- Email: [email protected]
High prevalence of cardiac involvement in patients with myotonic dystrophy type 1: A cross-sectional study
International Journal of Cardiology 174 (2014) 31–36
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
High prevalence of cardiac involvement in patients with myotonic dystrophy type 1: A cross-sectional study Helle Petri a,⁎,1, Nanna Witting c,1, Mads Kristian Ersbøll a,1, Ahmad Sajadieh e,1, Morten Dunø d,1, Susanne Helweg-Larsen c,1, John Vissing c,1, Lars Køber a,1, Henning Bundgaard b,1 a
Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark Unit for Inherited Cardiac Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark d Department of Clinical Genetics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark e Department of Cardiology, University Hospital of Bispebjerg, Copenhagen, Denmark b c
a r t i c l e
i n f o
Article history: Received 29 October 2013 Received in revised form 10 March 2014 Accepted 14 March 2014 Available online 20 March 2014 Keywords: Myotonic dystrophy Arrhythmia Conduction disorders Sudden cardiac death
a b s t r a c t Background: Patients with myotonic dystrophy type 1 (DM1) have a three-fold higher risk of sudden cardiac death (SCD) than age-matched healthy controls. Despite numerous attempts to define the cardiac phenotype and natural history, existing literature suffers from low power, selection-bias and lack of controls. Thus, the optimal strategy for assessing cardiac involvement in DM1 is unclear. Method: In this large single-centre study, we evaluated 129 unselected DM1 patients (49.6% men), mean (SD) age 44 (14.7) years with family history, physical examination, electrocardiogram (ECG), echocardiography, Holtermonitoring and muscle strength testing. Results: Cardiac involvement was found in 71 patients (55%) and included: 1) Conduction abnormalities: atrioventricular block grade I (AVB grade I) (23.6%), AVB grade II (5.6%), right/left bundle branch block (5.5/3.2%) and prolonged QTc (7.2%); 2) arrhythmias: atrial fibrillation/flutter (4.1%), other supraventricular tachyarrhythmia (7.3%) and non-sustained ventricular tachycardia (4.1%); and 3) structural abnormalities: left ventricular systolic dysfunction (20.6%) and reduced global longitudinal strain (21.7%). A normal ECG was not significantly associated with normal findings on Holter-monitoring or echocardiography. Patients with abnormal cardiac findings had weaker muscle strength than those with normal cardiac findings: ankle dorsal flexion (median (range) 4.5 (0–5) vs. 5.0 (2.5–5), p = 0.004) and handgrip (median 4.0 (0–5) vs. 4.50 (2–5), p = 0.02). Conclusion: The cardiac phenotype of DM1 includes a high prevalence of conduction disorders, arrhythmias and risk factors of SCD. Systematic cardiac screening with ECG, Holter-monitoring and echocardiography is needed in order to make a proper characterization of cardiac involvement in DM1. © 2014 Elsevier Ireland Ltd. All right reserved.
1. Introduction Myotonic dystrophy type 1 (DM1) is an autosomal dominantly inherited neuromuscular disorder caused by an unstable expansion of a tri-nucleotide (CTG) repeat on chromosome 19 in the 3′ untranslated region of the myotonic dystrophy protein kinase gene [1]. Cardiac involvement in patients with DM1 is a major concern and includes an increased risk of conduction disturbances, arrhythmias, compromised systolic and diastolic function and sudden cardiac death
⁎ Corresponding author at: Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 9, 2100 Copenhagen, Denmark. Tel.: +45 28 26 44 32; fax: +45 35 45 77 05. E-mail address: [email protected] (H. Petri). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.03.088 0167-5273/© 2014 Elsevier Ireland Ltd. All right reserved.
(SCD) [2,3]. Nevertheless, cardiac involvement is mainly described in smaller studies with low power or limited to studies of electrocardiographic or studies of echocardiographic abnormalities [2–8]. This restricted focus also applies for the existing longitudinal studies [3,9–16] e.g. the study by Groh and co-authors investigating specific ECGpredictors of SCD [2]. The cardiac phenotype of DM1 is complex with unpredictable progression and is not necessarily correlated with the severity of neuromuscular involvement [2,3]. The cardiac conduction disorders are probably caused by myocardial fibrosis, fat infiltration and hypertrophy frequently identified in autopsies from patients with DM1 [17–19]. These changes may also be a substrate for supraventricular and ventricular arrhythmias and also play a key role in the development of the observed systolic dysfunction [4,17,20]. Patients with DM1 rarely report cardiac symptoms despite cardiac involvement. This may mainly be due to the reduced cardiac demand caused by the impaired skeletal muscle function. Consequently,
32
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
arrhythmias may remain unnoticed e.g. atrial tachyarrhythmia, which is an independent predictor of SCD [2,15,21]. Early detection of cardiac involvement is pivotal in order to optimize early intervention which may consequently reduce cardiac symptoms and manifestations. In this single-centre study, we performed a systematic cardiac assessment including cardiac symptoms, ECG, Holter-monitoring and echocardiography. With this systematic approach, we aimed at investigating the complete cardiac phenotype of DM1 and the association between abnormal findings on each of the used modalities. Secondly, short term follow-up was performed exclusively for selected major events (cardiac and all-cause mortality) to assess if a possible minimum time between cardiac follow-up could be recommended. 2. Methods 2.1. Study design The study was conducted at the Department of Cardiology and the Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark and approved by the regional scientific ethics committee (reference number H-d-2008-077). Patients were evaluated by family-history, physical examination, 12-lead ECG, transthoracic echocardiography, 48-hour ECG-monitoring (Holter-monitoring) and muscle strength testing. Blood samples were analyzed for plasma levels of NT-proBNP, creatinine kinase (CK) and myoglobin and screened for liver, renal and thyroid diseases. All bloodsamples were analyzed in the same laboratory and results were adjusted for age and gender, when appropriate. SCDs in relatives below or above the age of 50 years were registered. Lastly, we collected short-term follow-up data exclusively regarding selected major events (cardiac and all-cause mortality) and registered number and cause of deaths. 2.2. Study population All adult patients, age ≥18 years, with genetically confirmed DM1 were invited to participate and enrolled from December 2010 to December 2012. None of the patients were selected according to age, gender or cardiac symptoms and those included provided written informed consent. A thorough telephone interview was performed in those patients who did not want to participate. Patients were asked about cardiac symptoms and informed about the necessity of seeking medical advice in case of heart specific symptoms. Holter-parameters were compared to a healthy control group from The Copenhagen Holter study [22]. These controls were enrolled from April 1998 to June 2000 and enrolment was not designed to specifically match the DM1 cohort. 2.3. Genetic testing Genetic testing had been performed earlier as part of the diagnostic work-up and the heterozygous presence of an abnormally expanded CTG repeat allele was confirmed either by Southern-blot analysis and/or by triplet repeat primed polymerase chain reaction [1, 23]. 2.4. Electrocardiography A 12 lead ECG was performed using a Burdick Atria 6100 ECG. An ECG was considered abnormal in the presence of: atrial flutter/fibrillation (AFL/AF), other supraventricular arrhythmia including atrial and re-entry tachycardia, atrio-ventricular block (AVB) grades I–III (AVB I: PR-interval N220 ms), right and left bundle branch block (RBBB/LBBB), incomplete right bundle branch block (IRBBB) and prolonged QTc (N450 ms in men and N470 ms in women) using Bazett's formula (QTc = QT/√RR). Additionally, we assessed the presence of cardiac abnormalities separately in patients with PR-interval N 240 ms. All ECGs were analyzed by one investigator (HP) and in case of uncertainty discussed with other members of the study group.
(ASE) [24]. Left ventricular systolic dysfunction (LVSD) was defined as a biplane left ventricular ejection fraction (LVEF) ≤50%. Doppler recordings of mitral inflow were performed by placing a 2.5 mm sample volume at the tip of the MV leaflets during diastole and recording the pulsed-waved Doppler signal. Peak velocities of early (E) and atrial (A) diastolic filling and MV deceleration time were measured and the E/A ratio was calculated. Continuous-wave Doppler recordings of the LV outflow tract were obtained and aortic valve opening and closure times were measured. Pulsed-wave Doppler tissue imaging recordings were performed at the lateral and medial mitral annulus with measurements of myocardial peak early (E′). The E/E′ ratio was calculated from the lateral values of E/E′. Diastolic dysfunction was assessed and graded in accordance with ASE's recommendations based on the following parameters: e′ lateral, LA volume, MV deceleration time, E/A ratio and E/e′ [25]. 2.6. Left ventricular longitudinal strain analysis LV longitudinal function was assessed by global longitudinal strain (GLS) using a semiautomatic algorithm (Automated Functional Imaging (AFI); GE). Briefly, manual positioning of three points was performed in each of the three apical projections, enabling the software to semi-automatically track the myocardium throughout the heart cycles. Careful inspection of tracking and manual correction, if needed, was performed and in case of unsatisfactory tracking the segment was excluded from speckle tracking analysis. The algorithm then calculated the overall GLS as the average value of all three projections. Normal values of GLS between −22.1% and −15.9% (mean, −19.7%; 95% CI (−20.4% to −18.9%)) have been reported previously [26]. Abnormal GLS was defined as GLS above −15.9%. 2.7. Holter-monitoring A 48-hour Holter-monitoring was performed with a 3-electrode Lifecard CF (Spacelabs Healthcare). Holter-monitoring was considered abnormal in the presence of; AVB grades I–III, AF/AFL, other supraventricular tachyarrhythmia (N30 SVES/h or runs of ≥20 SVES), frequent VPCs (≥30/h) and non-sustained VT (NSVT) (minimum of 3 beats at ≥100 bpm). Recordings were evaluated by a single observer (HP) and discussed in the study group in case of uncertainty. Holter-parameters were compared to a healthy control group comprising 285 healthy individuals (74.4% men, mean (SD) age 57.6 (2.5) years). This control-group and the method used for the evaluation of Holter-results have been previously described in detail [27]. 2.8. Cardiac involvement and neuromuscular affection Muscle strength was graded from 0 to 5 using the Medical Research Council scale (MRC) (0 = no ability to contract muscle, 5 = normal strength). In patients with DM1, early affection occurs primarily in the distal muscles. Therefore, we investigated the association between handgrip (dominant hand) and ankle dorsal flexion and abnormal cardiac findings. Neurological assessments were performed by experienced neurologists (JV, NW, SHL). 2.9. Statistics Data were analyzed with IBM SPSS Statistics version 19. A p-value ≤ 0.05 were considered statistically significant. Normally distributed values are expressed as means ± SD. Data with skewed distribution is given as median (range). Categorical variables were summarized by frequency counts (percentage) and differences among groups were evaluated using chi-square test. Results of continuous variables are presented as median (range) and comparisons between categories were made with Mann-Whitney U test. Correlation analyses were performed using Spearman Correlation. The prevalence of any given parameter was first assessed in the total study cohort. Secondly, we corrected all parameters for familiar relations using chi-square test, comparing the prevalence of any given parameter from all included patients vs. the prevalence of the equivalent parameter from the oldest representative in each family.
