Arrhythmogenic cardiomyopathy (ACM) is a genetic disorder affecting the myocardium. Initially, it was termed arrhythmogenic right ventricular cardiomyopathy (ARVC) as it was thought to predominantly affect the right ventricle (RV) with predisposition to fatal arrythmias. This is characterised histologically by progressive replacement of myocytes by fibrous tissue, with preponderance in the RV free wall in a region known as the triangle of dysplasia (between anterior part of pulmonary infundibulum, the infero-posterior wall and RV apex) [1]. These lesions typically extend from epicardium to endocardium. In light of developments in imaging, genotyping and our overall understanding of the disease, there is emerging recognition that this disease process is not exclusive to RV with left ventricular (LV) and biventricular ACM being recognised as phenotypic variants. The change in nomenclature from ARVC to ACM is a reflection of our improved understanding of this disease.

Estimated prevalence of ACM is between 1:2500 to 1:5000 but this is likely to be an underestimate as it does not include biventricular or LV dominant variants [2]. As sudden cardiac death (SCD) can be an initial manifestation of ACM, the true prevalence is likely to be higher than reported. The variation in the presentation frequently means that alternate differentials of myocarditis or dilated cardiomyopathy are initially considered before ACM is identified, often at the point of detailed imaging acquired through the use of cardiac magnetic resonance imaging (CMR).

In this review, we aim to provide a succinct overview of genetics and presentation with respect to electrophysiological characteristics. In recent years, there have been several changes and updates to the diagnostic criteria for ACM which we will summarise along with pertinent updates in imaging and management.

ACM is a hereditary cardiovascular disorder with complex genetic underpinning. Understanding the genetic basis of ACM is crucial for accurate diagnosis, risk stratification and management of these patients.

In 2000, a study by McKoy et al. [3] in patients with Naxos disease revealed genetic mutations in a gene (Junctional Plakoglobin (JUP)) that encodes for plakoglobin (a protein that is a component of desmosomes). Desmosomes are specialised structures that facilitate cell-to-cell adhesion and are crucial for maintaining the structural integrity of cardiac tissue. Mutations in desmosomal genes disrupt these adhesion complexes, compromising the mechanical strength of myocardial cells which eventually leads to cell death and fibrofatty replacement [3]. This discovery led to focused efforts to identify other genes that code for desmosomal proteins and this identified mutations in the desmoplakin (DSP) gene, plakophilin-2 (PKP2), desmoglein-2 (DSG2) and desmocollin (DSC2) genes [4, 5, 6]. Non-desmosomal proteins have also been implicated in ACM, including those encoding for adherens junction proteins, ion channels and their modulators and cytoskeleton structures [7].

To date, there are 15 genes that have been implicated in pathogenesis and include both desmosomal genes (PKP2, DSP, DSC2, DSG2 and JUP) and nondesmosomal genes (transmembrane protein 43 (TMEM43), desmin (DES), titin (TTN), phospholamban (PLN), and ryanodine receptor-2 (RYR2)) [7, 8]. Of these 15 genes, 8 genes (PKP2, DSP, DSG2, DSC2, JUP, TMEM43, PLN, and DES) have at least moderate evidence to be considered ACM causative [9]. It is important to note that many have no detectable pathological variants; these patients may reflect unknown pathological variants, or may harbour mutations in genes not currently associated with ACM [10].

Up to 50% of ACM cases are thought to be caused by mutations affecting desmosomal proteins [8], and of these mutations, the PKP2, encoding Plakophilin-2, is implicated in up to 20–46% of cases [7]. Plakophilin-2 plays a critical role in stabilizing desmosomes, and its mutations contribute to the breakdown of these structures initiating the pathological changes observed in ACM. Mutations in DSP and DSG2 genes are thought to account for 10% of cases of ACM whilst mutations in DSC2 account for approximately 5% of cases. Mutations in JUP gene are thought to account for less than 1% of cases [7]. Desmosomal mutations tend to have an autosomal dominant (AD) inheritance pattern with incomplete penetrance leading to an isolated cardiac phenotype with PKP2 mutations being mostly AD in inheritance and most likely to lead to conventional phenotype than other desmosomal mutations [1, 11]. In a study performed by Biernacka et al. [12], ACM patients with PKP2 mutations were less likely to present with LV involvement and heart failure symptoms and overall had a favourable prognosis compared to other mutations. Patients with DSP mutations present with variable phenotypes such as biventricular cardiomyopathy, or isolated LV arrhythmogenic cardiomyopathy with Rigato et al. [13] presenting higher degree of LV involvement in patients with DSP mutations. DSP mutation phenotypes of ACM have also been associated with higher risk of ventricular arrythmias and SCD [14]. Homozygous or compound heterozygous mutations in JUP, DSP and DSC2 can lead to cardio-cutaneous syndromes such as Naxos and Carvajal syndromes [11].

