Background
The overt stage of arrhythmogenic right ventricular (RV) dysplasia/cardiomyopathy (ARVD/C) is preceded by a concealed stage with minor or no signs of disease. However, sudden death may occur in this early phase. Deformation imaging may contribute to early diagnosis. The aims of this study were to compare the diagnostic accuracy of the conventional (1994) versus the recently published (2010) new echocardiographic criteria for ARVD/C and to evaluate the additional value of echocardiographic tissue deformation imaging to detect subclinical RV functional abnormalities in asymptomatic carriers of pathogenic ARVD/C mutations.
Methods
Fourteen asymptomatic first-degree relatives of ARVD/C probands (the ARVD/C-r group; mean age, 38.0 ± 13.2 years) with a pathogenic plakophilin-2 mutation and a group of age-matched controls ( n = 56; mean age, 38.2 ± 12.7 years) were included at a 1:4 ratio. A complete echocardiographic evaluation (dimensions, global systolic parameters, and visual assessment and deformation imaging of the RV free wall including Doppler tissue imaging and two-dimensional strain echocardiography) was obtained. Peak systolic strain less negative than −18% and/or postsystolic shortening (postsystolic index > 15%) in any RV segment was considered abnormal.
Results
RV dimensions in the ARVD/C-r group were similar to those in controls (RV outflow tract, 15.4 ± 2.9 vs 14.4 ± 1.9 mm/m 2 , P = NS; RV inflow tract, 18.6 ± 2.6 vs 19.1 ± 2.6 mm/m 2 , P = NS), and global systolic parameters were moderately reduced (tricuspid annular plane systolic excursion, 20.0 ± 3.2 vs 23.9 ± 2.8 mm, P = .001; RV fractional area change, 40.3 ± 8.4 vs 40.6 ± 7.1, P = NS). According to task force criteria, 57% of the ARVD/C-r group and 29% of controls were classified as abnormal when applying the 1994 criteria and 29% and 4% when applying the 2010 criteria, respectively. Doppler tissue imaging and two-dimensional strain deformation (and strain rate) values were reduced in the ARVD/C-r group in the basal and mid RV segments compared with controls ( P < .001). In the ARVD/C-r group, peak systolic strain less negative than −18% was seen in six patients (43%), postsystolic strain in nine (64%), and either abnormality in 10 (71%), almost exclusively in the basal segment; these findings were observed in none of the controls.
Conclusions
The 2010 criteria for ARVD/C improve specificity, whereas sensitivity is significantly reduced in this asymptomatic population. In contrast, echocardiographic deformation imaging detects functional abnormalities in the subtricuspid region in 71% of asymptomatic carriers of a pathogenic plakophilin-2 mutation, while regional deformation was normal in all control subjects, indicating superiority of both sensitivity and specificity with these new modalities.
Arrhythmogenic right ventricular (RV) dysplasia/cardiomyopathy (ARVD/C) is a disease histopathologically characterized by fibrofatty alteration predominantly in the RV myocardium. A molecular-genetic origin of this disease can be identified in up to 70% of patients with familial ARVD/C, with an autosomal dominant mode of inheritance with incomplete penetrance. Potentially lethal ventricular arrhythmias may be the first manifestation of this disease in previously healthy young individuals. This emphasizes the importance of early recognition of this disease, for instance, in the family members of ARVD/C probands. Currently, the diagnosis is established using a composite of criteria proposed by a task force including electrocardiographic (ECG) disturbances, functional and morphologic abnormalities, and family history. These criteria have recently been modified to increase the sensitivity of ARVD/C diagnosis. Unfortunately, because the pathognomonic characteristics on noninvasive imaging are often not present in the early stages of the disease, early detection remains cumbersome. This phase is often referred to as the concealed phase of the disease, but it nevertheless carries the risk for sudden cardiac death.