3. Results 2.5. Echocardiography Echocardiography was performed using a Vivid e9 (General Electric, Horten, Norway) and the examinations were obtained and analyzed by a single operator/observer (HP). The examinations were discussed with other members of the study group in case of uncertainty or suboptimal image quality. Three consecutive heart cycles were recorded and images were obtained at a frame rate of ≥60 frames/s and analyzed with EchoPac BT 11.1.0; General Electric, Horten, Norway. Two-dimensional parasternal images were used to determine left ventricular (LV) cavity dimensions and wall thickness. LV volumes were determined using the biplane Simpson model. Left atrial volume was calculated from the biplane area-length method with maximum volume before mitral valve (MV) opening (LAmax) and minimum volume just before MV closure (LAmin). Volumetric and dimensional measurements of the left ventricle and left atrium were indexed to body surface area when appropriate. All volumetric analyses were performed in accordance with the recommendations from the European Association of Echocardiography and the American Society of Echocardiography
3.1. Study population We invited 171 patients with DM1 to participate in this study. Fortytwo patients did not participate: two due to severe neuromuscular involvement and “lacking energy”, 25 due to lack of interest in scientific research projects and long distances to our hospital, 13 did not respond to repeated invitations and two died of non-cardiac causes prior to inclusion. A thorough telephone interview regarding cardiac symptoms was performed in the 27 patients who refused to participate and none of the patients had any cardiac complaints. Thus, we included 129 patients with genetically confirmed DM1 from 73 families (64 men (49.6%), mean (SD) age 44.0 (14.7) years. Gender specific
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
demographics are presented in Table 1. Blood-samples revealed no signs of liver, renal or thyroid disease. The most predominant complaint in our patients was mental and/or physical fatigue: n = 77 (59.7%). Possible cardiac-specific symptoms were only present in few patients including: palpitations (n = 11 (8.5%)), blurred vision (n = 10 (7.8%)), near-syncope and syncope (n = 4 (3.1%)), dyspnea (n = 9, (7.1%)), chest pain (n = 6 (4.7%)) and peripheral oedema (n = 4 (3.1%)). 3.2. Electrocardiography
33
Table 2 Abnormal ECG- and echocardiographic findings in patients with DM1.
ECGa AVB grade I IRBBB RBBB LBBB Prolonged QTc
n (%)
Men (n)
Women (n)
p-Value
30 (23.6) 11 (8.8) 7 (5.5) 4 (3.2) 9 (7.2)
14 7 1 1 5
16 4 6 3 4
0.68 0.53 0.06 0.37 1.00
n (%)
ECG was available in 127/129 patients. The most predominant conduction disorder was AVB grade I (PR-interval N220 ms) (n = 30 (23.6%)) (Table 2). AVB grade I was associated with: longer QRSinterval (median (range) 119 ms (80–200) vs. 90 ms (70–175), p b 0.001), higher GLS-values (median (range) − 16.9% (− 24 to − 10) vs. − 19.0% (− 24 to − 12), p = 0.04) and a higher number of VPCs/48 h (median (range) 20 (0–6724) vs. 4 (0–20,220), p = 0.023). Additionally, a PR-interval N 240 ms was present in 7/30 (23.3%) and these patients all had additional cardiac abnormalities, such as reduced LVEF, RBBB/LBBB, AVB grade II, frequent VES or NSVT. A normal ECG was not significantly associated with normal findings on 1) Holter-monitoring (i.e. 12/47 patients had concomitant abnormal ECG and Holter-monitoring while 9/76 patients with normal ECG had abnormal Holter-monitoring, p = 0.08), or 2) echocardiography (i.e. 12/48 had concomitant abnormal ECG and echocardiography while 16/78 patients with abnormal echocardiography had normal ECG, p = 0.66). PR-interval was associated with age despite the low mean age in our cohort (Spearman correlation, r = 0.32, p b 0.001). 3.3. Echocardiography Echocardiography was performed in 126/129, and speckle tracking analysis in 93/126 patients. LVSD was found in 26 patients (20.6%) with a significantly higher prevalence in men (Table 2). LVSD was not associated with abnormal findings on 1) ECG (i.e. 12/26 patients had concomitant LVSD and abnormal ECG while 36/100 patients with normal LVEF had abnormal ECG, p = 0.37), or 2) Holter-monitoring (i.e. 6/23 patients had
Echocardiographyb LVEF ≤ 50% 26 (20.6) IVSD N 11 mm 14 (11.4) LVIDD N 55 mm 6 (4.9) 2 LAEDVi N 34 ml/m 9 (8.3) GLS N −15.9% 20 (21.7)
Median (range)
Men (n)
Women (n)
p-value
48 (33–50) 13 (11–17) 57 (56–60) 36 (35–40) −14 (−15 to −10)
18 9 6 7 13
8 5 0 2 7
0.05 0.27 0.01 0.09 0.08
GLS: global longitudinal strain (available in 92 patients). IVSD: interventricular septum in diastole. IRBBB: incomplete right bundle branch block. LAEDVi: left atrial end diastolic volume indexed. LBBB/RBBB: left/right bundle branch block. LVEF: left ventricular ejection fraction. LVIDD: left ventricular internal diameter in diastole. a ECG was available in 127 patients. b Echocardiography was performed in 125 patients. Parameters were not available in all patients; percent is calculated from the actual number of observations.
concomitant LVSD and abnormal Holter-monitoring while 15/99 patients with normal LVEF had abnormal Holter-monitoring, p = 0.23). LVEF did not correlate with age (Spearman correlation r = 0.09, p = 0.30). Nine patients (8.3%) had abnormal LAEDVi N34 ml/m2 and six (4.8%) had diastolic dysfunction (grade 1 (n = 3), grade 2 (n = 3). No major valvular abnormalities were observed. A large proportion of patients (n = 20 (21.7%)) had abnormal GLS above − 15.9% and 12/20 had preserved LVEF N50% (Table 2). There was a trend towards a positive correlation between GLS and PRinterval (Spearman correlation r = 0.183, p = 0.08). 3.4. Holter-monitoring
Table 1 Clinical characteristics in patients with DM1.
Patients, n Age, mean (SD) BPs mm Hg, mean (SD) BPd mm Hg, mean (SD) NT-proBNP (pmol/l), median (range) CK (U/l) , median (range) Myoglobin (μg/l), median (range) Muscle strength sum-score, median (range)b Increased NT-proBNP, n (%), median (range)c Increased CK, n (%), median (range)c Increased myoglobin, n (%), median (range)c
DM1 cohorta
Men
129 44 (15) 125 (19) 77 (13) 9 (1–76)
64 (50) 44 (16) 125 (17) 77 (13) 6 (1–76)
65 (50) 44 (14) 125 (21) 76 (13) 13 (1–44)
1.0 0.78 0.47 0.55 b0.001
164 (38–952) 83 (27–258) 9 (1–10)
182 (56–780) 106 (32–258) 9 (1–10)
131 (38–952) 71 (27–190)
0.007
9 (2–10)
0.72
14 (13)
18 (11–41)
29 (16–44)
0.44
24 (21) 76(66)
455 (312–780) 118 (82–258)
Women
318 (250–952) 86 (50–190)
p-Value
0.004
0.01 b0.001
BPs/BPd: blood pressure systolic/diastolic. CK: creatine kinase. a Not all parameters were obtainable in all patients; percent is calculated from the actual number of observations. b Muscle strength sum-score for ankle dorsal flexion and handgrip (MRC-score). c Increased value after adjustment for age and gender.
Holter-monitoring was performed in 122/129 patients and parameters compared to a healthy control group (285 individuals (74.4% men, mean (SD) age 57.6 (2.5) years). Patients with DM1 had a significantly higher prevalence of AVB grade II than the total control-group (7/122 vs. 4/285, p = 0.02). We found no difference between DM1 patients and controls concerning: AF/AFL (5/122 vs. 4/285, p = 0.13); SVT-episodes (9/122 vs. 24/285, p = 0.68); frequent VPC (7/122 vs. 17/285, p = 1.00) and NSVT (5/ 122 vs. 21/285, p = 0.27) (Table 3). The observed arrhythmias were generally asymptomatic except in a 21-year-old man with AF/AFL and palpitations during exercise and in a 63-year-old woman with dyspnea, Table 3 Holter-findings in patients with DM1 and healthy controls.
Age, mean (SD) (years) AVB grade II, n (%) AF/AFL, n (%) SVT, n (%) VPC N30/h, n (%) NSVT, n (%)
DM1 (n = 122)
Controls (n = 285)
p-Value
44 (14.7) 7 (5.6) 5 (4.1) 9 (7.3) 7 (5.8) 5 (4.1)
58 (2.5) 4 (1.4) 4 (1.4) 24 (8.4) 17 (6.0) 21 (7.3)
b0.001 0.02 0.13 0.68 1.00 0.27
AF/AFL: atrial fibrillation/flutter. AVB grade I/II: atrio-ventricular block grade I/II. NSVT: non-sustained ventricular tachycardia. SVT: supraventricular tachycardia. VPC: ventricular premature contractions.
