Cardiovascular magnetic resonance imaging characteristics in patients with methamphetamine-associated cardiomyopathy

The growth in methamphetamine usage worldwide continues to present increasing societal and healthcare challenges. With the escalation of its usage in a variety of social demographics, methamphetamine-associated cardiomyopathy (MA-CMP) has become an increasingly recognised aetiology of cardiomyopathy, particularly amongst younger patients presenting to emergency departments with heart failure (HF) symptoms and are found to have newly detected ventricular dysfunction [1‐4]. The consequence of this disease process is development of HF with reduced ejection fraction (HFrEF) from severe ventricular dilatation and left ventricular (LV) dysfunction [5]. This clinical entity has substantially increased in its recognition and incidence over the past 15 years with a substantial increase in hospitalisations [6].

Although there has been a number of observational studies of this cohort [4, 5, 7‐10], the complex aetiology of MA-CMP remains unclear, and the cardiovascular magnetic resonance (CMR) findings have yet to be characterised. CMR is a specialised cardiac imaging modality commonly used in assessment of cardiomyopathy with the potential for accurate biventricular volume measurement, calculation of ejection fraction, tissue characterisation and detection of intracardiac thrombus. We aimed to characterise the clinical and CMR features associated with MA-CMP. This is of particular importance as a significant burden of MA-CMP may be reversible with early intervention and complete drug abstinence [11‐13]

A total of 4,700 consecutive patients with a history of HF were identified during the study period. After exclusion criteria were applied, 30 patients were included in the cohort. Cohort selection flowchart is shown in Fig. 1.

The year of diagnosis was recorded with an increase in incidence evident across the data collection period, both within the population with CMR imaging performed, and those without, as seen in Fig. 2.

Table 1 outlines the baseline information for this cohort. Most MA-CMP patients were in their fifth decade of life and predominantly male. The average duration of methamphetamine use was 13.4 years, with a comparable portion who were chronic intravenous users (43.3%); whilst others smoked (53.3%), inhaled (6.7%) or had a mixture of consumption methods (13.3%).

Many MA-CMP patients had concurrent heavy tobacco use (80%), and a portion had a previous or current history of excess alcohol use (40%) and/or marijuana use (50%). Comorbidities were present within the cohort, with overall 33.3% possessing a history of hypertension, 10% diabetes mellitus, 10% hyperlipidaemia, 3.3% chronic kidney disease and 3.3% with prior ischaemic heart disease. No patients had a history of cardiac rhythm disorders.

CMR parameters are presented in Table 2. The mean LV volume obtained of the participants included was consistent with severe LV dilatation (LVEDV 334 ± 99 ml; LVESV: 269 ± 98 ml), whilst the mean RV volume indicated moderate-to-severe RV dilatation (RVEDV: 272 ± 91 ml; RVESV: 173 ± 82 ml). LVEF (20.9 ± 9.2%) indicated severe LV dysfunction, with moderate-to-severe RV dysfunction also detected (RVEF: 29.4 ± 13.4%) in all MA-CMP patients.

Table 3 demonstrates the findings related to presence and location of LGE on CMR. Twenty-two patients (73.3%) had myocardial LGE, of which 59.1% were located in the mid-wall, with all of these involving the interventricular septum. 22.7% displayed localised regions of sub-endocardial LGE in a variety of locations and 18.2% had transmural regions of LGE that were located in the inferior and inferolateral segments. Several case studies are shown in Fig. 3 to highlight these features.

Six patients (20%) had intracardiac thrombus. Of these cases, 66.7% were in the LV and 33.3% were biventricular. Figure 4 demonstrates one of the cases in which the burden was limited to the LV. Multiple, irregular large LV thrombi appeared on the horizontal long-axis view at the apex and were circular appearing at the apex on the vertical long-axis view. They had an irregular appearance involving the basal-to-mid anterior segments, with some tracking along the mid anterior segment.

Follow-up data were available on 24 of the 30 patients (80%) included in this cohort using echocardiography. Complete cessation of methamphetamine use and compliance to HF pharmacotherapy was advised to all included participants in this cohort during their index admission. Of the cohort of 11 patients who ceased methamphetamine, all (100%) complied with their medication regime. This compared with only 5 out of the 13 patients taking prescribed medications who continued to use methamphetamine following diagnosis. Over a median follow-up of 6 months (IQR 3.0 to 6.5 months) from index admission, improvement in cardiac function, specifically a LVEF increase of > 5%, was observed in participants who achieved complete cessation of their methamphetamine use compared to those who reduced or remained consistent in their consumption (p-value = 0.003). The presence of baseline LGE on CMR, however, did not show a significant impact on change in follow up LVEF. Additionally, whilst there were variable cessation rates in both patients with and without LGE, a two-way ANOVA test indicates that only cessation is statistically significant (p-value = 0.009), independent of LGE. These figures are presented in Tables 4, 5 and 6.

Given the potential of alcohol abuse being a confounder and contributor to cardiomyopathy in this population, the relationship of alcohol abuse on ventricular volumes, LVEF and RVEF (at baseline), as well as recovery in ventricular function was measured. Whilst there was a small trend to increased volumes in those with a history of alcohol abuse (Table 7), this was not significant. In addition, Fig. 5 highlights that a history of prior alcohol use did not affect rates of improvement in ventricular function with methamphetamine cessation, as patients who ceased methamphetamine use did not also cease alcohol use.

