- Email: [email protected]
Malignant ventricular arrhythmias in alcoholic cardiomyopathy
International Journal of Cardiology 199 (2015) 99–105
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Malignant ventricular arrhythmias in alcoholic cardiomyopathy☆ Gonzalo Guzzo-Merello a,b, Fernando Dominguez a, Esther González-López a,c, Marta Cobo-Marcos a, Manuel Gomez-Bueno a, Ignacio Fernandez-Lozano d, Isabel Millan e, Javier Segovia a, Luis Alonso-Pulpon a, Pablo Garcia-Pavia a,c,⁎ a
Heart Failure and Inherited Cardiac Diseases Unit, Department of Cardiology, Hospital Universitario Puerta de Hierro, Madrid, Spain Department of Cardiology, Hospital General de Villalba, Madrid, Spain Myocardial Biology Programme, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain d Arrhythmia Unit, Department of Cardiology, Hospital Universitario Puerta de Hierro, Madrid, Spain e Biostatistics Unit, Hospital Universitario Puerta de Hierro, Madrid, Spain b c
a r t i c l e
i n f o
Article history: Received 14 May 2015 Received in revised form 24 June 2015 Accepted 6 July 2015 Available online 11 July 2015 Keywords: Alcohol Arrhythmia Alcoholic cardiomyopathy Sudden cardiac death
a b s t r a c t Background: Excessive alcohol consumption is a well-known aetiology of atrial arrhythmias but there is little information concerning the prevalence or incidence of malignant ventricular arrhythmias in alcoholic cardiomyopathy (ACM). This study sought to investigate incidence and predictive factors of ventricular arrhythmias in ACM. Methods: Retrospective observational study of the clinical characteristics and long-term arrhythmic events in 282 consecutive patients with ACM (94 individuals) and idiopathic dilated cardiomyopathy (IDCM) (188 individuals) evaluated between 1993 and 2011. Results: During a median follow-up of 38 months (IQR:12–77), 42 patients died and 79 underwent heart transplantation [31 (33%) with ACM vs 90 (48%) with IDCM; p = 0.017]. A total of 37 (13%) patients [18 (19%) ACM vs 20 (11%) IDCM; p = 0.048] suffered malignant ventricular arrhythmias. On multivariate analysis, left bundle branch block (LBBB) (OR 2.4; CI95%: 1.2–5; p = 0.015) and alcoholic aetiology (OR 2.3; CI95%: 1.1–4.5; p = 0.026) were the only independent predictors of malignant ventricular arrhythmic events. A total of 18 (19%) ACM patients experienced 20 malignant ventricular arrhythmic events (4 aborted SCD, 8 SCD and 8 appropriate ICD therapies). At baseline evaluation, the only independent predictor of malignant ventricular arrhythmias in ACM patients was LBBB (OR 11.2; CI95%: 2.6–50; p = 0.001). No malignant ventricular arrhythmias were recorded during follow-up in ACM patients if left ventricular ejection fraction (LVEF) had increased or remained ≥40%. Conclusions: Malignant ventricular arrhythmias are more frequent in ACM than in IDCM. LBBB identifies ACM patients with increased risk of SCD. No malignant ventricular arrhythmias were found during follow-up in ACM patients when LVEF was ≥40%. © 2015 Elsevier Ireland Ltd. All rights reserved.
Introduction Light to moderate alcohol intake has beneficial effects on cardiovascular health [1–3]. In contrast, chronic heavy alcohol consumption could ☆ Funding: This work was supported in part by the Instituto de Salud Carlos III [grants PI11/0699, and RD12/0042/0066] through the National Plan for Scientific Research, Development and Technological Innovation 2008–2011 and 2013–2016 — European Regional Development Fund (FEDER) “A way of making Europe”. ⁎ Corresponding author at: Department of Cardiology, Hospital Universitario Puerta de Hierro, Manuel de Falla, 2. Majadahonda, Madrid 28222, Spain. E-mail addresses: [email protected] (G. Guzzo-Merello), [email protected] (F. Dominguez), [email protected] (E. González-López), [email protected] (M. Cobo-Marcos), [email protected] (M. Gomez-Bueno), [email protected] (I. Fernandez-Lozano), [email protected] (I. Millan), [email protected] (J. Segovia), [email protected] (L. Alonso-Pulpon), [email protected] (P. Garcia-Pavia).
http://dx.doi.org/10.1016/j.ijcard.2015.07.029 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
lead to systolic dysfunction and heart failure [3–5]. Excessive alcohol intake has been implicated in up to 40% of cases of dilated cardiomyopathy (DCM), and DCM due to alcohol abuse is known as alcoholic cardiomyopathy (ACM) [5–9]. Ventricular arrhythmias are a possible complication in ACM and epidemiological data from numerous studies show that the risks of ventricular tachycardia (VT) and sudden cardiac death (SCD) are increased among individuals with high alcohol intake [10]. Moreover, multiple mechanisms contributing to arrhythmias have been described in relation to excessive alcohol consumption, and ethanol is a well-known cause of supraventricular arrhythmias [10,11]. We have recently shown that ACM patients die more frequently from SCD than patients with idiopathic DCM (IDCM) despite better transplant-free survival among the ACM group [12]. However, only one study to date has investigated the long-term prognosis and incidence of SCD and ventricular arrhythmias in ACM [13].
100
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Furthermore, this study was conducted in the era before modern pharmacotherapy. The purpose of this study was two-fold: to better determine the incidence of malignant ventricular arrhythmias among ACM patients in the current era, and to identify clinical characteristics associated with arrhythmic events in this disease. Methods Data from all consecutive ACM patients referred for evaluation to the Heart Failure and Heart Transplant Units of the Hospital Universitario Puerta de Hierro (Madrid, Spain) between January 1993 and December 2011 were retrospectively analysed. The study was approved by our institution's local review board and conforms to the principles outlined in the Declaration of Helsinki. DCM was defined according to World Health Organization criteria [14] in the absence of coronary artery disease, severe systemic hypertension, congenital disease and primary valve disease [14]. Heavy alcohol consumption was defined as a self-reported history of alcohol abuse of N80 g per day (8 standard drinks) over a period of at least 5 years [6–9]. Complete abstinence from alcohol was recommended to all patients but a specific programme for alcohol discontinuation and patient's self-care was not provided. During follow-up, patients were classified as abstainers if they reported complete discontinuation of alcohol consumption during all follow-up visits and as non-abstainers if they reported continued alcohol consumption in any follow-up visit. Non-abstainers were subsequently classified as moderate drinkers if they had reduced consumption to b80 g/day of alcohol, and as heavy-persistent drinkers if they continued abusing alcohol (N80 g/day). In order to have a reliable and contemporary control IDCM group for comparison, for each case of ACM, we selected the next 2 consecutive new patients with IDCM evaluated at our Unit at the same setting (hospitalization or outpatient clinic) as the index ACM subject. Initial assessment of all patients included physical examination, blood tests, 12-lead electrocardiogram (ECG) and echocardiography. Additional studies including 24-hour ECG monitoring, 6-minute walk test, upright exercise testing, right-heart catheterization, electrophysiological study, cardiac magnetic resonance imaging and endomyocardial biopsy were performed as ordered by the treating physician. Successive tests were performed as required. ICD indications evolved through time and ICD implantation was decided by the treating physician. Coronary angiography to rule out coronary heart disease was performed in all but 3 patients. Of these 3 cases, 2 patients had normal coronary CT scans and the other was a 30-year-old male without coronary risk factors who completed a normal exercise test. The study began after the completion of the baseline evaluation and terminated at the latest available follow-up or at the patient's death or transplantation. Most patients were regularly followed at our centre at least once per year. Follow-up data from patients under surveillance at other institutions were also collected. Information on each patient's final status in December 2011 was obtained from their medical records or by telephone calls to the patient or the referring physician. Death was classified as: 1) due to progressive heart failure, 2) SCD, or 3) non-cardiac. Malignant ventricular arrhythmias were defined as SCD, cardiopulmonary resuscitation or appropriate ICD therapy. SCD was defined as witnessed sudden cardiac death with or without documented ventricular fibrillation (VF) or death within 1 h of acute symptoms or nocturnal deaths with no previous history of worsening symptoms. Cardiopulmonary resuscitation was defined as a successful basic
life support for a cardiac arrest. Appropriate ICD therapies were considered in case of an ICD discharge terminating VF or VT. ICD interrogation was performed every 6–8 months or earlier if clinically indicated. All devices had the capacity to store cardiac electrograms (Fig. 1), and therapies were independently classified by 3 of the authors.
