Arrhythmogenic right ventricular dysplasia/cardiomyopathy (ARVD/C) is characterized by structural and functional abnormalities of the right ventricular (RV) leading to ventricular arrhythmias and progressive RV failure. The most striking morphological feature of ARVD/C is diffuse or segmental replacement of myocardium in the RV free wall by fibro-fatty tissue (1- 3). Diagnosis of ARVD/C at its early stages remains a clinical challenge (4- 5), more so in patients with minimal RV abnormalities at echocardiographic or angiographic examination. Endomyocardial biopsy has the potential for in vivo demonstration of typical fibro-fatty replacement of the RV myocardium. However, sensitivity of this test is low because, for reasons of safety, biopsy samples are usually taken from the septum, a region uncommonly involved by the disease (6).

Magnetic resonance imaging (MRI) is a promising technique for delineation of RV anatomy and function as well as for characterizing the composition of the RV wall, especially with regard to the presence of fatty tissue (7- 8). However, fat visualization on MRI has not been found to be specific for ARVD (9- 10), and there is poor inter-reader agreement in reporting of fat (11). Recently, myocardial delayed enhancement (MDE) after intravenous administration of a gadolinium-based contrast agent has been shown in dysfunctional areas of the left ventricle in patients with prior myocardial infarction and fibrous scar (12- 13). We hypothesized that the RV in ARVD/C patients will demonstrate increased signal on MDE-MRI because of the fibrotic nature of the disease process.

The study population included 30 patients who were prospectively evaluated for possible ARVD/C because of either a family history or left bundle branch block (LBBB) morphology ventricular arrhythmia. Each of these patients was evaluated with a detailed clinical history, physical examination, electrocardiogram (ECG), signal-averaged ECG, two-dimensional echocardiography, and contrast-enhanced cardiac MRI. Invasive testing, including electrophysiologic (EP) testing, right ventriculography, and endomyocardial biopsy, was performed guided by the results of noninvasive testing. Final diagnosis of ARVD/C was made independently of the MRI findings. The results of delayed-enhancement MRI were then correlated with the final diagnoses and the results of endomyocardial biopsy. The study was approved by our institutional review board, and informed consent was obtained from the study subjects.

The MRI examination was performed with a 1.5-T MR imager (CV/I, General Electric Medical Systems, Waukesha, Wisconsin) using a dedicated phased-array cardiac coil. Transaxial black blood images were acquired using a double-inversion recovery (blood suppression) fast-spin echo sequence (time to repetition [TR] = 1 R-R interval, time to echo [TE] = 5 ms, slice thickness = 5 mm, interslice gap = 5 mm, and field of view [FOV] = 24 to 28 cm). Following this, the same sequence was repeated with the chemical shift fat suppression manually tuned to the fat peak to generate fat-suppressed axial black blood images. Bright blood cine imaging in the axial and short-axis planes was acquired by a steady-state, free-precession pulse sequence (TR = 3.5, TE = 1.2 ms, flip angle = 45°, slice thickness = 8 mm, interslice gap = 4 mm, and FOV = 36 to 40 cm, 10 views per segment, 35 ms temporal resolution per cine phase). After intravenous administration of an MRI contrast agent (0.2 mmol/kg of gadodiamide [Omniscan, Amersham Health, Princeton, New Jersey]), inversion recovery prepared breath-hold cine gradient-echo images were obtained 20 min after contrast agent injection. Initially, three-dimensional inversion recovery prepared breath-hold, fat-suppressed, gradient-echo imaging was performed in the short axis and axial planes. The use of a variable sampling in time segmentation (14) scheme enabled the acquisition of three-dimensional volumes in a single 24-heartbeat breath-hold. For three-dimensional imaging, the parameters were as follows: TR/TE = 4.7 ms/1.7 ms, flip angle = 20°, slice thickness = 5 to 7 mm, matrix = 256 × 160, half Fourier acquisition, 12 slices interpolated to 24, FOV = 360 × 270 mm. A single three-dimensional slab encompassed both ventricles. This was followed by breath-hold two-dimensional imaging (7.2/3.2; inversion time optimized 150 to 200 ms; flip angle = 25°; slice thickness = 8 mm; slice gap = 2 mm; number of excitations = 2; matrix = 256 × 192; and FOV = 360 × 270 mm). The two-dimensional MDE MRI scans were obtained in the short axis and axial planes at 10-mm intervals covering the entire right and left ventricles.

