Arrhythmic risk management after acute myocarditis: never too early, only too late

This article refers to ‘Post-discharge arrhythmic risk stratification of patients with acute myocarditis and life-threatening ventricular tachyarrhythmias’ by P. Gentile et al., published in this issue on pages 2045–2054. Prediction of the sudden cardiac death (SCD) is the Holy Grail of arrhythmology, and the search for reliable predictors of SCD has pushed the investigations into the connected fields within cardiology.1 Also in acute myocarditis, ventricular arrhythmias (VA) and SCD can be the initial manifestation.2 Myocarditis is an inflammatory process of the heart muscle, with a clinical course ranging from an asymptomatic finding to severe cardiogenic shock and SCD. The important association between undiagnosed myocarditis and SCD is emphasized by post-mortem data, which have implicated myocarditis in SCD of young adults at rates of 8.6–44%.2 Currently the non-invasive gold-standard method for the diagnosis of myocarditis is cardiac magnetic resonance (CMR) imaging. Two multicentre studies with significant patient populations have demonstrated the importance of CMR with late gadolinium enhancement (LGE) in the diagnosis and prognosis of patients with acute myocarditis.3, 4 Patterns of LGE, the potential progression of LGE and the extent of focal fibrosis on CMR have been shown to predict the risk of hospitalization and adverse cardiovascular events in patients with suspected myocarditis.3-5 Arrhythmias during acute myocarditis range from mild conduction abnormalities to life-threatening ventricular tachycardia (VT) and ventricular fibrillation (VF), requiring medication and/or device therapy. The risk of VA has been found to be independent of left ventricular ejection fraction in patients with myocarditis.6 Several mechanisms have been postulated to explain the occurrence of VA during and after myocarditis, but the risk of SCD seems to correlate poorly with the severity of the acute inflammation6, 7 and can very well manifest later as a scar-related VA, regardless of left ventricular ejection fraction.8 Implantable cardioverter-defibrillators (ICD) for the management of VA should be evaluated after the resolution of the inflammatory process, generally 3–6 months after the onset of the acute phase.7, 9 In a contemporary cohort of recent onset non-ischaemic cardiomyopathy (IMAC2),10 early ICD implantation did not impact survival. The 3–6 months generally accepted as healing interval may still have a high arrhythmic risk, despite potential complete recovery. This period may therefore be bridged by using a wearable cardioverter-defibrillator (WCD).9, 11 In this issue of the Journal, Gentile et al.12 provide an elegant study of this rather rare, but clinically significant disease spectrum. The aim was to address the patient population who present with significant VA during the acute phase of myocarditis and may be subject to preventive actions. The authors showed that during a long-term follow-up, the incidence of life-threatening VA was as high as 37%, with 60% of the events occurring in the first year. The median time of occurrence of VA after discharge in this international cohort was 8 months. Current follow-up and recommendations for the use of WCD, however, suggest a delay of 3–6 months.3, 9 This leaves a significant number of this young patient population uncovered. To better stratify the risk of VA recurrences and avoid on one hand unnecessary ICD implantation, and on the other hand precocious discharge from follow-up, the authors propose an early discharge risk prediction based on clinical presentation (sustained VT) and early CMR characteristics (absence of positive short-tau inversion recovery and extensive LGE involving more than two segments). Cardiac magnetic resonance has been extensively used in the diagnosis and prognostic assessment of acute myocarditis.4, 13, 14 In the acute inflammatory stage, there is an increase in vascular permeability with local hyperaemia and oedema, leading to an increase in extracellular volume, detected by enhanced signal and/or prolonged decay time in T2-weighted sequences (suggesting water/oedema), and also prolonged relaxation time in T1-weighted sequences (suggesting inflammation). Also in the acute phase, myocyte injury and necrosis lead to myocardial damage and an increase in the extracellular volume, resulting in positive LGE images.5, 14 Interestingly, while the increase in extracellular water/local oedema is a crucial diagnostic finding,15 it is precisely the absence of oedema that proves to be deleterious in the short-term risk for VA. Most of the studies on the role of CMR in myocarditis have been directed towards the prognostic role of LGE, the extension of which is proportional to the risk of late complications (development of heart failure and occurrence of VA).13 The ITAMY study3 already raised awareness for the negative prognostic effect of early disappearance of oedema (T2-weighted negative) and its replacement with fibrosis (LGE positive), by using a 6-month CMR study. The fact that only 64% of the patients in the current study by Gentile et al. had positive T2-weighted images at the admission CMR may reflect a somewhat later presentation in a more subacute phase when oedema already started to recede. This has already been described in the MyoRacer-Trial,15 and reflected in a poorer diagnostic accuracy of the T1 and T2-weighted sequences in case of later presentation, because of diminishing oedema and replacement with fibrosis. This may also partially explain the worse long-term prognosis in patients presenting with sustained VT (irreversible damage already present) as compared to VF (extensive oedema, with potential for recovery). Unfortunately, Gentile et al. do not provide T1 mapping in their study, mainly because of the heterogeneous and retrospective nature of this multicentre database. The early risk stratification by the CMR strategy proposed by Gentile et al. may allow for completely bypassing the observation period of 6 months for a number of patients, and assigning them from the start to either ICD implantation, or alternatively medical treatment associated or not with WCD as bridge to healing. It is tempting to believe that we might be able to accurately predict the patients who would completely recover without any need for treatment, and to set apart the ones who would benefit from early ICD implantation, as well as those needing long watchful waiting periods, with the help of WCD. However, the matter is not so simple, and as always, directing the prediction score only towards the substrate (myocardial lesions and scarring) is probably an oversimplification. Some additional efforts are required for building up prospective international registries and larger studies, which would be directed at disentangling the remaining challenges: (i) to assess the best timing for ICD implantation; (ii) to investigate the role and optimal duration for WCD; and (iii) to develop an effective patient stratification model for the risk of SCD. Consider herein clinical/electrophysiological data, CMR, but also newer clinical echo parameters (left ventricular global longitudinal strain, atrial strain), or even novel parameters which showed promise in translational studies (like the non-invasive myocardial stiffness assessment using high frame rate ultrasound and shear wave imaging16). In the end, the proposed strategy may include a baseline stratification at discharge including clinical and CMR parameters, coverage with WCD up to 3 months, re-stratification based on new CMR data,3, 4, 14 and possibly systematic prolonged use of WCD up to 1 year in patients without definitive ICD indication but with persistent left ventricular abnormalities on CMR (Figure 1). In these patients a new evaluation and prognostic stratification at 6 months may be performed. The obvious disadvantage of this approach would be mainly related to patient compliance (repetitive exams, WCD use) and costs. Patient adherence to treatment may be enhanced by appropriate education, and the costs are probably counter-balanced by the gain in survival and quality of life in this young productive patient population. Conflict of interest: none declared.