Relevance to identify patients with uncomplicated presentation at the index hospitalization for suspected acute myocarditis to better plan follow-up

Acute myocarditis (AM), a generally self-resolving inflammation of the myocardium, mainly affects adolescents and young adults with a male prevalence (60–80%).1 The estimated incidence is between 4 and 14 per 100 000 individuals each year with an overall estimated mortality rate between 1 and 7%.1 Nevertheless, most patients have an uneventful course after the initial hospitalization for AM.2,3 A previous retrospective Italian multicentre registry (Lombardy registry) studied 443 patients with AM diagnosed by either endomyocardial biopsy (EMB) or increased troponin plus oedema and late gadolinium enhancement (LGE) at cardiac magnetic resonance (CMR). At the time of hospital admission, 118 patients (26.6%) had left ventricular ejection fraction (LVEF) < 50%, sustained ventricular arrhythmias (VAs), or a low cardiac output syndrome, whereas 325 (73.4%) had no such complications (uncomplicated presentation).3 At the 5-year follow-up, excluding the events during the acute phase, major AM-related cardiac events [post-discharge death and heart transplantation (HTx), sustained VA treated with electric shock or ablation, and symptomatic heart failure (HF) needing device implantation] occurred in 2.8% of the overall population, with a higher incidence in patients with complicated forms at baseline presentation (10.8% vs. 0% in uncomplicated AM; log-rank P < 0.0001; Table 1).3 Other studies have identified potential markers of events after the acute phase of AM, including patients with normal or nearly normal LVEF.4,5 In particular, it has been demonstrated that among 374 patients with AM diagnosed by CMR, the evidence of septal inflammatory involvement was associated with future events (sudden cardiac death, VA, and HF hospitalizations).4 Other studies identified, in a small proportion of patients (<5%), the presence of pathogenic or likely pathogenic desmosomal gene variants.5–7 These patients had an increased risk of VA or myocarditis recurrence despite a normal or nearly normal LVEF.5,7 Thus, a proportion of patients with AM initially classified at low risk based on clinical and echocardiographic criteria may have a specific localization of inflammation, detectable at CMR, or a genetic background that put them at higher risk of adverse events at follow-up. Main differences between the current population with uncomplicated acute myocarditis in the Swiss registry and in the previously published Lombardy registry AM, acute myocarditis; CMRI, cardiac magnetic resonance imaging; EMB, endomyocardial biopsy; HTx, heart transplantation; LVEF, left ventricular ejection fraction; Q, quartile; VA, ventricular arrhythmia. aEchocardiogram performed on admission or in emergency department. Main differences between the current population with uncomplicated acute myocarditis in the Swiss registry and in the previously published Lombardy registry AM, acute myocarditis; CMRI, cardiac magnetic resonance imaging; EMB, endomyocardial biopsy; HTx, heart transplantation; LVEF, left ventricular ejection fraction; Q, quartile; VA, ventricular arrhythmia. aEchocardiogram performed on admission or in emergency department. In the current issue of the European Heart Journal – Acute Cardiovascular Care, Schelldorfer et al.8 report the occurrence of cardiac events after AM in a retrospective Swiss registry during a 9-year period (2013–20). The authors identified 1439 patients who were admitted with AM based on ICD-10 administrative data that also included the codes for rheumatic myocarditis (I01.2 and I09.0). The median follow-up was 39 months (∼3.3 years), and the comparator was a matched population undergoing laparoscopic appendectomy hospitalization for the first time. The authors provided retrospectively their own definition of uncomplicated presentation as the absence of cardiac events during index hospitalization in the discharge report. Overall, the rate of diagnostic assessment was low, or simply not reported, as we can presume for transthoracic echocardiography, which was performed only in 20.8% of patients (n = 300). Cardiac magnetic resonance was performed only in 24.9% (n = 360) of patients during hospitalization that had a median duration of 3 days (first to third quartile: 2–5 days). It is likely that CMR was performed after discharge, but the results were not available. In particular, we do not know whether CMR would have led to reclassification of the diagnosis, for instance in patients with myocardial infarction and no obstructive coronary disease (MINOCA).9 Schelldorfer et al. report that 203 (14.1%) patients with uncomplicated AM had a rehospitalization due to cardiovascular events during follow-up with an estimated annual incidence rate of 4.4 per 100 person-years, a figure significantly higher compared with a matched population who underwent laparoscopic appendectomy (hazard ratio of 42.3). This was based on a composite outcome including hospitalization for myocarditis recurrence (6.0%, n = 87), HF (0.8%, n = 11), diagnosis of dilated cardiomyopathy (0.1%, n = 1), VA (0.7%, n = 10), cardiac arrest (0.4%, n = 6), cardiogenic shock (0.3%, n = 4), supraventricular tachycardia (0.3%, n = 4), atrial fibrillation or flutter (0.4%, n = 6), second- or third-degree atrioventricular block (0), sick sinus syndrome (0.1%, n = 1), or pericardial disease including acute pericarditis (8.6%, n = 124), cardiac device implantation (0.3%, n = 4), or HTx (0.1%, n = 1). In addition, it is worth noting that during follow-up, 126 patients (8.8%) were admitted with a diagnosis of ischaemic heart disease, likely due to a previous misdiagnosis of AM. Another surprising and counterintuitive finding is that among uncomplicated AM, the only two independent protective factors for future adverse events were female sex and older age. These two factors have been identified as risk factors for cardiac death or HTx after AM.10,11 It could be possible that the composition of the endpoint that included recurrence of myocarditis and pericardial disease can explain the different association between events and age and sex. Unfortunately, the authors did not have data on potential deaths that could have occurred without hospitalization, and for this reason, they did not provide this piece of information. The investigators should be commended for the collection of administrative data that enrich our knowledge about the risk of rehospitalizations after a suspected AM. The main message that can be derived from this study is that it is fundamental to reach the most accurate diagnosis of AM during the index hospitalization, and CMR should be performed timely even in presumed myocarditis with uncomplicated presentation to avoid misdiagnoses. If we compare the outcomes after AM discharge in the Lombardy registry, there were only 2 patients (0.6%) with hospitalization due to ischaemic heart disease in the follow-up vs. 126 (8.8%) in the Swiss registry (Table 1). It could be possible that patients with AM deserve a longer duration of hospitalization than 3 days to reduce the risk of early rehospitalization due to misdiagnosis or undertreatment of symptoms, especially chest pain derived from myocarditis or pericardial involvement. When the authors blanked the events that occurred within the first month of follow-up, there were an estimated 3.5-fold decrease in the total endpoint events that patients had during follow-up. This estimation was based on the drop in the hazard ratio for the risk of the primary endpoint occurrence from 42.3 to 12.4 excluding the events that occurred in the first month and graphically from the comparison of the Kaplan–Meier failure estimate curves of cardiovascular composite outcome including or excluding the events that occurred during the first month (Figure 1 in the manuscript). In the same direction, the authors reported a significant reduction of events after discharge in patients with immediate cardiac diagnostics vs. those without immediate cardiac diagnostics, 111 events out of 590 patients (18.8%) vs. 93 out of 851 patients (10.9%), with an estimated annual incidence rate of 3.2 per 100 person-years among those with immediate diagnostics vs. 6.3 per 100 person-years among those without immediate diagnostics. We believe that the simple criterion of uncomplicated presentation to identify patients at very low risk of major cardiac events including cardiac death and HTx proposed in the Lombardy registry are still valid among patients with definite diagnosis of AM.1 The main differences between the current Swiss registry and the previous Lombardy registry are presented in Table 1.3,8 Having a persistent LVEF < 50% (even without HF symptoms), a variable not captured by the Swiss registry, at index hospitalization can be a risk factor for future development of HF or VA, and these patients should undergo a stricter follow-up planning and periodic outpatient visits to monitor cardiac function and the risk of VA.12–14 Furthermore, it must be recognized that even among uncomplicated patients with AM, about 4% have an associated autoimmune disorders that could explain a potentially unfavourable outcome in the long term.3 Searching for potential associated autoimmune disorder is thus relevant in patients with first diagnosis of AM.11 Another relevant finding that the Swiss registry has pointed out is the risk of myocarditis recurrence that was observed in 6% in this population, in line with previous observations.3,15 Identification of patients with myocarditis recurrence should steer physician’s attention towards a potential genetic background for pathogenic or likely pathogenic variants in desmosomal genes.5,7 The coexistence of a large extent of fibrosis on CMR involving the ventricular septum, a high ventricular arrhythmic burden defined by a high burden of premature ventricular complexes or non-sustained ventricular tachycardia, and a family history of myocarditis beyond myocarditis recurrence can increase the likelihood of a positive genetic testing among patients with AM.5,7 These patients with AM and a genetic background of arrhythmogenic cardiomyopathy are at risk of developing biventricular or left prevalent cardiomyopathies in the long-term and could explain the excess in VA and resuscitated cardiac death observed in the current registry. As a final message, we believe that this registry can increase our awareness of a hidden subgroup of patients with AM at increased risk of arrhythmic events, even if at first glance they could have been classified as low risk, but the identification of these patients requires that key investigations such as CMR have been obtained since the index hospitalization. No new data were generated or analysed in support of this research.