Arrhythmogenic cardiomyopathy

Arrhythmogenic Cardiomyopathy (AC, ACM), Arrhythmogenic RV Cardiomyopathy (ARVC), Arrhythmogenic RV Cardiomyopathy/Dysplasia (ARVC/D), Arrhythmogenic RV Dysplasia (ARVD).

More than 20 years elapsed since the 1995 WHO definition and classification of cardiomyopathies [12], the genetic background has been discovered and new diagnostic tools are now available. Thanks to genotype-phenotype correlation, we know that the disease spectrum is wider than previously thought, with LV involvement present and even dominant at early stages. Thus, nowadays the original disease definition could be updated as: “AC is characterized by progressive fibro-fatty replacement of ventricular myocardium, including RV and LV, with relative sparing of the septum. Presentation with arrhythmias and SD is common, particularly in the young. Mutations in genes mostly encoding for desmosomal proteins are found in about half of probands”.

The estimated prevalence of AC in the general population ranges from 1:2000 to 1:5000 [3, 5]. Considered in the past an endemic disease in North East Italy (“Venetian disease”), AC is now well recognized in human populations of different ethnicity, but no data are available on its prevalence across various countries. AC affects more frequently males than females (up to 3:1), despite a similar prevalence of carrier status between sexes, and becomes clinically overt most often in the second-fourth decade of life [3, 5]. More rarely, symptoms and signs can appear before puberty or in the elderly. However, occasionally the first clinical manifestations arise even in patients >70, but the diagnosis is often missed because clinicians do not take it into consideration this morbid entity in this older age-group [3].

AC is a “structural” cardiomyopathy characterized by dystrophy of the ventricular myocardium with replacement by fibro-fatty tissue [1, 2] (Fig. 1). Myocardial atrophy occurs progressively with time, starts from the epicardium and eventually extends to become transmural. This entity should not be confound with Uhl’s disease, a congenital heart defect in which the RV myocardium fails to develop during embryonic life [9, 13]. The gross pathognomonic features of AC consist of RV aneurysms, whether single or multiple, located in the so-called “triangle of dysplasia” (i.e. inflow, apex and outflow tract) [2, 14]. Nevertheless, grossly normal hearts have been reported in whom only a careful histopathology investigation can reveal AC features. Even cases with isolated or predominant LV involvement [4] are no so rare. Indeed, up to 76 % of the AC hearts studied at post-mortem disclosed a LV involvement, usually limited to the subepicardium or midmural layers of the postero-lateral free wall [2, 15]. Hearts with end-stage disease and congestive heart failure usually show multiple RV aneurysms and huge chamber dilatation, with a higher prevalence of biventricular involvement, while the ventricular septum is mostly spared [2].

Histological examination reveals islands of surviving myocytes, interspersed with fibrous and fatty tissue [1, 2]. Fatty infiltration of the RV is not a sufficient morphologic hallmark of AC [16, 17] and replacement-type fibrosis and myocyte degenerative changes should be always searched for. Myocyte necrosis is seldom evident and may be associated with inflammatory infiltrates [2, 18]. Myocardial inflammation has been reported in up to 75 % of hearts at autopsy. An apoptotic mechanism of myocyte death has been also proven [19, 20]. Rather than being a continuous, ongoing process, disease progression may occur through periodic “acute bursts” of an otherwise stable disease, as to mimic myocarditis or myocardial infarction with normal coronary arteries. In a desmoglein-2 (dsg2) transgenic animal model, myocyte necrosis was demonstrated to be the key initiator of myocardial injury, triggering progressive myocardial damage, followed by an inflammatory response [21]. The detection of viral genomes let to advance an infective viral etiology, but it is most likely that either viruses are innocent bystanders or that myocardial cell degeneration may serve as a milieu favoring viral settlement [22].

In adolescents or young adults, AC usually presents with palpitations, syncope, or cardiac arrest. Then, premature ventricular complexes (PVC) or ventricular tachycardia (VT) with LBBB morphology and T-wave inversion in V_1- V₃ leads on basal electrocardiogram (ECG) are the most common index of suspicion. Less-common presentations are RV or biventricular dilatation, with or without heart failure symptoms, mimicking dilated cardiomyopathy. Clinical manifestations vary with age and stage of disease. Men usually develop a more severe phenotype, most likely because of the effect of vigorous sport activity on disease onset and progression and influence of sex hormones.