34
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
which might as well have been attributed to her reduced physical ability. The prevalence of arrhythmias was independent of gender. All analyses in this study showed similar results after correction for familiar relation. 3.5. Cardiac involvement and neuromuscular affection Overall, 71 patients (55%) had abnormal findings on ECG, Holtermonitoring and/or echocardiography. Patients with abnormal findings had a lower MRC-score on ankle dorsal flexion (median (range) 4.5 (0–5) vs. 5.0 (2.5–5), p = 0.004) and handgrip (median 4.0 (0–5) vs. 4.50 [2–5], p = 0.02) compared to patients with normal cardiac findings (Fig. 1). 3.6. Implantable cardiac devices Five patients (3.9%) had an ICD at study inclusion and were all diagnosed with DM1 prior to ICD-implantation: three family-members (a father and his son and sister: 51, 18 and 45-year-old, respectively) had prophylactic ICDs due to the SCD of the father's 11-year-old daughter, who was diagnosed with DM1 post-mortem [28]. A 45-year-old
p = 0.004
A
Ankle dorsal flexion (MRC)
3.7. Invasive procedures Radiofrequency ablation for AFL was performed in a 45-year-old man. The patient had AFL on initial ECG and Holter-monitoring, abnormal NT-proBNP (40.6 pmol/l) and normal echocardiography. After isthmus-ablation, the patient had sinus rhythm with AVB grade I and episodes with AVB grade II (type 1). At latest follow-up (21 months after ablation), Holter-results showed episodes with atrial tachycardia and the patient had received a PM. Invasive electrophysiological evaluation was performed in a 21-year-old man with palpitations during exercise, IRBBB and episodes with AF/AFL. Atypical AFL and atrial tachycardia were induced and terminated with burst-pacing. There was no indication for ablation. 3.8. Family history and sudden cardiac death Family history revealed a total of 37 patients (28.7%) with SCDs in relatives, i.e. SCD in 20 families out of a total of 73 families (27.4%). Suspected SCD below or above the age of 50 years accounted for 25 (67.6%) and 12 (32.4%) cases, respectively. The majority of SCDs (22/37 (64.7%)) were in first degree relatives. Patients with SCD in relatives did not have a higher prevalence of abnormal cardiac examinations compared to the remaining patients (23/37 vs. 48/92, p = 0.33).
5
4
3
2
3.9. Short term follow-up for cardiac and all-cause mortality
1
During short-term follow-up regarding mortality (median (range) 31 (15–51) months), six patients died (4 men (age 31, 45, 53 and 77 years) and 2 women (age 56 and 59 years). Four of these patients had cardiac involvement at inclusion: the 55-year-old woman had frequent VES, the 53-year-old man had abnormal GLS (− 13%) and a family-history of SCD, the 31-year-old man had RBBB and the 45-year-old man had a PR-interval of 250 ms and LVEF 45%. The lastmentioned patient was hospitalized due to manifest heart failure shortly before he died, where he was optimized in heart failure treatment but failed to comply this treatment. He suffered from sudden death few months after his last admission. The remaining patients died of noncardiac causes (three patients due to sepsis (primary pulmonary or urinary infection), one patient due to chronic pancreatitis and one due to abdominal cancer).
0
Normal ECG, Holter and TTE
B
Abnormal ECG, Holter and/or TTE p = 0.02
5
4
Handgrip (MRC)
woman had an ICD due to syncope, AVB grade II (type II), RBBB, left posterior fascicular block and prolonged QTc (519 ms). The patient also carried a previously described disease-causing KCNE1-mutation (D76N, c.226GNA) associated with prolonged QTc-interval. A 51-year-old woman had a prophylactic ICD due to a highly malignant family history of SCD. A 19-year-old woman had an implantable loop recorder (ILR) due to syncopes. None of the patients had a pacemaker (PM).
4. Discussion
3
2
1
0
Normal ECG, Holter and TTE
Abnormal ECG, Holter and/or TTE
Fig. 1. Cardiac involvement and peripheral muscle strength. Patients with abnormal cardiac findings had significantly weaker muscle strength compared to patients with normal cardiac findings.
In this single-centre study comprising 129 DM1 patients, 55% had cardiac involvement including conduction abnormalities, arrhythmias and/or reduced pump function. A systematic approach to detect cardiac involvement is pivotal as a normal ECG could not rule out abnormal findings on Holter or echocardiography. SCD was reported in almost one third of the families, mainly in first-degree relatives at younger age. During short term follow-up (median 31 months) six patients died, five due to cardiac causes and one suffered from sudden death. We identified a high prevalence of AVB grade I (24%). In comparison, AVB grade I had a prevalence of 1.6% in a large prospective, communitybased cohort study including 7.575 healthy individuals (mean age 47 years, 54% women) and was associated with increased risk of atrial fibrillation, PM-implantation and all-cause mortality [11]. As atrial tachyarrhythmia is an independent predictor of SCD in patients with
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
DM1 [2], the association between AVB grade I and increased risk of atrial fibrillation plays a central role in risk stratification. A PR-interval above 240 ms is also an independent predictor of SCD in DM1 patients [2,12]. In our study, those with a PR-interval above 240 ms all had additional cardiac abnormalities pointing towards a more severe cardiac phenotype. Ultimately, AVB grade I was associated with longer QRS-interval, higher GLS-values and a higher number of VPCs. Taken together, these findings strongly indicate that AVB grade I is essential in risk stratification and should not be considered a normal variant. Progression of AVB in DM1 can be sudden and unpredictable and PM-implantation has proved beneficial in asymptomatic patients with DM1 and evidence of infra-Hisian conduction delay [13]. These findings challenge the perception that AVB grade I has a benign prognosis and emphasizes the need of special attention in patients with DM1. The prevalence of LBBB and RBBB in our cohort was at least 11- and 6-fold higher, respectively, than observed in healthy age-matched individuals [29–32]. RBBB or LBBB should alert the clinician because of the association with all-cause mortality and increased risk of cardiovascular morbidity, and because LBBB seems to be marker of progressing, degenerative cardiac disease affecting both the conduction system and the myocardium [33–35]. Long term studies are needed to document whether LBBB is a predictive marker of heart failure also in patients with DM1. In contrast, IRBBB is not associated with cardiovascular risk factors or adverse outcomes in healthy individuals [33]. Nevertheless, the relatively high prevalence of IRBBB in DM1 (9%) is notable due to the unpredictable progression of conduction disorders in DM1, which is not comparable with healthy individuals. DM1 patients had a 2- to 3-fold higher prevalence of AF/AFL compared to controls despite the significantly lower mean age in the DM1 patients (44 vs. 57 years). This finding is of major importance due to the association between AF/AFL and increased risk of SCD [2]. AF/AFLepisodes (present in 4.1% of our patients) were generally asymptomatic substantiating the need of a systematic cardiac approach to detect and to evaluate whether treatment with anti-arrhythmia or anticoagulants is indicated. In addition, DM1 patients had a higher prevalence of AVB grade II. Overall, DM1-patients did not differ from a significantly older control group, emphasizing the progressive nature of cardiac pathology in DM1 [2,13,36]. The prevalence of reduced pump function (LVSD) was 21%, i.e. seven times higher than in a background population (3%) with a mean age of 50 years [37]. This high prevalence is consistent with prior studies with a prevalence of LVSD in DM1 ranging from 12 to 90%, though different inclusion criteria and imaging techniques make comparisons difficult and may explain the vast range [3,4,17,20,38,39]. LVSD in patients with DM1 is strongly associated with cardiac and all-cause death [2,4]. LVSD was not significantly associated with abnormal findings on ECG or Holter and did not correlate with age as expected, emphasizing that repeated echocardiography is mandatory to detect cardiac involvement also in young patients with DM1. Abnormal GLS above −15.9% was found in 22% of the patients and 12/22 had preserved LVEF. To our knowledge, GLS in patients with DM1 has only been assessed in one previous study, demonstrating a significant correlation between GLS and PR-interval [8]. Accordingly, we found a trend towards a positive correlation between GLS and PRinterval, and significantly higher GLS values in patients with AVB grade I. Abnormal GLS is associated with myocardial fibrosis in patients with e.g. Fabry disease and type II diabetes mellitus [40,41]. Moreover, it is related to all-cause mortality or heart failure admissions in patients with myocardial infarction and preserved LVEF [42]. These findings suggest a link between the conduction disorder and the subclinical abnormal myocardial function but longitudinal studies are needed to evaluate the prognostic value of this parameter. We identified a significant association between cardiac involvement and reduced muscle strength. Generally, the frequency of arrhythmic events and conduction disturbances has been reported to increase
35
with the severity of the muscular disease [2,43–45] but this is not at consistent finding [14,20,46]. Although our findings revealed weaker muscle strength in patients with cardiac involvement, the small difference indicates that the degree of peripheral muscular involvement cannot confidently be used to identify patients at increased risk of cardiac events, which strongly emphasizes the need of a systematic cardiac approach irrespective of the degree of peripheral muscular involvement. Several studies have assessed the association between CTG repeat length and cardiac involvement with ambiguous results [3,16]. A recent study revealed a large inter- and intra-tissue CTG length variation in adult DM1 tissue with the largest expansion in the heart and cerebral cortex [47]. On this basis we omitted analyses of CTG-repeat length in our study. 4.1. Strengths and limitations To our knowledge, this study is the largest single-centre study evaluating patients with DM1 systematically and concomitantly with family-history, physical examination, ECG, Holter-monitoring, echocardiography and muscle strength testing. As this is a single-centre study, our results may not be generalizable. However, we invited all patients with DM1 in eastern Denmark to participate, independent of cardiac symptoms or signs of cardiac involvement. Despite the large difference in age between DM1 patients and controls, we observed a higher prevalence of specific cardiac abnormalities in DM1, adding important knowledge to the literature concerning the progression of cardiac involvement in DM1. So far, recommendations for cardiac work-up in patients with DM1 are sparse. We recognize that long-term follow-up including cardiac assessment with ECG, Holter-monitoring and echocardiography is necessary to establish optimal guidelines for cardiac work-up and risk stratification for SCD. Nevertheless, we believe that our results, together with the existing literature and current guidelines, are sufficient to cautiously recommend proposals for cardiac work-up. 5. Proposals for cardiac work-up in patients with DM1 Based on our findings that 1) conduction abnormalities and 2) arrhythmias are the most frequent cardiac abnormalities in DM1 and that 3) reduced LVEF is prevalent and not age-dependent, cardiac screening, treatment and follow-up should focus on these three aspects. In our study, a normal ECG could not rule out the existence of abnormal findings on Holter-monitoring or echocardiography. Therefore, all three aspects need to be assessed systematically and we suggest the following recommendations: 1. Recommendations for cardiac evaluation at time of diagnosis Risk-reduction of SCD by timely cardiac treatment necessitates cardiac screening. ECG, 48-hour Holter-monitoring and echocardiography at time of diagnosis seems appropriate. Patients need to be educated in seeking medical advice in case of heart specific symptoms and understand the relevance of repeated cardiac assessment due to the generally progressive nature of the disease. 2. Recommendations for cardiac follow-up We found that conduction abnormalities and arrhythmias are the most frequent findings. No SCDs or deaths from cardiac causes were observed during short-term follow-up (median 31 months). Taken together, and in agreement with the Danish National Guidelines (Danish Society of Cardiology), we consider it appropriate to suggest repeated cardiac assessment with ECG and 48-hour Holter-monitoring every second year and additionally echocardiography every fourth year in asymptomatic patients with no cardiac abnormalities. However, longterm studies are indeed needed to define the optimal timely follow-up. Special attention is needed in patients with symptoms such as syncope or specific predictors of SCD as defined by Groh and co-authors