Non-desmosomal genes contribute to development of ACM in several ways; TMEM43 gene encodes for a nuclear envelope protein and mutations in this gene have been associated with high penetrance and risk of SCD especially in young men [15]. DES genes encode for intermediate filament desmin and have been identified in patients with ACM (with right predominant or biventricular involvement). DES mutations have also been linked to dilated cardiomyopathies (DCM), suggesting an overlap syndrome [16]. TTN genes encode for sarcomeric protein titin and have been found in high proportion of patients with ACM (18%) [11].

Filamin C, an actin binding protein (encoded by FLNC gene) plays a vital role in anchoring membrane proteins to cytoskeleton, thereby contributing to sarcomere maintenance. Truncating mutations in FLNC have been linked with phenotypically left dominant ACM with high risk for sudden cardiac death [17]

Finally, given the hereditary component to the disease, family screening and genetic counselling forms an integral part of diagnosis. Indeed, family screening has been a component of the diagnostic criteria in all its iterations [18, 19, 20]. Current guidelines recommend that first-degree relatives undergo clinical evaluation every 1–3 years with 12-lead electrocardiogram (ECG), ambulatory ECG and imaging [21]. However, disease expression of ACM is variable even in the same family or those carrying the same pathogenic mutation, therefore, Muller et al. [22] recently undertook a study to determine if there are predictors of disease development among at-risk relatives. In this study, symptomatic relatives, those 20 to 30 years of age and those with one minor task force criteria had higher hazard for developing ACM. This may help clinicians risk stratify patients that may benefit from more frequent follow-up.

Understanding the electrophysiological perturbations implicated in ACM is essential for screening, diagnosis, risk stratification and guiding therapeutic interventions. The nature of structural changes within the myocardium modulates transmission of electrical signals and can result in conduction defects that may be visible on a standard 12-lead ECG (Fig. 1, Ref. [20]).

The most common ECG changes include:

$\bullet$ T-wave abnormalities: Fibrofatty replacement leads to delayed repolarisation and altered activity during the repolarisation phase (phase 3). This is represented by T wave inversion. Negative t-waves in V${}_1$–V${}_3$ (in the absence of right bundle branch block) is observed in RV dominant phenotypes whilst negative t-waves in V${}_4$–V${}_6$ (in the absence of left bundle branch block) is observed in LV dominant phenotype.

$\bullet$ Depolarisation abnormalities manifesting as a low amplitude signal which is situated between the end of the QRS and start of the T wave, most discernible in V${}_1$–V${}_3$. Terminal activation duration of QRS $>$55 ms measured from nadir of S wave to end of QRS R’ in V${}_1$–V${}_3$ in the absence of complete right bundle branch block.

$\bullet$ Prolonged S-wave upstroke in V${}_1$–V${}_3$, which reflects delayed right ventricular activation.

$\bullet$ Epsilon waves, small positive deflections at the end of QRS complex in leads V${}_1$–V${}_3$; indicative of delayed depolarisation in the right ventricular free wall.

$\bullet$ Low QRS voltage (0.5 mV) in limb leads indicating LV involvement

Often, subtle ECG changes may not be visible on 12-lead surface ECG; therefore, this requires more sophisticated techniques for identification. Signal-averaged electrocardiogram (SAECG) is a non-invasive signal-processing technique to detect subtle abnormalities in the surface ECG. It can identify low-amplitude signals at the terminus of the QRS complex, referred to as ‘ventricular late potentials’. These represent delayed activation and predispose to re-entrant arrythmias. SAECG uses computerised averaging of ECG complexes during sinus rhythm to detect small amplitude signals which occur later than rapid ventricular activation. In the 2010 International Task Force (ITF) diagnostic criteria, the presence of late potentials on SAECG was considered a minor criterion for diagnosis based on studies supporting their increased sensitivity and specificity for diagnosis of ACM [18, 23]. Interestingly, in the newer 2020 ‘Padua Criteria’, SAECG are not considered part of the diagnostic work up with the authors reporting low diagnostic accuracy [19].

Additionally, invasive electrophysiological studies (EPS) can detect and provide direct visualisation and assessment of arrhythmogenic substrate. It is noted that identifying the genotype can help confer the phenotype relating to the expected location of fibrosis, e.g., variants of the lamin A/C genotype manifest substrate foci within subaortic, mid and basalseptal regions [24].

Changes that can be observed on EPS include:

$\bullet$ Delayed activation: ACM often leads to delayed electrical activation in either/both ventricles. This delayed activation can be seen in the form of prolonged endocardial activation duration.