Statistical analysis Categorical data were expressed as percentages and compared using the chi-square test or Fisher's exact test. Normally distributed variables were expressed as the mean and standard deviation while non-normally distributed variables were given as the median and interquartile range (IQR). For statistical analysis, Student's t test and Mann–Whitney nonparametric test were used in two-group comparisons. To predict “sudden cardiac death or appropriate ICD discharge” from baseline variables of ACM patients and the overall patient cohort, initially univariate screening of all parameters at enrolment was made. In order to identify independent predictors, all variables that differed between groups in both cohorts with p b 0.05 in univariate analysis were entered into a backward stepwise selection procedure with removal based on the probability of the likelihood-ratio statistic on the maximum partial likelihood estimated. Persistence of alcohol consumption was also entered into the ACM cohort multivariate analysis due to its clinical importance and its previous association with ventricular arrhythmias in ACM. Peak oxygen uptake, 6-minute test and right heart haemodynamic parameters were not included in the multiple logistic regression analyses because these tests were performed in b 50% of patients. The level of statistical significance was p b 0.05. All hypothesis tests were two-sided. The entire analysis was performed using the SPSS package, version 14.0 (SPSS Inc.). Results A total of 282 patients with DCM (94 ACM patients and 188 with IDCM) were studied (Table 1). Among the overall study cohort (ACM and IDCM) and during a median follow-up of 38 months (IQR: 12–77), 42 (15%) patients died and 79 (28%) underwent heart transplantation [31 (33%) in ACM group vs 90 (48%) in IDCM group; p = 0.017]. A total of 37 (13%) patients [18 (19%) with ACM vs 20 (11%) with IDCM; p = 0.048] suffered malignant ventricular arrhythmic events. A total of 12 patients suffered an aborted
Fig. 1. Examples of malignant ventricular arrhythmias in patients with alcoholic cardiomyopathy. A. Stored ICD electrogram of a 24-beat VT at a rate of 230 to 250 bpm. The arrhythmia was classified as VF due to the elevated heart rate and a shock was delivered, restoring a normal 70 bpm sinus rhythm. B. Stored ICD electrogram which shows the onset of VF, successfully terminated with electric shock and followed by atrial fibrillation rhythm at a heart rate of 90–120 bpm.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
101
Table 1 Clinical, electrocardiographic, echocardiographic and haemodynamic characteristics at first evaluation of 282 patients with and without malignant ventricular arrhythmias. Entire cohort (n = 282) Malignant ventricular arrhythmias (n = 37) No malignant ventricular arrhythmias (n = 245) p value Mean age (years) Heart failure symptoms (years) Sex (%) Male (%) Female (%) Baseline NYHA class I–II (%) III–IV (%) Comorbidities, (%) Hypertension (%) Dyslipidaemia (%) Diabetes (%) Smoker (%) COPD (%) History of heavy alcohol consumption (%) Blood pressure and pulse Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse (beats/min) Blood test Haemoglobin (g/dl) Creatinine (mg/dl) Bilirubin (mg/dl) ALAT (U/L) ASAT (U/L) ECG Sinus rhythm, (%) Atrial fibrillation, (%) QRS N 120 ms (%) Left bundle branch block QRS duration (ms) Echocardiography LVEF (%) LVEDD (mm) Right heart catheterization⁎ SPAP (mm Hg) DPAP (mm Hg) MPAP (mm Hg) PCW (mm Hg) CO (l/min) CI (l/min/m2) Exercise tests 6-minute walking test† (m) Peak oxygen uptake‡ (l/kg/min) Treatment ACEI or ARB (%) Digoxin (%) Loop diuretics (%) Spironolactone/eplerenone (%) Beta-blockers (%) Amiodarone (%) ICD (%) CRT (%)
50 ± 13 2.9 ± 4
50 ± 11 3.7 ± 3
50 ± 13 2.8 ± 4
0.975 0.237 0.779
83 17
81 19
83 17
43 57
54 46
41 58
34 30 18 27 19 33
35 38 13 32 30 49
34 29 19 26 17 31
0.919 0.251 0.407 0.711 0.068 0.034
116 ± 21 75 ± 14 87 ± 22
115 ± 17 75 ± 11 86 ± 20
117 ± 21 75 ± 14 88 ± 22
0.597 0.851 0.737
14 ± 2 1.2 ± 0.6 1.4 ± 2 50 ± 133 38 ± 59
14 ± 2 1.1 ± 0.3 1.7 ± 2.4 41 ± 56 43 ± 99
14 ± 2 1.2 ± 0.6 1.3 ± 2.1 51 ± 140 37 ± 51
0.970 0.729 0.381 0.735 0.652
73 27 43 32 111 ± 31
65 35 54 49 123 ± 29
74 26 42 29 110 ± 31
0.266 0.266 0.193 0.019 0.016
27 ± 8 67 ± 9
27 ± 8 69 ± 10
27 ± 9 67 ± 9
0.788 0.142
42 ± 17 22 ± 11 30 ± 12 21 ± 11 4.4 ± 2.2 2.3 ± 0.5
44 ± 20 22 ± 12 31 ± 15 20 ± 12 4.0 ± 1 2.3 ± 0.6
42 ± 17 22 ± 11 29 ± 12 21 ± 10 4.5 ± 2 2.3 ± 0.6
0.762 0.776 0.604 0.483 0.414 0.859
363 ± 81 18.7 ± 14
354 ± 73 16 ± 6
364 ± 82 19 ± 15
0.684 0.561
87 45 79 48 63 18 32 14
86 54 73 46 51 27 59 19
88 43 79 48 65 17 28 13
0.151
0.841 0.216 0.373 0.803 0.111 0.136 b0.001 0.336
Values are means ± SD. NYHA indicates New York Heart Association; COPD, chronic obstructive pulmonary disease; ECG, electrocardiogram; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction. SPAP, systolic pulmonary artery pressure; DPAP, diastolic pulmonary artery pressure; MPAP, mean pulmonary artery pressure; PCW, pulmonary capillary wedge pressure; CO, cardiac output; CI, cardiac index; ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ICD, implantable cardiac defibrillator; CRT, cardiac resynchronisation therapy. ⁎ 22 patients with malignant ventricular arrhythmias and 130 patients without malignant ventricular arrhythmias underwent right heart catheterization. † 13 patients with malignant ventricular arrhythmias and 119 patients without malignant ventricular arrhythmias underwent 6-min walking test. ‡ 13 patients with malignant ventricular arrhythmias and 114 patients without malignant ventricular arrhythmias underwent exercise test with O2 consumption.
SCD (8 IDCM and 4 ACM), 13 died suddenly (5 IDCM and 8 ACM) and 20 received appropriate ICD therapies (12 IDCM and 8 ACM). No differences between ACM and IDCM patients were observed in terms of age, ejection fraction, ECG and treatment [12]. The ACM group showed a higher prevalence of men, smokers, liver disease and chronic obstructive pulmonary disease. ACM patients exhibited worse NYHA class and higher body mass index [12]. Baseline clinical, echocardiographic and haemodynamic characteristics of patients with and without malignant ventricular arrhythmias are presented in Table 1. No differences between groups were observed in
terms of age, ejection fraction, ECG rhythm and heart failure treatment. Patients with malignant ventricular arrhythmias showed longer duration of QRS, left bundle branch block (LBBB), and alcoholic aetiology (Table 1). In multivariate analysis, the only independent predictors of malignant ventricular arrhythmias were LBBB (OR 2.4; CI95%: 1.2–5; p = 0.015) and alcoholic aetiology (OR 2.3; CI95%: 1.1–4.5; p = 0.026) (Table 2).
102
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Table 2 Multiple logistic regression analysis showing the independent predictors of malignant ventricular arrhythmias in the entire study cohort (n = 282).
Alcoholic aetiology Left bundle branch block
OR
95% CI
p value
2.3 2.4
1.1–4.5 1.2–5.0
0.026 0.015
Variables entered into multiple regression analysis: alcoholic aetiology (p = 0.034), Left bundle branch block (0.019).
ACM cohort In the ACM cohort, during a median follow-up of 59 months (interquartile range [IQR]: 25 to 107 months), 14 patients (15%) underwent orthotropic heart transplantation and 17 died (18%). Eight deaths were due to SCD, 6 from heart failure and 3 from malignancies. A total of 18 (19%) patients experienced 20 malignant ventricular arrhythmic events: 4 patients were resuscitated from documented VF, 8 received appropriate ICD therapies and 8 patients died suddenly (Table 3 and Fig. 2). An ICD was implanted for primary or secondary prophylaxis in 31 (33%) ACM patients; of these, 8 (26%) received appropriate ICD therapies. Rate of appropriate ICD therapy was 12 per 100 person–years (median followup period 20 months, IQR 7 to 34 months). One patient with an ICD–CRT implanted for primary prophylaxis died suddenly at home during sleep 9 months after device implantation. No autopsy was performed and the device was not checked. Among the 4 patients (4%) who received successful cardiopulmonary resuscitation, only 1 underwent subsequent ICD implantation for secondary prophylaxis and received an appropriate ICD discharge 39 months after device implantation. The other 3 patients did not receive an ICD due to severe neurologic impairment after the first SCD episode, other comorbidities or patient decision. After surviving a first malignant ventricular arrhythmia (n = 11 patients), 1 patient died due to SCD that occurred 1 month after the initial event (no ICD had been implanted), 2 patients died due to end-stage heart failure, 1 received an ICD discharge and 1 patient underwent cardiac transplantation (Fig. 2). Baseline clinical, echocardiographic and haemodynamic characteristics of ACM patients with and without malignant ventricular arrhythmias are presented in Table 4. No differences between groups were observed in terms of age, alcohol abstinence, ejection fraction, ECG rhythm and heart failure treatment. Although the use of certain heart failure therapies at baseline evaluation was lower than currently desired (beta-blockers were not proved to be beneficial at the start of study), heart failure treatment at last clinical evaluation was in accordance with modern practice. The rate of ACEI/ARB, beta-blockers and aldosterone antagonists at last follow-up was 84%, 76% and 57%, respectively. Ventricular tachycardia ablation procedures were not performed in any patient and no antiarrhythmic drugs, except for amiodarone, were prescribed. In baseline univariate analysis, ACM patients with ventricular arrhythmic events showed higher prevalence of QRS width N120 ms
Table 3 Malignant ventricular arrhythmias and end-stage heart failure in 94 ACM patients. Malignant ventricular arrhythmias Cardiopulmonary resuscitation Appropriate ICD therapy (n = 31) Sudden cardiac death End-stage heart failure Heart Transplant Death due to end-stage heart failure Other causes of death ICD = implantable cardiac defibrillator.