Syncope, palpitations and ventricular arrhythmias are the usual presenting symptoms also in the paediatric age [23]. However, non-specific clinical features not infrequently consist of myocarditis or myocardial infarction like-picture with chest pain, dynamic ST-T wave changes on basal 12- lead ECG and myocardial enzymes release with normal coronary arteries [3].

Four phases are recognizable along the natural history of the classic AC variant: 1) “concealed”, with subtle RV structural changes, with or without ventricular arrhythmias; 2) “overt electrical disorder”, with symptomatic life-threatening ventricular arrhythmias associated with clear cut RV morpho-functional abnormalities; 3) “RV failure”, due to progression and extension of RV disease; and 4) “biventricular failure”, caused also by pronounced LV disease [24]. Electrical instability may lead to arrhythmic SD any time during the course of the disease [2, 5, 6, 25‐27]. AC has been reported as the second cause of SD in the young and the first cause in competitive athletes in the Veneto Region of Italy [1, 6, 7]. The incidence of SD ranges from 0,08 · to 3,6 % per year in adults with AC [3, 5, 25‐27]. While patients with an overt disease phenotype more often experience scar-related re-entrant VT, those with an early stage or a “hot phase” of the disease may present with ventricular fibrillation (VF) due to ongoing myocyte death and reactive inflammation [27]. More recently, gap junction remodelling and sodium channel interference have been advanced in experimental models as alternative substrates for life-threatening arrhythmias even in the pre-phenotypic disease stage [28, 29].

No single gold-standard is available for AC diagnosis. Multiple criteria are needed, combining different sources of diagnostic information, such as morpho-functional abnormalities (by echocardiography and/or angiography and/or cardiac magnetic resonance-CMR), histopathological features on endomyocardial biopsy (EMB), ECG, arrhythmias and family history, including genetics (Fig. 2). The diagnostic criteria, originally put forward in 1994 [30], have been revised in 2010 to improve diagnostic sensitivity, but with the important prerequisite of maintaining diagnostic specificity (Table 1) [8].

Quantitative parameters have been included and abnormalities were defined, based on the comparison with normal subject data. Moreover, T-wave inversion in V₁-V₃, and VT with a LBBB morphology with superior or indeterminate QRS axis (either sustained or no sustained), have become major diagnostic criteria; and T-wave inversion in V₁-V₂, in the absence of right bundle branch block (RBBB), and in V₁-V₄, in the presence of complete RBBB, has been included among the minor criteria. Finally, in the family history category, the confirmation of AC in a first-degree relative, by either meeting current criteria or pathologically (at autopsy or surgery), and the identification of a pathogenic mutation, categorized as associated or probably associated with AC, in the patient under evaluation are considered major criteria. Because of the diagnostic implications, however, caution is highly recommended since the pathogenic significance of a mutation is increasingly questioned (see genetic section) [8].

AC diagnosis is exceptionally made below the age of 10 [23]. The diagnostic criteria in adults have been demonstrated to be also valid in the pediatric age group, with the exception of inverted T wave on right precordial leads in children < 12 years of age, which may be normal. However, negative results are quite common before adolescent growth is completed, due to absent or limited morpho-functional phenotype, since AC usually shows an age-related penetrance [3, 5]. Follow-up by non-invasive clinical investigation of children who have a family and/or personal history suspicious for AC or healthy gene carriers is recommended on a regular basis, to monitor the pending disease onset in the pubertal period.

The 2010 criteria acknowledge that RV AC is the best recognized variant of a broad disease spectrum, that includes also LV and biventricular subtypes [8]. While the revised criteria easily catch the classical RV variant, there is lack of specific diagnostic guidelines for non-classical disease patterns, thus explaining the under-recognition of the LV one. The only step forward with respect to the 1994 criteria [30] is that patients with a moderate-to-severe LV dysfunction can now be diagnosed as affected by AC. ECG abnormalities such as lateral or inferolateral T-wave inversion (leads V₅, V₆, L_I, and aVL), low voltage QRS complex on peripheral leads and RBBB/polymorphic ventricular arrhythmias suggest a left-side involvement [31‐33] (Fig. 3 and Table 2). Contrast-enhanced CMR is the more sensitive imaging tool to catch LV involvement by non-invasive tissue characterization [33‐35]. Late gadolinium enhancement is in fact a far more-sensitive indicator of even early or minor left-sided disease, and is frequently detected in a wall segment without a concomitant morphofunctional abnormality, thus preceding the onset of LV dysfunction or dilatation. Typically, LV late gadolinium enhancement involves the inferolateral and inferoseptal regions, and affects the subepicardial or midwall layers (Fig. 4). Differential diagnosis with dilated cardiomyopathy is mandatory for risk-stratification and familial evaluation purposes. The propensity to electrical instability that exceeds the degree of ventricular dysfunction is typical of left dominant AC, at difference from dilated cardiomyopathy where life-threatening ventricular arrhythmias usually occur in the context of systolic dysfunction with low ejection fraction (<35 %). Moreover, a regional rather than global involvement is more in keeping with AC, particularly when RV abnormalities are also prominent. Thus, a revision of the 2010 Task Force criteria is needed to fill the gap by incorporating features suggesting LV involvement.