The RV was divided into three levels in the short-axis plane: basal, mid, and apical. Each of these three levels was further divided into three segments (superior, midwall, inferior), resulting in a nine-segment model for the RV. Gadolinium enhancement was assessed in each of these segments by observers who had no knowledge of clinical information, using the following scale: 0 = none and 1 = presence of delayed enhancement. The segment scores were summed, yielding a range per patient of 0 (no enhancement in any slice) to 9.

The contrast-to-noise ratio (CNR) of the RV free wall was evaluated as follows: a region of interest (ROI) with a size of 0.1 to 0.2 cm² was placed within the RV wall in the zone of increased signal intensity. A second ROI of identical size was placed within the nearest zone of the RV free wall showing normal-appearing myocardium. The difference between the mean signal intensities of both ROIs was then divided by the standard deviation of the background noise signal (ROI = 2 cm²) measured anteriorly to the RV.

Global ventricular volumes were calculated from the short-axis cine images using a summation of disks method (“Simpson's Rule”), with integration over the image slices using the software program MASS (Medis, Leiden, The Netherlands).

Biopsy specimens were taken from the myocardial border of the interventricular septum, with the use of modified Stanford-Caves bioptome (Scholten Surgical, Redwood City, California), by way of the right internal jugular vein. At least five specimens were obtained from each patient, immediately fixed in 10% formalin, embedded in paraffin, sectioned, stained with hematoxylin and eosin, and reviewed at a minimum of four section levels by one cardiac pathologist. The reviewing pathologist was not aware of the clinical characteristics of the patients. A percentage of fat >3% and of fibrous tissue >40% with <45% myocytes was considered diagnostic of ARVD/C (6).

A standardized EP study was performed in all patients. A conventional stimulation protocol was used with one, two, and three extrastimuli delivered at three drive train cycle lengths at the RV apex and RV outflow tract. Burst pacing at cycle lengths from 600 to 300 ms was also used. Testing was performed in the baseline state and during a 2 μ/min infusion of isoproterenol. A study was considered positive if sustained monomorphic ventricular tachycardia (VT) was induced (VT with uniform QRS configuration and cycle length >200 ms lasting 30 s or requiring termination because of hemodynamic compromise). The EP study was negative if the stimulation protocol was completed without induction of sustained VT. Induction of non-sustained ventricular tachycardia (NSVT) and ventricular fibrillation were considered nonspecific findings.

The data are presented as mean ± standard deviation. The Mann-Whitney U test and Fisher exact test was performed where appropriate. Spearman correlation coefficient was used for correlation analyses. A p value of <0.05 was considered statistically significant.

The mean age was 35 ± 12 years, and 60% (18) of the patients were female. Nine patients (30%) were evaluated because of a family history of ARVD/C, and the others presented with LBBB morphology VT (sustained [n = 6] and nonsustained [n = 15]). None of the patients had a prior history of coronary artery disease, diabetes, or hypertension. Coronary angiography performed in 14 (47%) of the 30 patients upon discretion of their attending cardiologist revealed no significant coronary artery disease. Left ventricular function assessed on echocardiography was normal in all patients. None of the patients had evidence of pulmonary hypertension on Doppler echocardiography. Twelve (40%) of the 30 patients met the Task Force criteria and were eventually diagnosed with ARVD/C. The clinical characteristics of the 12 patients with ARVD/C are shown in (Table 1). The remaining 18 patients had structurally normal hearts, and 12 of these were diagnosed with idiopathic RV outflow tract tachycardia.