We hypothesized that an objective and regional approach to evaluate changes in RV function would improve the diagnostic accuracy of echocardiography and thus identify the phenotypic expression of the disease earlier in its course. We have previously shown that regional assessment using echocardiographic deformation imaging is able to accurately identify patients in whom ARVD/C has been diagnosed previously. Using a cutoff value of −18% for peak systolic strain (ϵ) in any RV segment showed sensitivity of 97% and specificity of 91.2% in differentiating between patients with ARVD/C and healthy controls. The role of this technique in detecting functional abnormalities as an early manifestation of ARVD/C, however, remains unknown. The aim of this study was to evaluate the additional value of tissue deformation imaging included in the echocardiographic examination to detect subclinical RV functional abnormalities in asymptomatic carriers of a pathogenic ARVD/C mutation. A secondary goal of this study was to compare the recently published updated guidelines with the conventional 1994 task force criteria (TFC) for ARVD/C in this specific population.
Methods
Study Population
A total of 38 consecutive relatives, not fulfilling criteria for ARVD/C at the time of analysis, of ARVD/C probands referred for echocardiographic evaluation at our tertiary center were prospectively enrolled. All included individuals were first-degree relatives of 21 index patients in whom the diagnosis of ARVD/C was established as previously described. All relatives underwent full echocardiographic examinations, as well as electrocardiography at rest, typically performed within 2 months of echocardiography. All included individuals were offered deoxyribonucleic acid analysis as part of our clinical workup if a disease-causing mutation had been found in the proband. All probands were tested for pathogenic desmosomal mutations in plakophilin-2 ( PKP2 ), desmoglein-2, desmocollin-2, desmoplakin, and junction plakoglobin, whereas relatives were tested only for the mutation that was found in the proband. Of the initial cohort, 24 individuals were excluded because (1) no pathogenic mutation was found in the proband ( n = 11) and thus no genetic testing was performed in the relative, (2) no pathogenic mutation was detected in the relative ( n = 6), or (3) the subject refused to undergo genetic testing ( n = 7). Thus, the study population consisted of 14 ARVD/C first-degree relatives of nine index-patients with disease-causing mutations (asymptomatic mutation carriers; the ARVD/C-r group). At the time of echocardiographic examination and data analysis, no member of the ARVD/C-r group fulfilled diagnostic criteria for ARVD/C, as established by the TFC valid at that time. According to the modified TFC proposed in 2010, however, six of 14 did fulfill the diagnostic criteria for ARVD/C. This higher yield was due not to phenotypic criteria but exclusively to the presence of a pathogenic mutation as a major criterion. Nevertheless, all 14 members of the ARVD/C-r group were included for final analysis in this study.
As reference group, a total of 56 age-matched healthy controls free of cardiovascular disease were included (ratio of first-degree relatives to controls, 1:4 ) and subjected to the same echocardiographic protocol. No genetic analysis was performed in any of the control subjects. All included individuals were aged ≥ 18 years and in stable sinus rhythm. None of the controls were athletes, because of the impact of regular physical activity on RV geometry and function. The local ethics committee approved study protocol, and consent was obtained before echocardiographic examinations.
Standard Echocardiographic Examination
The echocardiographic examination was performed with the subject at rest, lying in the left lateral decubitus position. Ultrasound data were acquired using a Vivid 7 scanner (GE Vingmed Ultrasound AS, Horten, Norway) equipped with an M3S broadband transducer. A complete echocardiographic study was performed in two-dimensional (B-mode) and Doppler tissue imaging (DTI) mode. Both standard parasternal and apical views were obtained, as well as additional views as proposed by Foale et al. of both the left and right ventricles.
