Author + information
- Received November 2, 2016
- Revision received February 8, 2017
- Accepted February 10, 2017
- Published online May 29, 2017.
- Stephan Schürer, MDa,
- Karin Klingel, MDb,
- Marcus Sandri, MDa,
- Nicolas Majunke, MDa,
- Christian Besler, MDa,
- Reinhard Kandolf, MDb,
- Philipp Lurz, MDa,
- Michael Luck, MDa,
- Pia Hertel, MSca,
- Gerhard Schuler, MDa,
- Axel Linke, MDa and
- Norman Mangner, MDa,∗ ()
- aDepartment of Internal Medicine/Cardiology, University of Leipzig, Heart Center Leipzig, Leipzig, Germany
- bDepartment of Molecular Pathology, University of Tübingen, Tübingen, Germany
- ↵∗Address for correspondence:
Dr. Norman Mangner, University of Leipzig, Heart Center Leipzig, Struempellstrasse 39, D-04289 Leipzig, Germany.
Objectives This study aimed to assess characteristics including endomyocardial biopsy and outcome of patients with methamphetamine (MA)-associated cardiomyopathy in a series of patients treated in Germany.
Background MA abuse is an increasing problem worldwide.
Methods The cases of 30 consecutive MA-abusing patients with a left ventricular (LV) ejection fraction of <40% and endomyocardial biopsy performed at initial diagnosis were analyzed. Baseline characteristics were collected retrospectively, whereas follow-up was prospective. The primary endpoint was a composite of death, nonfatal stroke, and rehospitalization for heart failure.
Results Patients were 30.3 ± 1.9 years of age, predominantly male (93.3%), and highly symptomatic; 83.3% had New York Heart Association functional class III or IV dyspnea. Echocardiography revealed marked LV dilatation (mean LV end-diastolic diameter 67.1 ± 7.4 mm) and impaired LV ejection fraction (mean 19 ± 6%). One-third of the patients had intraventricular thrombi. Endomyocardial biopsy revealed markers of inflammation and fibrosis; the fibrosis correlated with the duration of MA abuse. At follow-up, discontinuation of MA abuse together with medical therapy partially improved cardiac function (LV ejection fraction, 19 ± 6 vs. 43 ± 13; p < 0.001) and symptoms (p = 0.056), whereas patients with continued abuse did not show any improvement. The improvement in cardiac function was independently associated with the extent of fibrosis. The primary endpoint occurred more often in patients with continued MA abuse (57.1% vs. 17.4%; p = 0.037).
Conclusions MA-associated cardiomyopathy is characterized by severe heart failure and depressed cardiac function. The extent of myocardial fibrosis seems to predict the recoverability of LV function. Cessation of MA abuse is associated with improvement in cardiac function and symptoms, whereas continued MA abuse leads to ongoing heart failure and worse outcome.
Drug abuse and its consequences are burdens on many societies worldwide. The United Nations World Drug Report of 2015 estimated that approximately 250,000,000 people between the ages of 15 and 64 years used an illicit drug (global prevalence 5.2%) (1). Methamphetamine (MA) and related substances have become the second most frequently used drugs in the world. In 2013, there were approximately 33,900,000 users worldwide (1).
Cardiac complications play a significant role in MA-related morbidity and include malignant hypertension, arrhythmias, aortic dissection, myocardial infarction secondary to vasospasm, stroke, and MA-associated cardiomyopathy (MACM) (2). Moreover, cardiovascular complications are the leading causes of death and were found in up to three-fourths of MA abusers in an Australian cohort (2). Various mechanisms (e.g., catecholamine excess and/or direct toxic effects of MA to the myocytes) are supposed to lead to cardiac complications, in particular MACM (3).
However, data on the characteristics of histopathological changes in the myocardium and the outcome of patients with MACM are rare. In particular, the reversibility of MACM after discontinuation of MA abuse has not been documented in detail (4–6). Thus we aimed to assess clinical characteristics, histopathological features, and clinical outcome in a well-defined cohort of patients presenting with MACM and to evaluate the impact of MA abuse discontinuation as opposed to uninterrupted MA abuse.