(Figure 1)A shows an example of delayed enhancement of the RV anterior wall on an axial MDE-MRI from an ARVD/C patient. Epicardial fat was hyperintense on a non-fat suppressed MDE-MRI, thus making it difficult to distinguish the enhanced RV myocardial signal from that of epicardial fat. Signal in the anterior chest wall close to the cardiac surface coil also contributed to high signal intensity in the anterior mediastinum. Using chemical shift fat-suppression (Figure 1B) and the body coil instead of the cardiac surface coil (Figure 1C), the increased signal due to the epicardial fat and cardiac coil were eliminated, and enhancement of the RV anterior wall was confirmed.

Eight of the 12 patients (67%) with a final diagnosis of ARVD/C demonstrated delayed enhancement with MDE-MRI compared with none (0%) of the 18 patients without the diagnosis of ARVD/C (p < 0.001). The CNR ratio was significantly higher in the eight patients who had ARVD/C with delayed enhancement compared with the patients who had ARVD/C without enhancement, and the non-ARVD/C patients (10.9 ± 4, −1.5 ± 3.6, and 0.8 ± 2.9, respectively; p < 0.01). The area of delayed enhancement also showed dyskinesis on cine imaging in six of the eight patients (75%).

In six of the eight ARVD/C patients with delayed enhancement, the enhancement was seen in the basal sub-tricuspid region, extending anteriorly into the RV outflow tract. One patient had enhancement of entire anterior wall and the apex; cine images in this patient demonstrated multiple aneurysmal outpouchings of the RV. The same patient also showed diffuse enhancement of the RV portion of the interventricular septum and of the lateral wall of the left ventricle (Figure 2). A localized area of delayed enhancement that correlated with a small bulging in the outflow tract was observed in one asymptomatic patient with a family history of ARVD/C.

Endomyocardial biopsy was performed in 9 of the 12 patients who had the final diagnosis of ARVD/C. The remaining three patients refused biopsy. Of these nine patients, four (44%) had fibro-fatty changes consistent with the diagnosis of ARVD/C, one (11%) had extensive myocarditis, and the remainder (44%) had normal biopsies. Each of the four patients with fibro-fatty changes on biopsy had delayed enhancement on MDE-MRI. Of the four ARVD/C patients with normal biopsies, only one had delayed enhancement limited to the antero-basal RV. Endomyocardial biopsy was also performed in 8 (44%) of the 18 patients who did not meet the Task Force criteria for ARVD/C. None of these patients had any abnormalities on histopathology. Also, none of these patients showed delayed enhancement on MDE-MRI. Presence of delayed enhancement in the RV showed an excellent correlation with fibrosis on biopsy (Figure 3).

Patients who had delayed enhancement on MRI and who had positive biopsy results (n = 4) had moderate-to-severe RV dysfunction (right ventricular ejection fraction [RVEF] <40%). Three patients with normal RVEF (>55%) who met ARVD/C Task Force criteria had negative biopsies, and one of them had delayed enhancement on MRI. This suggested the possibility of a limited or concealed form of ARVD/C in these patients.

The EP testing was performed in 10 of the 12 ARVD/C patients; 6 had inducible sustained monomorphic VT, and 4 were noninducible (Table 2). The six inducible patients showed delayed enhancement of the RV, compared to only one of the four ARVD/C patients who were noninducible (p = 0.01). Of the 18 non-ARVD/C patients, 12 had EP testing and none of them was inducible for sustained VT. Also, none of these patients had delayed enhancement on MDE-MRI.

The 30 patients were divided into three groups based on number of segments that showed delayed enhancement (<2 segments, 3 to 5 segments, and >5 segments), and the right ventricular end-diastolic volume (RVEDV) and the RVEF were compared between the groups. The mean RVEF decreased (R² = 0.82, p < 0.02), while the mean RVEDV showed a significant increase (R² = 0.61, p <0.02) with increasing extent of delayed enhancement (Figure 4A and B).

