Empty handed: a call for an international registry of risk stratification to reduce the 'sudden-ness' of death in patients with non-compaction cardiomyopathy

Isolated left ventricular (LV) non-compaction or non-compaction cardiomyopathy (NCCM) is a relatively new and rare clinico-pathologic entity, first described by Engberding and Bender in 1984. 1 Since then, many case reports and a few case series are published. 2 , 3 The clinical presentation includes congestive heart failure, supra- and ventricular arrhythmias, and thrombo-embolic events. It is recently classified as a primary cardiomyopathy. Morphologically, it is characterized by a prominent trabecular meshwork and deep intertrabecular recesses communicating with the LV cavity and is thought to be caused by an arrest of normal embryogenesis of the myocardium. 2 It is hypothesized that intrauterine arrest of the normal compaction in the embryonal heart results in a ‘spongy’ heart, characterized abnormal trabeculation, especially confined to the LV apical segments. Therefore, the diagnostic criteria fully rely on the imaging modality, especially transthoracic echocardiography, contrast echocardiography, or MRI. Most of the cases are familial and show genetically heterogenous patterns. However, there is strong evidence that at least a part of the NCCM population are caused by sarcomeric gene mutations. These data suggest that NCCM could be a part of spectrum of cardiomyopathies, consisting of dilated and hypertrophic cardiomyopathy. The clinical spectrum of NCCM shows a highly variable clinical presentation, ranging from asymptomatic to severely symptomatic with end-stage heart failure, lethal arrhythmias, and/or thrombo-embolic events. The first clinical case series described a high risk of ventricular tachy-arrhythmias and sudden cardiac death (SCD) 2 identifying malignant ventricular arrhythmias as a major clinical issue. However, recent studies show a more benign natural history, including lower risk for (malignant) ventricular arrhythmias. 3 This explains the continuing discussion about the real incidence and prevalence of malignant ventricular arrhythmias in NCCM patients, including sustained ventricular tachycardia (VT) and ventricular fibrillation (VF). 2–5 Based on the above-mentioned facts, the study published of the journal by Steffel et al.6 is interesting and it tries to assess the possible role of invasive electrophysiology testing. The authors collected the electrophysiologic study (EPS) results of 24 patients, in a retrospective fashion. Inducible ventricular tachyarrhythmias were found in 38% of the patients and only 17% of them had sustained VT or VF. During a follow-up period, only the minority of patients implanted with an implantable cardioverter–defibrillator (ICD) had appropriate therapy. Interestingly, none of the 12 patients in the non-inducible group had VT or VF occurred. The fact that the seven patients with inducible supraventricular tachycardias also included two with an accessory pathway [Wolf-Parkinson-White (WPW) syndrome] draws attention to the heterogeneity of the study population, in view of the frequent coincidence of congenital heart diseases such as Epstein's anomaly with WPW syndrome. In view of the frequent coincidence of congenital heart diseases such as Ebstein's anomaly with WPW syndrome, at least some patients with NCCM may have sarcomeric gene mutations. 7 The conclusion that EPS testing may be of limited value for risk stratification for SCD, sounds fair despite the limitations of the study. The authors recommend prophylactic ICD implantation only after individual clinical risk stratification. We can only agree with that and support the need for prospective long-term FU studies. Indeed, the only way for appropriate risk stratification is collecting relevant clinical data in a multi-centre, prospective registry and follow-up studies. The authors of this editorial strongly believe that this could, in this information technology area, easily be done in the form of a web-based national, and international registry. Additional problem in this yet rare clinical entity is that the underlying arrhythmogenic substrate is not known and therefore changes and progression of the arrhythmogenic substrate in the long term remains speculative. This is consistent with the findings of the authors, since most of their patients had relatively preserved LV function. This suggests that the degree of LV dysfunction, which is generally an indication of the severity of disease, could probably not be used to predict malignant ventricular arrhythmias. 