Study Design

Between 2006 and 2008, consecutive individuals aged >18 years were prospectively enrolled for echocardiographic examination: (1) those with either known or suspected ARVD/C, (2) family members of patients with ARVD/C, and (3) healthy control subjects.

Group classification was established according to major and minor criteria as defined by the current 2010 ARVD/C diagnostic task force and the results of deoxyribonucleic acid analysis of ARVD/C-associated genes, performed as described previously. Figure 1 demonstrates, in detail, the group classification on the basis of clinical assessment and genetic testing. Three groups were specified: (1) patients with ARVD/C ( n = 38), (2) relatives of patients with ARVD/C carrying pathogenic mutations ( n = 16), and (3) control subjects ( n = 55) free of any cardiovascular disease.

Flowchart of study group subdivisions. ARVD/C+ , presence of ARVD/C according to the revised 2010 task force criteria; ARVD/C- , absence of ARVD/C according to the revised 2010 task force criteria; ARVD/C-p , patients with ARVD/C; ARVD/C-r , relatives of patients with ARVD/C; Mutation+ , presence of mutation associated with ARVD/C (from the index patient); Mutation- , absence of mutation associated with ARVD/C (from the index patient); Mutation? , no genetic analysis was performed, resulting in exclusion from the study.

End point data were obtained directly from patients or relatives during periodic evaluations at the outpatient clinic or hospital admissions until September 2014. The outcome measure was a composite of end points: spontaneous sustained monomorphic ventricular tachycardia, sudden cardiac death, aborted sudden cardiac death, appropriate ICD intervention for a ventricular arrhythmia, and heart transplantation. Outcome definitions are available in Supplemental Table 1 . In case of reaching multiple clinical events during the follow-up period, the first event was considered the end point. Subsequent events were registered but were not included in further analysis.

The study protocol was carried out with the approval of the Ethics Committee of the University Medical Center Utrecht, and all patients gave informed consent.

Standard Echocardiographic Study

The echocardiographic examination was performed with the subject at rest, in the left lateral decubitus position, using a Vivid 7 scanner (GE Vingmed Ultrasound AS, Horten, Norway) equipped with an M3S broadband transducer. A complete echocardiographic study was performed, with two-dimensional (B-mode) and Doppler tissue imaging recorded in both parasternal and apical views. Additional recordings of the three conventional apical views were recorded with the implementation of dual focus to optimize wall motion assessment. Special care was taken to avoid recording the right ventricle in any of these views for optimal blinding during postprocessing. According to the 16-segment model of the American Society of Echocardiography, regional LV wall motion was designated as normokinetic, hypokinetic, akinetic, dyskinetic, or not interpretable, after consensus was reached by two blinded experienced observers. From this the wall motion score index was calculated as the sum of scores divided by the total number of analyzed segments.

Conventional echocardiographic measurement of both LV and RV dimensions was performed (see “Results”). All dimensions were corrected for body surface area. RV function was measured by tricuspid annular plane systolic excursion (TAPSE) and RV fractional area change (FAC). Global LV function was measured by LV ejection fraction (LVEF) using the Simpson biplane method. Pulsed Doppler imaging was used to measure diastolic function. A full description of the echocardiographic measurements is available as Supplemental Data .

Deformation Imaging

We previously described our methods for image acquisition and postprocessing with commercially available software (EchoPAC PC version 11.2; GE Vingmed Ultrasound AS) for two-dimensional speckle-tracking analysis. More details according to the acquisition and processing of tissue deformation imaging are available in the Supplemental Files . The following parameters were measured in the basal, middle, and apical segment of each wall (a total of 18 segments): systolic peak strain and strain rate, defined as the maximum negative value between aortic valve opening and closure (in case values were positive during systole, the end-systolic value was measured). Averaging systolic peak strain values over 18 LV segments resulted in the mean systolic peak strain. Postsystolic shortening was defined as 100 × [(peak strain value − end-systolic value)/end-systolic value].

CMR Imaging

A subgroup of 19 patients with ARVD/C and eight relatives underwent CMR imaging as part of their clinical workup on a 1.5-T magnetic resonance imaging scanner (Achieva; Philips Healthcare, Best, The Netherlands) according to standard ARVD/C protocols. All CMR examinations were performed within 12 months of the echocardiographic studies. LE of intravenously administered gadolinium was used to identify areas within the left ventricle with myocardial fibrosis using the 16-segment model of the American Society of Echocardiography. The presence of gadolinium LE on CMR imaging was determined by consensus reading of two blinded experienced observers and considered definite only if present in the same myocardial segment in two different imaging planes.