Conventional measurements included RV outflow tract (RVOT) end-diastolic diameter in the parasternal long-axis and short-axis views. In addition, parasternal long-axis RVOT end-systolic diameter was measured, with which the fractional change was calculated as the percentage change. Left ventricular internal diameter at end-diastole was measured using M-mode echocardiography. In the apical four-chamber view, left ventricular and RV short-axis inflow tract end-diastolic diameters were measured at the level of the valve leaflet tips, while by measuring RV end-diastolic and end-systolic areas, RV fractional area change was calculated. Right atrial and left atrial single-plane areas were measured at end-systole. Additionally, all dimensions were corrected for body surface area. In the four-chamber view, tricuspid annular plane systolic excursion (TAPSE) was measured using M-mode imaging. Pulsed Doppler imaging was used to interrogate transtricuspid and RVOT flow at end-expiration during breath hold for timing of cardiac events. By DTI, the systolic (s′) and diastolic, both early (e′) and late (a′), velocities were calculated in the basal segment of the RV free wall in addition to isovolumic acceleration. Color DTI was used to extract these velocity data.
Wall motion in the RV free wall was evaluated in the apical four-chamber view in the basal, mid, and apical segments. Wall motion was classified as normokinetic, hypokinetic, akinetic, dyskinetic, or uninterpretable. Finally, a major or a minor criterion was ascribed to the echocardiographic examination according to both the 1994 and 2010 modified TFC, using all available echocardiographic data, with the exception of the results of the deformation analysis. Visual assessment was performed by two experienced observers, blinded to group, who had to reach consensus for both regional wall motion abnormalities and the assignment of TFC point(s).
Tissue Deformation Imaging
Our protocol for image acquisition and postprocessing has previously been described in detail for both DTI and two-dimensional ϵ echocardiographic (2DSE) deformation imaging. For the DTI measurements, tissue Doppler myocardial imaging data from the RV free wall were recorded in the standard apical four-chamber view at a frame rate > 180 frames/sec. The Nyquist limit was adjusted to the lowest level avoiding aliasing. The image sector angle was set parallel to the investigated wall (ensuring optimal alignment of the Doppler beam to the myocardial wall) and set as narrow as possible to achieve the maximal tissue Doppler frame rate. Three consecutive cycles were recorded at end-expiration for offline analysis.
For the 2DSE measurements, real-time two-dimensional (B-mode), small-angle ultrasound data from the RV free wall were recorded for offline analysis. A frame rate of 78 to 110 frames/sec was implemented to ensure optimal speckle tracking.
Offline Analysis
Data were stored in a digital format and transferred to a computer workstation for offline analysis. Doppler flow curves of the cardiac valves were used for timing cardiac events, and all timing information was aligned through ECG traces. For the DTI-derived deformation parameters (EchoPAC PC version 6.0.1; GE Vingmed Ultrasound AS), all data were averaged over three consecutive cardiac cycles. For quantitative analysis, the RV free wall was divided into basal, mid, and apical segments. A sample volume (8 × 4 mm; offset length, 12 mm; temporal and spatial smoothing set to default values) was placed within each of these segments and manually adjusted throughout the cardiac cycle. Linear drift compensation was applied. Reverberations, dropout, and a deviation of the ultrasound insonation angle of >20° resulted in exclusion of the myocardial segment from analysis.
For 2DSE analysis, B-mode images (non-Doppler) of one cardiac cycle of the RV free wall were used to extract two-dimensional ϵ and ϵ rate curves. Dedicated software (EchoPAC PC 2D-strain version 59; GE Vingmed Ultrasound AS) using a two step-tracking algorithm was used (visually checked at 50% speed). A region of interest was manually traced along the endocardial border from base to apex at the end of systole, with the width set to match the wall thickness. The tracked region of interest was visually checked at 50% speed and adjusted if necessary. The RV free wall was then divided into three segments (basal, mid, and apical), matching the segments in which the samples of the DTI analysis were placed. The calculated values for ϵ and ϵ rate were averages over entire myocardial segments. Default configurations on the software package were used. Only inappropriate tracking and drop out from the image plane resulted in exclusion of the myocardial segment from analysis.