4 (4%) 8 (26%) 8 (9%) 14 (15%) 6 (6%) 3 (3%)
and LBBB, an increased left ventricular end-diastolic diameter and a lower rate of beta-blockers (Table 4). In multivariate analysis, the only independent predictor of malignant ventricular arrhythmias was LBBB (OR 11.2; CI95%: 2.6–50; p = 0.001) (Table 5). Although persistent exposition to alcohol was not significantly associated with malignant ventricular arrhythmias, it was close to significance (OR 4.3; CI95%: 0.99– 18.9; p = 0.051). Follow-up LVEF data were available in 92 patients (98%). At latest follow-up, 39 patients (41%) showed a LVEF ≥ 40%. None of them died from cardiac causes or underwent heart transplant. Only 1 patient with LVEF ≥ 40% died, but this was due to malignancy. Analysis of the relationship between LVEF and appearance of malignant ventricular arrhythmias revealed that there were no malignant ventricular arrhythmic events among patients with LVEF ≥ 40% or in NYHA I functional class. There were 34 ACM patients (36%) with LVEF b 40% at study entry and in whom LVEF recovered to ≥40% during follow-up. Four (12%) of these patients suffered malignant ventricular arrhythmias (all ICD discharges) while their LVEF was still b40%. During a median follow-up period of 34 months (IQR: 13–96) after achieving a LVEF ≥ 40%, all patients who recovered LVEF were free from malignant ventricular arrhythmias. Twenty-seven of the 34 (79%) ACM patients in whom LVEF recovered to ≥40% during followup had an echocardiogram performed within one year of baseline evaluation. Twenty patients (74%) exhibited a LVEF ≥ 40% on that evaluation, while the remaining 7 patients who did not recover LVEF during the first year exhibited a LVEF ≥40% in an echocardiogram performed after a median time of 29 months (IQR: 24–37) from baseline evaluation. Among the 7 patients who did not have LVEF examined during the first year, 6 of them had a LVEF ≥ 40% on the first echocardiographic evaluation performed after baseline examination that took place within a median time of 24 months (IQR: 19–43) from baseline visit. Discussion This study is the largest cohort of ACM patients described to date, the only one which includes modern therapies on heart failure, and the first focusing specifically on sudden death and ventricular arrhythmic complications. The study shows that malignant ventricular arrhythmias are a relatively frequent complication among ACM patients. In our series, ACM patients died more frequently from SCD and suffered more ICD discharges than patients with IDCM. Although LBBB was the only independent predictor of ventricular arrhythmias at baseline evaluation, our findings show that no patients suffered from ventricular arrhythmias when their LVEF was ≥ 40% or their NYHA was I. Furthermore, most patients who improved their LVEF to ≥40% did so during the first year of follow-up. Alcohol abuse and ventricular arrhythmias While low chronic alcohol intake reduces mortality, heart failure events and SCD, probably due to alcohol's anti-ischemic properties [15– 18], heavy alcohol consumption abuse has been shown to induce multiple cardiac changes that predispose to fatal arrhythmias. Those changes include ultrastructural changes, proarrhythmic electrolyte abnormalities, QT interval prolongation, adrenergic hyperactivity and a decreased heart rate variability and baroreceptor sensitivity [10,11]. Additionally, in animal models, ethanol reduces atrioventricular and intraventricular conduction, increases sinus node recovery time [19], causes shortening of the action potential and decreases contractility [10]. In line with such evidence, our study shows that malignant ventricular arrhythmias are a frequent complication in ACM. In fact, ACM patients died more frequently from SCD and suffered more ICD discharges than patients with IDCM.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
103
Fig. 2. Overview of clinical evolution and malignant ventricular arrhythmias in 94 ACM patients.
In spite of the epidemiological relevance of ACM in Western countries, little information is available on the incidence of ventricular arrhythmias in ACM because most of the available long-term prognostic studies in ACM did not specify the cause of death [6–9]. The only study addressing the incidence of malignant arrhythmias in ACM was conducted by Fauchier in France and was published in 2003 [13]. In this study, 75 patients were followed during 51 ± 42 months and 8% of them suffered from SCD or VF/sustained VT (16% of nonabstainers and 2% of abstainers). In our study, 18 patients (19%) suffered a malignant arrhythmic event, which is higher than the incidence reported before. This difference could be due to the inclusion of appropriate ICD therapies as end-point in our study, and to the increased transplant-free survival reported in our cohort (62% vs 50% and 63% vs 32% at 10 years among abstainers and non-abstainers, respectively) [8,12,13]. Increased end-stage heart failure survival increases the chance of suffering from arrhythmic complications, a phenomenon previously reported in other heart failure studies [20]. Predictors of ventricular arrhythmias Prediction of ventricular arrhythmic events in heart failure and DCM is still a complex and controversial issue. Several variables have been proposed, but an effective stratification of the risk of SCD has not yet been identified. Our results show that alcoholic aetiology should be considered as a risk factor in this context because ACM patients suffered a higher rate of malignant ventricular arrhythmias. Appropriate prediction of ventricular arrhythmic events in ACM patients is crucial to define which patients will benefit from prophylactic ICD implantation. In our study, the only independent predictor of malignant ventricular arrhythmias was LBBB. This factor is a well-known predictor of adverse prognosis in DCM [21,22], and was also shown to predict impaired transplant-free survival in the overall ACM cohort [12]. The only available data concerning long-term predictors of arrhythmic events in ACM come also from Fauchier's study [13]. He found that ACM patients with persistent moderate alcohol intake or abstinence had only 2% of arrhythmic events during follow-up [13]. In contrast, relative risk of arrhythmic events of ACM patients who retained a persistent heavy alcohol intake was 8.0 compared with ACM patients with moderate alcohol intake or abstinence [13]. In our study, persistent alcohol intake was not associated with malignant ventricular arrhythmic events in multivariate analysis although it was close to significance. In contrast to Fauchier's work, we decided to
include moderate drinkers and heavy drinkers in the non-abstainers group due to the low rate of heavy drinkers found in our study. Finally, although baseline LVEF and NYHA did not reach statistical significance as clinical predictors of arrhythmic events, none of the ICD appropriate therapies, SCD or recovered cardiac arrests took place in patients with baseline LVEF ≥ 40% or NYHA I. Moreover, in ACM patients with impaired LVEF at baseline, malignant arrhythmias did not occur once LVEF was over 40%. As LVEF recovery was associated with such a good prognosis in our study, and considering that LVEF could improve substantially in ACM after alcohol withdrawal, we sought to investigate the time required to reach the 40% threshold among improvers. To date, only a French study focusing on 26 ACM heart transplant candidates who recovered LVEF with abstinence has investigated the time required to achieve LVEF improvement in ACM [23]. In this study, the authors recommended not to place ACM patients on heart transplant waiting list until 3 months of absence of improvement with abstinence, although the reported average interval between abstinence and recovery in the study was 11.7 months [23]. In our study, 46 patients with LVEF b 40% at baseline evaluation did not have an event and underwent echocardiographic evaluation at one year. Among these patients, 20 (43%) showed LVEF recovery during their first year of follow-up while only 7 (15%) achieved it subsequently. Overall, our findings reflect that when LVEF has not improved to ≥40% during the first year of follow-up in ACM, the chance of a late improvement is low and therefore an ICD implant should be considered at that time.
Limitations Our study cohort was obtained from a single heart transplant centre and that may have resulted in a referral bias toward young and more affected ACM patients prone to abandon alcohol consumption. Also, because of the retrospective nature of the study, LVEF was not evaluated in all patients at the same time intervals. Additionally, and similar to all previous ACM studies, the identification of ACM patients and their allocation into abstainers or persistent drinker groups during follow-up was based on patients' self-reported alcohol intake, which may lead to underestimation. Finally, the multiple logistic regression analysis presented was based on a stepwise selection procedure from a larger set of candidate predictor variables found in univariate analysis. Therefore, there is some risk of identifying a false positive predictor due to this multiplicity issue.
104
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Table 4 Clinical, electrocardiographic, echocardiographic and haemodynamic characteristics at first evaluation of ACM patients with and without malignant ventricular arrhythmias.
