Identification of causative desmosomal mutations has resulted in ARVD to be increasingly viewed upon as a desmosomal disease and has provided insight into its pathogenesis. Mutations affecting components of the cardiac desmosome are believed to disrupt cell-cell adhesion, provoking detachment of myocytes particularly under conditions of mechanical stress. This, in turn, might lead to apoptosis and cell necrosis, predisposing to inflammation and repair by fibro fatty substitution.

Accordingly, early ARVD shows predilection for the regions of heart with relatively thinner walls and high shear stress – the subtricuspid portion, outflow tract, apex and mid-free wall of the RV ("triangle of dysplasia") and the inferior wall septal junction and infero-lateral wall of the LV. Young individuals and athletes are predisposed because of high level of physical activity and resultant greater strain on heart. Gap junction remodeling associated with impaired mechanical coupling at cell adhesion junctions may result in a predisposition to the arrhythmogenicity seen in ARVD. Furthermore, adrenergic stimulation during highly strenuous activity may precipitate arrhythmic events in patients with ARVD. Desmosomal mutations may induce nuclear protein dysfunction as desmosomal proteins are believed to be involved in regulation of transcription. Desmosomal mutations may also heighten the pathogenic role of viral infection speculated in the causation of ARVD 343637.

A pathogenic role of cytoplasmic calcium overload is shown by the association of mutations in RYR2 with ARVD2. This type of ARVD has a very different clinical presentation and is rather distinct from the more common types of ARVD. RYR2 has also been independently implicated in isolated familial catecholaminergic polymorphic VT (CPVT) reflecting possibly different mutations within the same gene and the probable role of effect-modification by individual genetic make-ups and environmental influences in different phenotypic expressions of the same genetic abnormality 38. Finally, alterations in cytokines such as TGFβ3 may also contribute to disease progression as TGFβ3 is known to induce fibrotic response by stimulating mesenchymal cells and has been shown to modulate expression of genes encoding desmosomal proteins in different cell types and regulate stability of cell-cell junctional complexes 20. A link between the mutations in the genes involved in ARVD and induction of apoptotic death of myocardiocytes is believed to exist but remains speculative and is yet to be proven 39.

There have been a number of studies addressing the issue of prevalence of PKP2 mutations in different groups of ARVD patients 625262840. Both Dalal et al in the United States 25 and van Tintelen et al in the Netherlands 26 working independently found the prevalence of mutations in PKP2 among unrelated ARVD probands to be 43%. Tintelen et al additionally discovered that the rate of PKP2 mutations was as high as 70% in cases with familial ARVD and no PKP2 mutations were identifiable in cases with non-familial sporadic phenotype. The high prevalence of PKP2 mutations among ARVD patients points towards a potential role of mutation analysis in PKP2 mutation positive ARVD cases and their family members 25. Cases with PKP2 mutations did not differ significantly from those without PKP2 mutations with regard to clinical characteristics and outcomes, except for earlier age of first clinical presentation and first arrhythmia and more frequent negative T waves in V₂ and V₃ 29. Additionally, patients with a PKP2 mutation undergo ICD interventions irrespective of the classic risk factors determining ICD intervention in ARVD patients 25. Although molecular genetic analysis is unlikely to contribute significantly to prediction of clinical phenotype and risk stratification, it has potential for family screening and preclinical diagnosis, particularly for patients with early/incomplete disease expression 2529.

In keeping with variable penetrance and multi-factorial causation of ARVD, the mere inheritance of a mutation previously identified in a proband does not imply a diagnosis of ARVD in a family member, in the absence of clinical diagnostic criteria. Conversely, a fully negative family history could suggest a case of novel mutation or alternatively, a different disease mimicking ARVD such as sarcoidosis, chagas disease, enterovirus infection, chronic right-side myocarditis or sub-acute bartonella infection 39. The prognostic value of early identification of asymptomatic gene carriers may be limited because a reliable algorithm for primary prevention of sudden death in familial ARVD is lacking and risk stratification in asymptomatic relatives has not been systematically addressed 2941.