$\bullet$ Fractionated potentials and complex electrogram (EGM) signals: These can be observed during EPS and are indicative of heterogenous conduction, which occurs when electrical impulses are slowed down or rerouted due to abnormal tissue. Fractionated EGMs are those that show multiple spikes within a single cardiac cycle, suggesting that the electrical impulse is meandering through the heart tissue rather than following the conventional path. Complex EGMs may include late potentials and local abnormal ventricular activities (LAVA), which are signals that occur after the main ventricular activation and are associated with areas of slow conduction.

$\bullet$ Low voltage areas: The presence of fibrous and fatty tissue can dampen the amplitude of electrical signals because these tissues are less electrically active than normal myocardial tissue. During an electrophysiological (EP) study, low voltage signals are used as criteria to identify areas of the myocardium which are affected [25].

$\bullet$ Ventricular tachycardia (VT): ACM is associated with ventricular arrhythmias, including VT. EPS can induce these arrhythmias to understand origin and mechanism. VT patterns in ACM can originate based on two predominant mechanisms and typically manifest at different times in the disease course. Early phase or rapid VT occurs at a younger age with the trigger being exercise. Late phase VT presents at a later point and is triggered by scar tissue. Anatomical mapping shows low voltage signals associated with fatty infiltrate and scar tissue. Early phase VT often presents with exertional exercise whilst late phase VT presents later and at rest with no neuronal triggers. Early phase VT responds to beta blockers given its neurohormonal trigger. Late phase VT may respond to catheter ablation though long-term outcome benefit is unclear.

$\bullet$ VT entrainment utilises catheters with multiple electrodes to record electrical signals across the heart, to pinpoint origin and mechanism of tachycardia. Activation mapping identifies the critical isthmus consisting of the entry, mid-isthmus and exit pathways. It is important to acknowledge the role of the epicardial approach in these studies, given the fact that arrythmogenic material is often situated in the subepicardial or middle layers of the heart muscle. This approach is not performed in all EP studies [25].

Mapping the heart’s electrical activity in three dimensions allows identification of fixed lines of conduction block associated with reentrant circuits. This approach allows construction of a 3D map that can help plan catheter ablation approaches [26]. Further work mediated through simultaneous endocardial and epicardial delineation allowed for construction of 3D structures derived from both small and large substrate areas [27].

Diagnosis of ACM can be challenging as it mimics other cardiomyopathies in presentation and imaging. However, accurate diagnosis is vital as it allows early treatment initiation and cascade screening. The 2010 revised International Task Force (ITF) criteria provided an update to the Original Task Force guidelines in the diagnosis of ACM [18]. It is a scoring system containing six disease characteristics: structural alterations on imaging, tissue characterisation, repolarisation abnormalities, depolarisation abnormalities, arrhythmia and family history. In each category, there is major, minor or no criteria that can be fulfilled. A diagnosis is confirmed if 2 major, or 1 major and 2 minor, or 4 minor criteria are fulfilled from different disease categories. A criticism of the 2010 ITF criteria has been that it focused on RV phenotypic manifestations without recognising cohorts with biventricular or LV dominant variants. For instance, the major criteria for structural alteration exclusively describes RV regional abnormalities with no reference to LV structural abnormalities. Although the 2010 ITF criteria increased diagnostic yield of ACM, it cannot be applied to ACM with LV involvement [28].

In 2020, the International Expert consensus document (‘the Padua Criteria’) provided an update to diagnostic criteria to address some of the limitations in the 2010 ITF criteria [19]. This newer classification categorises ACM into three phenotypic variants: the dominant-right, biventricular and dominant-left. Although differences between these criteria have been described well elsewhere in the literature, it is important to note some key differences [29]. As endomyocardial biopsy (EMB) is invasive with potentially serious sequelae, there is less emphasis on EMB for tissue characterisation. In the 2020 ITF criteria, this is reserved for probands with negative genotyping where histology can exclude phenocopies such as cardiac sarcoidosis. Recent advancements in use of electroanatomic voltage mapping (EMV) guided EMB has been proven to be a promising tool for targeted EMB, thereby improving diagnostic yield and reducing complication rates and should be considered when EMB is being explored [30]. There is also more emphasis on the use of contrast-enhanced cardiac magnetic resonance imaging (CE-CMR) for morphological assessment and tissue characterisation. With respect to LV involvement, there is addition of new repolarisation, depolarisation and ventricular arrythmia criteria for this subgroup. Family history and genetics remain unchanged between the two criteria.