Mean age (years) Sex (%) Male Female Baseline NYHA class I (%) II (%) III (%) IV (%) Comorbidities, (%) Hypertension (%) Dyslipidaemia (%) Diabetes (%) Smoker (%) Body mass index (kg/m2) COPD (%) Alcohol consumption Duration of alcohol abuse (years) Mean alcohol consumption (g/day) Previous alcohol consumption (groups) b120 g/day (%) 120–160 g/day (%) N160 g/day (%) Alcohol consumption during follow-up Alcohol abstinence (%) Persistent alcohol intake (%) Blood pressure and pulse Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse (beats/min) Blood test Haemoglobin (g/dl) Creatinine (mg/dl) Sodium (mg/dl) Bilirubin (mg/dl) ALAT (U/L) ASAT (U/L) GGT (U/L) ECG Sinus rhythm, (%) Atrial fibrillation, (%) QRS ≥ 120 msec (%) QRS duration (ms) Left bundle branch block (%) Echocardiogram LVEF (%) LVEDD (mm) LVESD (mm) Right heart catheterization⁎ SPAP (mm Hg) DPAP (mm Hg) MPAP (mm Hg) PCW (mm Hg) CO (l/min) CI (l/min/m2) Exercise tests 6-minute walking test† (m) Peak oxygen uptake‡ (l/kg/min) Treatment ACEI or ARB (%) Digoxin (%) Loop diuretics (%) Spironolactone/eplerenone (%) Beta-blockers (%) Amiodarone (%) ICD (%) CRT (%)
Malignant ventricular arrhythmias (n = 18)
No malignant ventricular arrhythmias (n = 76)
p value
52 ± 7
49 ± 10
0.259 0.039
94 6
100 0
11 39 28 22
6 23 39 32
39 44 22 72 29 ± 6 39
35 26 24 45 28 ± 4 29
0.789 0.130 0.895 0.110 0.382 0.412
21 ± 13 143 ± 58
24 ± 12 135 ± 66
0.463 0.634 0.665
44 19 37
51 24 26
56 44
65 35
113 ± 15 75 ± 10 90 ± 23
121 ± 21 78 ± 17 92 ± 25
0.216 0.559 0.805
14 ± 1.5 1.4 ± 0.3 137 ± 3 1.6 ± 1 62 ± 79 70 ± 151 94 ± 89
15 ± 1.4 1.2 ± 0.3 137 ± 4 2±4 93 ± 247 53 ± 85 165 ± 268
0.478 0.157 0.893 0.706 0.697 0.603 0.433
56 44 69 123 ± 24 59
68 32 32 108 ± 30 26
0.430 0.430 0.020 0.063 0.032
25 ± 9 73 ± 11 58 ± 12
26 ± 9 67 ± 9 57 ± 9
0.726 0.024 0.696
48 ± 26 26 ± 14 37 ± 19 24 ± 15 3.5 ± 0.7 2 ± 0.4
46 ± 16 24 ± 10 31 ± 12 24 ± 15 4.2 ± 1.3 2.3 ± 0.5
0.801 0.644 0.322 0.970 0.155 0.274
283 ± 156 11 ± 8
373 ± 74 16 ± 6
0.041 0.166
83 61 72 50 39 17 50 16
92 45 77 49 64 20 30 28
0.255 0.208 0.668 0.918 0.047 0.730 0.112 0.235
0.371
0.482
Values are means ± SD. NYHA indicates New York Heart Association; COPD, chronic obstructive pulmonary disease; ECG, electrocardiogram; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction. SPAP, systolic pulmonary artery pressure; DPAP, diastolic pulmonary artery pressure; MPAP, mean pulmonary artery pressure; PCW, pulmonary capillary wedge pressure; CO, cardiac output; CI, cardiac index; ACEI, angiotensin converting enzyme inhibitors; ARB, angiotensin receptor blockers; ICD, implantable cardiac defibrillator; CRT, cardiac resynchronisation therapy. ⁎ 9 ACM patients with malignant ventricular arrhythmias and 31 ACM patients without malignant ventricular arrhythmias underwent right heart catheterization. † 6 ACM patients with malignant ventricular arrhythmias and 26 ACM patients without malignant ventricular arrhythmias underwent 6-min walking test. ‡ 5 ACM patients with malignant ventricular arrhythmias and 23 ACM patients without malignant ventricular arrhythmias underwent exercise test with O2 consumption.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105 Table 5 Multiple logistic regression analysis showing the independent predictors for malignant ventricular arrhythmias in ACM.
Left bundle branch block Persistent alcohol intake
OR
95% CI
P value
11.2 4.3
2.6–50 0.99–18.9
0.001 0.051
OR = odds ratio; CI = confidence interval. Variables entered into multiple regression analysis: Absence of beta-blockers (p = 0.047), female gender (p = 0.039), persistent alcohol intake (p = 0.482), left bundle branch block (p = 0.032) left-ventricle end-diastolic diameter (p = 0.024) and QRS width N120 ms (p = 0.020)
Conclusions Malignant ventricular arrhythmias are a frequent complication in ACM. SCD and ICD discharges are more frequent in ACM than in IDCM. LBBB was identified as the only independent predictor of ventricular arrhythmic events in ACM at baseline evaluation. In our study, no patient suffered an arrhythmic event if LVEF was ≥40% and most ACM patients in whom LVEF recovered did so during the first year of follow-up. The results of our study could be of help when deciding ICD placement in this complex condition. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. Acknowledgements We gratefully acknowledge Kenneth McCreath for English editing. References [1] C.L. Bryson, K.J. Mukamal, M.A. Mittleman, et al., The association of alcohol consumption and incident heart failure: the Cardiovascular Health Study, J. Am. Coll. Cardiol. 48 (2006) 305–311. [2] J.L. Abramson, S.A. Williams, H.M. Krumholz, et al., Moderate alcohol consumption and risk of heart failure among older persons, JAMA 285 (2001) 1971–1977. [3] R. Movva, V.M. Figueredo, Alcohol and the heart: to abstain or not to abstain? Int. J. Cardiol. 164 (2013) 267–276.
105
[4] I. Laonigro, M. Correale, M. Di Biase, E. Altomare, Alcohol abuse and heart failure, Eur. J. Heart Fail. 11 (2009) 453–462. [5] G. Guzzo-Merello, M. Cobo-Marcos, M. Gallego-Delgado, P. Garcia-Pavia, Alcoholic cardiomyopathy, World J. Cardiol. 6 (2014) 771–781. [6] C.J. McKenna, M.B. Codd, H.A. McCann, D.D. Sugrue, Alcohol consumption and idiopathic dilated cardiomyopathy: a case control study, Am. Heart J. 135 (1998) 833–837. [7] A. Gavazzi, R. De Maria, M. Parolini, M. Porcu, Alcohol abuse and dilated cardiomyopathy in men, Am. J. Cardiol. 85 (2000) 1114–1118. [8] L. Fauchier, D. Babuty, P. Poret, et al., Comparison of long-term outcome of alcoholic and idiopathic dilated cardiomyopathy, Eur. Heart J. 21 (2000) 306–314. [9] P. Prazak, M. Pfisterer, S. Osswald, P. Buser, F. Burkart, Differences of disease progression in congestive heart failure due to alcoholic as compared to idiopathic dilated cardiomyopathy, Eur. Heart J. 17 (1996) 251–257. [10] A. George, V.M. Figueredo, Alcohol and arrhythmias: a comprehensive review, J. Cardiovasc. Med. (Hagerstown) 11 (2010) 221–228. [11] M. Kupari, P. Koskinen, Alcohol, cardiac arrhythmias and sudden death, Novartis Found. Symp. 216 (1998) 68–79. [12] G. Guzzo-Merello, J. Segovia, F. Dominguez, et al., Natural history and prognostic factors in alcoholic cardiomyopathy, JACC Heart Fail. 3 (2015) 78–86. [13] L. Fauchier, Alcoholic cardiomyopathy and ventricular arrhythmias, Chest 123 (2003) 1320. [14] P. Richardson, W. McKenna, M. Bristow, et al., Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies, Circulation 93 (1996) 841–842. [15] G. Wannamethee, A.G. Shaper, Alcohol and sudden cardiac death, Br. Heart J. 68 (1992) 443–448. [16] K. Kramer, L. Kuller, R. Ficher, The increasing mortality attributed to cirrhosis and fatty liver, in Baltimore (1957–1966), Ann. Intern. Med. 69 (1968) 273–282. [17] B. Randall, Sudden death and hepatic fatty metamorphosis. A North Carolina survey, JAMA 243 (1980) 1723–1725. [18] C.M. Albert, J.E. Manson, N.R. Cook, U.A. Ajani, J.M. Gaziano, C.H. Hennekens, Moderate alcohol consumption and the risk of sudden cardiac death among US male physicians, Circulation 100 (1999) 944–950. [19] M.J. Goodkind, N.H. Gerber Jr., J.R. Mellen, J.B. Kostis, Altered intracardiac conduction after acute administration of ethanol in the dog, J. Pharmacol. Exp. Ther. 194 (1975) 633–638. [20] J. Sengupta, R. Abdelhadi, Approach to reduction of ventricular arrhythmias and implantable cardioverter–defibrillator therapies in patients with heart failure, Curr. Opin. Cardiol. 28 (2013) 337–343. [21] M. Hofmann, R. Bauer, R. Handrock, G. Weidinger, L. Goedel-Meinen, Prognostic value of the QRS duration in patients with heart failure: a subgroup analysis from 24 centers of Val-HeFT, J. Card. Fail. 11 (2005) 523–528. [22] I. Sipahi, S.P. Chaudhry, M. Issa, M. Ambati, S. Dagdelen, J.C. Fang, Impact of QRS duration on survival benefit with prophylactic implantable cardioverter–defibrillators: a meta-analysis of randomized controlled trials, J. Cardiovasc. Med. (Hagerstown) (Jul 10 2014) (Epub ahead of print). [23] E. Coignard, B. Blanchard, F. Jault, et al., Alcoholic cardiomyopathy and heart transplantation, Arch. Mal. Coeur Vaiss. 91 (1998) 45–51.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Malignant ventricular arrhythmias in alcoholic cardiomyopathy☆ Gonzalo Guzzo-Merello a,b, Fernando Dominguez a, Esther González-López a,c, Marta Cobo-Marcos a, Manuel Gomez-Bueno a, Ignacio Fernandez-Lozano d, Isabel Millan e, Javier Segovia a, Luis Alonso-Pulpon a, Pablo Garcia-Pavia a,c,⁎ a