In order to account for the broader spectrum of disease in familial ARVD, it has been suggested that the Task Force criteria be modified such that the presence of any minor criterion in first degree relative of a proven case of ARVD be regarded as clinical disease expression 4243. Syrris et al 28 found that the application of modified criteria does expand the diagnostic yield but also may lead to false positives because several family members assumed to be clinically affected were subsequently found to be gene negative. Thus to date the interpretation of minor repolarisation abnormalities in first-degree relatives remains inconclusive.

In summary, a positive genetic result can only be a part of a more comprehensive and multi-disciplinary diagnostic protocol and work-up involving ECG, arrhythmic, morpho-functional, histo-pathological, and clinical/molecular genetic findings 29. More complete understanding of the underlying genetic basis and the pathogenesis of ARVD will facilitate diagnosis, prognostication and therapeutic decision-making in patients which would be of great significance since adverse clinical outcomes such as SCD occur without premonitory symptoms in many cases.

Sensitivity and specificity of CMR detection of RV intra-myocardial fat in diagnosis of ARVD is variable, ranging from 22% to 100% 4448525462. Identifying fat can be challenging because the RV is a thin structure and areas of affected myocardium can be quite small. It may be particularly difficult to distinguish pathologic adipose infiltration in areas where adjacent epicardial fat is normally present such as in AV groove and antero-apical portion of RV.

Castillo et al 61 demonstrated that even state-of-the-art clinical CMR protocols using double inversion-recovery fast SE sequences may be limited in both spatial and contrast resolution for detecting fine intra-myocardial structural detail in ARVD. The presence of signal intensity increase seen in T1 weighted images is not specific for fat. Proximity to the surface coil, truncation band artifacts, and various motion-related artifacts may cause projection of high signal intensities onto the myocardium and be mistaken for fat. Similarly, artifacts related to arrhythmia, breathing motion, and blood flow can also produce high signal intensity in the free wall of the RV 63.

The varying sensitivities that have been reported for CMR to detect intra-myocardial fat can be ascribed to patient selection, which includes not only those with overt forms of the disease with extensive morpho-functional alterations and inducible arrhythmias, but also those with concealed forms without global dysfunction and without inducible arrhythmias 62. Isolated areas of fat replacement have also been reported in elderly patients, long-term steroid use, other cardiomyopathies and in idiopathic RVOT tachycardia, which forms an important differential diagnosis of ARVD 52. Obesity may be associated with fatty replacement of the RV wall 64. Macedo et al 65 reported that lipomatous infiltration of RV on cine CMR may occur without global or regional functional abnormalities suggestive of ARVD. Thus RV fat replacement may be a distinct disorder in itself and does not always imply ARVD.

The reliability of detecting increased intra-myocardial T1 signal is probably low. Bluemke et al 55 reported that identification of high T1 fat signal is less reliable than RV measurements or other morphological changes in ARVD.

Abbara at al 63 reported that applying fat-suppressed sequences to conventional fast spin-echo (FSE) imaging leads to an increase in the inter-reader reproducibility and confidence for intra-myocardial fat infiltration, also allowing identification of even subtle to mild degrees of intra-myocardial fatty infiltration (although the significance of such minor fatty changes with respect to their association with ARVD is not clear). Tandri at al 54 similarly found good inter-observer agreement for fat and excellent reliability for other qualitative and quantitative variables among experienced CMR readers by the use of a uniform imaging protocol and application of fat-suppression on breath-hold double-inversion recovery blood suppression pulses. However both Bluemke et al 55 and Tandri et al 54 concluded that the identification of fat within the myocardium is the least specific and least reproducible of any of the other parameters in the CMR evaluation of ARVD. Furthermore, it should be noted that fat signal by imaging tests is not a Task Force criterion for the diagnosis of ARVD.

By the application of CMR, Harper et al 66 confirmed that RV free wall thinning is independently predictive of the presence of ARVD. They quantified RV free wall thinning in ARVD and reported that RV free wall thickness (RVFWT) of less than 5 mm is suspicious for ARVD, between 5 and 8 mm is indeterminate and greater than 8 mm is highly improbable to represent ARVD. Thinning of the RV free wall has been noted as a significant abnormality in ARVD. However, like fat, qualitative evaluation of CMR RVFWT in ARVD may be compromised by thinness of the normal RV free wall, limited in-plane resolution, adjacent high epicardial fat signal and motion artifacts. Like fat, free-wall thinning is also not a part of the Task Force criteria.