Common dilemmas in differential diagnosis include myocarditis, sarcoidosis, RV infarction, dilated cardiomyopathy, Chagas disease, Brugada syndrome, pulmonary hypertension and congenital heart disease with right chambers overload [3]. Moreover, one of the main clinical challenge is still differentiation of AC from idiopathic RV outflow tract VT, which is usually benign and non-familial (Table 2) [36]. EMB can be crucial both to rule out the phenocopies such as myocarditis and sarcoidosis, especially when dealing with sporadic forms, and in the setting of negative or doubtful CMR and/or electroanatomic voltage mapping [37]. The latter is an invasive electrophysiological tool that can be performed in selected patients with suspected AC, in the setting of ventricular arrhythmias of RV origin, and/or when contrast-enhanced CMR is negative or doubtful in terms of RV involvement (Fig. 5) [36, 38]. The abnormal low-voltage areas found in AC patients correspond to the loss of electrically active myocardium caused by fibrofatty replacement (“electrical scars”). Noteworthy, because of the wavefront progression of RV fibro-fatty replacement from the epicardium to the endocardium, scar tissue in non-advanced stages may be confined to epicardial/midmural layers, sparing (or reaching focally) the endocardial region. Thus, bipolar endocardial voltage mapping of the RV free wall may underestimate or miss nontransmural low-voltage areas [39].

Finally, in people performing sport activity, AC should be distinguished from so-called “athlete heart”, i.e. physiological adaptation to training with hemodynamic overload. RV enlargement, ECG abnormalities and arrhythmias are well documented in endurance athletes, reflecting the increased hemodynamic load during exercise [3, 40, 41]. Global RV systolic dysfunction and/or regional wall motion abnormalities, such as bulgings or aneurysms, are in keeping more with AC than physiologic ventricular enlargement. The absence of overt structural changes of the RV, frequent PVCs or inverted T waves in precordial leads all support a benign nature.

In 1982, Marcus et al. advocated the inherited nature of AC, with the description of two affected adult cases in the same family [14]. The Padua group then demonstrated that the classical inheritance pattern of AC is autosomal dominant with variable expression and age-dependent penetrance [5, 42]. Inter- and intra-familial phenotype diversity is frequent in AC, with co-existence of both the classic RV and dominant LV pattern in the same family and/or life-threatening ventricular arrhythmias in probands vs a more favorable prognosis in relatives [5, 31, 32]. The first disease-causing gene [43], i.e. the plakoglobin gene (JUP), was identified in a fully penetrant autosomal-recessive form of AC associated with palmoplantar keratosis, also known as Naxos or cardiocutaneous syndrome [44]. Subsequently, mutations of the desmoplakin (DSP) gene were found to cause another autosomal recessive cardiocutaneous syndrome, i.e. Carvajal syndrome [45]. Soon after, heterozygous mutations in the same gene were identified in a dominant form of AC without hair/skin abnormalities [46]. To-date, 11 disease genes have been linked to the AC phenotype, highlighting genetic heterogeneity. Most of mutations in dominant forms have been identified in desmosomal genes including DSP, plakophilin-2 (PKP2), desmoglein-2 (DSG2), desmocollin-2 (DSC2) and JUP [43, 46‐50] (Table 3). Only isolated reports showed causal mutations in non-desmosomal genes, such as transmembrane protein 43 (TMEM43), desmin (DES), titin (TTN), Lamin A/C (LMNA), phospholamban (PLN) and αT-catenin (CTNNA3), sometimes with a clinical phenotype similar but not identical to AC, as to be considered phenocopies or overlap syndromes [51‐56]. Moreover, mutations in the regulatory region of transforming growth factor beta-3 gene have also been reported [57], but their pathogenicity is still controversial. Finally, ryanodine receptor 2 gene mutations are associated with catecholaminergic polymorphic VT (CPVT-see) rather than with AC, as originally considered [58].