Heart Failure and Inherited Cardiac Diseases Unit, Department of Cardiology, Hospital Universitario Puerta de Hierro, Madrid, Spain Department of Cardiology, Hospital General de Villalba, Madrid, Spain Myocardial Biology Programme, Centro Nacional de Investigaciones Cardiovasculares (CNIC), Madrid, Spain d Arrhythmia Unit, Department of Cardiology, Hospital Universitario Puerta de Hierro, Madrid, Spain e Biostatistics Unit, Hospital Universitario Puerta de Hierro, Madrid, Spain b c
a r t i c l e
i n f o
Article history: Received 14 May 2015 Received in revised form 24 June 2015 Accepted 6 July 2015 Available online 11 July 2015 Keywords: Alcohol Arrhythmia Alcoholic cardiomyopathy Sudden cardiac death
a b s t r a c t Background: Excessive alcohol consumption is a well-known aetiology of atrial arrhythmias but there is little information concerning the prevalence or incidence of malignant ventricular arrhythmias in alcoholic cardiomyopathy (ACM). This study sought to investigate incidence and predictive factors of ventricular arrhythmias in ACM. Methods: Retrospective observational study of the clinical characteristics and long-term arrhythmic events in 282 consecutive patients with ACM (94 individuals) and idiopathic dilated cardiomyopathy (IDCM) (188 individuals) evaluated between 1993 and 2011. Results: During a median follow-up of 38 months (IQR:12–77), 42 patients died and 79 underwent heart transplantation [31 (33%) with ACM vs 90 (48%) with IDCM; p = 0.017]. A total of 37 (13%) patients [18 (19%) ACM vs 20 (11%) IDCM; p = 0.048] suffered malignant ventricular arrhythmias. On multivariate analysis, left bundle branch block (LBBB) (OR 2.4; CI95%: 1.2–5; p = 0.015) and alcoholic aetiology (OR 2.3; CI95%: 1.1–4.5; p = 0.026) were the only independent predictors of malignant ventricular arrhythmic events. A total of 18 (19%) ACM patients experienced 20 malignant ventricular arrhythmic events (4 aborted SCD, 8 SCD and 8 appropriate ICD therapies). At baseline evaluation, the only independent predictor of malignant ventricular arrhythmias in ACM patients was LBBB (OR 11.2; CI95%: 2.6–50; p = 0.001). No malignant ventricular arrhythmias were recorded during follow-up in ACM patients if left ventricular ejection fraction (LVEF) had increased or remained ≥40%. Conclusions: Malignant ventricular arrhythmias are more frequent in ACM than in IDCM. LBBB identifies ACM patients with increased risk of SCD. No malignant ventricular arrhythmias were found during follow-up in ACM patients when LVEF was ≥40%. © 2015 Elsevier Ireland Ltd. All rights reserved.
Introduction Light to moderate alcohol intake has beneficial effects on cardiovascular health [1–3]. In contrast, chronic heavy alcohol consumption could ☆ Funding: This work was supported in part by the Instituto de Salud Carlos III [grants PI11/0699, and RD12/0042/0066] through the National Plan for Scientific Research, Development and Technological Innovation 2008–2011 and 2013–2016 — European Regional Development Fund (FEDER) “A way of making Europe”. ⁎ Corresponding author at: Department of Cardiology, Hospital Universitario Puerta de Hierro, Manuel de Falla, 2. Majadahonda, Madrid 28222, Spain. E-mail addresses: [email protected] (G. Guzzo-Merello), [email protected] (F. Dominguez), [email protected] (E. González-López), [email protected] (M. Cobo-Marcos), [email protected] (M. Gomez-Bueno), [email protected] (I. Fernandez-Lozano), [email protected] (I. Millan), [email protected] (J. Segovia), [email protected] (L. Alonso-Pulpon), [email protected] (P. Garcia-Pavia).
http://dx.doi.org/10.1016/j.ijcard.2015.07.029 0167-5273/© 2015 Elsevier Ireland Ltd. All rights reserved.
lead to systolic dysfunction and heart failure [3–5]. Excessive alcohol intake has been implicated in up to 40% of cases of dilated cardiomyopathy (DCM), and DCM due to alcohol abuse is known as alcoholic cardiomyopathy (ACM) [5–9]. Ventricular arrhythmias are a possible complication in ACM and epidemiological data from numerous studies show that the risks of ventricular tachycardia (VT) and sudden cardiac death (SCD) are increased among individuals with high alcohol intake [10]. Moreover, multiple mechanisms contributing to arrhythmias have been described in relation to excessive alcohol consumption, and ethanol is a well-known cause of supraventricular arrhythmias [10,11]. We have recently shown that ACM patients die more frequently from SCD than patients with idiopathic DCM (IDCM) despite better transplant-free survival among the ACM group [12]. However, only one study to date has investigated the long-term prognosis and incidence of SCD and ventricular arrhythmias in ACM [13].
100
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Furthermore, this study was conducted in the era before modern pharmacotherapy. The purpose of this study was two-fold: to better determine the incidence of malignant ventricular arrhythmias among ACM patients in the current era, and to identify clinical characteristics associated with arrhythmic events in this disease. Methods Data from all consecutive ACM patients referred for evaluation to the Heart Failure and Heart Transplant Units of the Hospital Universitario Puerta de Hierro (Madrid, Spain) between January 1993 and December 2011 were retrospectively analysed. The study was approved by our institution's local review board and conforms to the principles outlined in the Declaration of Helsinki. DCM was defined according to World Health Organization criteria [14] in the absence of coronary artery disease, severe systemic hypertension, congenital disease and primary valve disease [14]. Heavy alcohol consumption was defined as a self-reported history of alcohol abuse of N80 g per day (8 standard drinks) over a period of at least 5 years [6–9]. Complete abstinence from alcohol was recommended to all patients but a specific programme for alcohol discontinuation and patient's self-care was not provided. During follow-up, patients were classified as abstainers if they reported complete discontinuation of alcohol consumption during all follow-up visits and as non-abstainers if they reported continued alcohol consumption in any follow-up visit. Non-abstainers were subsequently classified as moderate drinkers if they had reduced consumption to b80 g/day of alcohol, and as heavy-persistent drinkers if they continued abusing alcohol (N80 g/day). In order to have a reliable and contemporary control IDCM group for comparison, for each case of ACM, we selected the next 2 consecutive new patients with IDCM evaluated at our Unit at the same setting (hospitalization or outpatient clinic) as the index ACM subject. Initial assessment of all patients included physical examination, blood tests, 12-lead electrocardiogram (ECG) and echocardiography. Additional studies including 24-hour ECG monitoring, 6-minute walk test, upright exercise testing, right-heart catheterization, electrophysiological study, cardiac magnetic resonance imaging and endomyocardial biopsy were performed as ordered by the treating physician. Successive tests were performed as required. ICD indications evolved through time and ICD implantation was decided by the treating physician. Coronary angiography to rule out coronary heart disease was performed in all but 3 patients. Of these 3 cases, 2 patients had normal coronary CT scans and the other was a 30-year-old male without coronary risk factors who completed a normal exercise test. The study began after the completion of the baseline evaluation and terminated at the latest available follow-up or at the patient's death or transplantation. Most patients were regularly followed at our centre at least once per year. Follow-up data from patients under surveillance at other institutions were also collected. Information on each patient's final status in December 2011 was obtained from their medical records or by telephone calls to the patient or the referring physician. Death was classified as: 1) due to progressive heart failure, 2) SCD, or 3) non-cardiac. Malignant ventricular arrhythmias were defined as SCD, cardiopulmonary resuscitation or appropriate ICD therapy. SCD was defined as witnessed sudden cardiac death with or without documented ventricular fibrillation (VF) or death within 1 h of acute symptoms or nocturnal deaths with no previous history of worsening symptoms. Cardiopulmonary resuscitation was defined as a successful basic
life support for a cardiac arrest. Appropriate ICD therapies were considered in case of an ICD discharge terminating VF or VT. ICD interrogation was performed every 6–8 months or earlier if clinically indicated. All devices had the capacity to store cardiac electrograms (Fig. 1), and therapies were independently classified by 3 of the authors.