Patients and follow-up
Overall, 30 patients were recruited for this study: 15 patients from the Heart Center Leipzig in Leipzig, Germany; and 15 patients identified by the Department of Molecular Pathology at the University of Tübingen in Tübingen, Germany. The initial diagnosis of MACM was established between 2007 and 2016. For inclusion, patients had to have a left ventricular (LV) ejection fraction (LVEF) of <40% with endomyocardial biopsy (EMB) performed at the initial diagnosis. Baseline characteristics were collected retrospectively, whereas follow-up was prospective with visits in Leipzig for clinical and echocardiographic evaluation. Time from diagnosis until inclusion in this study was 23 ± 23 months, and prospective follow-up was 12 ± 7 months. Outcome data were available in all patients; however, echocardiographic data were available in only 27 patients (90%) because of the death of 1 patient and refusal by 2 patients. The local ethics committee approved the study, and all patients gave written informed consent at follow-up.
The primary clinical endpoint was a composite of death, nonfatal stroke, and rehospitalization for heart failure. Secondary endpoints included the composite of death and rehospitalization for heart failure. Furthermore, the development of LVEF and symptoms were evaluated according to the status of continued or discontinued MA abuse at follow-up.
Every subject underwent a standardized 2-dimensional echocardiographic examination using commercial ultrasound systems and analyzed according to current recommendations (7). Diagnostic coronary angiography was performed according to local standards. EMB specimens were taken from the left ventricle, right ventricle, or both in 18, 9, and 3 patients, respectively.
Analysis of endomyocardial biopsies
EMB specimens were fixed in 4% buffered formaldehyde for immunohistological examination. Another sample was fixed in RNAlater (Ambion, Inc., Foster City, California) for detection of viral genomes by nested real-time polymerase chain reaction. Paraffin-embedded EMB specimens were stained with Masson’s trichrome reflecting myocyte necrosis, as well as interstitial fibrosis, and were analyzed by light microscopy (8). The extent of myocardial fibrosis in EMB specimens was defined as an index, as described previously (8), and patients were classified as having no or mild, moderate, or severe fibrosis.
The analysis of inflammation was performed as described previously (9), and we refined the degree of inflammation according to a modified scheme as described (9) (grade 0 = no inflammation; grade 1 = single inflammatory cells [T lymphocytes and macrophages ≥14/mm2]; grade 2 = a few foci of inflammation; grade 3 = several foci of inflammation; and grade 4 = pronounced inflammation).
To elucidate the effect of MA on inflammation and fibrosis, we compared the findings in patients with MACM with 31 age-, sex-, and LVEF matched patients with dilated cardiomyopathy (DCM).
Categorical variables are expressed as numbers and percentages and were compared with the use of the Fisher exact test. Continuous variables are expressed as mean ± SD and were compared using the 2-sided, Student t test or the Mann-Whitney U test, as appropriate. Predictors of improvement of LVEF at follow-up were analyzed using a multiple linear regression model including 3 clinical meaningful factors that showed the highest predictor importance in an automatic linear modeling using the Akaike information criterion. Time-to-event curves were analyzed according to the method of Kaplan-Meier, and group comparisons were made by applying the log-rank test. Significance was accepted as p < 0.05. All analysis was performed with the use of SPSS software version 21 (IBM, Armonk, New York).
The baseline characteristics of all patients (n = 30) and of patients who continued (n = 7) and discontinued MA abuse (n = 23) at follow-up are shown in Table 1. Patients were 30.3 ± 1.9 years of age and predominantly male (93.3%). Patients were highly symptomatic; 83.3% of patients had New York Heart Association functional class III or IV dyspnea. Other symptoms included angina pectoris, palpitations, cough, and hemoptysis, in descending order. Three patients presented with cardiogenic shock, and 1 patient presented with endocarditis leading to severe aortic regurgitation. Another patient presented with sustained ventricular tachycardia. During initial hospitalization, 1 patient had a minor stroke resulting from an LV thrombus. Pre-existing comorbidities were infrequent; however, approximately 15% of the patients had a positive family history for DCM or sudden cardiac death.
All patients used MA, but only 3 patients (10%) used exclusively MA. The other patients abused the following drugs, in descending order: nicotine; alcohol; cannabis; heroin; and cocaine. The mean duration of MA abuse was 5.7 years, and it ranged from 1 to 15 years.
Overall, there were no significant differences in baseline characteristics between patients with continued and discontinued MA abuse (Table 1, Online Table 1). However, patients with continued MA abuse were numerically younger and abused MA for a longer time.