Late-enhancement (LE) CMR has the potential to identify myocardial fibro-fatty changes in both RV and LV in ARVD 5167. CMR findings suggesting fibrous tissue additionally show correlation with histo-pathological findings and also show association with RV dysfunction and inducibility of VT on electrophysiological (EP) testing 51. Fibro-fatty variety of ARVD is more common than fatty ARVD. Documentation of fibrous tissue has been suggested to be diagnostically more important than finding fat alone. Fat replacement of RV without fibrosis probably represents a distinct clinico-pathological entity that may not be considered synonymous with ARVD. Fibrous tissue replacement of the myocardium is also believed to be more arrhythmogenic than fat alone 68.

A potential role of CMR lies in its ability to detect early and subtle cases of ARVD which are otherwise insensitive to diagnosis by Task Force criteria, posing a real diagnostic dilemma. In patients with structurally severe disease, all imaging modalities are likely to provide abnormal findings and the ability of CMR to do so therefore does not stand out 69. CMR can uniquely detect regional and diastolic (right and left) ventricular dysfunction 4870 which may represent earliest manifestations of ARVD, a continually evolving and progressive disease.

Regional functional abnormalities of RV in ARVD include focal hypokinesis (systolic wall thickening of <40%), akinesis (systolic wall thickening of <10%), dyskinesis (movement of myocardial segment outward in systole), and aneurysms (segments with persistent bulging in diastole and dyskinetic in systole) 52. Myocardial tissue tagging may potentially improve the sensitivity of CMR to detect regional dysfunction in ARVD because visual analysis is often inadequate due to the complex contraction pattern of the RV 71. However, tissue tagging is not easily applied due to the very thin RV free wall and poor signal to noise ratio 52.

Regional functional abnormalities have been shown to correlate with areas of signal abnormality. The presence of regional abnormal wall motion associated with signal abnormality is more suggestive of ARVD than either of them alone 52. Abnormal diastolic dysfunction has been shown to correlate with morphologic abnormalities, indicating that global diastolic dysfunction accompanies more extensive morphologic alterations in patients with ARVD 4872. In our experience, regional delayed and dyssynchronous contraction of the RV (mostly at the base) is present in about one-third of phenotypically confirmed cases of ARVD with no other CMR findings of ARVD. We have also found that dyssynchronous contraction of the RV basal free wall may represent an early phenotypic expression of desmosomal mutations (especially PKP2) in the RV in asymptomatic first-degree relatives of ARVD pro-bands. This finding showed good correlation with electrocardiographic minor criteria and inducibility for ventricular arrhythmias (Figure 6).

MR imaging is a useful tool in the evaluation of patients with suspected ARVD. However, it is important to recognize that CMR is not a "gold standard" for the diagnosis of ARVD. But rather, CMR is one of a number of tests which should be employed in the evaluation of patients with suspected ARVD. Other tests which should be performed include a 12 lead electrocardiogram (ECG), signal-averaged ECG (SAECG), Holter monitor, stress test, and echocardiogram. If these tests suggest ARVD, we at our institution, typically recommend biopsy to confirm the diagnosis and exclude other potential etiologies (such as sarcoidosis) mainly in sporadic cases as well as EP study as a part of risk stratification 73. Great caution must be employed when the only abnormalities which are seen are observed with CMR. In our experience, CMR imaging can result in an over-diagnosis of ARVD. As noted above, many abnormalities which can be seen with CMR, particularly wall thinning and the presence of fat are nonspecific findings. This is important as misdiagnosis of ARVD has great implications not only for the patient but for the family as well. It is extremely uncommon to have an entirely normal 12 lead ECG and have ARVD. Therefore, CMR abnormalities observed in a patient with a normal 12 lead ECG should be interpreted with great caution 74.

Currently, two large clinical trials are under-way that use CMR for ARVD diagnosis. The European ARVD registry 75 attempts prospective validation of criteria for clinical diagnosis of ARVD, evaluation of accuracy of clinical diagnosis and assessment of the natural course of disease 52. The Multidisciplinary Study of Right Ventricular Dysplasia (U.S. ARVD study) 76 aims to prospectively enroll 100 ARVD patients with 500 first-degree relatives with the following scientific goals: 1) to study a large group of newly diagnosed patients with this uncommon disorder, 2) to elucidate the genetic etiology of this disease, 3) to evaluate treatment of ARVD in these patients. These studies may help better define the role of CMR in ARVD in cases and family members.

