Myocardial fibrosis by late gadolinium enhancement cardiovascular magnetic resonance in myotonic muscular dystrophy type 1: highly prevalent but not associated with surface conduction abnormality

We retrospectively evaluated consecutive patients from the Muscular Dystrophy Clinic of The Ohio State University Medical Center, with genetically confirmed diagnosis of MMD1 referred for a clinical CMR between April 2012 and March 2017. Subsequent decision-making regarding PPM implantation was done at the clinical discretion of the referring provider, typically based on ECG findings, and did not incorporate CMR data. Standard 12-lead ECG performed within 6 months of CMR exam was necessary for inclusion. Exclusion criteria included standard contraindications to perform a CMR exam: severe claustrophobia, presence of a pacing device, ferromagnetic foreign body, active implant, known allergy to gadolinium-based contrast, or advanced kidney disease (e.g. glomerular filtration rate < 30 mL/min/1.73 m2). The Ohio State University Institutional Review Board approved this retrospective study and waived informed consent.

All patients underwent a comprehensive clinical evaluation and physical examination during their clinical visit. The severity and extent of MMD was quantified using a validated Muscular Impairment Rating Scale (MIRS) related to the clinical distribution and extent of muscle weakness and myotonia [15]. The New York Heart Association (NYHA) classification system was used to assess exercise capacity. Patient’s charts were reviewed up to 12 months after CMR to record date of PPM implant or occurrence of any other major adverse cardiac events defined as heart failure, ventricular arrhythmias, or death.

Electrical parameters were assessed from a standard 12-lead surface ECG. Based on current American College of Cardiology Foundation/ American Heart Association/Heart Rhythm Society (ACCF/AHA/HRS) guidelines for device-based therapy of cardiac rhythm abnormalities [5], we defined surface conduction abnormality as the presence of ECG alterations meeting criteria for PPM implant (class I or II indications) in subjects with MMD. Accordingly, surface conduction abnormality was defined as evidence of PR interval > 200 ms (type I AV block) and/or mono or bifascicular block (QRS > 120 ms), or evidence of an advanced AV block [5]. Left ventricular (LV) hypertrophy (LVH) was assessed by Cornell criteria [16]. According to current consensus documents for Standardization and Interpretation of the Electrocardiogram from the ACCF/AHA/HRS, standard criteria were adopted to define the presence of left anterior fascicular block (LAFB), left posterior fascicular block (LPFB), left bundle branch block (LBBB), right bundle branch block (RBBB), and non-specific intraventricular conduction delay (IVCD) [17]. The QT interval was obtained via automated ECG analysis in leads II, V2, and V5. The QTc interval was calculated using the Bazett formula. Electrophysiological (EP) study data and pacing device information were retrieved from patients’ charts. Referrals for intracardiac EP studies were performed in the presence of symptomatic or asymptomatic AV conduction abnormalities (first-degree or higher AV block) with or without QRS interval ≥ 120 ms, and/or presence of palpitations, syncope, near syncope or documented arrhythmias.

All images were acquired on a single 1.5T CMR scanner with a 12-element phased array chest coil (MAGNETOM Avanto, Siemens Healthineers, Erlangen, Germany). Balanced steady-state free precession sequences (bSSFP) were used for assessment of LV volumes, ejection fraction (LVEF) and LV mass. LGE imaging was performed using a gradient-echo inversion recovery sequence with magnitude and phase sensitive inversion recovery reconstructions 10 min after 0.2 mmol/kg of gadolinium based contrast agent [18]. LV volumes and LVEF were measured from contiguous LV short-axis cine images using semi-automated software for endocardial segmentation using endocardial and epicardial contours at end-systole and end-diastole with Simpson’s rule. LV mass was calculated from the total end-diastolic myocardial volume multiplied by the specific gravity of the myocardium (1.05 g/ml) [19]. Presence of LGE was assessed by 2 expert operators blinded to clinical data (AC, KZ), with a third providing adjudication if necessary (SVR). The LGE (grams) mass was quantified by a blinded operator (AC) using a dedicated software implementing the full-width at half-maximum technique (CMR42, Circle Cardiovascular Imaging Inc. Calgary, Alberta, Canada) and indexed as a percentage of LV mass [20]. MOdified Look-Locker Inversion Recovery (MOLLI) acquisition schemes were used to acquire T1 maps produced using vendor software before and 15 min after administration of contrast. The MOLLI 3(3)3(3)5 sequence was used in the first 23 patients and subsequently the MOLLI 5(3)3 sequence was used in 29 patients. T1 values and extracellular volume fraction (ECV) were measured and calculated utilizing interventricular septal values from the mid short axis view. The region of interest was placed in the mid myocardium with manual tracing to avoid partial volume effects and regions of LGE if present [21, 22]. Myocardial ECV was calculated as previously described [23]. Reference values for normal pre- and post-contrast myocardial T1 values and ECV for the MOLLI 3(3)3(3)5 sequence are based on data of 18 healthy subjects at our institution (45 ± 18 years, 39% female), and are as follows: native T1 940 ± 28 ms, post-contrast T1 403 ± 42 ms, and ECV 24.3 ± 2.3%. Reference values for the myocardium utilizing the MOLLI 5(3)3 sequence based on 44 healthy subjects at our institution (36 ± 15 years, 59% female) are as follows: native T1 999 ± 31 ms, post-contrast T1 453 ± 30 ms, and ECV 23.8 ± 2.6%. In line with previous literature, the abnormal value for ECV was 29%, corresponding to 2SD above the reference value [24].