The implementation of ‘Padua criteria’ improved diagnostic yield of ACM particularly in relation to LV involvement [29]. A refinement to the 2020 criteria was recently published this year (2024 European Task Force consensus report) with changes to major criteria for diagnosis of arrhythmogenic left ventricular cardiomyopathy (ALVC) based on cardiac magnetic resonance imaging (CMR), late-gadolinium enhancement (LGE) and ECG features (Table 1, Ref. [20]). A ‘ring-like’ pattern of LGE (subepicardial or midmyocardial involvement in $\ge$3 Bull’s Eye segments) is now a considered a major criteria as it was highly specific for ALVC. Low voltage QRS (0.5 mV peak to peak) in the absence of other causes was upgraded from minor to major in this update as it is rarely observed in healthy individuals and is specific to myocardial scarring in ACM [31]. There has been not been any study to date comparing the 2020 ‘Padua criteria’ with the 2024 European task force update but it is anticipated that this will improve diagnostic yield of ACM even further.

Non-invasive imaging plays a key role in diagnosis. Both transthoracic echocardiogram (TTE) and CMR feature in the 2010 ITF criteria and the 2020 ‘Padua Criteria’ for the diagnosis of ACM. Recently, computed tomography (CT) has also emerged as a robust tool in RV assessment [32].

The recent scrutiny and updates to diagnostic parameters meant that our inclusion criteria has changed redefining the nature of ACM. This improved phenotypic understanding of ACM provides several avenues that require future consideration including risk stratification (for the biventricular and LV phenotypes), earlier detection and pharmacological therapies to target underlying disease mechanism.

Earlier detection and diagnosis is a challenge as most patients are identified as part of a cascade screening process or after life threatening arrythmia. Although CMR provides excellent tissue characterisation and aids diagnosis, it is challenging to decide who should undergo this investigation as it is not universally accessible and can be expensive. One option to navigate the latter issue would be the use of artificial intelligence (AI) and machine learning into CMR analysis to enhance speed and accuracy of image interpretation and risk assessment [64]. Another benefit of CMR is the ability to detect myocardial scar. LV scar burden has been noted to be a predictor of arrhythmogenesis and ICD therapy in observational studies [65]. In a recent study by Gutman et al. [66], 452 patients with non-ischaemic cardiomyopathy and LV ejection fraction $\le$35% who fulfilled criteria for primary prevention ICD were stratified according to LV scar burden to assess mortality benefit. In this study, the mortality benefit was higher in patients with primary prevention ICD and LV scar (hazard ratio 0.4; 95% CI 0.21–0.76). Although this has not been explored in the context of ALVC, with the increased use of CMR in the diagnostic pathway, this could be a valuable tool in patient selection for ICD therapy in ACM that should be explored.

Given the genetic basis for ACM, there is an appreciation that gene therapy may serve as a potential option for treatment. A mouse model was generated with the PKP2 mutation to replicate the ACM features [67]. Adeno-associated viral therapy containing the corrected gene variant AAV-PKP2 was found to prevent desmosomal pathological deficits. Other successful endpoints including resolution of CMR changes and length of survival were associated with AAV-PKP2 gene therapy.

The role of cytokines, particularly TGF$\beta$3, has been explored in the context of ACM [68]. Mutations in the TGF$\beta$3 gene, particularly in non-coding regions, have been associated with ACM, suggesting mechanisms involving paracrine and autocrine signalling. Likewise, molecular targets such as Glycogen synthase kinase-3 (GSK-3) have been explored given their role in cardiac electrophysiology [69]. In the context of ACM, this protein when inhibited in murine models mediated an improvement in histology, arrythmia burden at rest and on exertion [70]. This emphasises an area that has not been investigated thoroughly and may serve as a platform for future interventional therapy.

In relation to symptom management, bilateral cardiac sympathetic denervation (BCSD) has been used in the context of refractory ventricular arrythmia for its antiarrhythmic effect [71]. There is emerging evidence from small studies that this approach can be used in the context of ACM to reduce ICD shocks or sustained VT [72]. At present, this is a treatment option only for refractory cases with failed conventional therapy and is available in a few select centres with unclear long-term benefit.

ACM is a hereditary cardiomyopathy that has deleterious consequences as it may present with life threatening arrythmias and sudden cardiac death. Over the last few years, our improved understanding of the disease process has enabled refinement of the diagnostic process, allowing categorisation based on phenotypic manifestations (RV, LV or bi-ventricular predominance). There have been several updates to the diagnostic criteria for ACM which we have summarised in this review, including discussion on imaging techniques and risk stratification. We have explored current therapeutic options including pharmacological management and device therapy. Although gene therapy and immune modulation show early promise, it is still in its infancy and formal trial data is lacking.

ZZ, NS and ND were involved in methodology, data synthesis, write up and review of the manuscript. PAP contributed to conceptualisation, review, and overall supervision as senior author. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.