For both DTI and 2DSE deformation, the following parameters were measured in all three segments of the RV free wall: peak systolic ϵ and ϵ rate, defined as the maximum negative value between pulmonic valve opening and closure. In case values were positive during systole, the end-systolic value was measured. The postsystolic ϵ index (PSI) was calculated as the amount of shortening after pulmonic valve closure (PSI = peak ϵ value − ϵ/peak ϵ value × 100%). RV deformation was classified abnormal if ϵ less negative than −18% was measured and/or if a PSI of >15% was present in any of the analyzed segments and the most abnormal segment was identified. An illustrative example of these measurements for DTI and 2DSE deformation is shown in Figures 1 A and 1 B.
Blinding
Because unblinding occurred after all echocardiographic offline measurements, all 38 relatives (14 in the ARVD/C-r group and 24 excluded family members) and 56 controls were randomly analyzed. Analyses were performed blinded and at four different time points with ≤2 weeks between each analysis: (1) the standard echocardiographic measurements, (2) DTI-derived parameters (velocity and deformation), (3) 2DSE offline analysis, and (4) echocardiographic analysis according to the current guidelines. For the DTI and 2DSE analyses, only the small-angle recording and the RVOT Doppler flow (for timing) were available for the offline deformation analysis.
Statistical Analysis
All continuous data are presented as mean ± SD and categorical variables as numbers or percentages. Significant differences between the groups were calculated using independent Student’s t test for continuous data and χ 2 or Fisher’s exact tests for categorical data. We implemented the previously proposed cutoff values for conventional parameters of global RV function and size and deformation analysis to calculate the value of deformation compared with conventional echocardiography. The diagnostic accuracy of each separate technique was expressed as sensitivity and specificity. The presence of a pathogenic desmosomal mutation was used as a “gold standard” for this analysis. The ARVD/C-r group was reanalyzed for both the DTI and 2DSE deformation analysis. The same digitally stored images were used for the entire postprocessing procedure, described above. The lowest ϵ value and the presence of a PSI > 15% were identified. Bias and agreement between the two measurements were analyzed according to Bland and Altman, and the correlation of the peak values was assessed. In addition, each deformation analysis (both DTI and 2DSE analyses) in the ARVD/C-r group was classified as normal or abnormal in this second analysis (as described above). From this, the κ coefficient was calculated, indicating the strength of agreement. A P value < .05 was used to indicate significant differences. Statistical calculations were performed using SPSS version 16.0 for Windows (SPSS, Inc., Chicago, IL).
Results
Study Population
A total of 14 asymptomatic mutation carriers (the ARVD/C-r group) and 56 healthy control subjects were used in the final analysis. The baseline characteristics are summarized in Table 1 . Groups were not matched for gender, which resulted in more men in the control group. In all subjects in the ARVD/C-r group, pathogenic mutations in the PKP2 gene were identified in concordance with the genetic mutations in the index patients; in two relatives, an unclassified variant of uncertain pathogenecity was found in PKP2 in addition to the pathogenic PKP2 mutation. No other mutations were encountered in this study population. One subject in the ARVD/C-r group was on antiarrhythmic medication. None of the other included individuals were on antiarrhythmic drugs (or any other relevant medications) or had reported complaints of palpitations or syncope. None had an implantable cardioverter-defibrillator, and all were in New York Heart Association functional class I.