36
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
[2]: rhythm other than sinus, PR-interval ≥240 ms, QRS ≥120 ms, second/third degree AVB) and/or a diagnosis with atrial tachyarrhythmia. Such findings necessitate more frequent and more thorough follow-up including e.g. extended Holter-monitoring, implantable loop-recorders and in selected cases invasive electrophysiological studies. Higher survival rates and lower incidence of sudden death has been reported in DM1 patients with conduction disease using an invasive strategy based on systematic electrophysiological studies and prophylactic permanent pacing compared with a non-invasive strategy [48]. If syncopes are associated with tachy- and/or bradyarrhythmia, PM/ ICD-implantation is appropriate. Medical treatment of supraventricular or ventricular arrhythmias does not differ from routine guidelines. In accordance with the 2012 ACCF/AHA/HRS “Focused Update concerning the 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities”, PM-implantation in patients with neuromuscular disease is considered a class IB indication for patients with AVB grade III and advanced AVB grade II, and a class IIB indication for any degree of AVB (including AVB grade I) [49]. These recommendations are regardless of symptoms because there may be unpredictable progression of AV conduction disease. ICDs should be considered rather than PMs due to the increased risk of VT/VF and SCD in accordance with the recommendations from a large longitudinal study by Bhakta and co-authors [9]. The indications for ICD-implantation are similar to the indications in other patients. Heart-failure should be managed according to current guidelines [50]. We acknowledge that treatment of echocardiographic abnormalities has so far not proven any benefit among patients with DM1 — to document this would necessitate a randomized clinical trial. Nevertheless, we have no evidence or reason to believe that treatment of structural cardiac manifestations should respond differently to treatment in patients with DM1 compared to other patients with such manifestations. 6. Conclusion This single-centre study documents the complete cardiac phenotype in patients with DM1. Conduction abnormalities, arrhythmias or reduced cardiac pump function were observed in more than 50% of the patients. The additive value of this study is that a systematic cardiac assessment including ECG, Holter-monitoring and echocardiography is useful in order to make a proper characterization of cardiac involvement in DM1. The unpredictable progressive nature of cardiac abnormalities in DM1 must be considered in the clinical work-up. Cardiac symptoms and manifestations may be minimized by early intervention, which necessitates repeated, systematic cardiac screening and follow-up programs. In asymptomatic patients without cardiac involvement followup at least every second year seems appropriate but long-term followup is needed before more accurate recommendations and guidelines for DM1-specific risk stratification can be established. Funding We thank The Research Foundation of Rigshospitalet (ref. nr. R40A1247), The Danish Heart Foundation (ref. nr. 11-04-R83-A335322622), The Lundbeck Foundation (ref. nr. R83-A7909) and The Stibo Foundation for the financial support.
References [1] Brook JD, McCurrach ME, Harley HG, et al. Cell Feb 21 1992;68(4):799–808. [2] Groh WJ, Groh MR, Saha C, et al. N Engl J Med Jun 19 2008;358(25):2688–97. [3] Petri H, Vissing J, Witting N, Bundgaard H, Kober L. Int J Cardiol 2012 Oct 4;160(2): 82–8. [4] Bhakta D, Groh MR, Shen C, Pascuzzi RM, Groh WJ. Am Heart J Dec 2010;160(6): 1137–41 (1141). [5] Dhand UK, Raja F, Aggarwal K. Neurol Int Feb 11 2013;5(1):e5. [6] Magri D, Piccirillo G, Bucci E, et al. Int J Cardiol Nov 2010;26. [7] Morner S, Lindqvist P, Mellberg C, et al. J Intern Med Jul 2010;268(1):59–65. [8] Wahbi K, Ederhy S, Becane HM, et al. Int J Cardiol Nov 3 2011;152(3):375–6. [9] Bhakta D, Shen C, Kron J, et al. J Cardiovasc Electrophysiol Dec 2011;22(12): 1369–75. [10] Breton R, Mathieu J. Can J Cardiol Feb 2009;25(2):e23–7. [11] Cheng S, Keyes MJ, Larson MG, et al. JAMA Jun 24 2009;301(24):2571–7. [12] Colleran JA, Hawley RJ, Pinnow EE, Kokkinos PF, Fletcher RD. Am J Cardiol Dec 1 1997;80(11):1494–7. [13] Lazarus A, Varin J, Babuty D, et al. J Am Coll Cardiol Nov 6 2002;40(9):1645–52. [14] Mammarella A, Paradiso M, Antonini G, et al. Adv Ther Sep 2000;17(5):238–51. [15] Mathieu J, Allard P, Potvin L, Prevost C, Begin P. Neurology May 12 1999;52(8): 1658–62. [16] Sansone VA, Brigonzi E, Schoser B, et al. Int J Cardiol Dec 2012;22. [17] Bhakta D, Lowe MR, Groh WJ. Am Heart J Feb 2004;147(2):224–7. [18] Phillips MF, Harper PS. Cardiovasc Res Jan 1997;33(1):13–22. [19] Nguyen HH, Wolfe III JT, Holmes Jr DR, Edwards WD. J Am Coll Cardiol Mar 1988; 11(3):662–71. [20] De AL, Raisaro A, Marchiano V, Radice S, Meola G. Eur Heart J Jul 1995;16(7): 1007–10. [21] Antonini G, Giubilei F, Mammarella A, et al. Neurology Oct 24 2000;55(8):1207–9. [22] Sajadieh A, Nielsen OW, Rasmussen V, Hein HO, Hansen JF. J Intern Med Oct 2006; 260(4):377–87. [23] Warner JP, Barron LH, Goudie D, et al. J Med Genet Dec 1996;33(12):1022–6. [24] Lang RM, Bierig M, Devereux RB, et al. J Am Soc Echocardiogr Dec 2005;18(12): 1440–63. [25] Nagueh SF, Appleton CP, Gillebert TC, et al. J Am Soc Echocardiogr Feb 2009;22(2): 107–33. [26] Yingchoncharoen T, Agarwal S, Popovic ZB, Marwick TH. J Am Soc Echocardiogr Feb 2013;26(2):185–91. [27] Sajadieh A, Nielsen OW, Rasmussen V, et al. Eur Heart J Mar 2004;25(5):363–70. [28] Christensen AH, Bundgaard H, Schwartz M, Hansen SH, Svendsen JH. Circ Arrhythm Electrophysiol Oct 2008;1(4):317–20. [29] De BD, De BG, Kornitzer M. Heart Dec 2000;84(6):625–33. [30] Hingorani P, Natekar M, Deshmukh S, et al. Indian J Med Res Mar 2012;135:322–30. [31] HISS RG, LAMB LE. Circulation 1962 Jun;25:947–61. [32] Pelliccia A, Culasso F, Di Paolo FM, et al. Eur Heart J Aug 2007;28(16):2006–10. [33] Bussink BE, Holst AG, Jespersen L, et al. Eur Heart J Jan 2013;34(2):138–46. [34] Schneider JF, Thomas Jr HE, Kreger BE, McNamara PM, Kannel WB. Ann Intern Med Mar 1979;90(3):303–10. [35] Zannad F, Huvelle E, Dickstein K, et al. Eur J Heart Fail Jan 2007;9(1):7–14. [36] Nazarian S, Wagner KR, Caffo BS, Tomaselli GF. Pacing Clin Electrophysiol Feb 2011; 34(2):171–6. [37] Nielsen OW, Hilden J, Larsen CT, Hansen JF. Heart Aug 2001;86(2):172–8. [38] Hartwig GB, Rao KR, Radoff FM, et al. Neurology May 1983;33(5):657–60. [39] Tokgozoglu LS, Ashizawa T, Pacifico A, et al. JAMA Sep 13 1995;274(10):813–9. [40] Kosmala W, Przewlocka-Kosmala M, Szczepanik-Osadnik H, et al. JACC Cardiovasc Imaging Dec 2011;4(12):1239–49. [41] Kramer J, Niemann M, Liu D, et al. Eur Heart J Jun 2013;34(21):1587–96. [42] Ersboll M, Valeur N, Mogensen UM, et al. J Am Coll Cardiol Jun 11 2013;61(23): 2365–73. [43] Cudia P, Bernasconi P, Chiodelli R, et al. J Neurol Neurosurg Psychiatry Jul 2009; 80(7):790–3. [44] Hawley RJ, Milner MR, Gottdiener JS, Cohen A. Neurology Feb 1991;41(2 (Pt 1)): 259–62. [45] Nishioka SA, Martinelli FM, Marie S, Zatz M, Costa R. Arq Bras Cardiol Apr 2005; 84(4):330–6. [46] Florek RC, Triffon DW, Mann DE, Ringel SP, Reiter MJ. West J Med Jul 1990;153(1): 24–7. [47] Lopez CA, Nakamori M, Tome S, et al. Hum Mol Genet Jan 1 2011;20(1):1–15. [48] Wahbi K, Meune C, Porcher R, et al. JAMA Mar 28 2012;307(12):1292–301. [49] Epstein AE, DiMarco JP, Ellenbogen KA, et al. J Am Coll Cardiol Jan 22 2013;61(3): e6-75. [50] Hunt SA, Abraham WT, Chin MH, et al. J Am Coll Cardiol Apr 14 2009;53(15):e1-90.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
High prevalence of cardiac involvement in patients with myotonic dystrophy type 1: A cross-sectional study Helle Petri a,⁎,1, Nanna Witting c,1, Mads Kristian Ersbøll a,1, Ahmad Sajadieh e,1, Morten Dunø d,1, Susanne Helweg-Larsen c,1, John Vissing c,1, Lars Køber a,1, Henning Bundgaard b,1 a
Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark Unit for Inherited Cardiac Diseases, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark d Department of Clinical Genetics, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark e Department of Cardiology, University Hospital of Bispebjerg, Copenhagen, Denmark b c
a r t i c l e
i n f o
Article history: Received 29 October 2013 Received in revised form 10 March 2014 Accepted 14 March 2014 Available online 20 March 2014 Keywords: Myotonic dystrophy Arrhythmia Conduction disorders Sudden cardiac death
a b s t r a c t Background: Patients with myotonic dystrophy type 1 (DM1) have a three-fold higher risk of sudden cardiac death (SCD) than age-matched healthy controls. Despite numerous attempts to define the cardiac phenotype and natural history, existing literature suffers from low power, selection-bias and lack of controls. Thus, the optimal strategy for assessing cardiac involvement in DM1 is unclear. Method: In this large single-centre study, we evaluated 129 unselected DM1 patients (49.6% men), mean (SD) age 44 (14.7) years with family history, physical examination, electrocardiogram (ECG), echocardiography, Holtermonitoring and muscle strength testing. Results: Cardiac involvement was found in 71 patients (55%) and included: 1) Conduction abnormalities: atrioventricular block grade I (AVB grade I) (23.6%), AVB grade II (5.6%), right/left bundle branch block (5.5/3.2%) and prolonged QTc (7.2%); 2) arrhythmias: atrial fibrillation/flutter (4.1%), other supraventricular tachyarrhythmia (7.3%) and non-sustained ventricular tachycardia (4.1%); and 3) structural abnormalities: left ventricular systolic dysfunction (20.6%) and reduced global longitudinal strain (21.7%). A normal ECG was not significantly associated with normal findings on Holter-monitoring or echocardiography. Patients with abnormal cardiac findings had weaker muscle strength than those with normal cardiac findings: ankle dorsal flexion (median (range) 4.5 (0–5) vs. 5.0 (2.5–5), p = 0.004) and handgrip (median 4.0 (0–5) vs. 4.50 (2–5), p = 0.02). Conclusion: The cardiac phenotype of DM1 includes a high prevalence of conduction disorders, arrhythmias and risk factors of SCD. Systematic cardiac screening with ECG, Holter-monitoring and echocardiography is needed in order to make a proper characterization of cardiac involvement in DM1. © 2014 Elsevier Ireland Ltd. All right reserved.
