Two bad actors: can cardiac magnetic resonance distinguish idiopathic granulomatous from giant cell myocarditis?

This editorial refers to ‘Cardiac magnetic resonance in giant cell myocarditis: a matched comparison with cardiac sarcoidosis’, by P. Pöyhönen et al., https://doi.org/10.1093/ehjci/jeac265. Myocarditis is usually a self-limited disorder, but a significant minority of patients develop fulminant cardiogenic shock and/or intractable arrhythmias. Non-specific clinical history, presenting symptoms, and overlapping electrocardiogram and echocardiographic features usually prevent the accurate identification of specific causes of myocarditis. Multi-modality imaging, specifically with cardiac magnetic resonance (CMR) imaging, plays a critical role in identifying the underlying pathology, with its strength primarily in differentiating ischaemic vs. non-ischaemic presentations. Limited data exist on the specific non-ischaemic patterns on injury in giant cell myocarditis (GCM) compared with cardiac sarcoidosis (CS). A combination of clinical, imaging, and pathologic findings is primarily used in the acute setting. Pöyhönen et al.1 provide the first CMR comparison study of patients with GCM and CS. Most often, myocarditis results from viral infections, but specific cardiac inflammatory conditions such as GCM and CS frequently can present with acute life-threatening inflammatory cardiomyopathies that require early recognition for achieving the best outcomes.2 Giant cell myocarditis is a rare fulminant myocarditis that carries high mortality risk, with a reported incidence of <0.05%, with true incidence being likely underestimated.3,4 Cardiac Sarcoidosis and giant cell myocarditis frequently present with acute cardiogenic shock, intractable ventricular arrhythmias, or high-grade atrioventricular block. Multi-modality imaging is helpful, but often difficult to complete because of critical illness.3 Early diagnosis is critical, as most cases result in death or require urgent heart transplantation with improved outcomes from early multi-drug immunosuppression. Specific CMR patterns of injury in GCM are limited to case reports. Cardiac sarcoidosis is typically associated with a multi-system granulomatous disease process with clinical cardiac involvement in ∼5–10% of patients with systemic sarcoidosis.4,5 Although the aetiology of sarcoidosis is unknown, it is postulated as an immunologic response to a specific antigenic trigger in susceptible patients.4,5 Patients with cardiac involvement typically present with sub-acute to chronic heart failure symptoms associated with ventricular arrhythmias and high-grade conduction abnormalities. Unlike in GCM, endomyocardial biopsy (EMB) generally has low sensitivity due to patchy myocardial involvement, as low as 20–30%.6 FDG (fluorodeoxyglucose F 18)-positron emission tomography (PET) is a useful adjunct test in assessing for active inflammation and potential sources of extracardiac biopsy to confirm diagnosis. Current multi-society guidelines recommend non-invasive imaging along with cardiac or more frequent extracardiac biopsy to confirm CS.7 In their article, Pöyhönen et al. study comparative findings with CMR in age, sex, echocardiographic, and clinical history–matched patients with GCM and CS. As described by the authors, there are limited data on CMR findings in GCM patients having likely related rapidly progressive cardiogenic shock symptoms. The authors used the Myocardial Inflammatory Disease in Finland database to identify clinically matched patient groups with GCM and CS. A total of 18 patients in both groups were identified to have undergone CMR at the time of inflammatory cardiomyopathy diagnosis. These patients were then evaluated by two CMR-trained cardiologists who were blinded to the diagnosis and clinical data. The two patient groups were appropriately matched for clinical and echocardiographic criteria. The more diffuse involvement in GCM allows for a significantly higher sensitivity for EMB in GCM over CS, and the authors are commended in comparing all but one CS patient with EMB-confirmed disease. The analysis demonstrates that the myocardial late gadolinium enhancement (LGE) imaging is indistinguishable between CS and GCM, except for a comparatively decreased amount of LGE for a comparable ejection fraction in the GCM group. More importantly, their analysis reveals that previously noted non-ischaemic scar patterns thought of as ‘imaging biomarkers’ for CS are seen in both groups.8 Specifically, the septal/right ventricle (RV) insertion site of LGE, which is thought to be more common in CS and with more subendocardial involvement in GCM, is seen with similar frequency in both groups. T2-weighted images were completed in 67 and 72% of GCM and CS