| Variable | ARVD/C-r group ( n = 14) | Controls ( n = 56) | P |
|---|---|---|---|
| Men | 14.3% | 48.2% | .021 |
| Age (y) | 38.0 ± 13.2 | 38.2 ± 12.7 | NS |
| Height (cm) | 175.4 ± 6.2 | 176.5 ± 9.6 | NS |
| Weight (kg) | 72.0 ± 10.9 | 71.8 ± 12.7 | NS |
| Body surface area (m 2 ) | 1.86 ± 0.16 | 1.88 ± 0.20 | NS |
| Heart rate (beats/min) | 62.6 ± 8.8 | 59.3 ± 11.2 | NS |
| Systolic blood pressure (mm Hg) | 130.0 ± 18.3 | 127.4 ± 15.8 | NS |
| Diastolic blood pressure (mm Hg) | 80.0 ± 4.1 | 75.9 ± 11.2 | NS |
ECG Findings
Depolarization abnormalities (prolonged terminal activation duration) were recorded in three members of the ARVD/C-r group. Repolarization abnormalities (negative T waves in the right precordial leads) were recorded in four (negative T waves in leads V 1 and V 2 in two and in leads V 1 and V 3 in two). In total, six subjects in the ARVD/C-r group had abnormal ECG findings with signs of ARVD/C. In the remaining eight members of the ARVD/C-r group and in all controls, diagnostic ECG criteria for ARVD/C were absent.
Standard Echocardiographic Findings
Results for conventional quantitative echocardiographic parameters are shown in Table 2 . Left ventricular and in particular RV dimensions showed no significant differences between the ARVD/C-r and control groups. In line with these findings, the ratio of left ventricular inflow tract to RV inflow tract, a marker of RV dilatation, was comparable in both groups. This ratio indicated RV dilation (<1.0) in 21% of the ARVD/C-r group and in 11% of controls. Also consistent with these findings, no significant differences were found in end-diastolic RVOT measurements, using our previously proposed cutoff value : parasternal long-axis RVOT was abnormal (≥16 mm/m 2 ) in 43% of the ARVD/C-r group and in 18% of controls ( P = NS).
| Parameter | ARVD/C-r group | Controls | P |
|---|---|---|---|
| Dimensions corrected for body surface area | |||
| Parasternal long-axis RVOT (mm/m 2 ) | 15.4 ± 2.9 | 14.4 ± 1.9 | NS |
| Parasternal short-axis RVOT (mm/m 2 ) | 16.7 ± 2.9 | 15.6 ± 1.9 | NS |
| Left ventricular internal diastolic diameter (mm/m 2 ) | 25.5 ± 2.8 | 26.1 ± 2.2 | NS |
| RVIT (mm/m 2 ) | 18.6 ± 2.6 | 19.1 ± 2.6 | NS |
| LVIT (mm/m 2 ) | 22.0 ± 2.0 | 22.7 ± 2.0 | NS |
| LVIT/RVIT ratio | 1.22 ± 0.26 | 1.21 ± 0.17 | NS |
| RV end-diastolic area (cm 2 /m 2 ) | 11.3 ± 1.9 | 11.2 ± 2.1 | NS |
| RV end-systolic area (cm 2 /m 2 ) | 7.1 ± 2.2 | 6.7 ± 1.7 | NS |
| Right atrium (cm 2 /m 2 ) | 8.1 ± 1.4 | 9.0 ± 1.7 | NS |
| Left atrium (cm 2 /m 2 ) | 8.5 ± 1.3 | 9.5 ± 1.9 | NS |
| Global RV systolic parameters | |||
| RV fractional area change (%) | 40.3 ± 8.4 | 40.6 ± 7.1 | NS |
| RV DTI s′ (cm/sec) | 9.1 ± 1.6 | 11.1 ± 1.7 | <.001 |
| Isovolumetric acceleration (cm/sec 2 ) | 1.89 ± 0.84 | 2.00 ± 0.52 | NS |
| TAPSE (mm) | 20.0 ± 3.2 | 23.9 ± 2.8 | <.001 |
| RVOT fractional change (%) | 25.9 ± 6.6 | 27.0 ± 8.1 | NS |
The conventional quantitative parameters for global RV function showed reductions of mean values for TAPSE and DTI s′, while isovolumic acceleration and RVOT fractional change were unchanged ( Table 2 ). Using our previously proposed cutoff values for these parameters in ARVD/C, abnormalities were found in the ARVD/C-r and control groups, respectively, in 64% and 14% for DTI s′ (<9.35 cm/sec; P < .001), in 29% and 9% for isovolumic acceleration (<1.25 cm/sec 2 ; P = NS), in 36% and 0% for TAPSE (<18.5 mm; P < .001), and in 14% and 14% for RVOT fractional change (<17.2%; P = NS). RV fractional area change was abnormal (<32% ) in 14% of the ARVD/C-r group and 13% of controls ( P = NS).