Diagnostic findings at baseline
At baseline (Table 1), echocardiography revealed severely impaired LVEF (mean LVEF 19 ± 6%) and LV dilatation (mean LV end-diastolic diameter 67.1 ± 7.4 mm). Furthermore, left atrial and right ventricular dilatation was evident. Right ventricular function was impaired. LV and/or right ventricular thrombi were found in 10 patients (33.3%). Pleural effusion (56.7%) and pericardial effusion (40%) were frequent, and ascites was found in 14.3%. Significant valve regurgitation (≥grade 2 on a scale of 3) was found in 12 patients (40%) in the mitral valve, in 6 patients (26.1%) in the tricuspid valve, and in 1 patient (4.3%) in the aortic valve.
Elevated N-terminal pro–B-type natriuretic peptide was found in all patients with data on this substance (n = 22), with a mean value of 5,553 ± 6,112 ng/l. Other altered parameters included troponin, creatinine, and liver enzymes (Online Table 1).
All patients underwent cardiac catheterization on initial diagnosis, and coronary artery disease was excluded in 29 patients. In 1 patient, high-grade stenosis of a small right ventricular branch without indication for revascularization was found. All patients underwent EMB on initial diagnosis.
Initial therapy and in-hospital outcome
The 3 patients in cardiogenic shock were successfully treated by extracorporeal membrane oxygenation, intra-aortic balloon counterpulsation, and medical treatment, respectively. The patient presenting with infective endocarditis and severe aortic regurgitation received mechanical aortic and mitral valve replacement. The patient with sustained ventricular tachycardia received an implantable cardioverter-defibrillator during the initial hospitalization, whereas 10 patients were discharged with a wearable cardioverter-defibrillator. Medical therapy included guideline-supported medication; patients with persistent intraventricular thrombus received oral anticoagulation (Table 1). None of the 30 patients died during the initial hospitalization.
Follow-up and outcome
At follow-up, 23 patients stopped using MA, and 7 patients continued MA abuse. Mean follow-up was 35 ± 22 months, with no significant difference between both groups (Table 2). Continued MA abuse was associated not only with longer exposure to MA at the time of follow-up (p = 0.003) but also with persistent, severely impaired LVEF and dilatation; conversely, MA discontinuation led to LV remodeling and improved LVEF (Figure 1, Table 2, Online Table 2). Furthermore, discontinuation of MA was related to improved symptoms, whereas patients with continued MA abuse remained predominantly in New York Heart Association functional class II or III (Figure 2).
After a mean follow-up period of 35 ± 22 months, 3 patients had died, 1 patient with discontinued MA abuse and 2 patients with continued MA abuse (p = 0.128). The composite endpoints of death, nonfatal stroke, and rehospitalization for heart failure (p = 0.037) and death and rehospitalization for heart failure (p = 0.016) occurred more frequently in patients with continued MA abuse (Table 2, Figure 3).
Histological analysis of endomyocardial biopsies
Major histological findings included markers of inflammation, fibrosis, and myocyte damage (Figures 4A to 4D and 5). Severity of inflammation was equally distributed over the duration of MA abuse. In contrast, fibrosis (r = 0.57; p = 0.002) was related to the duration of MA abuse and occurred in a more severe form after longer abuse. When comparing patients with continued and discontinued MA abuse at follow-up, patients with continued abuse already had more severe fibrosis at baseline, whereas no difference was found for inflammation (Figures 4A to 4D). Only 3 patients tested positive for a viral genome, including Epstein-Barr virus, parvovirus B-19, and human herpesvirus-6. According to the traditional Dallas criteria, 4 patients (13.3%) had acute necrosis (definitive myocarditis), including only 1 patient tested positive for human herpesvirus-6. The other 26 patients (86.7%) were classified as having borderline myocarditis.
Compared with age-, sex-, and LVEF-matched patients with DCM (age 29.5 ± 1.2 years, 93.5% male, LVEF 22 ± 6%), patients with MACM showed more severe inflammation (no or mild inflammation, 100% vs. 46.6%; moderate inflammation, 0% vs. 40:0%; severe inflammation, 0% vs. 13.3%; p < 0.001), whereas fibrosis was more pronounced in patients with DCM (no or mild fibrosis, 0% vs. 33.3%; moderate fibrosis, 51.6% vs. 40:0%; severe fibrosis, 48.4% vs. 26.7%; p = 0.002).
