We would like to discuss a recently published article on a retrospective Italian study, which compared outcomes of 132 patients with acute myocarditis with 55 patients with fulminant myocarditis. Disease onset was < 1 month in all patients and fulminant myocarditis had been defined as need for either mechanical support or inotropes.
Background
A broad range of symptoms and complaints characterizes patients with myocarditis. Clinical presentation, however, is not reliable in predicting patient outcome [1] [2]. Histology and immunohistochemistry of endomyocardial biopsies allows at least recognizing subgroups of patients with distinct disease phenotypes, such as giant cell myocarditis for example, who take advantage of aggressive immunosuppression [3].
Previous studies suggested that patients with fulminant myocarditis, defined as need for inotrope treatment or mechanical support have a better long-term outcome compared to patients with acute myocarditis with impaired cardiac function but stable circulatory function [4] [5] [6] [7]. Consequently, patients with fulminant myocarditis were categorized as suffering from a distinct disease entity. Pathologically, all patients with fulminant myocarditis in these previous studies showed infiltrates with lymphocytes and macrophages, as well as cardiomyocyte necrosis in heart biopsies. Cases with giant cell myocarditis or eosinophilic myocarditis were not included. This meanwhile well accepted “outcome paradoxon” has now been questioned by a most recent study of Ammirati et al. [8]
Methods
In two Italian hospitals patients with myocarditis and symptoms < 1 month before admission were identified according to their ICD diagnostic codes in hospital discharge forms. Patients with ischemic heart disease were excluded. According to these criteria Ammiriti et al selected 187 patients, of which 55 required circulatory support with either parenteral inotropes or mechanic devices. These patients were classified as fulminant myocarditis (FM). The remaining 132 patients without inotropic or mechanic support were classified as non-fulminant myocarditis (NFM). A primary subanalysis further excluded patients with age < 15 years, non-lymphocytic myocarditis in endomyocardial biopsies, as well as patients with findings or patient history pointing to systemic diseases (“viral” FM (vFM) versus “viral” NFM (vNFM).
The diagnosis of myocarditis relied in most patients on medical history, symptoms, ECG analysis, echocardiography, biomarkers, exclusion of coronary arterial disease and cardiac MRI (CMR) analysis. CMR was performed according to established standard protocols in 1.5 Tesla scanners. CMR images were then analyzed offline by experienced cardiologists.
Endomyocardial biopsies were taken only in cases with severe hemodynamic compromise or suspicion for infectious, or autoimmune systemic disorders. Histological diagnosis of myocarditis relied on revised Dallas criteria. In addition immunohistochemistry was performed including staining for CD3/CD8 T cells, CD68+ macrophages, and HLA-DR+ cells. Presence of virus was assessed in selected cases only.
Patient management in stable patients followed current guidelines, immunosuppressive treatments were given on a basis of individual patients characteristics. Unstable patients received either inotropes or got support from intra-aortic ballon pump alone or together with venous-arterial extracorporal membrane oxygenation.
Patients with NFM and FM were then compared with respect to demographics, baseline clinical data, echocardiographic parameters, CMR, and pathology findings. Overall survival and HTx-free survival (in-hospital and at the last follow-up) as well as echocardiographic LVEF changes between discharge and last follow-up were assessed. The same analyses were applied to the defined subgroup of adult patients with “viral” myocarditis (v-FM versus v-NFM).
Outcome analysis based on reviews of medical records at follow-up visits, on phone contact with either the patient including his/her family, or the general practitioner; or on review of electronic vital statistics. Follow-up was completed in all except 3 (1.6%) patients with NFM and preserved LVEF.
Results
In contrast to former studies, Ammirati et al found that patients with fulminant myocarditis had worse outcome compared to patients with nonfulminant myocarditis. In fact, in-hospital death or transplantation occurred significantly more often in FM patients (25.5% versus 0% for FM vs NFM) and long-term survival was also lower in FM patients (64.5% versus 100% for FM vs NFM). In cases with FM, LVEF improvement during hospitalization was greater compared to NFM forms (median, 32% [interquartile range, 20%–40%] versus 3% [0%–10%], respectively; P<0.0001), the proportion of patients with LVEF <55% at last follow-up, however, was higher in FM versus NFM (29% versus 9%; relative risk, 3.32; 95% confidence interval, 1.45–7.64, P=0.003). Similar results for survival end LVEF changes were reported for the subgroup with “viral” myocarditis.
Conclusion
The findings of this study apparently contradict earlier studies, and suggest worse transplantation free survival in patients with fulminant myocarditis compared to patients with non-fulminant myocarditis. Despite the fact that LVEF improved more markedly during hospitalization in FM compared to NFM patients, those with fulminant presentation are more likely to have worse left ventricular function on follow-up.
Comments
The findings of Ammirati et al question findings of previous retrospective case series [4]. Despite the fact, that the study design was also retrospective, Ammirati et al provide an updated view on the clinical relevance and outcome of patients with fulminant and non-fulminant myocarditis. Their insights will certainly impact work-up standards for patients with a presumed myocarditis diagnosis in many centers. So far, and mainly based on the highly cited work of McCarthy et al [4], many institutions were reluctant in performing endomyocardial biospies in patients with fulminant myocarditis weighing interventional risks against presumably favorable outcomes.
Nevertheless, we must be careful in comparing the Ammirati data with the results of the former case series. Divergent outcomes may be due to important differences in study design. In the McCarthy series, diagnosis was confirmed by conventional histology according to the Dallas criteria. Accordingly, more severe cases could have been selected in the previous studies. In contrast, contemporary immunohistochemistry criteria are much more sensitive and are most likely positive in a less severe range of cases as well. Clinically, most acute non-fulminant myocarditis patients in the McCarthy series initially showed a dilated cardiomyopathy phenotype suggesting a longer period of disease progression compared to the Ammirati study, where more patients presented with chest pain symptoms or palpitations.
Most important, however, many patients in the Ammirati study were diagnosed with cardiac MRI and clinical criteria without biopsy. Despite the fact that Ammirati et al reported similar findings in their subanalysis of the “viral” myocarditis cohort after exclusion of giant cell myocarditis and eosinophilic myocarditis, there were 17 patients in the FM group, which were diagnosed without biopsy. Some of these FM patients may have had a more aggressive histological form of myocarditis. Myocarditis [9]. Finally, advances in heart failure treatment may also have impacted outcomes of NFM patients in current series compared to previous studies.
The study of Ammirati et al does not only question former reports on case series. Given the fact that data acquisition was again a retrospective analysis, the findings are also subject to many biases, mainly due to inclusion criteria. Consequently, the present study again emphasizes the need for prospective multi-center databases and uniform diagnostic criteria. Given the fact that cardiac MRI – despite its added value in supporting a diagnosis of myocarditis – does not yet provide a specificity comparable to endomyocardial biopsy, every prospective study must rely on diagnostic criteria based on systematic histological, immunohistochemical, and virological work-up of biopsy samples [10] [11].
Taken together, Ammirati et al provide a remarkable retrospective analysis that reflect our current difficulties in defining reliable prognostic criteria for patients with myocarditis based on clinical findings and/or clinical findings and cardiac MRI. Personally, I believe that this study indirectly supports the policy in many centers to perform endomyocardial biopsies on all patients with impaired cardiac function and suspicion for myocarditis. Endomyocardial biopsy is still the gold standard for myocarditis diagnosis and certainly a prerequisite for any prospective study in the future, which must include detailed immunohistochemical and virological work-up of samples, as well as standardized cardiac MRI protocols in well-defined cohorts [3].
Our mission: To reduce the burden of cardiovascular disease.