8 This is in line with our clinical experience that among high-risk NCCM patients with severe LV dysfunction and heart failure who receive an ICD for primary prevention, none had appropriate shock, unlike those who received an ICD for secondary prevention. 9 To put into perspective and to understand the reason for significant limitations of EPS-based risk stratification, we have to look into this issue of risk determination from a broader perspective. EPS in clinical decision-making in patients at risk of SCD has dramatically changed over the past decades. Although the annual incidence of SCD in the general population is 0.1–0.2%, specific subgroups of patients (coronary artery disease and reduced LV function, dilated cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, hypertrophic cardiomyopathy, Brugada syndrome, long QT syndrome, and non-compaction cardiomyopathy) are at higher risk. 10 Programmed ventricular stimulation (PES) with intracardiac electrical stimulation and recording began in 1972, when Wellens et al.11 had introduced this technique for the evaluation of VT. Programmed ventricular stimulation is used for the induction of VT, as the rhythm generally has re-entry because of the scar tissue. The ease of VT induction is related to the substrate, and to the aggressiveness of the used protocol. Sustained or haemodynamically unstable monomorphic VT, sustained polymorphic VT and VF induced with less than three extrastimuli are considered positive result. 12 The general concept of EPS was to test the electrical vulnerability of the myocardium, thus the inducibility of VT with anti-arrhythmic medication: patients whose arrhythmia remained inducible with drug therapy had a worse outcome. In the 1980s, it became clear that anti-arrhythmic medication may harm more patients than they benefit, and the development and successful implantation of ICD reduced the number of serial EPS. Two landmark trials, the MADIT and the MUSTT, demonstrated in the late 1990s that patients with coronary artery disease who are at high risk (reduced LV function and inducible ventricular arrhythmias) benefit from ICD implantation as a primary prevention. EPS, as a risk stratification tool, showed modest positive and poor negative predictive values in these studies. Furthermore, recent data from the Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation (DEFINITE) trial has proven that EPS does not provide data that are useful in assigning risk in patients with non-ischaemic-dilated cardiomyopathy. For high-risk groups of patients with coronary artery disease, the annual incidence of SCD can be 10–20%, so the majority of these patients will never develop fatal arrhythmias. On the other hand, in absolute numbers, due to a greater population at risk, most SCD occurs in patients who are apparently at lower risk. Consequently, there is a need to better identify patient who are at the highest risk of SCD, and will benefit from ICD implantation, which is not without complications. Using PES, the specificity of induced arrhythmias among non-ischaemic patients is controversial. The performed studies had small number of patients, and on the top of that the stimulation protocols were different. The most common finding with PES in 38% of these patients is inducible polymorphic VT or VF, but it has limited clinical significance. This suggests that the positive predictive value of induced polymorphic ventricular arrhythmias is low in the population with non-ischaemic-dilated cardiomyopathy. For routine EPS for risk stratification in patients with NCCM, currently there is no evidence in the daily practice, in spite of the article by Steffel et al. Until we have larger, prospective studies with long-term follow-up, EPS probably should be reserved for research protocols, including long-term FU studies. Certainly, in specific clinical situations such as unexplained syncope or palpitations, the EPS remains mandatory. Until we have reliable prospective data, it remains reasonable to use the current guidelines for management of patients with ventricular arrhythmias and the prevention of SCD in non-ischaemic cardiomyopathy patients. It is, likely that the ‘classical ICD trials’ also included non-compacted patients, because this diagnosis was frequently overlooked or misclassified because of the unawareness of clinicians and less-advanced imaging modalities. In conclusion, these findings support that EPS as a risk stratification tool is most accurate in patients with dense scar as a stable arrhythmogenic substrate. Based on the above-mentioned facts and the results of Brilakis et al. , 13 invasive EPS in its present form is insufficient to identify high-risk cardiomyopathy patients. Conflict of interest: none declared.