Predictors of LVEF development at follow-up
In univariate analysis, histological markers of fibrosis, duration of MA abuse, and continued MA abuse at the time point of follow-up were related to LVEF at follow-up. In multivariate analysis, only fibrosis remained independently associated with LVEF at follow-up (Table 3). Exclusion of patients with continued MA abuse at follow-up slightly attenuated the correlation between fibrosis and improvement of LVEF, but it remained significant (Online Table 3).
The purpose of this registry was to determine the characteristics and clinical course of MACM. In this study, all patients had an invasive evaluation of coronary artery disease and EMB. Specifically, the major findings of this study include the following: 1) MA abuse was associated with severe heart failure with severely depressed LVEF and dilatation of the cardiac chambers; 2) MACM was associated with the development of intracardiac thrombi in one-third of the study subjects; 3) discontinuation of MA abuse at follow-up was associated with an improvement of symptoms and cardiac function, whereas continued abuse was related to a persistently impaired functional status and cardiac function leading to 4) an increased risk of death, nonfatal stroke, and rehospitalization for heart failure; 5) fibrosis was closely related to a longer exposure of MA and was more pronounced in patients with continued MA abuse at follow-up; and 6) fibrosis, assessed in the EMB, seemed to be an independent predictor of the development of LVEF at follow-up.
Clinical characteristics of MACM
Several case reports and some larger case series, mainly from the United States and Australia, have described MACM as a potential cardiac complication of MA abuse (6,10–12). With the more recent spread of crystal methamphetamine, MACM and other MA-associated complications have become enormous health care problems in Germany, particularly in the eastern regions bordering the Czech Republic (13).
According to demographics, MACM is a disease of young, predominantly male patients. In our study, the mean age was 30 years, comparable to the mean age in other published registries (14). In contrast, the onset of alcoholic cardiomyopathy (49 ± 11 years) and idiopathic DCM (52 ± 13 years) occurs more frequently in middle-aged patients (15).
The most prevalent symptoms in our cohort included dyspnea and angina pectoris. Whereas dyspnea is associated with severely impaired LVEF and consequently elevated filling pressures, angina pectoris is also common in MA abusers, and it is frequently attributed to transient coronary artery spasm (16). There is a significant association between MA abuse and acute myocardial infarction (17), and U.S. data suggest a 25% incidence of acute coronary syndrome among MA abusers with chest pain (18).
MACM has been described as DCM (10,11), hypertrophic cardiomyopathy (19), or Takotsubo-like cardiomyopathy (20,21). In our study, only patients with DCM were included. LVEF was severely depressed, and the left ventricle was markedly dilated at baseline. Hence mitral regurgitation and tricuspid regurgitation were frequent. In a study by Ito et al. (10), MA abusers had echocardiographic findings of more severe DCM compared with nonabusers, including larger parameters of the left atrium, left ventricle, and right ventricle.
The more severe dilatation may be an important factor contributing to the high prevalence of intracardiac thrombi affecting one-third of our patients. Two published case reports described thrombotic complications in MA abusers, 1 report of a patient presenting with biventricular thrombi, and the other report describing a patient with simultaneous thrombosis of the left anterior descending artery and right coronary artery (22,23). In a retrospective analysis including different cardiomyopathies, LV thrombi were found in ∼20% of the patients, and the occurrence of these thrombi was predicted by ischemic cardiomyopathy and LV end-diastolic diameter >60 mm (24). However, one may assume that factors other than low flow and dilatation (e.g., disturbances in the coagulation system) may also contribute to the high prevalence of intracardiac thrombi in MA abusers. Further studies are warranted to examine those aspects of MA abuse.
Treatment and outcome of MACM
Treatment included supportive measures and guideline-supported medical therapy (25). One patient received an automatic implantable cardioverter-defibrillator during the index hospitalization, whereas one-third of the patients in this study were discharged with a wearable cardioverter-defibrillator. At follow-up, 8 more patients had received an automatic implantable cardioverter-defibrillator or a cardiac resynchronization therapy defibrillator, with no difference between continued and discontinued MA abusers.
