The demographically similar study population included 38 DMD-FC (mean age 39.1 ± 8.8 years) and 22 CG (mean age 39.9 ± 12.6 years). All eligible subjects who fulfilled the inclusion criteria (age over 18 years, signed informed consent, absence of CMR contraindications, and cardiovascular pathology besides dystrophic cardiomyopathies) were enrolled. The CG were healthy patients without any pathological findings on CMR, no anamnesis of cardiac disease, and no other pathological test. Exclusion criteria for both groups were renal insufficiency (estimated glomerular filtration rate < 30 mL/min/1.73 m2), CMR contraindications, or limited life expectancy. Following the Declaration of Helsinki (2000) of the World Medical Association, the Faculty Hospital St. Anne's Ethics Committee (reference number 55V/2016) approved the study.
Cmr Acquisition
CMR was performed on a 1.5T scanner (Ingenia, Philips Medical Systems, Best, The Netherlands) according to our standard protocol. It was equipped with 5- and 32-element phased-array receiver coils allowing parallel acquisition techniques in the supine position in repeated breath-hold. Functional imaging using balanced turbo field echo (b-TFE) cine sequences included four-chamber, two-chamber, and LV outflow tract long axis views and a short axis (SAX) stack. Wall motion abnormalities were assessed. LV functional and morphological parameters were calculated from the SAX stack using the summation-of-disc method following the recommendations on post-processing evaluation from the Society for Cardiovascular Magnetic Resonance (SCMR)(19).
LGE images in all long-axis views and the SAX view were acquired 10 min after an intravenous bolus of 0.2 mmol/kg of the gadolinium-based contrast agent gadobutrol (Gadovist, Bayer-Schering Pharma, Germany) using an inversion recovery (IR-TFE) sequence, and, if uncertain, by phase-sensitive inversion recovery TFE (PSIR-TFE). Both two- and three-dimensional acquisitions were performed in mid-diastole. LGE was defined as positive if the visual enhancement was higher than the mean signal intensity of the reference myocardium.
T1 native mapping was performed as described previously(17) using a Modified Look-Locker Inversion recovery sequence (MOLLI) with a 5(3)3 scheme to measure T1 native (pre-contrast) and a 4(1)3(1)2 for T1 post-contrast (15 min after contrast agent administration). The measurements were done at the mid-ventricular level in the SAX plane.
Mr Data Analysis
T1 native and T1 enhanced maps were constructed pixel-by-pixel by two readers (MLMP, TH) using cvi42 (release 5.13.9, Circle Cardiovascular Imaging, Calgary, Canada). Manual epi- and endocardial contours were drawn in the mid-ventricular slice of SAX using 10% of the myocardial wall as border cutting. A motion correction algorithm was integrated into the analysis. A semi-quantitative approach was followed to determine the presence of LGE according to the American Heart Association (AHA) 17-segment model(20). Finally, the ECV was calculated according to (1-haematocrit) (1/T1myo,post – 1/T1myo,native)/(1/T1blood,post – 1/T1blood,native) for each segment, and the global ECV was the average of the values in those segments. The haematocrit was obtained on the same day(21). A DMD-FC with LGE in at least one myocardial segment was considered to be LGE-positive. If no enhancement was observed, the DMD-FC was identified as LGE-negative.
Statistics
Variables in both groups were compared using the Student's t-test for unpaired data. The normal distribution was checked by the Kolmogorov-Smirnov test and a visual inspection of histograms. The data are presented as mean ± standard deviation (SD).
The power of the tests was considered for T1 native relaxation time and ECV values to determine the magnitude of the effect that could be identified as statistically significant. The following assumptions were employed: α = 0.05; power 0.8 (0.6, respectively); normal distribution; standard deviation corresponding with the actual distribution in the CG.
The interobserver agreement was assessed with the intraclass correlation coefficient (ICC, two-way mixed-effects model), which was determined from eight randomly selected cases analysed by two readers (MLMP, TH).
All analyses were performed using Statistica (version 14.0. TIBCO Software Inc., 2020) and in R (v4.2.1) with RStudio IDE (v2022.7.1.554, RStudio, PBC) software. The value of α = 0.05 was used as a threshold for statistical significance throughout the analyses.