Statistical analysis Categorical data were expressed as percentages and compared using the chi-square test or Fisher's exact test. Normally distributed variables were expressed as the mean and standard deviation while non-normally distributed variables were given as the median and interquartile range (IQR). For statistical analysis, Student's t test and Mann–Whitney nonparametric test were used in two-group comparisons. To predict “sudden cardiac death or appropriate ICD discharge” from baseline variables of ACM patients and the overall patient cohort, initially univariate screening of all parameters at enrolment was made. In order to identify independent predictors, all variables that differed between groups in both cohorts with p b 0.05 in univariate analysis were entered into a backward stepwise selection procedure with removal based on the probability of the likelihood-ratio statistic on the maximum partial likelihood estimated. Persistence of alcohol consumption was also entered into the ACM cohort multivariate analysis due to its clinical importance and its previous association with ventricular arrhythmias in ACM. Peak oxygen uptake, 6-minute test and right heart haemodynamic parameters were not included in the multiple logistic regression analyses because these tests were performed in b 50% of patients. The level of statistical significance was p b 0.05. All hypothesis tests were two-sided. The entire analysis was performed using the SPSS package, version 14.0 (SPSS Inc.). Results A total of 282 patients with DCM (94 ACM patients and 188 with IDCM) were studied (Table 1). Among the overall study cohort (ACM and IDCM) and during a median follow-up of 38 months (IQR: 12–77), 42 (15%) patients died and 79 (28%) underwent heart transplantation [31 (33%) in ACM group vs 90 (48%) in IDCM group; p = 0.017]. A total of 37 (13%) patients [18 (19%) with ACM vs 20 (11%) with IDCM; p = 0.048] suffered malignant ventricular arrhythmic events. A total of 12 patients suffered an aborted
Fig. 1. Examples of malignant ventricular arrhythmias in patients with alcoholic cardiomyopathy. A. Stored ICD electrogram of a 24-beat VT at a rate of 230 to 250 bpm. The arrhythmia was classified as VF due to the elevated heart rate and a shock was delivered, restoring a normal 70 bpm sinus rhythm. B. Stored ICD electrogram which shows the onset of VF, successfully terminated with electric shock and followed by atrial fibrillation rhythm at a heart rate of 90–120 bpm.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
101
Table 1 Clinical, electrocardiographic, echocardiographic and haemodynamic characteristics at first evaluation of 282 patients with and without malignant ventricular arrhythmias. Entire cohort (n = 282) Malignant ventricular arrhythmias (n = 37) No malignant ventricular arrhythmias (n = 245) p value Mean age (years) Heart failure symptoms (years) Sex (%) Male (%) Female (%) Baseline NYHA class I–II (%) III–IV (%) Comorbidities, (%) Hypertension (%) Dyslipidaemia (%) Diabetes (%) Smoker (%) COPD (%) History of heavy alcohol consumption (%) Blood pressure and pulse Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse (beats/min) Blood test Haemoglobin (g/dl) Creatinine (mg/dl) Bilirubin (mg/dl) ALAT (U/L) ASAT (U/L) ECG Sinus rhythm, (%) Atrial fibrillation, (%) QRS N 120 ms (%) Left bundle branch block QRS duration (ms) Echocardiography LVEF (%) LVEDD (mm) Right heart catheterization⁎ SPAP (mm Hg) DPAP (mm Hg) MPAP (mm Hg) PCW (mm Hg) CO (l/min) CI (l/min/m2) Exercise tests 6-minute walking test† (m) Peak oxygen uptake‡ (l/kg/min) Treatment ACEI or ARB (%) Digoxin (%) Loop diuretics (%) Spironolactone/eplerenone (%) Beta-blockers (%) Amiodarone (%) ICD (%) CRT (%)
50 ± 13 2.9 ± 4
50 ± 11 3.7 ± 3
50 ± 13 2.8 ± 4
0.975 0.237 0.779
83 17
81 19
83 17
43 57
54 46
41 58
34 30 18 27 19 33
35 38 13 32 30 49
34 29 19 26 17 31
0.919 0.251 0.407 0.711 0.068 0.034
116 ± 21 75 ± 14 87 ± 22
115 ± 17 75 ± 11 86 ± 20
117 ± 21 75 ± 14 88 ± 22
0.597 0.851 0.737
14 ± 2 1.2 ± 0.6 1.4 ± 2 50 ± 133 38 ± 59
14 ± 2 1.1 ± 0.3 1.7 ± 2.4 41 ± 56 43 ± 99
14 ± 2 1.2 ± 0.6 1.3 ± 2.1 51 ± 140 37 ± 51
0.970 0.729 0.381 0.735 0.652
73 27 43 32 111 ± 31
65 35 54 49 123 ± 29
74 26 42 29 110 ± 31
0.266 0.266 0.193 0.019 0.016
27 ± 8 67 ± 9
27 ± 8 69 ± 10
27 ± 9 67 ± 9
0.788 0.142
42 ± 17 22 ± 11 30 ± 12 21 ± 11 4.4 ± 2.2 2.3 ± 0.5
44 ± 20 22 ± 12 31 ± 15 20 ± 12 4.0 ± 1 2.3 ± 0.6
42 ± 17 22 ± 11 29 ± 12 21 ± 10 4.5 ± 2 2.3 ± 0.6
0.762 0.776 0.604 0.483 0.414 0.859
363 ± 81 18.7 ± 14
354 ± 73 16 ± 6
364 ± 82 19 ± 15
0.684 0.561
87 45 79 48 63 18 32 14
86 54 73 46 51 27 59 19
88 43 79 48 65 17 28 13
0.151
0.841 0.216 0.373 0.803 0.111 0.136 b0.001 0.336
Values are means ± SD. NYHA indicates New York Heart Association; COPD, chronic obstructive pulmonary disease; ECG, electrocardiogram; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction. SPAP, systolic pulmonary artery pressure; DPAP, diastolic pulmonary artery pressure; MPAP, mean pulmonary artery pressure; PCW, pulmonary capillary wedge pressure; CO, cardiac output; CI, cardiac index; ACEI, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; ICD, implantable cardiac defibrillator; CRT, cardiac resynchronisation therapy. ⁎ 22 patients with malignant ventricular arrhythmias and 130 patients without malignant ventricular arrhythmias underwent right heart catheterization. † 13 patients with malignant ventricular arrhythmias and 119 patients without malignant ventricular arrhythmias underwent 6-min walking test. ‡ 13 patients with malignant ventricular arrhythmias and 114 patients without malignant ventricular arrhythmias underwent exercise test with O2 consumption.
SCD (8 IDCM and 4 ACM), 13 died suddenly (5 IDCM and 8 ACM) and 20 received appropriate ICD therapies (12 IDCM and 8 ACM). No differences between ACM and IDCM patients were observed in terms of age, ejection fraction, ECG and treatment [12]. The ACM group showed a higher prevalence of men, smokers, liver disease and chronic obstructive pulmonary disease. ACM patients exhibited worse NYHA class and higher body mass index [12]. Baseline clinical, echocardiographic and haemodynamic characteristics of patients with and without malignant ventricular arrhythmias are presented in Table 1. No differences between groups were observed in
terms of age, ejection fraction, ECG rhythm and heart failure treatment. Patients with malignant ventricular arrhythmias showed longer duration of QRS, left bundle branch block (LBBB), and alcoholic aetiology (Table 1). In multivariate analysis, the only independent predictors of malignant ventricular arrhythmias were LBBB (OR 2.4; CI95%: 1.2–5; p = 0.015) and alcoholic aetiology (OR 2.3; CI95%: 1.1–4.5; p = 0.026) (Table 2).
102
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Table 2 Multiple logistic regression analysis showing the independent predictors of malignant ventricular arrhythmias in the entire study cohort (n = 282).
Alcoholic aetiology Left bundle branch block
OR
95% CI
p value
2.3 2.4
1.1–4.5 1.2–5.0
0.026 0.015
Variables entered into multiple regression analysis: alcoholic aetiology (p = 0.034), Left bundle branch block (0.019).
ACM cohort In the ACM cohort, during a median follow-up of 59 months (interquartile range [IQR]: 25 to 107 months), 14 patients (15%) underwent orthotropic heart transplantation and 17 died (18%). Eight deaths were due to SCD, 6 from heart failure and 3 from malignancies. A total of 18 (19%) patients experienced 20 malignant ventricular arrhythmic events: 4 patients were resuscitated from documented VF, 8 received appropriate ICD therapies and 8 patients died suddenly (Table 3 and Fig. 2). An ICD was implanted for primary or secondary prophylaxis in 31 (33%) ACM patients; of these, 8 (26%) received appropriate ICD therapies. Rate of appropriate ICD therapy was 12 per 100 person–years (median followup period 20 months, IQR 7 to 34 months). One patient with an ICD–CRT implanted for primary prophylaxis died suddenly at home during sleep 9 months after device implantation. No autopsy was performed and the device was not checked. Among the 4 patients (4%) who received successful cardiopulmonary resuscitation, only 1 underwent subsequent ICD implantation for secondary prophylaxis and received an appropriate ICD discharge 39 months after device implantation. The other 3 patients did not receive an ICD due to severe neurologic impairment after the first SCD episode, other comorbidities or patient decision. After surviving a first malignant ventricular arrhythmia (n = 11 patients), 1 patient died due to SCD that occurred 1 month after the initial event (no ICD had been implanted), 2 patients died due to end-stage heart failure, 1 received an ICD discharge and 1 patient underwent cardiac transplantation (Fig. 2). Baseline clinical, echocardiographic and haemodynamic characteristics of ACM patients with and without malignant ventricular arrhythmias are presented in Table 4. No differences between groups were observed in terms of age, alcohol abstinence, ejection fraction, ECG rhythm and heart failure treatment. Although the use of certain heart failure therapies at baseline evaluation was lower than currently desired (beta-blockers were not proved to be beneficial at the start of study), heart failure treatment at last clinical evaluation was in accordance with modern practice. The rate of ACEI/ARB, beta-blockers and aldosterone antagonists at last follow-up was 84%, 76% and 57%, respectively. Ventricular tachycardia ablation procedures were not performed in any patient and no antiarrhythmic drugs, except for amiodarone, were prescribed. In baseline univariate analysis, ACM patients with ventricular arrhythmic events showed higher prevalence of QRS width N120 ms
Table 3 Malignant ventricular arrhythmias and end-stage heart failure in 94 ACM patients. Malignant ventricular arrhythmias Cardiopulmonary resuscitation Appropriate ICD therapy (n = 31) Sudden cardiac death End-stage heart failure Heart Transplant Death due to end-stage heart failure Other causes of death ICD = implantable cardiac defibrillator.