The chance to improve symptoms and cardiac function was significantly higher in patients discontinuing MA abuse. Consequently, patients with discontinued MA abuse had a lower incidence of the primary endpoint including death, nonfatal stroke, and rehospitalization for heart failure. There was no difference in medical therapy between patients with continued and discontinued MA abuse, but we cannot rule out possible noncompliance in patients who continued MA abuse, thereby influencing clinical and echocardiographic findings. Data about compliance with medical therapy for heart failure in MA abusers are missing, but MA abusers with human immunodeficiency virus infection are known to have poor adherence to antiretroviral therapy (26). Conversely, the improvement of symptoms and cardiac function seen in our cohort is in line with the findings of a study from California that showed improved New York Heart Association functional class (p = 0.001) and LVEF (p = 0.06) in patients who stopped MA abuse at follow-up (5).
Predictors of outcome in MACM
The only independent predictor of the development of LVEF at follow-up was the evidence of moderate or severe fibrosis in the EMB specimen at baseline. Fibrosis was already supposed to be a main prognostic factor for future development of LVEF in MA abusers (5,6), but this study showed that the degree of fibrosis, as assessed in an EMB, is associated with continuation and longer duration of MA abuse and leads to irreversible changes of the left ventricle and poor cardiac function at follow-up. The degree of fibrosis was also correlated with LVEF at follow-up, if patients with continued MA abuse were excluded. This observation indicated that possible noncompliance with medical therapy in continued MA abusers seemed not to be the main contributor to this finding and suggested that fibrosis itself is a predictor of LVEF development. In a case report by Lopez et al. (27), the use of late gadolinium-enhanced (LGE) cardiac magnetic resonance (CMR) was described in an MA-abusing patient, who had no relevant LGE and full recovery of cardiac function at 6 months. Therefore one may hypothesize that LGE CMR is clinically relevant, and the extent of fibrosis measured by magnetic resonance imaging may predict the development of LVEF in those patients (6). This hypothesis is supported by a study in patients with newly diagnosed DCM that showed that the extent of myocardial fibrosis assessed by magnetic resonance imaging was also independently associated with a lack of response to medical therapy (28). Unfortunately, we could not evaluate LGE CMR in our cohort because of missing data and different CMR protocols used.
Possible pathological mechanism in MACM
The pathology of MACM is complex and multifactorial. Catecholamine excess, direct toxic effects, vasospasm, ischemia, reactive oxygen species, and mitochondrial and metabolic alterations are possible mechanisms of myocardial injury (3,4,29). Atrophy, hypertrophy, patchy cellular infiltration, eosinophilic degeneration and disarray, edema, myolysis, fibrosis, and the appearance of vacuoles were histological features in rats treated with MA for several weeks. Withdrawal of MA in those rats was associated with a gradual recovery of the myocardial changes (4,30). Similar findings were made in autopsy studies of MA abusers (31). Furthermore, these findings are comparable to those in other hyperadrenergic states (e.g., pheochromocytoma). A recent publication by Ferreira et al. (32) described focal and diffuse myocardial fibrosis and myocardial dysfunction in patients with pheochromocytoma. These investigators supposed that the catecholamine toxicity to the heart may lead to a form of catecholamine-induced myocarditis and consecutively to myocardial fibrosis (32). This concept is supported by our comparison with matched patients with DCM showing that MACM was associated with inflammation and fibrosis, whereas fibrosis in DCM developed without relevant inflammation. Furthermore, 4 patients were classified as having “definitive myocarditis” according to the classical Dallas criteria, whereas only 1 of these patients was virus positive, thus supporting the hypothesis that MA itself or its consequences led to inflammation, necrosis, and subsequent fibrosis. Overall, a viral genome was detected in only 3 patients. This finding suggests that MA abuse and not masked myocarditis was responsible for the histological alterations.
We used retrospective data for the baseline characteristics, but follow-up was prospective. Therefore we had to rely on the available documentation; in particular, urine tests were not available for all patients. However, at follow-up all patients confirmed former or current MA abuse. Furthermore, sample size, in particular for patients with continued MA abuse, is small. Therefore our multiple regression analysis should be interpreted as hypothesis generating, and the findings need to be confirmed in larger cohorts that evaluate the prognostic impact of fibrosis in MA abusers. Moreover, quantifying fibrosis from biopsy samples is subject to several potential biases that limit the precision of estimates of total myocardial fibrosis. The most precise methodology would require an analysis of the entire heart (e.g., applying LGE CMR for prediction of outcome in patients with MACM).