Regional wall motion by visual assessment was considered hypokinetic in the ARVD/C-r and control groups, respectively, in 29% and 11% for the basal RV free wall ( P = NS), in 7% and 12.5% for the mid RV free wall ( P = NS), and in 0% and 5% for the apical RV free wall ( P = NS). Akinesia or dyskinesia was seen in 21% and 2% for the basal RV free wall ( P = .023), in 21% and 0% for the mid RV free wall ( P = .007), and in 21% and 5% for the apical RV free wall ( P = NS) in the ARVD/C-r and control groups, respectively. Overall, hypokinesia was seen in 29% of the ARVD/C-r group and in 23% of controls ( P = NS), while akinesia or dyskinesia was seen in 29% in the ARVD/C-r group and 5% in the control group ( P = NS).
Deformation Imaging
Mean values for peak systolic deformation and deformation rate were significantly reduced in all RV segments using DTI, which was less pronounced using 2DSE imaging ( Table 3 ). In particular the basal RV free wall and the lowest ϵ and ϵ rate (i.e., the most abnormal RV segment) showed the most pronounced difference between the groups. In line with the reductions in TAPSE and DTI s′, global ϵ and ϵ rate using 2DSE imaging were reduced.
| Parameter | ARVD/C-r group | Controls | P |
|---|---|---|---|
| DTI-derived parameters | |||
| Peak systolic ϵ (%) | |||
| Basal | −16.5 ± 9.8 | −25.7 ± 4.4 | <.001 |
| Mid | −25.2 ± 5.4 | −32.8 ± 6.1 | <.001 |
| Apical | −27.6 ± 7.5 | −31.3 ± 5.2 | .046 |
| Lowest peak systolic ϵ (%) | −16.4 ± 9.8 | −25.2 ± 3.9 | <.001 |
| Peak systolic SR (sec −1 ) | |||
| Basal | −1.21 ± 0.42 | −1.63 ± 0.38 | <.001 |
| Mid | −1.33 ± 0.41 | −1.86 ± 0.52 | .001 |
| Apical | −1.68 ± 0.47 | −1.98 ± 0.53 | NS |
| Lowest peak systolic SR (sec −1 ) | −1.08 ± 0.43 | −1.52 ± 0.38 | <.001 |
| 2DSE parameters | |||
| Peak systolic ϵ (%) | |||
| Basal | −19.5 ± 8.0 | −25.2 ± 5.3 | .003 |
| Mid | −23.8 ± 5.8 | −28.2 ± 4.4 | .003 |
| Apical | −28.4 ± 7.2 | −31.2 ± 4.9 | NS |
| Lowest peak systolic ϵ (%) | −17.8 ± 6.7 | −24.6 ± 4.5 | <.001 |
| Global systolic ϵ (%) | −25.0 ± 5.3 | −29.0 ± 3.9 | .004 |
| Peak systolic SR (sec −1 ) | |||
| Basal | −1.39 ± 0.31 | −1.59 ± 0.47 | NS |
| Mid | −1.31 ± 0.30 | −1.55 ± 0.38 | .032 |
| Apical | −1.66 ± 0.49 | −1.74 ± 0.37 | NS |
| Lowest peak systolic SR (sec −1 ) | −1.20 ± 0.27 | −1.43 ± 0.31 | .017 |
| Global systolic SR (sec −1 ) | −1.31 ± 0.31 | −1.48 ± 0.29 | NS |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