4 (4%) 8 (26%) 8 (9%) 14 (15%) 6 (6%) 3 (3%)
and LBBB, an increased left ventricular end-diastolic diameter and a lower rate of beta-blockers (Table 4). In multivariate analysis, the only independent predictor of malignant ventricular arrhythmias was LBBB (OR 11.2; CI95%: 2.6–50; p = 0.001) (Table 5). Although persistent exposition to alcohol was not significantly associated with malignant ventricular arrhythmias, it was close to significance (OR 4.3; CI95%: 0.99– 18.9; p = 0.051). Follow-up LVEF data were available in 92 patients (98%). At latest follow-up, 39 patients (41%) showed a LVEF ≥ 40%. None of them died from cardiac causes or underwent heart transplant. Only 1 patient with LVEF ≥ 40% died, but this was due to malignancy. Analysis of the relationship between LVEF and appearance of malignant ventricular arrhythmias revealed that there were no malignant ventricular arrhythmic events among patients with LVEF ≥ 40% or in NYHA I functional class. There were 34 ACM patients (36%) with LVEF b 40% at study entry and in whom LVEF recovered to ≥40% during follow-up. Four (12%) of these patients suffered malignant ventricular arrhythmias (all ICD discharges) while their LVEF was still b40%. During a median follow-up period of 34 months (IQR: 13–96) after achieving a LVEF ≥ 40%, all patients who recovered LVEF were free from malignant ventricular arrhythmias. Twenty-seven of the 34 (79%) ACM patients in whom LVEF recovered to ≥40% during followup had an echocardiogram performed within one year of baseline evaluation. Twenty patients (74%) exhibited a LVEF ≥ 40% on that evaluation, while the remaining 7 patients who did not recover LVEF during the first year exhibited a LVEF ≥40% in an echocardiogram performed after a median time of 29 months (IQR: 24–37) from baseline evaluation. Among the 7 patients who did not have LVEF examined during the first year, 6 of them had a LVEF ≥ 40% on the first echocardiographic evaluation performed after baseline examination that took place within a median time of 24 months (IQR: 19–43) from baseline visit. Discussion This study is the largest cohort of ACM patients described to date, the only one which includes modern therapies on heart failure, and the first focusing specifically on sudden death and ventricular arrhythmic complications. The study shows that malignant ventricular arrhythmias are a relatively frequent complication among ACM patients. In our series, ACM patients died more frequently from SCD and suffered more ICD discharges than patients with IDCM. Although LBBB was the only independent predictor of ventricular arrhythmias at baseline evaluation, our findings show that no patients suffered from ventricular arrhythmias when their LVEF was ≥ 40% or their NYHA was I. Furthermore, most patients who improved their LVEF to ≥40% did so during the first year of follow-up. Alcohol abuse and ventricular arrhythmias While low chronic alcohol intake reduces mortality, heart failure events and SCD, probably due to alcohol's anti-ischemic properties [15– 18], heavy alcohol consumption abuse has been shown to induce multiple cardiac changes that predispose to fatal arrhythmias. Those changes include ultrastructural changes, proarrhythmic electrolyte abnormalities, QT interval prolongation, adrenergic hyperactivity and a decreased heart rate variability and baroreceptor sensitivity [10,11]. Additionally, in animal models, ethanol reduces atrioventricular and intraventricular conduction, increases sinus node recovery time [19], causes shortening of the action potential and decreases contractility [10]. In line with such evidence, our study shows that malignant ventricular arrhythmias are a frequent complication in ACM. In fact, ACM patients died more frequently from SCD and suffered more ICD discharges than patients with IDCM.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
103
Fig. 2. Overview of clinical evolution and malignant ventricular arrhythmias in 94 ACM patients.
In spite of the epidemiological relevance of ACM in Western countries, little information is available on the incidence of ventricular arrhythmias in ACM because most of the available long-term prognostic studies in ACM did not specify the cause of death [6–9]. The only study addressing the incidence of malignant arrhythmias in ACM was conducted by Fauchier in France and was published in 2003 [13]. In this study, 75 patients were followed during 51 ± 42 months and 8% of them suffered from SCD or VF/sustained VT (16% of nonabstainers and 2% of abstainers). In our study, 18 patients (19%) suffered a malignant arrhythmic event, which is higher than the incidence reported before. This difference could be due to the inclusion of appropriate ICD therapies as end-point in our study, and to the increased transplant-free survival reported in our cohort (62% vs 50% and 63% vs 32% at 10 years among abstainers and non-abstainers, respectively) [8,12,13]. Increased end-stage heart failure survival increases the chance of suffering from arrhythmic complications, a phenomenon previously reported in other heart failure studies [20]. Predictors of ventricular arrhythmias Prediction of ventricular arrhythmic events in heart failure and DCM is still a complex and controversial issue. Several variables have been proposed, but an effective stratification of the risk of SCD has not yet been identified. Our results show that alcoholic aetiology should be considered as a risk factor in this context because ACM patients suffered a higher rate of malignant ventricular arrhythmias. Appropriate prediction of ventricular arrhythmic events in ACM patients is crucial to define which patients will benefit from prophylactic ICD implantation. In our study, the only independent predictor of malignant ventricular arrhythmias was LBBB. This factor is a well-known predictor of adverse prognosis in DCM [21,22], and was also shown to predict impaired transplant-free survival in the overall ACM cohort [12]. The only available data concerning long-term predictors of arrhythmic events in ACM come also from Fauchier's study [13]. He found that ACM patients with persistent moderate alcohol intake or abstinence had only 2% of arrhythmic events during follow-up [13]. In contrast, relative risk of arrhythmic events of ACM patients who retained a persistent heavy alcohol intake was 8.0 compared with ACM patients with moderate alcohol intake or abstinence [13]. In our study, persistent alcohol intake was not associated with malignant ventricular arrhythmic events in multivariate analysis although it was close to significance. In contrast to Fauchier's work, we decided to
include moderate drinkers and heavy drinkers in the non-abstainers group due to the low rate of heavy drinkers found in our study. Finally, although baseline LVEF and NYHA did not reach statistical significance as clinical predictors of arrhythmic events, none of the ICD appropriate therapies, SCD or recovered cardiac arrests took place in patients with baseline LVEF ≥ 40% or NYHA I. Moreover, in ACM patients with impaired LVEF at baseline, malignant arrhythmias did not occur once LVEF was over 40%. As LVEF recovery was associated with such a good prognosis in our study, and considering that LVEF could improve substantially in ACM after alcohol withdrawal, we sought to investigate the time required to reach the 40% threshold among improvers. To date, only a French study focusing on 26 ACM heart transplant candidates who recovered LVEF with abstinence has investigated the time required to achieve LVEF improvement in ACM [23]. In this study, the authors recommended not to place ACM patients on heart transplant waiting list until 3 months of absence of improvement with abstinence, although the reported average interval between abstinence and recovery in the study was 11.7 months [23]. In our study, 46 patients with LVEF b 40% at baseline evaluation did not have an event and underwent echocardiographic evaluation at one year. Among these patients, 20 (43%) showed LVEF recovery during their first year of follow-up while only 7 (15%) achieved it subsequently. Overall, our findings reflect that when LVEF has not improved to ≥40% during the first year of follow-up in ACM, the chance of a late improvement is low and therefore an ICD implant should be considered at that time.
Limitations Our study cohort was obtained from a single heart transplant centre and that may have resulted in a referral bias toward young and more affected ACM patients prone to abandon alcohol consumption. Also, because of the retrospective nature of the study, LVEF was not evaluated in all patients at the same time intervals. Additionally, and similar to all previous ACM studies, the identification of ACM patients and their allocation into abstainers or persistent drinker groups during follow-up was based on patients' self-reported alcohol intake, which may lead to underestimation. Finally, the multiple logistic regression analysis presented was based on a stepwise selection procedure from a larger set of candidate predictor variables found in univariate analysis. Therefore, there is some risk of identifying a false positive predictor due to this multiplicity issue.
104
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105
Table 4 Clinical, electrocardiographic, echocardiographic and haemodynamic characteristics at first evaluation of ACM patients with and without malignant ventricular arrhythmias.
