Heart failure in arrhythmogenic cardiomyopathy: is phenotypic variability justamatter of genetics?

Abstract

This article refers to ‘High risk of heart failure associated with desmoglein-2 mutations compared to plakophilin-2 mutations in arrhythmogenic right ventricular cardiomyopathy/dysplasia’ by A. Hermida et al., published in this issue on pages 792–800. Arrhythmogenic cardiomyopathy (AC) is a genetically determined heart muscle disease typically characterized by electrical instability at risk of sudden death.1-3 A disease-causing genetic variant can be identified in about 50% of AC cases, mainly in genes encoding for proteins of the desmosome, such as plakophilin-2 (PKP2), desmoplakin (DSP), desmoglein-2 (DSG2), desmocollin-2 and plakoglobin.2 At difference from all other cardiomyopathies, the presence of a pathogenic variant in disease-related genes is considered a major criterion to establish the diagnosis. However, incomplete penetrance and the difficulty to determine genetic variant pathogenicity are making challenging prognosis and differential diagnosis. Although ventricular arrhythmias and sudden death are the usual clinical manifestation of AC, less frequent presentation is congestive heart failure (HF) that may even mimic dilated cardiomyopathy. In a previous analysis of published studies without genetic data, the incidence of HF and heart transplantation in AC was largely variable, ranging from 0% to 14% depending mostly on the selection criteria of patients, whether referred for arrhythmias or HF.3 In the era of widespread genetic testing, multiple attempts to correlate genetic background with clinical phenotype, including HF, have been undertaken. To this aim, in the current issue of the Journal, Hermida and colleagues compared 27 DSG2 (20 index cases and 7 family members) to 91 PKP2 (58 index cases and 33 family members) gene mutation carriers of a large multicentre AC registry with a 10-year follow-up,4 providing compelling evidence of a higher risk of end-stage HF in missense DSG2 mutation carriers and in patients with a more complex genetic status branded by multiple heterozygosis/homozygosis. Noteworthy, in the first investigation demonstrating DSG2 gene mutations in a significant number of AC-unrelated index cases, cardiac phenotype was characterized clinically by frequent left ventricular involvement; the proband, diagnosed at the age of 55, eventually underwent heart transplantation due to HF at the age of 63.5 Data by Hermida et al.4 are different from those published by the US–The Netherlands6 and Italy7, 8 study groups, who failed to determine HF correlation with the genetic status. Variable prevalence of PKP2 vs. DSP or DSG2 mutation carriers in different study populations, referral bias, variable cohort size and follow-up period might be some of the reasons explaining the differences in the published series. More recently, Gilotra et al.9 reported again lack of correlation between genetics and HF in a large US cohort, despite a similar 10-year follow-up. The role of mutation type (radical vs. missense) in the clinical outcome of AC is still a matter of debate.2, 10 The current study, assigning a higher impact of missense mutations in disturbing functionally important DSG2 protein domains, the lack of copy-number variation evaluation and the assessment of rare variants in other disease-related genes are major limitations.11 Previous studies have reported a significant correlation of PKP2 and DSP radical mutations (premature truncating and slice site genetic variants) with the development of left ventricular dysfunction as well as the absence of correlation between desmosomal mutations and HF/arrhythmic outcome in AC patients.2 Altogether literature data are still scanty and suggest the need for systematic genotype–phenotype correlation studies on large registries to generate unbiased findings. Multiple desmosomal gene mutation carriers have been reported in about 10–25% of AC patients, predicting an adverse arrhythmic outcome and/or symptomatic HF.7, 8 A recent study on a DSG2 founder mutation demonstrated that homozygote DSG2 patients exhibited full disease penetrance, though clinical presentation varied from HF to late-onset symptoms.12 Similarly, Hermida and colleagues confirmed the prognostic value of a more complex genetic status in determining end-stage HF (either death or transplantation) independent of a DSG2 mutation presence, reflecting a variable inheritance disease pattern.4 Furthermore, the authors suggest that histopathologic data of extensive myocardial damage coming from experimental DSG2 animal models, could be in keeping with DSG2-specific pathophysiologic mechanisms predisposing AC DSG2 mutation carriers to HF.13 However, similar structural changes have been described also in desmosomal mouse models other than DSG2, which are characterized as well by biventricular dysfunction, dilatation and aneurysms,14 thus supporting the premise of a general role of desmosomal components in HF. It is a pity that, being a retrospective and multicentre study, the authors did not include data concerning cardiac magnetic resonance in their study population, to assess whether a peculiar distribution of myocardial atrophy and fibro-fatty tissue replacement exists in DSG2 mutation carriers. Finally, the lack of a longitudinal morphological study and of data concerning physical exercise are also major limitations, as recognized by the authors themselves. In particular, it is well known that endurance sports and frequent exercise increase age-related penetrance, the risk of ventricular tachycardia, and the occurrence of HF in AC desmosomal gene carriers, as suggested by both clinical and experimental data.15 In summary, as for other inherited cardiomyopathies, AC patients exhibit a variety of cardiac phenotypes. This is most probably due to the diversity of the disease-causing genes and mutations as well as to the effects of modifier genes and environmental factors. Elucidation of the molecular bases for the various AC phenotypes could lead to more accurate diagnosis, personalized risk stratification and therapeutic management. Supported by Registry for Cardio-Cerebro-Vascular Pathology, Veneto Region, Venice, Italy; Ministry of Health grant RF-2013-02356762 and RF-2014-00000394; Veneto Region Target Research, Venice 933/2015; and University Research grants CPDA144300 and BIRD162733, Padua, Italy; PRIN Ministry of Education, University and Research 2015ZLNETW_001, Rome, Italy, and the CARIPARO Foundation, Padua, Italy. Conflict of interest: none declared.