CMR has evolved as an essential investigative tool in evaluation of cardiomyopathy due to its accurate quantification of biventricular volumes and ejection fraction as well as tissue characterisation [15]. In addition, detection of specific patterns of LGE and assessment of its burden does assist in assessment of risk for arrhythmias and prognosis in non-ischaemic cardiomyopathy. In this retrospective study of an important HF group, we identified a number of adverse prognostic markers on CMR in patients with MA-CMP.

LV dilation and contractile impairment is a significant predictor of adverse outcomes in dilated cardiomyopathy [16]. In addition, RV dilatation and left atrial dilatation also confer significant prognostic value in cardiomyopathy [17, 18]. Our retrospective study has demonstrated a substantial degree of LV and RV dilatation as measured on CMR. A follow-up CMR study assessing the effects of medical therapy and prognosis relating to degree of ventricular dilatation would provide further insights into prognosis of MA-CMP.

The mechanisms of ventricular dilatation specifically in MA-CMP warrant discussion. MA-CMP development is postulated to be a secondary to an interplay of direct and indirect mechanistic pathways. Directly, amphetamines promote free radical production resulting in direct cellular injury, apoptosis via increased p53 activity, mitochondrial dysfunction, altered gene expression and defective intracellular haemostasis. Contractile dysfunction develops from disrupted electrical–mechanical coupling [19]. In susceptible individuals, long-term use results low-grade inflammation and consequent myocyte loss with replacement fibrosis. Indirectly, amphetamine does potentiate heightened sympathetic nervous system activity, which can also induce coronary and microvascular spasm, resulting in acute ischaemia [20‐22]. These mechanisms consequently may result in the development of significantly dilated cardiac chambers, systolic dysfunction, and clinical presentation of the heart failure syndrome [5, 23, 24].

LGE is an effective, well-validated and reproducible method of detecting myocardial fibrosis and other pathologies such as infarction, infiltration and inflammation. An abundance of CMR clinical studies have demonstrated correlation with adverse events (including sudden cardiac death in non-ischaemic cardiomyopathy [25‐27]. The incidence of LGE in our population of MA-CMP was quite high (76.3%), and the pattern of LGE identified in our cohort was heterogenous. This implies varied pathological processes that may occur with MA-CMP including chronic stress precipitating mid-wall fibrosis, micro-embolic phenomena and possibly areas of low-grade chronic inflammation. The majority of patients (56.5%) had evidence of mid-wall LGE, indicating that chronic inflammation and resultant mid-wall fibrosis is likely to be the predominant mechanism for developing of cardiomyopathy. However, 21.7% of the patients included had evidence of sub-endocardial LGE and 21.7% had transmural LGE, both of which can typically be indicative of prior myocardial infarction. In the population included, all patients had obstructive coronary artery disease excluded at coronary angiography which supports the hypothesis of these patients having possible coronary arterial spasm or myocardial infarction with non-obstructive coronary arteries (MINOCA) as the aetiology. The presence and burden of LGE detected on CMR in the non-ischaemic population has been demonstrated to be a powerful imaging biomarker in predicting adverse outcomes and response to medical therapy [28]. In addition, a previous study of 30 patients undergoing endomyocardial biopsy in MA-CMP has reported the presence of fibrosis on biopsy representing an adverse relationship with recovery of ventricular function [5]. Therefore, in the population of MA-CMP where individual clinical factors may render some patients challenging for device implantation, the further prognostic information from LGE assessment may assist in important clinical decision making with regard to implanted cardiodefibrillator (ICD) placement and may also assist with pharmacotherapy selection as well as discussions with patients regarding long-term prognosis.

The incidence of ventricular thrombi was also a highly relevant clinical finding in the population studied. Intra-cardiac thrombi carry significant clinical risk with embolization and its detection has a profound impact in management by mandating therapeutic anticoagulation [29]. Intra-cardiac thrombi have been observed in other studies assessing the MA-CMP population [5]. Our study revealed a 20% incidence of ventricular thrombus with MA-CMP, observed in mixed patterns of LGE. Given the potential clinical catastrophic consequences, this finding necessitates an adequately thorough exclusion for ventricular thrombus with patients with MA-CMP that undergo cardiac imaging.

CMR sequences that were not available in our cohort but would be of value in a future study include native T1 mapping, post contrast T1 mapping and patterns in tissue tracking measuring LV, RV and left atrial strain. Data from these sequences would provide further mechanistic and prognostic insights into the entity of MA-CMP, specifically around the incidence of diffuse fibrosis, long-axis LV function and left atrial function.

Dedicated left atrial volume acquisition may also have provided additional data in this patient population given the prognostic value of atrial dilatation in cardiomyopathy.

To our knowledge, this is the largest series of MA-CMP that has been characterised with CMR. This study demonstrated that MA-CMP is associated with significant biventricular dilatation and systolic dysfunction, a high incidence of LGE and a 20% incidence of intra-cardiac thrombus, which was sometimes biventricular. The MA-CMP population possessed heterogenous patterns of LGE detected, most frequently observed in the mid-wall and interventricular septum. Follow-up data of this cohort revealed that abstinence from methamphetamine use following diagnosis of MA-CMP was associated with a greater improvement in LVEF. Future studies with larger cohorts, as well as outpatient cohorts, are required to corroborate these findings and to investigate their follow-up imaging for further recovery of cardiac function.