1. Introduction Myotonic dystrophy type 1 (DM1) is an autosomal dominantly inherited neuromuscular disorder caused by an unstable expansion of a tri-nucleotide (CTG) repeat on chromosome 19 in the 3′ untranslated region of the myotonic dystrophy protein kinase gene [1]. Cardiac involvement in patients with DM1 is a major concern and includes an increased risk of conduction disturbances, arrhythmias, compromised systolic and diastolic function and sudden cardiac death
⁎ Corresponding author at: Department of Cardiology, Rigshospitalet, Copenhagen University Hospital, Blegdamsvej 9, 2100 Copenhagen, Denmark. Tel.: +45 28 26 44 32; fax: +45 35 45 77 05. E-mail address: [email protected] (H. Petri). 1 This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.
http://dx.doi.org/10.1016/j.ijcard.2014.03.088 0167-5273/© 2014 Elsevier Ireland Ltd. All right reserved.
(SCD) [2,3]. Nevertheless, cardiac involvement is mainly described in smaller studies with low power or limited to studies of electrocardiographic or studies of echocardiographic abnormalities [2–8]. This restricted focus also applies for the existing longitudinal studies [3,9–16] e.g. the study by Groh and co-authors investigating specific ECGpredictors of SCD [2]. The cardiac phenotype of DM1 is complex with unpredictable progression and is not necessarily correlated with the severity of neuromuscular involvement [2,3]. The cardiac conduction disorders are probably caused by myocardial fibrosis, fat infiltration and hypertrophy frequently identified in autopsies from patients with DM1 [17–19]. These changes may also be a substrate for supraventricular and ventricular arrhythmias and also play a key role in the development of the observed systolic dysfunction [4,17,20]. Patients with DM1 rarely report cardiac symptoms despite cardiac involvement. This may mainly be due to the reduced cardiac demand caused by the impaired skeletal muscle function. Consequently,
32
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
arrhythmias may remain unnoticed e.g. atrial tachyarrhythmia, which is an independent predictor of SCD [2,15,21]. Early detection of cardiac involvement is pivotal in order to optimize early intervention which may consequently reduce cardiac symptoms and manifestations. In this single-centre study, we performed a systematic cardiac assessment including cardiac symptoms, ECG, Holter-monitoring and echocardiography. With this systematic approach, we aimed at investigating the complete cardiac phenotype of DM1 and the association between abnormal findings on each of the used modalities. Secondly, short term follow-up was performed exclusively for selected major events (cardiac and all-cause mortality) to assess if a possible minimum time between cardiac follow-up could be recommended. 2. Methods 2.1. Study design The study was conducted at the Department of Cardiology and the Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark and approved by the regional scientific ethics committee (reference number H-d-2008-077). Patients were evaluated by family-history, physical examination, 12-lead ECG, transthoracic echocardiography, 48-hour ECG-monitoring (Holter-monitoring) and muscle strength testing. Blood samples were analyzed for plasma levels of NT-proBNP, creatinine kinase (CK) and myoglobin and screened for liver, renal and thyroid diseases. All bloodsamples were analyzed in the same laboratory and results were adjusted for age and gender, when appropriate. SCDs in relatives below or above the age of 50 years were registered. Lastly, we collected short-term follow-up data exclusively regarding selected major events (cardiac and all-cause mortality) and registered number and cause of deaths. 2.2. Study population All adult patients, age ≥18 years, with genetically confirmed DM1 were invited to participate and enrolled from December 2010 to December 2012. None of the patients were selected according to age, gender or cardiac symptoms and those included provided written informed consent. A thorough telephone interview was performed in those patients who did not want to participate. Patients were asked about cardiac symptoms and informed about the necessity of seeking medical advice in case of heart specific symptoms. Holter-parameters were compared to a healthy control group from The Copenhagen Holter study [22]. These controls were enrolled from April 1998 to June 2000 and enrolment was not designed to specifically match the DM1 cohort. 2.3. Genetic testing Genetic testing had been performed earlier as part of the diagnostic work-up and the heterozygous presence of an abnormally expanded CTG repeat allele was confirmed either by Southern-blot analysis and/or by triplet repeat primed polymerase chain reaction [1, 23]. 2.4. Electrocardiography A 12 lead ECG was performed using a Burdick Atria 6100 ECG. An ECG was considered abnormal in the presence of: atrial flutter/fibrillation (AFL/AF), other supraventricular arrhythmia including atrial and re-entry tachycardia, atrio-ventricular block (AVB) grades I–III (AVB I: PR-interval N220 ms), right and left bundle branch block (RBBB/LBBB), incomplete right bundle branch block (IRBBB) and prolonged QTc (N450 ms in men and N470 ms in women) using Bazett's formula (QTc = QT/√RR). Additionally, we assessed the presence of cardiac abnormalities separately in patients with PR-interval N 240 ms. All ECGs were analyzed by one investigator (HP) and in case of uncertainty discussed with other members of the study group.
(ASE) [24]. Left ventricular systolic dysfunction (LVSD) was defined as a biplane left ventricular ejection fraction (LVEF) ≤50%. Doppler recordings of mitral inflow were performed by placing a 2.5 mm sample volume at the tip of the MV leaflets during diastole and recording the pulsed-waved Doppler signal. Peak velocities of early (E) and atrial (A) diastolic filling and MV deceleration time were measured and the E/A ratio was calculated. Continuous-wave Doppler recordings of the LV outflow tract were obtained and aortic valve opening and closure times were measured. Pulsed-wave Doppler tissue imaging recordings were performed at the lateral and medial mitral annulus with measurements of myocardial peak early (E′). The E/E′ ratio was calculated from the lateral values of E/E′. Diastolic dysfunction was assessed and graded in accordance with ASE's recommendations based on the following parameters: e′ lateral, LA volume, MV deceleration time, E/A ratio and E/e′ [25]. 2.6. Left ventricular longitudinal strain analysis LV longitudinal function was assessed by global longitudinal strain (GLS) using a semiautomatic algorithm (Automated Functional Imaging (AFI); GE). Briefly, manual positioning of three points was performed in each of the three apical projections, enabling the software to semi-automatically track the myocardium throughout the heart cycles. Careful inspection of tracking and manual correction, if needed, was performed and in case of unsatisfactory tracking the segment was excluded from speckle tracking analysis. The algorithm then calculated the overall GLS as the average value of all three projections. Normal values of GLS between −22.1% and −15.9% (mean, −19.7%; 95% CI (−20.4% to −18.9%)) have been reported previously [26]. Abnormal GLS was defined as GLS above −15.9%. 2.7. Holter-monitoring A 48-hour Holter-monitoring was performed with a 3-electrode Lifecard CF (Spacelabs Healthcare). Holter-monitoring was considered abnormal in the presence of; AVB grades I–III, AF/AFL, other supraventricular tachyarrhythmia (N30 SVES/h or runs of ≥20 SVES), frequent VPCs (≥30/h) and non-sustained VT (NSVT) (minimum of 3 beats at ≥100 bpm). Recordings were evaluated by a single observer (HP) and discussed in the study group in case of uncertainty. Holter-parameters were compared to a healthy control group comprising 285 healthy individuals (74.4% men, mean (SD) age 57.6 (2.5) years). This control-group and the method used for the evaluation of Holter-results have been previously described in detail [27]. 2.8. Cardiac involvement and neuromuscular affection Muscle strength was graded from 0 to 5 using the Medical Research Council scale (MRC) (0 = no ability to contract muscle, 5 = normal strength). In patients with DM1, early affection occurs primarily in the distal muscles. Therefore, we investigated the association between handgrip (dominant hand) and ankle dorsal flexion and abnormal cardiac findings. Neurological assessments were performed by experienced neurologists (JV, NW, SHL). 2.9. Statistics Data were analyzed with IBM SPSS Statistics version 19. A p-value ≤ 0.05 were considered statistically significant. Normally distributed values are expressed as means ± SD. Data with skewed distribution is given as median (range). Categorical variables were summarized by frequency counts (percentage) and differences among groups were evaluated using chi-square test. Results of continuous variables are presented as median (range) and comparisons between categories were made with Mann-Whitney U test. Correlation analyses were performed using Spearman Correlation. The prevalence of any given parameter was first assessed in the total study cohort. Secondly, we corrected all parameters for familiar relations using chi-square test, comparing the prevalence of any given parameter from all included patients vs. the prevalence of the equivalent parameter from the oldest representative in each family.