groups, respectively, with myocardial oedema present in nearly all cases in both groups. On limited, scout and cine images, radiologists blinded to the clinical history assessed for the presence of extracardiac disease, specifically, they assessed for any enlarged mediastinal or hilar lymph nodes. Cardiac magnetic resonance scout and cine images are inherently limited for the assessment of extracardiac disease, and despite a slightly larger number of CS patients with enlarged lymph nodes, there was no statistical significance between the groups. Rare diagnosis, along with critical presentations, limits evidence to registry data even for treatment in GCM. Unsuprisingly, results in a small sample size of this study, despite which the authors do admirable work to add more insight into this challenging imaging and pathologic diagnosis. The incorporation of myocardial histopathologic diagnosis into the CMR imaging analysis is the strength of this study. This is a difficult task given the yield of CS diagnosis with EMB is as low as 20% due to patchy myocardial involvement. As noted by the authors, CMR in this patient group was completed over a 10-year span, with some of the earlier imaging having limited availability with parametric mapping or myocardial strain evaluation. Novel techniques with T2 parametric mapping and myocardial strain imaging are increasingly being studied with various cardiomyopathies.9,10 T2-weighted imaging specifically has variable inter-reader variability due to a subjective evaluation of oedema, and quantitative T2 mapping has increased the sensitivity and specificity for the detection of inflammation that is supportive of myocarditis. The advantage of parametric mapping or myocardial strain is more uniform reporting with improved inter- and intra-reader variability in comparison with subjective assessment for oedema.11,12 The contemporary data on these novel techniques are the added prognostic value along with improved serial follow-up in patients on immunosuppressive therapy. For example, a recent study of 69 patients with acute myocarditis describes the prognostic role of myocardial strain on CMR over LGE, specifically, abnormalities in the left ventricular longitudinal strain as a predictor of recovery.9 It is important to note that these novel techniques have been studied in the myocarditis population, and the generalizability for specific inflammatory cardiomyopathies such as GCM or CS is less defined. Like the indistinguishable LGE patterns, parametric mapping or strain is unlikely to add value with specific patterns to help distinguish different aetiologies of myocarditis. FDG-PET provides valuable information of cardiac inflammation, but the cardiac findings alone add little to distinguish inflammatory entities. However, FDG-PET with hybrid magnetic resonance imaging (MRI) or computed tomography is helpful in identifying extracardiac areas of inflammation more commonly seen in CS.3 The strength of metabolic imaging with FDG lies in serial imaging, primarily described in the CS group to help titrate immunosuppression, but case reports have described similar utility of serial imaging in GCM patients.13 Non-invasive imaging modalities including cardiac MRI and FDG-PET are useful adjuncts for the evaluation of non-ischaemic cardiomyopathies. This study provides a valuable insight into the lack of specific LGE characteristics to help distinguish CS and GCM and highlights the under-recognized frequency of RV involvement in GCM (82%). Subendocardial LGE was present in 100% of both GCM and CS but was much less common in routine lymphocytic myocarditis. Novel CMR techniques with parametric mapping and strain are useful adjuncts for prognosis and serial follow-up, but likely will have similar limitations to the LGE pattern described in this study. Several clinical comparative studies in GCM and CS have postulated whether these two conditions sharing a common pathway with T-cell-mediated inflammation are a spectrum of the same disease.14,15 Although controversy exists among the experts, the more frequent fulminant course with GCM suggests a distinct pathology. We acknowledge that the branch point is a decision related to multi-drug immunosuppression, more commonly prescribed in GCM patients than in CS patients. There is no definitive imaging or clinical characteristics to non-invasively characterize patients with rapidly progressive heart failure and intractable arrhythmias. Endomyocardial biopsy remains crucial for decision-making in tailored immunosuppression, which can significantly alter the clinical course in these patients, and likely should be done immediately after coronary disease is excluded early in the presentation. No new data were generated or analysed in support of this research.