Thus, most of pathogenic mutations address structural proteins that are involved in the organization of the intercalated disc, which has been described as containing a mixed-type junctional structure (the so-called area composita) (Fig. 6) [59]. Comprehensive exonic sequence analysis of the known desmosomal AC-related genes currently identifies approximately 50 % of AC probands [60‐63]. The most commonly defective AC gene is PKP2 (10–45 %), followed by DSP (10–15 %), DSG2 (7–10 %) and DSC2, JUP (1–2 %). About 10–25 % of AC patients are compound and heterozygous mutation carriers [60, 65, 66] (Fig. 7). Founder mutations in both desmosomal and non-desmosomal encoding genes have been reported [53, 64, 65].

The inheritance pattern of AC is more complex than previously appreciated, with frequent requirement for more than one ‘hit’ for fully penetrant disease [60, 61, 66, 67]. The low penetrance of AC may be explained by a “recessive-like” inheritance pattern, based on the fact that AC probands often carry homozygous or compound heterozygous variants in the same gene, or digenic/oligogenic variants in a cluster of desmosomal genes.

Nowadays, the pathogenicity of missense and radical (nonsense, frameshifts, splice sites etc) mutations in cardiac diseases and specifically in AC is a matter of debate. Recent studies have shown that stop-coding in disease-causing genes are more pathogenic than missense mutations, since the former are causing alteration of protein length and conformation, leading to haploinsufficiency due to protein instability [67‐70]. Entire PKP2 exons or even whole gene deletions have been recently described in AC families with a frequency of approximately 2 % [71‐73]. These data are further stressing the question whether haploinsufficiency is enough to determine the disease phenotype.

Genotyping success rate in AC varies according to cohort location and ethnicity, sequencing techniques, selection criteria and the stringency of the standards by which mutations are considered causal. An allelic frequency lower than 0.02–0.05% is considered pathogenic or likely pathogenic. With the routine use of Next Generation Sequencing, the analysis of large panel of genes may lead to the identification of a high number of sequence variants with uncertain clinical significance. Thus, genetic testing and its interpretation should be performed in dedicated AC cardio-genetic centers, with pre- and post-counseling facilities. Given that the prevalence of causal genes and mutations has yet to be determined, a negative genetic test does not exclude a genetic predisposition. The major advantage of a positive genetic test in the affected AC proband consists in the possibility to identify early asymptomatic carriers by cascade genetic screening of family members. Finally, although prenatal diagnosis through amniocentesis is feasible, its application in many countries is subjected to ethical and legal issues; this is even more for pre-implantation diagnosis, which is a long procedure restricted to severe and untreatable diseases.

The most important goals of clinical management of AC patients comprise: 1) reduction of mortality, either by arrhythmic SD or death due to heart failure; 2) prevention of disease progression leading to RV, LV or biventricular dysfunction and heart failure; 3) attenuation of symptoms and improvement of quality of life by decreasing or suppressing palpitations, VT recurrences or implantable cardioverter defibrillator (ICD) discharges (either appropriate or inappropriate); and 4) reducing heart failure symptoms and increasing exercise capacity [74]. The AC management options regard life-style modifications, pharmacologic treatment, catheter ablation, ICD implantation, and exceptionally heart transplantation [75, 76]. Figure 8 summarizes the flow chart for clinical treatment of AC.

Before any therapy is undertaken, life-style modification should be pursued. Sport activity is associated with an increased risk of SD, thus supporting the concept that avoiding effort is “per se” life-saving [1, 6, 7]. Recently, it has been demonstrated that endurance sports and frequent exercise increase age-related penetrance, risk of VTs, and occurrence of heart failure in AC desmosomal gene carriers [77, 78]. Pregnancy is generally well tolerated but a pre-conception evaluation is mandatory for individualized arrhythmic risk stratification and prescription of the best antiarrhythmic therapy [79]. β--blockers treatment is better since no teratogen effects are known but they may be associated with intrauterine growth retardation and neonatal bradycardia or hypoglycemia [75].