Fifty-two patients with genetically confirmed MMD1 (median CTG repeats = 500, IQR = 200–1163), 38% male, 41 ± 14 years, met criteria for inclusion in the study. Overall, thirty-one patients (60%) demonstrated surface conduction abnormality: 20 (38%) with a prolonged PR interval (mean 232 ± 51 ms); 5 (9.6%) with a prolonged PR interval associated with RBBB; 2 (3.8%) patients had a prolonged PR interval associated with LBBB; 2 (3.8%) had LBBB alone, one patient had RBBB alone, and one had Mobitz AV block type I. Baseline demographic, clinical and ECG characteristics according to presence or absence of surface conduction abnormality are presented in Table 1.

Subjects with conduction abnormality tended to be older, had a longer disease length (15.5 vs. 7.8 years, p = 0.015) and demonstrated higher MIRS scores (p = 0.041) compared to subjects without conduction abnormalities (Table 1). As expected they also demonstrated a longer PR and QRS intervals as compared to subjects without conduction abnormality. Alternatively, no differences were found between the two groups in the prevalence of major cardiovascular risk factors or medications, CTG repeats, or functional status (Table 1).

After a median time of 57 days from the CMR exam, 15 patients (29%) received a PPM: 7 subjects were implanted due to coexistence of first degree AV and fascicular block, 4 for presence of bifascicular block associated with either a first degree AV block or a prolonged HV interval, 2 for a prolonged HV interval, one for presence of second degree AV block (Mobitz type I), and one for presence of I degree AV block (Table 2). No cardiac death or major cardiovascular events occurred during the 12 months of observation.

Overall, twenty-two subjects demonstrated midwall myocardial fibrosis by CMR-LGE, corresponding to a prevalence of 42%. The LVEF, and LV volumes were within normal range for the entire population (LVEF 60 ± 6%, LV end-diastolic volume index 65 ± 15 ml/m², LV end-systolic volume index 26 ± 8 ml/m²) (Table 3).

The median time between the CMR exam and the ECG was 21 days (IQR 7–54). Subjects with conduction abnormality demonstrated similar LVEF, LV volumes, LV mass, and left atrial (LA) volumes compared to those without conduction abnormality (Table 3). Importantly, high prevalence of myocardial fibrosis by CMR-LGE was detected in subjects with and without surface conduction abnormality but no significant difference was noted between the two groups (42% vs. 43%, p = 0.999). Further, LGE mass and LGE% were similar between the two groups. ECV tended to be higher in the conduction abnormality group, albeit values were within the normal range in both groups (26 ± 3 vs. 24 ± 3, p = 0.050). Data on native myocardial and post-contrast T1 values according to different MOLLI sequences and presence vs. absence of conduction abnormality are presented in Additional file 1: Table S1.

Of 22 LGE-positive subjects, all demonstrated midwall fibrosis; in 17 subjects (77%) the interventricular septum was involved, the inferior and inferolateral segments were involved in four subjects (18%), and the anterior wall was involved in one subject (5%). An analysis of clinical and ECG data according to the presence vs. absence of LGE was also conducted (Additional file 1: Tables S2 and S3). There were no significant differences in ECG characteristics in LGE positive vs.LGE negative patients. Fig. 1 demonstrates representative ECG and CMR findings in 2 patients. Additionally, we have provided a Additional file 1: Figure. S1 which includes representative ECG and T1 maps in 2 patients.

Disease length and Cornell voltage were associated with surface ECG conduction abnormality on univariate analysis. Disease length was the only independent variable associated with conduction abnormality in the multivariate model (OR 1.071, 95%CI 1.003–1.144, p = 0.040; Table 4). Importantly, the presence of myocardial fibrosis by CMR-LGE, LGE mass, or LGE% was not associated with surface conduction abnormality (full regression analysis is shown in Additional file 1: Table S4). Additionally, myocardial fibrosis by CMR-LGE was not associated with surface conduction abnormality by Spearman correlation analysis (ρ = − 0.009, p = 0.949) nor was LGE% (ρ = 0.004, p = 0.976) or LGE mass (ρ = − 0.022, p = 0.878). Similar findings were also true for the association between CMR-LGE and PPM implant (ρ = − 0.03, p = 0.834).

This is a single center, retrospective study and the results should be confirmed in a large, multicenter, prospective study powered for both bradyarrhythmias as well as tachyarrhythmic SCD events.

It could be speculated that lack of correlation between myocardial fibrosis by CMR-LGE with surface conduction abnormality could be related to relatively early course of disease from the point of view of myocardial structural abnormality; however our findings rather support the opposite conclusion for the following reasons: i) there is a high prevalence of myocardial fibrosis which demonstrates that myocardial structural abnormality was frequent in this population ii) LV mass index was in the low normal range, which in line with previous studies, confirms evident muscle wasting typical for MMD1 iii) finally the length of MMD1 (median > 12 years in the whole population) excludes the possibility of an early stage of the disease where myocardial structural abnormality could have been absent or not detectable with CMR. Although we did not see trends towards significance with respect to myocardial fibrosis by CMR-LGE between our two cohorts, and our study is one of the largest CMR based studies in MMD1, our acceptance of the null hypothesis may relate to our sample size.

As the intent of our manuscript was to determine the association between myocardial fibrosis by CMR-LGE and surface conduction abnormality, we retrospectively included all consecutive patients with contrasted CMR exams and ECGs performed within 6 months. Our study spanned nearly 5 years during which two different MOLLI sequences were used at our institution, MOLLI 3(3)3(3)5 initially and then MOLLI 5(3)3. This is a limitation of our study.

ECV values were calculated from ROIs placed in the interventricular septum from the mid short axis view, hence they are not representative of the whole myocardium. However, according to recent literature and consensus documents, septal segments have proven to be the most reproducible as they are less frequently affected by off-resonance artifacts [21, 22]. Additionally, basal short axis T1 maps were not obtained at the time of the clinical scan, thus our data is only representative of the mid-short axis slice.