Myocardial delayed enhancement-magnetic resonance imaging is a new but well-validated technique for assessing fibrosis following myocardial infarction (12- 13). The identification of fibrosis using this method has not been previously described in ARVD/C. The presence of delayed enhancement showed a high degree of correlation with endomyocardial biopsy and predicted induction of VT during EP testing. Furthermore, there was a strong association between the extent of delayed enhancement and RV dysfunction. These data suggest that MDE-MRI may have an important role in the evaluation and diagnosis of ARVD/C.

Prior studies in ischemic and other nonischemic cardiomyopathies have concluded that the mechanism of delayed enhancement is nonspecific, but may be related to an increase in volume distribution of gadolinium secondary to interstitial space expansion, which occurs in fibrous scarring or inflammation. The histopathologic specimens obtained in our patients with delayed enhancement showed predominantly replacement fibrosis and fat infiltration in the absence of significant inflammation. Most frequently, delayed enhancement was observed in the antero-basal region and in the RV outflow tract, consistent with the previously described anatomic distribution of fibro-fatty infiltration in ARVD/C. In six of the eight ARVD/C patients, regional wall-motion abnormalities were observed in the same region, suggesting lack of functioning myocardial cells in the same area. These observations lead us to conclude that the main mechanism of delayed enhancement in ARVD/C may be localization of gadolinium to areas of fibrosis within the RV myocardium. Myocarditis may be yet another mechanism of delayed enhancement in ARVD/C, as in the case of one patient who showed extensive myocarditis with minimal interstitial fibrosis.

Most cases of sustained monomorphic VT are associated with a myocardial scar. Re-entry during EP testing occurs through surviving myocyte bundles in and around the fibrotic areas, which are visualized as contrast-enhanced areas on MDE-MRI. Turrini et al. (15) have previously reported an association between fibrosis on endomyocardial biopsy and the occurrence of sustained ventricular arrhythmias in ARVD/C. Abnormal signal-averaged ECG and reduced RVEF were surrogates for extent of fibrosis in ARVD/C. Consistent with their findings, each of the six ARVD/C patients who had inducible monomorphic VT on EP testing demonstrated myocardial delayed enhancement of the RV. Myocardial delayed enhancement correlated better than abnormal signal-averaged ECG (5 of 6) or RV dysfunction (4 of 6) in predicting VT on EP testing. None of the ARVD/C patients without delayed enhancement was inducible. The above finding highlights the potential of MDE-MRI not only to noninvasively visualize scar tissue, but also to predict the results of EP testing.

In our study, only two-thirds of the ARVD/C patients showed delayed enhancement. The reason for the lack of delayed enhancement in the remainder may be the presence of a pure fatty form of ARVD/C, or may reflect the insensitivity of current MRI techniques to detect a small amount of fibrosis in early disease.

For reasons of safety, biopsy specimens were obtained from an endovascular approach. As such, the biopsy specimens are representative of subendocardial histology only. The location of the biopsies could not be systematically correlated with the MRI findings on a one-to-one basis. The low yield of biopsy (44%) in our study may in fact reflect sampling error or may also be due to the focal nature of the disease process. Another important limitation of our study is the small sample size. Further studies with a large sample size are needed to confirm our results.

This study suggests for the first time that fibrosis of the RV in ARVD/C can be noninvasively visualized using MRI. The presence of delayed enhancement in ARVD/C was associated with inducibility during EP testing and may also be useful in risk stratification. Abnormal RV enhancement on MRI may help improve the specificity of MRI for ARVD/C diagnosis. The absence of delayed enhancement in each of the patients with idiopathic VT is reassuring, and consistent with the pathophysiology of ventricular arrhythmias in this disease.