was used to estimate left ventricular Ecc at the mid-ventricular level as well as the twist, torsion, and θCL between basal and apical LV short-axis slices. A two-tailed t-test with unequal variance was used to test group-wise differences. Multiple comparisons were performed with Holm–Sidak post hoc correction. Multiple-regression analysis was used to test for correlations among biomarkers. A binomial logistic regression model assessed each biomarker’s ability to distinguish the following: (1) healthy volunteers vs. DMD patients, (2) healthy volunteers vs. LGE(−) DMD patients, and (3) LGE(−) DMD patients vs. LGE(+) DMD patients. Results: There was a significant impairment in the peak mid-wall Ecc [−17.0 ± 4.2% vs. −19.5 ± 1.9%, p < 7.8 × 10−3], peak LV twist (10.4 ± 4.3° vs. 15.6 ± 3.1°, p < 8.1 × 10−4), and peak LV torsion (2.03 ± 0.82°/mm vs. 2.8 ± 0.5°/mm, p < 2.6 × 10−3) of LGE(−) DMD patients when compared to healthy volunteers. There was a further significant reduction in the Ecc, twist, torsion, and θCL for LGE(+) DMD patients when compared to LGE(−) DMD patients. In the LGE(+) DMD patients, age significantly correlated with LVEF (r2 = 0.42, p = 9 × 10−3), peak mid-wall Ecc (r2 = 0.27, p = 0.046), peak LV Twist (r2 = 0.24, p = 0.06), peak LV torsion (r2 = 0.28, p = 0.04), and peak LV θCL (r2 = 0.23, p = 0.07). In the LGE(−) DMD patients, only the peak mid-wall Ecc was significantly correlated with age (r2 = 0.25, p = 0.006). The peak LV twist outperformed the peak mid-wall LV Ecc and EF in distinguishing DMD patients from healthy volunteer groups (AUC = 0.88, 0.80, and 0.72), as well as in distinguishing LGE(−) DMD patients from healthy volunteers (AUC = 0.83, 0.74, and 0.62). The peak LV twist and peak mid-wall LV Ecc performed similarly in distinguishing the LGE(−) and LGE(+) DMD cohorts (AUC = 0.74, 0.77, and 0.79). Conclusions: The peak mid-wall LV Ecc, peak LV twist, peak LV torsion, and peak LV θCL were significantly impaired in advance of the decreased LVEF and the development of focal myocardial fibrosis in boys with DMD and therefore were apparent prior to significant irreversible injury.