Another major limitation is possible noncompliance to medical therapy, in particular by patients with continued MA abuse. We had to rely on patients’ ability to give accurate information without having an objective control.
We included only patients with heart failure and severely impaired cardiac function. Therefore we have no information on the evolution of MACM and possible early, asymptomatic phases. Future prospective studies are needed to describe this form of cardiomyopathy with respect to the possible toxic dose, early asymptomatic phases, prevalence, and possible individual risk factors for the development of MACM.
This study describes the deleterious effects of MA on the myocardium, including a full data set on histological changes, which included findings such as myocyte damage, fibrosis, and inflammation in patients with severely reduced LVEF and signs and symptoms of heart failure. Discontinuation of MA abuse together with guideline-based medical therapy for heart failure partially improved cardiac function and symptoms, whereas patients with continued MA abuse did not show any improvement. The improvement of cardiac function seemed to depend on the extent of fibrosis as a marker of irreversible myocardial damage.
With the increasing abuse of MA worldwide, MACM will become a growing cause of heart failure in young adults. Because of the chance to recover cardiac function and symptoms at an early stage of the disease, early detection of MACM in patients with MA abuse could prevent further deterioration of this cardiomyopathy.
COMPETENCY IN MEDICAL KNOWLEDGE: This registry revealed that MACM is characterized by severe heart failure and depressed cardiac function. Cessation of MA abuse led to improved cardiac function and symptoms, whereas continued MA abuse led to ongoing heart failure and worse outcome. The extent of myocardial fibrosis noted on EMB specimens predicted the recoverability of LV function.
TRANSLATIONAL OUTLOOK: Future studies are necessary to identify patients with MA abuse at an early stage of the disease to prevent severe heart failure. Noninvasive imaging modalities such as MRI showing the extent of myocardial fibrosis may be helpful to predict the outcome of patients with MACM.
For supplemental tables, please see the online version of this article.
Dr. Linke has received speaker and consultant fees from Medtronic, St. Jude Medical, Edwards Lifesciences, Boston Scientific, Symetis, Bard, and Claret Medical; and has received stock option fees from Claret Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Schürer and Klingel contributed equally to this work.
- Abbreviations and Acronyms
- cardiac magnetic resonance
- dilated cardiomyopathy
- endomyocardial biopsy
- late gadolinium-enhanced
- left ventricular
- left ventricular ejection fraction
- methamphetamine-associated cardiomyopathy
- Received November 2, 2016.
- Revision received February 8, 2017.
- Accepted February 10, 2017.
- 2017 American College of Cardiology Foundation
- United Nations Office on Drugs and Crime
- Paratz E.D.,
- Cunningham N.J.,
- MacIsaac A.I.
- Sliman S.,
- Waalen J.,
- Shaw D.
- Voskoboinik A.,
- Ihle J.F.,
- Bloom J.E.,
- Kaye D.M.
- Lang R.M.,
- Badano L.P.,
- Mor-Avi V.,
- et al.
- Zuern C.S.,
- Walker B.,
- Sauter M.,
- et al.
- Wetzel W.,
- Rothe A.
- Fauchier L.,
- Babuty D.,
- Poret P.,
- et al.
- Janardhanan R.,
- Kannan A.
- Khaheshi I.,
- Mahjoob M.P.,
- Esmaeeli S.,
- Eslami V.,
- Haybar H.
- Ponikowski P.,
- Voors A.A.,
- Anker S.D.,
- et al.
- Montoya J.L.,
- Georges S.,
- Poquette A.,
- Depp C.A.,
- Atkinson J.H.,
- Moore D.J.
- Leong D.P.,
- Chakrabarty A.,
- Shipp N.,
- et al.
- Lord K.C.,
- Shenouda S.K.,
- McIlwain E.,
- Charalampidis D.,
- Lucchesi P.A.,
- Varner K.J.
- Islam M.N.,
- Jesmine K.,
- Kong Sn M.A.,
- Hasnan J.
- Ferreira V.M.,
- Marcelino M.,
- Piechnik S.K.,
- et al.