LV involvement in ARVD has been increasingly described, with prevalence reported as 16% 42 to 76% 77. Hebert et al 78 reported a subnormal LV ejection fraction (EF) by angiography in 20% of a group of ARVD patients. Hunold et al 67 detected LE in the LV in 2 of 6 ARVD patients. Peters and Reil 79 reported that despite a normal LVEF, 40% of their 60 ARVD patients had LV wall motion abnormalities. Bauce et al 31 found LV abnormalities in half of the patients in their study population with echocardiographic evidence of ARVD. Sen-Chowdhry et al 36 demonstrated LV abnormalities with preserved RV function in 40% of their ARVD cohort.

Lindstrom et al 80 demonstrated LV abnormalities in 93% of their ARVD patients by myocardial perfusion scintigraphy and echocardiography. Manyari et al 81 discovered latent LV dysfunction to be present in all cases of ARVD by uncovering it during exercise testing. Evolution of LV involvement in ARVD has been demonstrated in terms of appearance of new abnormalities or worsening of existing abnormalities or both, with progression of RV disease on long-term follow-up.

These cases of early and/or predominant LV involvement in ARVD have brought into question the traditional consideration of LV involvement as only an end-stage complication of progressive RV dilatation and dysfunction in ARVD. Although there is not a complete agreement in literature, LV involvement in ARVD may not stand for a separate disorder, but may be part of a larger disease process extending across the entire heart supporting the adoption of a broader term "arrhythmogenic cardiomyopathy" with appropriate sub-classifications 36.

The most common variant of ARVD has been postulated to be biventricular with LV involvement from its early stages. The rationale for this is the classification of ARVD as a desmosomal disease and desmosomal genes are expressed in both ventricles. Desmosomal mutations that primarily affect cell adhesion may affect the distensible, thin-walled RV earlier in the course of the disease. LV dysfunction in ARVD may also be secondary to the enlarged and poorly functioning RV because physiologically heart cannot be compartmentalized into functionally independent right and left sides.

LV involvement in classic ARVD is characterized by extension of T-wave inversion in the ECG lateral leads (V5, V6, L1, aVL), arrhythmia of LV or biventricular origin, and/or isolated LV dilatation and impairment 36. ARVD fatty/fibro-fatty changes of LV myocardium can extend across varying thickness of myocardial circumference, with a predilection for the sub-epicardial and mid-ventricular wall (Figure 7). Fibro-fatty changes can affect both the septum and more often, the LV free wall, either diffusely or regionally with a predilection for postero-septal and postero-lateral areas 77. Myocardial replacement of LV by adipose or fibro-adipose tissue, like the RV, is believed to progress inwards from the subepicardium to the trabecular myocardium 82.

Fibro-fatty replacement of the LV shows correlation with myocardial perfusion defects predominantly located in antero-septal and basal posterior segments of the LV, adjacent to the parts of the RV most frequently affected by the disease 80. Other quantitative and qualitative LV abnormalities that have reportedly been noted in ARVD, like RV, include increased LV end-diastolic pressure, decreased EF, AV block, regional wall motion abnormalities, aneurysms (most frequently at the apex), presence of asynergic areas (most frequently located at the apex or the infero-posterior wall) and dilatation 83.

LV involvement in ARVD appears histologically similar to the RV findings or is only fibrotic 83. Pinamonti et al found discrete small diverticula of the LV wall on angiography which were thought to represent some localized muscle defect as a component of the underlying ongoing dysplastic process 83. LV involvement has been detected even in asymptomatic patients and in those with a short duration of symptoms 808485. LV involvement is more common and more severe in families with several members affected with ARVD 8386. LV involvement is associated with increased myocardial mass, inflammatory infiltrates, clinical arrhythmic events and more severe RV wall thinning and heart failure. Frank left-sided heart failure is unusual.

When Task Force criteria were first proposed, minimal LV disease was a stipulation, intended to improve specificity of a then little known disease and ensure distinction from related disorders such as dilated cardiomyopathy. It appears likely that descriptions of LV involvement may be included in subsequent revisions of the diagnostic criteria for ARVD 41.