Different antiarrhythmic drugs have been employed, such as sodium blockers, β-blockers, sotalol, amiodarone, verapamil alone or combinations. Wichter et al. [75] reported various efficacy rates by demonstrating that sotalol in a dosage of 320-480 mg/day is the most powerful anti arrhythmic drug for inducible and non-inducible VT in AC patients; amiodarone is less effective and has a high risk of extracardiac side effects during long-term follow-up; β-blockers are effective only in non-inducible patients. Co-administration of more than one drug should be avoided. However, contradictory data regarding the effectiveness of empiric anti-arrhythmic drugs have been published, showing either a higher efficacy of amiodarone or inefficacy of anti-arrhythmic drugs against ICD intervention and SD [76, 80]. Differences in the study populations, drug doses, therapeutic approach (empiric vs. guided by electrophysiologic study/Holter) and follow-up duration may explain the divergence [80].

Catheter ablation of the re-entry circuit is a non pharmacological therapeutic option for AC patients who have VT. In fact, VT is the result of a scar-related macro-reentry circuit due to RV fibro-fatty replacement, which is suitable for mapping and interruption by catheter ablation [81‐84]. Catheter ablation may be guided by either conventional electrophysiology or substrate-based mapping. Linear ablation lesions connecting or encircling ventricular scar areas obtain the isolation of the re-entry circuit. In the presence of a large RV scar burden and/or in patients with VT recurrence, combined endo- and epi- substrate-based VT ablation, incorporating scar dechanneling technique, would increase the short- and long-term success rate. However, the epicardial approach has a significant procedural complication rate (up to 8 %) and should be always performed in high volume referral centers [84].

ICD therapy is the first line approach for the highest-risk patients, whose natural history is typically characterized by the risk of SD [27, 80, 85‐87]. Data obtained from either primary or secondary prevention studies indicate that ICD therapy improves long-term outcome of selected high-risk AC patients, with an estimated mortality reduction ranging from 20 to 30 %. Overall, 48–78 % of patients received appropriate ICD interventions during the mean follow-up period of 2–7 years after implantation [27, 87]. The published studies on ICD therapy in AC patients have provided also valuable information about the risk predictors for VF or VT triggering appropriate ICD discharges during follow-up. A stronger predictor of a life-saving ICD intervention was aborted SD due to VT/VF and syncope. Other risk factors associated with an increased risk of appropriate ICD interventions included hemodynamically well tolerated VT, either sustained or non-sustained, severe RV and/or LV dysfunction, young age, proband status, frequent polymorphic PVCs (≥1000/24 h), and inducibility at programmed ventricular stimulation (PVS). The presence of multiple risk factors increases the likelihood of appropriate ICD therapy. Most importantly, asymptomatic probands and relatives without relevant risk factors as well as healthy gene carriers showed a low rate of arrhythmic events over a long-term follow-up, regardless of family history of SD. Despite well-known ICD benefit on survival, disadvantages are related to the lead and device-related complications as well as the inappropriate ICD intervention, which occurs in 10–25 % of AC patients and is usually caused by sinus tachycardia or atrial tachyarrhythmia [27, 87]. The inappropriate interventions are painful and, especially in young patients, may have a severe psychological impact hindering the compliance to ICD therapy. When the incidence of inappropriate ICD discharges is too high, the patients can take advantage by appropriate ICD reprogramming and/or co-administration of β-blockers therapy.

Arrhythmic risk stratification relies on phenotypic predictors, such as previous cardiac arrest due to VF, sustained VT, unexplained syncope, severe RV or LV dilatation/dysfunction, compound and digenic heterozygosity of desmosomal gene mutations, low QRS amplitude, QRS fragmentation, male gender, young age at time of diagnosis, proband status, inducibility at PVS, burden of electroanatomic scar and scar-related fractioned electrograms, extent of T wave inversion across precordial and inferior leads on ECG. In a recent document on risk stratification and treatment of AC [87], indications for ICD implantation were determined by consensus taking into account the statistical risk, the general health, the socioeconomic factors, the psychological impact and the adverse effects of the device. The flow chart is represented in Fig. 9.

Noteworthy, these recommendations apply to the classical RV AC variant and prognostic data are not yet available in LV AC, which is increasingly detected by contrast-enhanced CMR.

Transgenic animal models that mimic the human AC phenotype (mice and zebrafish) and induced pluripotent stem cells (iPSCs) from affected patients are useful tools to explore how the mechanical and/or functional disruption of cell junctions by mutant desmosomal proteins leads to cardiomyocyte death and subsequent repair with fibrous and fatty tissue.