Definition of LV Involvement

To compare the presence of LV involvement between conventional echocardiography and deformation imaging, LV involvement according to conventional echocardiography was defined as LV dysfunction (LVEF <50%) and/or the presence of akinesia or dyskinesia (LV involvement–conventional). LV involvement according to deformation imaging was defined as the presence of abnormal systolic peak strain (<|−12.5%|) and/or postsystolic shortening >15% in two adjacent LV segments (LV involvement–deformation imaging). Details concerning the cutoff values used for deformation imaging–derived parameters are stated in “Statistical Analysis.”

Predictor Selection

All established phenotypic expressions of ARVD/C as described in the revised 2010 task force criteria were chosen as potential predictors of outcomes, as well as both our parameters for LV involvement (LV involvement–conventional and LV involvement–deformation imaging). To increase the potential applicability of our definitions of LV involvement, only abnormalities in the LV posterolateral wall were used for correlation with clinical outcomes. Other echocardiographic parameters were TAPSE, RV FAC, LVEF, abnormal diastolic function, and mean systolic peak strain. Although CMR imaging–derived parameters are part of the task force criteria, these were not used as potential predictors, because CMR was performed only in a subset of patients.

Statistical Analysis

Continuous data are presented as mean ± SD and categorical variables as numbers or percentages. Differences in continuous data between either patients with ARVD/C or their relatives and control subjects were calculated using the independent-samples Student’s t test and χ ² or Fisher exact tests for categorical data. Bivariate correlations between continuous variables were determined using Pearson correlation coefficients. P values < .05 and 95% CIs of hazard ratios not including 1 were considered to indicate statistical significance.

Normal values for deformation imaging parameters are limited. Therefore, to correct for variability due to different observers and different echocardiographic vendors, we used our control group to identify normal values. In the control group, the 95% range of observed systolic peak values for strain and strain rate was evaluated to identify the normal values’ lower limit. This resulted in single cutoff values of |−12.5%| for systolic peak strain and |−0.6%/sec| for peak strain rate, which were used to differentiate normal from abnormal regional deformation. Postsystolic shortening > 15% of the systolic peak strain value after aortic valve closure was considered abnormal.

Categorical parameters derived from conventional and deformation echocardiographic imaging were correlated with clinical outcomes by Kaplan-Meier survival analysis and tested for significance with a log-rank test. Kaplan-Meier 3-year estimates of event risk were calculated for all categorical parameters. To calculate the magnitudes of both categorical and continuous single predictors, we performed univariate Cox proportional hazards analysis to obtain hazard ratios for each parameter individually. Parameters showing statistical significance ( P < .05) were included in a multivariate Cox proportional hazards model to adjust for dependency relations between variables. If both the continuous and categorical variable of the same predictor were significantly related to clinical outcome, the continuous variable was selected. Multicollinearity was expected between different variables and was investigated by using correlation statistics and calculating variance inflation factors. Variables with correlation coefficients > 0.75 and/or variance inflation factors > 5 were not included simultaneously as predictors in a multivariate Cox regression model. Taking into account the small number of events, a backward elimination procedure was performed.

Statistical calculations were made using commercially available software (SPSS version 20.0 for Windows; SPSS, Inc, Chicago, IL).

All analyses of standard echocardiographic measurements, visual wall motion analysis, deformation imaging analysis, CMR imaging analysis were performed blinded to the other modality analyses and for clinical outcomes.

Study Population

The baseline characteristics of the study population are summarized in Table 1 . No significant differences were seen between the two patient groups and the control subjects, other than age, which was higher in patients with ARVD/C, and more women among the relatives. At baseline, 84.2% of the patients with ARVD/C had already experienced episodes of (non)sustained ventricular tachycardia. Electrocardiographic abnormalities, as described in the task force criteria, were present in 92.1% of the patients with ARVD/C. The mean duration of follow-up after the echocardiographic examination was 5.9 ± 2.3 years in the patients with ARVD/C and 6.7 ± 0.7 years in their relatives ( P = NS).

Pathogenic mutations in the plakophilin-2 gene were most commonly identified: 71% in patients with ARVD/C and 75% in relatives ( Supplemental Table 2 ). By study design, all 16 relatives of patients with ARVD/C carried pathogenic mutations associated with ARVD/C also identified in the index patients but did not fulfill the task force criteria.

The control group consisted of 55 healthy volunteers unrelated to any of the patients with ARVD/C or their relatives.

Conventional Echocardiographic Findings

All patients were in sinus rhythm during the echocardiographic examinations. All RV dimensions were significantly increased, and RV systolic function (TAPSE and RV FAC) was significantly reduced in patients with ARVD/C compared with control subjects ( Table 2 ). No significant RV dilatation was seen in relatives, whereas TAPSE was moderately but significantly reduced in relatives compared with control sujbects. Tricuspid regurgitation grade > 1 was present in 23.7% of the patients with ARVD/C and in none of the relatives or control subjects.