3. Results 2.5. Echocardiography Echocardiography was performed using a Vivid e9 (General Electric, Horten, Norway) and the examinations were obtained and analyzed by a single operator/observer (HP). The examinations were discussed with other members of the study group in case of uncertainty or suboptimal image quality. Three consecutive heart cycles were recorded and images were obtained at a frame rate of ≥60 frames/s and analyzed with EchoPac BT 11.1.0; General Electric, Horten, Norway. Two-dimensional parasternal images were used to determine left ventricular (LV) cavity dimensions and wall thickness. LV volumes were determined using the biplane Simpson model. Left atrial volume was calculated from the biplane area-length method with maximum volume before mitral valve (MV) opening (LAmax) and minimum volume just before MV closure (LAmin). Volumetric and dimensional measurements of the left ventricle and left atrium were indexed to body surface area when appropriate. All volumetric analyses were performed in accordance with the recommendations from the European Association of Echocardiography and the American Society of Echocardiography
3.1. Study population We invited 171 patients with DM1 to participate in this study. Fortytwo patients did not participate: two due to severe neuromuscular involvement and “lacking energy”, 25 due to lack of interest in scientific research projects and long distances to our hospital, 13 did not respond to repeated invitations and two died of non-cardiac causes prior to inclusion. A thorough telephone interview regarding cardiac symptoms was performed in the 27 patients who refused to participate and none of the patients had any cardiac complaints. Thus, we included 129 patients with genetically confirmed DM1 from 73 families (64 men (49.6%), mean (SD) age 44.0 (14.7) years. Gender specific
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
demographics are presented in Table 1. Blood-samples revealed no signs of liver, renal or thyroid disease. The most predominant complaint in our patients was mental and/or physical fatigue: n = 77 (59.7%). Possible cardiac-specific symptoms were only present in few patients including: palpitations (n = 11 (8.5%)), blurred vision (n = 10 (7.8%)), near-syncope and syncope (n = 4 (3.1%)), dyspnea (n = 9, (7.1%)), chest pain (n = 6 (4.7%)) and peripheral oedema (n = 4 (3.1%)). 3.2. Electrocardiography
33
Table 2 Abnormal ECG- and echocardiographic findings in patients with DM1.
ECGa AVB grade I IRBBB RBBB LBBB Prolonged QTc
n (%)
Men (n)
Women (n)
p-Value
30 (23.6) 11 (8.8) 7 (5.5) 4 (3.2) 9 (7.2)
14 7 1 1 5
16 4 6 3 4
0.68 0.53 0.06 0.37 1.00
n (%)
ECG was available in 127/129 patients. The most predominant conduction disorder was AVB grade I (PR-interval N220 ms) (n = 30 (23.6%)) (Table 2). AVB grade I was associated with: longer QRSinterval (median (range) 119 ms (80–200) vs. 90 ms (70–175), p b 0.001), higher GLS-values (median (range) − 16.9% (− 24 to − 10) vs. − 19.0% (− 24 to − 12), p = 0.04) and a higher number of VPCs/48 h (median (range) 20 (0–6724) vs. 4 (0–20,220), p = 0.023). Additionally, a PR-interval N 240 ms was present in 7/30 (23.3%) and these patients all had additional cardiac abnormalities, such as reduced LVEF, RBBB/LBBB, AVB grade II, frequent VES or NSVT. A normal ECG was not significantly associated with normal findings on 1) Holter-monitoring (i.e. 12/47 patients had concomitant abnormal ECG and Holter-monitoring while 9/76 patients with normal ECG had abnormal Holter-monitoring, p = 0.08), or 2) echocardiography (i.e. 12/48 had concomitant abnormal ECG and echocardiography while 16/78 patients with abnormal echocardiography had normal ECG, p = 0.66). PR-interval was associated with age despite the low mean age in our cohort (Spearman correlation, r = 0.32, p b 0.001). 3.3. Echocardiography Echocardiography was performed in 126/129, and speckle tracking analysis in 93/126 patients. LVSD was found in 26 patients (20.6%) with a significantly higher prevalence in men (Table 2). LVSD was not associated with abnormal findings on 1) ECG (i.e. 12/26 patients had concomitant LVSD and abnormal ECG while 36/100 patients with normal LVEF had abnormal ECG, p = 0.37), or 2) Holter-monitoring (i.e. 6/23 patients had
Echocardiographyb LVEF ≤ 50% 26 (20.6) IVSD N 11 mm 14 (11.4) LVIDD N 55 mm 6 (4.9) 2 LAEDVi N 34 ml/m 9 (8.3) GLS N −15.9% 20 (21.7)
Median (range)
Men (n)
Women (n)
p-value
48 (33–50) 13 (11–17) 57 (56–60) 36 (35–40) −14 (−15 to −10)
18 9 6 7 13
8 5 0 2 7
0.05 0.27 0.01 0.09 0.08
GLS: global longitudinal strain (available in 92 patients). IVSD: interventricular septum in diastole. IRBBB: incomplete right bundle branch block. LAEDVi: left atrial end diastolic volume indexed. LBBB/RBBB: left/right bundle branch block. LVEF: left ventricular ejection fraction. LVIDD: left ventricular internal diameter in diastole. a ECG was available in 127 patients. b Echocardiography was performed in 125 patients. Parameters were not available in all patients; percent is calculated from the actual number of observations.
concomitant LVSD and abnormal Holter-monitoring while 15/99 patients with normal LVEF had abnormal Holter-monitoring, p = 0.23). LVEF did not correlate with age (Spearman correlation r = 0.09, p = 0.30). Nine patients (8.3%) had abnormal LAEDVi N34 ml/m2 and six (4.8%) had diastolic dysfunction (grade 1 (n = 3), grade 2 (n = 3). No major valvular abnormalities were observed. A large proportion of patients (n = 20 (21.7%)) had abnormal GLS above − 15.9% and 12/20 had preserved LVEF N50% (Table 2). There was a trend towards a positive correlation between GLS and PRinterval (Spearman correlation r = 0.183, p = 0.08). 3.4. Holter-monitoring
Table 1 Clinical characteristics in patients with DM1.
Patients, n Age, mean (SD) BPs mm Hg, mean (SD) BPd mm Hg, mean (SD) NT-proBNP (pmol/l), median (range) CK (U/l) , median (range) Myoglobin (μg/l), median (range) Muscle strength sum-score, median (range)b Increased NT-proBNP, n (%), median (range)c Increased CK, n (%), median (range)c Increased myoglobin, n (%), median (range)c
DM1 cohorta
Men
129 44 (15) 125 (19) 77 (13) 9 (1–76)
64 (50) 44 (16) 125 (17) 77 (13) 6 (1–76)
65 (50) 44 (14) 125 (21) 76 (13) 13 (1–44)
1.0 0.78 0.47 0.55 b0.001
164 (38–952) 83 (27–258) 9 (1–10)
182 (56–780) 106 (32–258) 9 (1–10)
131 (38–952) 71 (27–190)
0.007
9 (2–10)
0.72
14 (13)
18 (11–41)
29 (16–44)
0.44
24 (21) 76(66)
455 (312–780) 118 (82–258)
Women
318 (250–952) 86 (50–190)
p-Value
0.004
0.01 b0.001
BPs/BPd: blood pressure systolic/diastolic. CK: creatine kinase. a Not all parameters were obtainable in all patients; percent is calculated from the actual number of observations. b Muscle strength sum-score for ankle dorsal flexion and handgrip (MRC-score). c Increased value after adjustment for age and gender.
Holter-monitoring was performed in 122/129 patients and parameters compared to a healthy control group (285 individuals (74.4% men, mean (SD) age 57.6 (2.5) years). Patients with DM1 had a significantly higher prevalence of AVB grade II than the total control-group (7/122 vs. 4/285, p = 0.02). We found no difference between DM1 patients and controls concerning: AF/AFL (5/122 vs. 4/285, p = 0.13); SVT-episodes (9/122 vs. 24/285, p = 0.68); frequent VPC (7/122 vs. 17/285, p = 1.00) and NSVT (5/ 122 vs. 21/285, p = 0.27) (Table 3). The observed arrhythmias were generally asymptomatic except in a 21-year-old man with AF/AFL and palpitations during exercise and in a 63-year-old woman with dyspnea, Table 3 Holter-findings in patients with DM1 and healthy controls.
Age, mean (SD) (years) AVB grade II, n (%) AF/AFL, n (%) SVT, n (%) VPC N30/h, n (%) NSVT, n (%)
DM1 (n = 122)
Controls (n = 285)
p-Value
44 (14.7) 7 (5.6) 5 (4.1) 9 (7.3) 7 (5.8) 5 (4.1)
58 (2.5) 4 (1.4) 4 (1.4) 24 (8.4) 17 (6.0) 21 (7.3)
b0.001 0.02 0.13 0.68 1.00 0.27
AF/AFL: atrial fibrillation/flutter. AVB grade I/II: atrio-ventricular block grade I/II. NSVT: non-sustained ventricular tachycardia. SVT: supraventricular tachycardia. VPC: ventricular premature contractions.
