T1 mapping in patients with Fabry disease predicts functional and morphological abnormalities in cardiac MRI

Abstract

Abstract Introduction About 50% of Fabry disease (FD) patients develop cardiac manifestation, also known as Fabry cardiomyopathy. Beside echocardiography, cardiac magnetic resonance imaging (CMR) is included in the regular assessment of these patients, allowing non-invasive tissue characterization via T1 Mapping but also evaluation of morphological and functional characteristics. Purpose Myocardial strain measured by feature-tracking CMR has been previously shown to be a sensitive indicator of mechanical dysfunction in different diseases. We therefore evaluated changes of these parameters over the time, potentially reflecting disease progression in patients with FD. Furthermore, we also assessed the association of T1 mapping with follow-up changes of myocardial strain. Thus, our data may provide novel insights about cardiac changes in the natural course of the disease and cardiac effects of specific therapies. Methods We used post processing software for CMR to analyze baseline and follow-up cardiac involvement in a cohort of FD patients who were enrolled in the KarMA study, which examines cardiac changes over time. We measured left ventricular mass (LVM), left ventricular end-systolic and end-diastolic volume (LVEDV, LVESV), left ventricular ejection fraction (LVEF), mean left ventricular T1 relaxation times (T1), global longitudinal strain (GLS), global regional strain (GRS) and global circumferential strain (GCS). Statistical analyses included T-test and Spearman's correlation. Results CMR Images of 33 Patients were analyzed (22 females; mean age 40.8±16.3 years; at baseline 17 were therapy naïve and 16 on ERT), of whom 31 underwent a second CMR. In 18 patients (10 females; 44.9±16.0 years) myocardial strain measurement by feature-tracking was possible in both examinations. T1, LVM, LVEDV and LVESV, LVEF as well as GLS, GRS and GCS showed no significant change during the mean follow-up period of 43 (±17) months in all patients. However, we found a significant correlation of baseline T1 (967±84.5ms) with follow up values of LVM (84.9±32.7 g/m2; p=0.02, ρ=−0.462), LVEDV (63.0±17.6 g/m2 p=0.049 ρ=−0.398) and GRS (70.2±14.2% p=0.011 ρ=−0.5). Baseline T1 also correlated with changes in LVM (p=0.036 ρ=−0.44) and changes of GCS (p=0.001 ρ=−0.858) from baseline to follow up. Differences in T1 from baseline to follow up correlated with follow-up values of GLS (−19.5±3.57 p=0.044, ρ=−0.414) and GRS (69.1±18.7 p=0.003, ρ=−0.588). Conclusion T1, reflecting glycosphingolipid accumulation might predict changes in LVM, LVEDV and strain. Moreover, changes in T1, reflecting myocardial remodeling, go in parallel with worse left ventricular function shown by more positive values of GLS, GCS and GRS. T1 mapping together with strain parameters are important markers for monitoring patients with FD. Funding Acknowledgement Type of funding source: None