LV end-diastolic volume was moderately reduced compared with control subjects in both ARVD/C groups. LV inflow tract diameter and end-diastolic LV posterior wall thickness in patients with ARVD/C were also found to be reduced ( Table 2 ). LVEF were moderately but significantly reduced in patients with ARVD/C. Reduced LVEFs (<50%) were present in 15.8% of patients with ARVD/C and none of the relatives or control subjects. Both medial and lateral annular peak velocities were found to be decreased in patients with ARVD/C compared with control subjects ( Supplemental Table 3 ). LV diastolic dysfunction was present only in patients with ARVD/C in 31.6%. All diastolic parameters are listed in Table 2 .

LV involvement according to conventional echocardiography was present in 31.6% of patients with ARVD/C and in none of the relatives or healthy control subjects ( Table 3 ).

Deformation Imaging

Deformation imaging by speckle-tracking was feasible in 83% of all segments ( n = 1,636 of 1,962). Feasibility did not differ significantly between patient groups.

Mean values of systolic peak strain were significantly reduced in most LV segments ( Table 2 and Supplemental Table 4 ). Deformation imaging detected a high incidence of abnormal systolic peak strain and postsystolic shortening in two or more adjacent segments in patients with ARVD/C. The middle and apical regions of the posterolateral wall were clear predilection sites within the left ventricle ( Figure 2 and Figure 3 ). LV involvement according to deformation imaging occurred in 68% of patients with ARVD/C ( Table 3 ). LV involvement was associated with a higher incidence of tricuspid valve regurgitation, 60%, compared with 31% in patients with ARVD/C without LV involvement. RV systolic pressure did not differ between patients with or without LV involvement ( Supplemental Table 5 ).

Graphical representation of abnormal segments by conventional echocardiography, deformation imaging (speckle-tracking), and CMR. Distribution and frequency of LV wall motion abnormalities in patients with ARVD/C detected with conventional echocardiography, deformation imaging (speckle-tracking), and CMR LE. Note the agreement between CMR LE and deformation imaging. Deformation imaging also detected wall motion abnormalities in 25% of the relatives of patients with ARVD/C, but a predilection site was less apparent because of small numbers of affected segments. CMR imaging was performed only in a subgroup of patients with ARVD/C ( n = 19). Visual-WMA , visual wall motion analyses.

LV abnormalities in a patient with ARVD/C. Example of a 16-year-old male patient presenting with ventricular tachycardia. Electrocardiography and conventional echocardiography showed typical phenotypic signs of ARVD/C. LV visual wall motion, LVEF, diastolic function, and LV dimensions were considered normal, indicating isolated RV disease. However, deformation imaging revealed abnormal deformation pattern of the LV lateral wall (A,B) . CMR imaging with LE was performed afterward, and the pattern of LE was in concordance with this finding (C) . (A) Deformation imaging of the septal wall. Normal deformation pattern and peak systolic value. (B) Deformation imaging of LV lateral wall. The lateral basal segment (red curve) shows decreased peak (end-)systolic-strain (asterisk) and postsystolic shortening (pound sign). Lateral midsegment (blue-green) shows postsystolic shortening (pound sign). Apical (pink) segment shows normal deformation. (C) CMR LE with LE in the basal and midlateral wall (yellow arrows) and a normal myocardium at the septal side. AVC , Aortic valve closure; AVO , aortic value opening.

Among relatives of patients with ARVD/C, mean systolic peak strain values were also significantly reduced ( Table 2 , Supplemental Table 4 ). Only postsystolic shortening was detected in this group, and predilection sites were less apparent because of the smaller number of affected segments. LV involvement was detected in 25% of relatives ( Table 3 ). No deformation imaging abnormalities were detected in control subjects.

CMR Imaging

CMR with LE was performed in a subgroup of 19 patients with ARVD/C and eight of their relatives. LE in the left ventricle was reported in four patients with ARVD/C; all others were negative. The distribution pattern of LE was subepicardial or midwall; no subjects had a subendocardial distribution pattern suggesting scar tissue due to coronary artery disease. In these four patients with ARVD/C, localization of LE corresponded well with the LV segments showing abnormal regional deformation. Less agreement regarding localization was seen between visual wall motion analyses and LE on CMR imaging. ( Figure 2 and Figure 3 ). LV involvement according to deformation imaging was seen in nine patients without LE on CMR. Six of the 19 patients with ARVD/C who underwent CMR imaging reached clinical end points during follow-up. Only two of these patients showed LE on CMR imaging, while five of these six had deformation imaging abnormalities in two adjacent segments.