34
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
which might as well have been attributed to her reduced physical ability. The prevalence of arrhythmias was independent of gender. All analyses in this study showed similar results after correction for familiar relation. 3.5. Cardiac involvement and neuromuscular affection Overall, 71 patients (55%) had abnormal findings on ECG, Holtermonitoring and/or echocardiography. Patients with abnormal findings had a lower MRC-score on ankle dorsal flexion (median (range) 4.5 (0–5) vs. 5.0 (2.5–5), p = 0.004) and handgrip (median 4.0 (0–5) vs. 4.50 [2–5], p = 0.02) compared to patients with normal cardiac findings (Fig. 1). 3.6. Implantable cardiac devices Five patients (3.9%) had an ICD at study inclusion and were all diagnosed with DM1 prior to ICD-implantation: three family-members (a father and his son and sister: 51, 18 and 45-year-old, respectively) had prophylactic ICDs due to the SCD of the father's 11-year-old daughter, who was diagnosed with DM1 post-mortem [28]. A 45-year-old
p = 0.004
A
Ankle dorsal flexion (MRC)
3.7. Invasive procedures Radiofrequency ablation for AFL was performed in a 45-year-old man. The patient had AFL on initial ECG and Holter-monitoring, abnormal NT-proBNP (40.6 pmol/l) and normal echocardiography. After isthmus-ablation, the patient had sinus rhythm with AVB grade I and episodes with AVB grade II (type 1). At latest follow-up (21 months after ablation), Holter-results showed episodes with atrial tachycardia and the patient had received a PM. Invasive electrophysiological evaluation was performed in a 21-year-old man with palpitations during exercise, IRBBB and episodes with AF/AFL. Atypical AFL and atrial tachycardia were induced and terminated with burst-pacing. There was no indication for ablation. 3.8. Family history and sudden cardiac death Family history revealed a total of 37 patients (28.7%) with SCDs in relatives, i.e. SCD in 20 families out of a total of 73 families (27.4%). Suspected SCD below or above the age of 50 years accounted for 25 (67.6%) and 12 (32.4%) cases, respectively. The majority of SCDs (22/37 (64.7%)) were in first degree relatives. Patients with SCD in relatives did not have a higher prevalence of abnormal cardiac examinations compared to the remaining patients (23/37 vs. 48/92, p = 0.33).
5
4
3
2
3.9. Short term follow-up for cardiac and all-cause mortality
1
During short-term follow-up regarding mortality (median (range) 31 (15–51) months), six patients died (4 men (age 31, 45, 53 and 77 years) and 2 women (age 56 and 59 years). Four of these patients had cardiac involvement at inclusion: the 55-year-old woman had frequent VES, the 53-year-old man had abnormal GLS (− 13%) and a family-history of SCD, the 31-year-old man had RBBB and the 45-year-old man had a PR-interval of 250 ms and LVEF 45%. The lastmentioned patient was hospitalized due to manifest heart failure shortly before he died, where he was optimized in heart failure treatment but failed to comply this treatment. He suffered from sudden death few months after his last admission. The remaining patients died of noncardiac causes (three patients due to sepsis (primary pulmonary or urinary infection), one patient due to chronic pancreatitis and one due to abdominal cancer).
0
Normal ECG, Holter and TTE
B
Abnormal ECG, Holter and/or TTE p = 0.02
5
4
Handgrip (MRC)
woman had an ICD due to syncope, AVB grade II (type II), RBBB, left posterior fascicular block and prolonged QTc (519 ms). The patient also carried a previously described disease-causing KCNE1-mutation (D76N, c.226GNA) associated with prolonged QTc-interval. A 51-year-old woman had a prophylactic ICD due to a highly malignant family history of SCD. A 19-year-old woman had an implantable loop recorder (ILR) due to syncopes. None of the patients had a pacemaker (PM).
4. Discussion
3
2
1
0
Normal ECG, Holter and TTE
Abnormal ECG, Holter and/or TTE
Fig. 1. Cardiac involvement and peripheral muscle strength. Patients with abnormal cardiac findings had significantly weaker muscle strength compared to patients with normal cardiac findings.
In this single-centre study comprising 129 DM1 patients, 55% had cardiac involvement including conduction abnormalities, arrhythmias and/or reduced pump function. A systematic approach to detect cardiac involvement is pivotal as a normal ECG could not rule out abnormal findings on Holter or echocardiography. SCD was reported in almost one third of the families, mainly in first-degree relatives at younger age. During short term follow-up (median 31 months) six patients died, five due to cardiac causes and one suffered from sudden death. We identified a high prevalence of AVB grade I (24%). In comparison, AVB grade I had a prevalence of 1.6% in a large prospective, communitybased cohort study including 7.575 healthy individuals (mean age 47 years, 54% women) and was associated with increased risk of atrial fibrillation, PM-implantation and all-cause mortality [11]. As atrial tachyarrhythmia is an independent predictor of SCD in patients with
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
DM1 [2], the association between AVB grade I and increased risk of atrial fibrillation plays a central role in risk stratification. A PR-interval above 240 ms is also an independent predictor of SCD in DM1 patients [2,12]. In our study, those with a PR-interval above 240 ms all had additional cardiac abnormalities pointing towards a more severe cardiac phenotype. Ultimately, AVB grade I was associated with longer QRS-interval, higher GLS-values and a higher number of VPCs. Taken together, these findings strongly indicate that AVB grade I is essential in risk stratification and should not be considered a normal variant. Progression of AVB in DM1 can be sudden and unpredictable and PM-implantation has proved beneficial in asymptomatic patients with DM1 and evidence of infra-Hisian conduction delay [13]. These findings challenge the perception that AVB grade I has a benign prognosis and emphasizes the need of special attention in patients with DM1. The prevalence of LBBB and RBBB in our cohort was at least 11- and 6-fold higher, respectively, than observed in healthy age-matched individuals [29–32]. RBBB or LBBB should alert the clinician because of the association with all-cause mortality and increased risk of cardiovascular morbidity, and because LBBB seems to be marker of progressing, degenerative cardiac disease affecting both the conduction system and the myocardium [33–35]. Long term studies are needed to document whether LBBB is a predictive marker of heart failure also in patients with DM1. In contrast, IRBBB is not associated with cardiovascular risk factors or adverse outcomes in healthy individuals [33]. Nevertheless, the relatively high prevalence of IRBBB in DM1 (9%) is notable due to the unpredictable progression of conduction disorders in DM1, which is not comparable with healthy individuals. DM1 patients had a 2- to 3-fold higher prevalence of AF/AFL compared to controls despite the significantly lower mean age in the DM1 patients (44 vs. 57 years). This finding is of major importance due to the association between AF/AFL and increased risk of SCD [2]. AF/AFLepisodes (present in 4.1% of our patients) were generally asymptomatic substantiating the need of a systematic cardiac approach to detect and to evaluate whether treatment with anti-arrhythmia or anticoagulants is indicated. In addition, DM1 patients had a higher prevalence of AVB grade II. Overall, DM1-patients did not differ from a significantly older control group, emphasizing the progressive nature of cardiac pathology in DM1 [2,13,36]. The prevalence of reduced pump function (LVSD) was 21%, i.e. seven times higher than in a background population (3%) with a mean age of 50 years [37]. This high prevalence is consistent with prior studies with a prevalence of LVSD in DM1 ranging from 12 to 90%, though different inclusion criteria and imaging techniques make comparisons difficult and may explain the vast range [3,4,17,20,38,39]. LVSD in patients with DM1 is strongly associated with cardiac and all-cause death [2,4]. LVSD was not significantly associated with abnormal findings on ECG or Holter and did not correlate with age as expected, emphasizing that repeated echocardiography is mandatory to detect cardiac involvement also in young patients with DM1. Abnormal GLS above −15.9% was found in 22% of the patients and 12/22 had preserved LVEF. To our knowledge, GLS in patients with DM1 has only been assessed in one previous study, demonstrating a significant correlation between GLS and PR-interval [8]. Accordingly, we found a trend towards a positive correlation between GLS and PRinterval, and significantly higher GLS values in patients with AVB grade I. Abnormal GLS is associated with myocardial fibrosis in patients with e.g. Fabry disease and type II diabetes mellitus [40,41]. Moreover, it is related to all-cause mortality or heart failure admissions in patients with myocardial infarction and preserved LVEF [42]. These findings suggest a link between the conduction disorder and the subclinical abnormal myocardial function but longitudinal studies are needed to evaluate the prognostic value of this parameter. We identified a significant association between cardiac involvement and reduced muscle strength. Generally, the frequency of arrhythmic events and conduction disturbances has been reported to increase
35