Mean age (years) Sex (%) Male Female Baseline NYHA class I (%) II (%) III (%) IV (%) Comorbidities, (%) Hypertension (%) Dyslipidaemia (%) Diabetes (%) Smoker (%) Body mass index (kg/m2) COPD (%) Alcohol consumption Duration of alcohol abuse (years) Mean alcohol consumption (g/day) Previous alcohol consumption (groups) b120 g/day (%) 120–160 g/day (%) N160 g/day (%) Alcohol consumption during follow-up Alcohol abstinence (%) Persistent alcohol intake (%) Blood pressure and pulse Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Pulse (beats/min) Blood test Haemoglobin (g/dl) Creatinine (mg/dl) Sodium (mg/dl) Bilirubin (mg/dl) ALAT (U/L) ASAT (U/L) GGT (U/L) ECG Sinus rhythm, (%) Atrial fibrillation, (%) QRS ≥ 120 msec (%) QRS duration (ms) Left bundle branch block (%) Echocardiogram LVEF (%) LVEDD (mm) LVESD (mm) Right heart catheterization⁎ SPAP (mm Hg) DPAP (mm Hg) MPAP (mm Hg) PCW (mm Hg) CO (l/min) CI (l/min/m2) Exercise tests 6-minute walking test† (m) Peak oxygen uptake‡ (l/kg/min) Treatment ACEI or ARB (%) Digoxin (%) Loop diuretics (%) Spironolactone/eplerenone (%) Beta-blockers (%) Amiodarone (%) ICD (%) CRT (%)
Malignant ventricular arrhythmias (n = 18)
No malignant ventricular arrhythmias (n = 76)
p value
52 ± 7
49 ± 10
0.259 0.039
94 6
100 0
11 39 28 22
6 23 39 32
39 44 22 72 29 ± 6 39
35 26 24 45 28 ± 4 29
0.789 0.130 0.895 0.110 0.382 0.412
21 ± 13 143 ± 58
24 ± 12 135 ± 66
0.463 0.634 0.665
44 19 37
51 24 26
56 44
65 35
113 ± 15 75 ± 10 90 ± 23
121 ± 21 78 ± 17 92 ± 25
0.216 0.559 0.805
14 ± 1.5 1.4 ± 0.3 137 ± 3 1.6 ± 1 62 ± 79 70 ± 151 94 ± 89
15 ± 1.4 1.2 ± 0.3 137 ± 4 2±4 93 ± 247 53 ± 85 165 ± 268
0.478 0.157 0.893 0.706 0.697 0.603 0.433
56 44 69 123 ± 24 59
68 32 32 108 ± 30 26
0.430 0.430 0.020 0.063 0.032
25 ± 9 73 ± 11 58 ± 12
26 ± 9 67 ± 9 57 ± 9
0.726 0.024 0.696
48 ± 26 26 ± 14 37 ± 19 24 ± 15 3.5 ± 0.7 2 ± 0.4
46 ± 16 24 ± 10 31 ± 12 24 ± 15 4.2 ± 1.3 2.3 ± 0.5
0.801 0.644 0.322 0.970 0.155 0.274
283 ± 156 11 ± 8
373 ± 74 16 ± 6
0.041 0.166
83 61 72 50 39 17 50 16
92 45 77 49 64 20 30 28
0.255 0.208 0.668 0.918 0.047 0.730 0.112 0.235
0.371
0.482
Values are means ± SD. NYHA indicates New York Heart Association; COPD, chronic obstructive pulmonary disease; ECG, electrocardiogram; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVEF, left ventricular ejection fraction. SPAP, systolic pulmonary artery pressure; DPAP, diastolic pulmonary artery pressure; MPAP, mean pulmonary artery pressure; PCW, pulmonary capillary wedge pressure; CO, cardiac output; CI, cardiac index; ACEI, angiotensin converting enzyme inhibitors; ARB, angiotensin receptor blockers; ICD, implantable cardiac defibrillator; CRT, cardiac resynchronisation therapy. ⁎ 9 ACM patients with malignant ventricular arrhythmias and 31 ACM patients without malignant ventricular arrhythmias underwent right heart catheterization. † 6 ACM patients with malignant ventricular arrhythmias and 26 ACM patients without malignant ventricular arrhythmias underwent 6-min walking test. ‡ 5 ACM patients with malignant ventricular arrhythmias and 23 ACM patients without malignant ventricular arrhythmias underwent exercise test with O2 consumption.
G. Guzzo-Merello et al. / International Journal of Cardiology 199 (2015) 99–105 Table 5 Multiple logistic regression analysis showing the independent predictors for malignant ventricular arrhythmias in ACM.
Left bundle branch block Persistent alcohol intake
OR
95% CI
P value
11.2 4.3
2.6–50 0.99–18.9
0.001 0.051
OR = odds ratio; CI = confidence interval. Variables entered into multiple regression analysis: Absence of beta-blockers (p = 0.047), female gender (p = 0.039), persistent alcohol intake (p = 0.482), left bundle branch block (p = 0.032) left-ventricle end-diastolic diameter (p = 0.024) and QRS width N120 ms (p = 0.020)
Conclusions Malignant ventricular arrhythmias are a frequent complication in ACM. SCD and ICD discharges are more frequent in ACM than in IDCM. LBBB was identified as the only independent predictor of ventricular arrhythmic events in ACM at baseline evaluation. In our study, no patient suffered an arrhythmic event if LVEF was ≥40% and most ACM patients in whom LVEF recovered did so during the first year of follow-up. The results of our study could be of help when deciding ICD placement in this complex condition. Conflict of interest The authors report no relationships that could be construed as a conflict of interest. Acknowledgements We gratefully acknowledge Kenneth McCreath for English editing. References [1] C.L. Bryson, K.J. Mukamal, M.A. Mittleman, et al., The association of alcohol consumption and incident heart failure: the Cardiovascular Health Study, J. Am. Coll. Cardiol. 48 (2006) 305–311. [2] J.L. Abramson, S.A. Williams, H.M. Krumholz, et al., Moderate alcohol consumption and risk of heart failure among older persons, JAMA 285 (2001) 1971–1977. [3] R. Movva, V.M. Figueredo, Alcohol and the heart: to abstain or not to abstain? Int. J. Cardiol. 164 (2013) 267–276.
105
[4] I. Laonigro, M. Correale, M. Di Biase, E. Altomare, Alcohol abuse and heart failure, Eur. J. Heart Fail. 11 (2009) 453–462. [5] G. Guzzo-Merello, M. Cobo-Marcos, M. Gallego-Delgado, P. Garcia-Pavia, Alcoholic cardiomyopathy, World J. Cardiol. 6 (2014) 771–781. [6] C.J. McKenna, M.B. Codd, H.A. McCann, D.D. Sugrue, Alcohol consumption and idiopathic dilated cardiomyopathy: a case control study, Am. Heart J. 135 (1998) 833–837. [7] A. Gavazzi, R. De Maria, M. Parolini, M. Porcu, Alcohol abuse and dilated cardiomyopathy in men, Am. J. Cardiol. 85 (2000) 1114–1118. [8] L. Fauchier, D. Babuty, P. Poret, et al., Comparison of long-term outcome of alcoholic and idiopathic dilated cardiomyopathy, Eur. Heart J. 21 (2000) 306–314. [9] P. Prazak, M. Pfisterer, S. Osswald, P. Buser, F. Burkart, Differences of disease progression in congestive heart failure due to alcoholic as compared to idiopathic dilated cardiomyopathy, Eur. Heart J. 17 (1996) 251–257. [10] A. George, V.M. Figueredo, Alcohol and arrhythmias: a comprehensive review, J. Cardiovasc. Med. (Hagerstown) 11 (2010) 221–228. [11] M. Kupari, P. Koskinen, Alcohol, cardiac arrhythmias and sudden death, Novartis Found. Symp. 216 (1998) 68–79. [12] G. Guzzo-Merello, J. Segovia, F. Dominguez, et al., Natural history and prognostic factors in alcoholic cardiomyopathy, JACC Heart Fail. 3 (2015) 78–86. [13] L. Fauchier, Alcoholic cardiomyopathy and ventricular arrhythmias, Chest 123 (2003) 1320. [14] P. Richardson, W. McKenna, M. Bristow, et al., Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies, Circulation 93 (1996) 841–842. [15] G. Wannamethee, A.G. Shaper, Alcohol and sudden cardiac death, Br. Heart J. 68 (1992) 443–448. [16] K. Kramer, L. Kuller, R. Ficher, The increasing mortality attributed to cirrhosis and fatty liver, in Baltimore (1957–1966), Ann. Intern. Med. 69 (1968) 273–282. [17] B. Randall, Sudden death and hepatic fatty metamorphosis. A North Carolina survey, JAMA 243 (1980) 1723–1725. [18] C.M. Albert, J.E. Manson, N.R. Cook, U.A. Ajani, J.M. Gaziano, C.H. Hennekens, Moderate alcohol consumption and the risk of sudden cardiac death among US male physicians, Circulation 100 (1999) 944–950. [19] M.J. Goodkind, N.H. Gerber Jr., J.R. Mellen, J.B. Kostis, Altered intracardiac conduction after acute administration of ethanol in the dog, J. Pharmacol. Exp. Ther. 194 (1975) 633–638. [20] J. Sengupta, R. Abdelhadi, Approach to reduction of ventricular arrhythmias and implantable cardioverter–defibrillator therapies in patients with heart failure, Curr. Opin. Cardiol. 28 (2013) 337–343. [21] M. Hofmann, R. Bauer, R. Handrock, G. Weidinger, L. Goedel-Meinen, Prognostic value of the QRS duration in patients with heart failure: a subgroup analysis from 24 centers of Val-HeFT, J. Card. Fail. 11 (2005) 523–528. [22] I. Sipahi, S.P. Chaudhry, M. Issa, M. Ambati, S. Dagdelen, J.C. Fang, Impact of QRS duration on survival benefit with prophylactic implantable cardioverter–defibrillators: a meta-analysis of randomized controlled trials, J. Cardiovasc. Med. (Hagerstown) (Jul 10 2014) (Epub ahead of print). [23] E. Coignard, B. Blanchard, F. Jault, et al., Alcoholic cardiomyopathy and heart transplantation, Arch. Mal. Coeur Vaiss. 91 (1998) 45–51.