Clinical Outcomes

During a mean follow-up period of 5.9 ± 2.3 years, 20 patients with ARVD/C (53%) reached clinical end points. The most common end point was an appropriate ICD intervention for ventricular tachycardia. Five patients (12.5%) died of sudden cardiac death or refractory end-stage heart failure, and one patient (2.5%) underwent heart transplantation because of progressive heart failure symptoms due to progression of the disease. All six patients had already reached earlier end points before death or heart transplantation. All relatives of patients with ARVD/C remained free of events during the follow-up period and were excluded from subsequent regression analysis.

Structural RV alteration and/or global RV dysfunction was the only established phenotypic expression, as mentioned in the task force criteria, that was significantly correlated with outcome. RV outflow tract (RVOT) end-diastolic dimension was the only conventional echocardiographic parameter indicating RV disease showing independent prognostic value ( Table 4 , Supplemental Table 6 ). The presence of electrocardiographic abnormalities and a history of ventricular arrhythmias did not differ between the event group and event-free survivors.

Conventional echocardiographic LV parameters, in particular LV dysfunction (LVEF < 50%), were also significantly correlated with outcomes. The presence of LV dysfunction (LVEF < 50%) at baseline (21.1% of patients with ARVD/C) was related to a significantly higher event rate during follow-up ( Figure 4 ). The Kaplan-Meier estimate for 1-year event risk was 50% for the presence of LVEF < 50% at baseline versus 16.7% with normal LV function. Systolic medial and lateral peak annular velocity was lower in the event group compared with the group remaining free of events but did not reach statical significance ( Supplemental Table 3 ).

Clinical course of patients with ARVD/C with LV involvement versus isolated RV disease. Kaplan-Meier survival plots for LVEF (left) and LV involvement–deformation imaging (right) . Signs of LV involvement are associated with adverse outcome.

Both LV involvement–conventional and LV involvement–deformation imaging were significantly correlated with adverse outcomes. However, after multivariate regression, LV involvement–deformation imaging and RVOT dimension remained the only independent predictors of clinical outcomes, indicating the adverse aspect of biventricular disease in ARVD/C. The presence of LV involvement at baseline, measured by deformation imaging, decreased the mean event-free survival time by threefold, compared with isolated RV disease (2.2 vs 6.6 years) ( Table 4 and Figure 4 ).

Study Population

The baseline characteristics of the study population are summarized in Table 1 . No significant differences were seen between the two patient groups and the control subjects, other than age, which was higher in patients with ARVD/C, and more women among the relatives. At baseline, 84.2% of the patients with ARVD/C had already experienced episodes of (non)sustained ventricular tachycardia. Electrocardiographic abnormalities, as described in the task force criteria, were present in 92.1% of the patients with ARVD/C. The mean duration of follow-up after the echocardiographic examination was 5.9 ± 2.3 years in the patients with ARVD/C and 6.7 ± 0.7 years in their relatives ( P = NS).

Table 1

Baseline characteristics

	Patients with ARVD/C ( n = 38)	Relatives ( n = 16)	Control subjects ( n = 55)
Patient characteristics
Male	57.9%	25.0% ∗	58.2%
Age (y)	47.1 ± 14.1 ∗	31.9 ± 13.8	37.9 ± 12.8
Length (cm)	178.6 ± 8.1	174.6 ± 8.1	178.0 ± 1.6
Weight (kg)	78.9 ± 13.7	73.8 ± 17.1	73.7 ± 15.6
Body surface area (m 2 )	1.98 ± 0.19	1.88 ± 0.22	1.91 ± 0.25
Heart rate (beats/min)	60.6 ± 10.6	59.9 ± 7.	57 ± 10.9
ICD	42.1%	0%	0%
Mean follow-up (y)	5.9 ± 2.3	6.7 ± 0.7	0
Task force criteria (major or minor)
Structural RV abnormalities	89.5%	18.8%	0%
Depolarization abnormalities	71.1%	18.8%	0%
Repolarization abnormalities	81.6%	18.8%	0%
History of ventricular arrhythmias	94.7%	6.3%	0%
Family history or pathogenic mutation †	94.7%	100%	NA

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Task Force 2: Pediatric Cardiology Fellowship Training in Noninvasive Cardiac Imaging
2. Predictors of Delayed Accreditation of Echocardiography Laboratories: An Analysis of the Intersocietal Accreditation Commission Database
3. Does Trastuzumab Cause Cardiotoxicity in the Treatment of Patients with Breast Cancer?
4. Highlights of the 26 thScientific Sessions’ Pediatric Congenital Track

Stay updated, free articles. Join our Telegram channel

Join