with the severity of the muscular disease [2,43–45] but this is not at consistent finding [14,20,46]. Although our findings revealed weaker muscle strength in patients with cardiac involvement, the small difference indicates that the degree of peripheral muscular involvement cannot confidently be used to identify patients at increased risk of cardiac events, which strongly emphasizes the need of a systematic cardiac approach irrespective of the degree of peripheral muscular involvement. Several studies have assessed the association between CTG repeat length and cardiac involvement with ambiguous results [3,16]. A recent study revealed a large inter- and intra-tissue CTG length variation in adult DM1 tissue with the largest expansion in the heart and cerebral cortex [47]. On this basis we omitted analyses of CTG-repeat length in our study. 4.1. Strengths and limitations To our knowledge, this study is the largest single-centre study evaluating patients with DM1 systematically and concomitantly with family-history, physical examination, ECG, Holter-monitoring, echocardiography and muscle strength testing. As this is a single-centre study, our results may not be generalizable. However, we invited all patients with DM1 in eastern Denmark to participate, independent of cardiac symptoms or signs of cardiac involvement. Despite the large difference in age between DM1 patients and controls, we observed a higher prevalence of specific cardiac abnormalities in DM1, adding important knowledge to the literature concerning the progression of cardiac involvement in DM1. So far, recommendations for cardiac work-up in patients with DM1 are sparse. We recognize that long-term follow-up including cardiac assessment with ECG, Holter-monitoring and echocardiography is necessary to establish optimal guidelines for cardiac work-up and risk stratification for SCD. Nevertheless, we believe that our results, together with the existing literature and current guidelines, are sufficient to cautiously recommend proposals for cardiac work-up. 5. Proposals for cardiac work-up in patients with DM1 Based on our findings that 1) conduction abnormalities and 2) arrhythmias are the most frequent cardiac abnormalities in DM1 and that 3) reduced LVEF is prevalent and not age-dependent, cardiac screening, treatment and follow-up should focus on these three aspects. In our study, a normal ECG could not rule out the existence of abnormal findings on Holter-monitoring or echocardiography. Therefore, all three aspects need to be assessed systematically and we suggest the following recommendations: 1. Recommendations for cardiac evaluation at time of diagnosis Risk-reduction of SCD by timely cardiac treatment necessitates cardiac screening. ECG, 48-hour Holter-monitoring and echocardiography at time of diagnosis seems appropriate. Patients need to be educated in seeking medical advice in case of heart specific symptoms and understand the relevance of repeated cardiac assessment due to the generally progressive nature of the disease. 2. Recommendations for cardiac follow-up We found that conduction abnormalities and arrhythmias are the most frequent findings. No SCDs or deaths from cardiac causes were observed during short-term follow-up (median 31 months). Taken together, and in agreement with the Danish National Guidelines (Danish Society of Cardiology), we consider it appropriate to suggest repeated cardiac assessment with ECG and 48-hour Holter-monitoring every second year and additionally echocardiography every fourth year in asymptomatic patients with no cardiac abnormalities. However, longterm studies are indeed needed to define the optimal timely follow-up. Special attention is needed in patients with symptoms such as syncope or specific predictors of SCD as defined by Groh and co-authors
36
H. Petri et al. / International Journal of Cardiology 174 (2014) 31–36
[2]: rhythm other than sinus, PR-interval ≥240 ms, QRS ≥120 ms, second/third degree AVB) and/or a diagnosis with atrial tachyarrhythmia. Such findings necessitate more frequent and more thorough follow-up including e.g. extended Holter-monitoring, implantable loop-recorders and in selected cases invasive electrophysiological studies. Higher survival rates and lower incidence of sudden death has been reported in DM1 patients with conduction disease using an invasive strategy based on systematic electrophysiological studies and prophylactic permanent pacing compared with a non-invasive strategy [48]. If syncopes are associated with tachy- and/or bradyarrhythmia, PM/ ICD-implantation is appropriate. Medical treatment of supraventricular or ventricular arrhythmias does not differ from routine guidelines. In accordance with the 2012 ACCF/AHA/HRS “Focused Update concerning the 2008 Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities”, PM-implantation in patients with neuromuscular disease is considered a class IB indication for patients with AVB grade III and advanced AVB grade II, and a class IIB indication for any degree of AVB (including AVB grade I) [49]. These recommendations are regardless of symptoms because there may be unpredictable progression of AV conduction disease. ICDs should be considered rather than PMs due to the increased risk of VT/VF and SCD in accordance with the recommendations from a large longitudinal study by Bhakta and co-authors [9]. The indications for ICD-implantation are similar to the indications in other patients. Heart-failure should be managed according to current guidelines [50]. We acknowledge that treatment of echocardiographic abnormalities has so far not proven any benefit among patients with DM1 — to document this would necessitate a randomized clinical trial. Nevertheless, we have no evidence or reason to believe that treatment of structural cardiac manifestations should respond differently to treatment in patients with DM1 compared to other patients with such manifestations. 6. Conclusion This single-centre study documents the complete cardiac phenotype in patients with DM1. Conduction abnormalities, arrhythmias or reduced cardiac pump function were observed in more than 50% of the patients. The additive value of this study is that a systematic cardiac assessment including ECG, Holter-monitoring and echocardiography is useful in order to make a proper characterization of cardiac involvement in DM1. The unpredictable progressive nature of cardiac abnormalities in DM1 must be considered in the clinical work-up. Cardiac symptoms and manifestations may be minimized by early intervention, which necessitates repeated, systematic cardiac screening and follow-up programs. In asymptomatic patients without cardiac involvement followup at least every second year seems appropriate but long-term followup is needed before more accurate recommendations and guidelines for DM1-specific risk stratification can be established. Funding We thank The Research Foundation of Rigshospitalet (ref. nr. R40A1247), The Danish Heart Foundation (ref. nr. 11-04-R83-A335322622), The Lundbeck Foundation (ref. nr. R83-A7909) and The Stibo Foundation for the financial support.
References [1] Brook JD, McCurrach ME, Harley HG, et al. Cell Feb 21 1992;68(4):799–808. [2] Groh WJ, Groh MR, Saha C, et al. N Engl J Med Jun 19 2008;358(25):2688–97. [3] Petri H, Vissing J, Witting N, Bundgaard H, Kober L. Int J Cardiol 2012 Oct 4;160(2): 82–8. [4] Bhakta D, Groh MR, Shen C, Pascuzzi RM, Groh WJ. Am Heart J Dec 2010;160(6): 1137–41 (1141). [5] Dhand UK, Raja F, Aggarwal K. Neurol Int Feb 11 2013;5(1):e5. [6] Magri D, Piccirillo G, Bucci E, et al. Int J Cardiol Nov 2010;26. [7] Morner S, Lindqvist P, Mellberg C, et al. J Intern Med Jul 2010;268(1):59–65. [8] Wahbi K, Ederhy S, Becane HM, et al. Int J Cardiol Nov 3 2011;152(3):375–6. [9] Bhakta D, Shen C, Kron J, et al. J Cardiovasc Electrophysiol Dec 2011;22(12): 1369–75. [10] Breton R, Mathieu J. Can J Cardiol Feb 2009;25(2):e23–7. [11] Cheng S, Keyes MJ, Larson MG, et al. JAMA Jun 24 2009;301(24):2571–7. [12] Colleran JA, Hawley RJ, Pinnow EE, Kokkinos PF, Fletcher RD. Am J Cardiol Dec 1 1997;80(11):1494–7. [13] Lazarus A, Varin J, Babuty D, et al. J Am Coll Cardiol Nov 6 2002;40(9):1645–52. [14] Mammarella A, Paradiso M, Antonini G, et al. Adv Ther Sep 2000;17(5):238–51. [15] Mathieu J, Allard P, Potvin L, Prevost C, Begin P. Neurology May 12 1999;52(8): 1658–62. [16] Sansone VA, Brigonzi E, Schoser B, et al. Int J Cardiol Dec 2012;22. [17] Bhakta D, Lowe MR, Groh WJ. Am Heart J Feb 2004;147(2):224–7. [18] Phillips MF, Harper PS. Cardiovasc Res Jan 1997;33(1):13–22. [19] Nguyen HH, Wolfe III JT, Holmes Jr DR, Edwards WD. J Am Coll Cardiol Mar 1988; 11(3):662–71. [20] De AL, Raisaro A, Marchiano V, Radice S, Meola G. Eur Heart J Jul 1995;16(7): 1007–10. [21] Antonini G, Giubilei F, Mammarella A, et al. Neurology Oct 24 2000;55(8):1207–9. [22] Sajadieh A, Nielsen OW, Rasmussen V, Hein HO, Hansen JF. J Intern Med Oct 2006; 260(4):377–87. [23] Warner JP, Barron LH, Goudie D, et al. J Med Genet Dec 1996;33(12):1022–6. [24] Lang RM, Bierig M, Devereux RB, et al. J Am Soc Echocardiogr Dec 2005;18(12): 1440–63. [25] Nagueh SF, Appleton CP, Gillebert TC, et al. J Am Soc Echocardiogr Feb 2009;22(2): 107–33. [26] Yingchoncharoen T, Agarwal S, Popovic ZB, Marwick TH. J Am Soc Echocardiogr Feb 2013;26(2):185–91. [27] Sajadieh A, Nielsen OW, Rasmussen V, et al. Eur Heart J Mar 2004;25(5):363–70. [28] Christensen AH, Bundgaard H, Schwartz M, Hansen SH, Svendsen JH. Circ Arrhythm Electrophysiol Oct 2008;1(4):317–20. [29] De BD, De BG, Kornitzer M. Heart Dec 2000;84(6):625–33. [30] Hingorani P, Natekar M, Deshmukh S, et al. Indian J Med Res Mar 2012;135:322–30. [31] HISS RG, LAMB LE. Circulation 1962 Jun;25:947–61. [32] Pelliccia A, Culasso F, Di Paolo FM, et al. Eur Heart J Aug 2007;28(16):2006–10. [33] Bussink BE, Holst AG, Jespersen L, et al. Eur Heart J Jan 2013;34(2):138–46. [34] Schneider JF, Thomas Jr HE, Kreger BE, McNamara PM, Kannel WB. Ann Intern Med Mar 1979;90(3):303–10. [35] Zannad F, Huvelle E, Dickstein K, et al. Eur J Heart Fail Jan 2007;9(1):7–14. [36] Nazarian S, Wagner KR, Caffo BS, Tomaselli GF. Pacing Clin Electrophysiol Feb 2011; 34(2):171–6. [37] Nielsen OW, Hilden J, Larsen CT, Hansen JF. Heart Aug 2001;86(2):172–8. [38] Hartwig GB, Rao KR, Radoff FM, et al. Neurology May 1983;33(5):657–60. [39] Tokgozoglu LS, Ashizawa T, Pacifico A, et al. JAMA Sep 13 1995;274(10):813–9. [40] Kosmala W, Przewlocka-Kosmala M, Szczepanik-Osadnik H, et al. JACC Cardiovasc Imaging Dec 2011;4(12):1239–49. [41] Kramer J, Niemann M, Liu D, et al. Eur Heart J Jun 2013;34(21):1587–96. [42] Ersboll M, Valeur N, Mogensen UM, et al. J Am Coll Cardiol Jun 11 2013;61(23): 2365–73. [43] Cudia P, Bernasconi P, Chiodelli R, et al. J Neurol Neurosurg Psychiatry Jul 2009; 80(7):790–3. [44] Hawley RJ, Milner MR, Gottdiener JS, Cohen A. Neurology Feb 1991;41(2 (Pt 1)): 259–62. [45] Nishioka SA, Martinelli FM, Marie S, Zatz M, Costa R. Arq Bras Cardiol Apr 2005; 84(4):330–6. [46] Florek RC, Triffon DW, Mann DE, Ringel SP, Reiter MJ. West J Med Jul 1990;153(1): 24–7. [47] Lopez CA, Nakamori M, Tome S, et al. Hum Mol Genet Jan 1 2011;20(1):1–15. [48] Wahbi K, Meune C, Porcher R, et al. JAMA Mar 28 2012;307(12):1292–301. [49] Epstein AE, DiMarco JP, Ellenbogen KA, et al. J Am Coll Cardiol Jan 22 2013;61(3): e6-75. [50] Hunt SA, Abraham WT, Chin MH, et al. J Am Coll Cardiol Apr 14 2009;53(15):e1-90.