Different Prognostic Value of Functional Right Ventricular Parameters in Arrhythmogenic Right Ventricular Cardiomyopathy/Dysplasia

The Zurich ARVC Program (www.arvc.ch) was established in 2011 to provide expert clinical care for patients with ARVC/D. Patients were recruited at 3 collaborative Swiss centers (University Hospital Zurich, University Hospital Bern, and Triemli Hospital Zurich). So far, the Zurich ARVC Program has recruited 131 patients ≥15 years of age with a possible, borderline, or definite diagnosis of ARVC/D according to the 2010 Revised Task Force Criteria (TFC).²⁴ For this study, only patients with a definite or borderline diagnosis, echocardiographic assessment at baseline including measurements of both FAC and TAPSE, and a follow-up duration of ≥3 months were included. Furthermore, patients in whom diagnosis was made by the inclusion of FAC were excluded from the current analysis to avoid interaction bias with disease probability within the Cox proportional hazard model. Thus, within this study, we report data from 70 patients that fulfilled the inclusion criteria (Figure 1). TTE at baseline was defined as the first TTE in which measurements of both FAC and TAPSE were available. There was no patient in whom diagnosis of ARVC/D was based on a reduced RVEF as detected by cardiac magnetic resonance (CMR) tomography. Echocardiography was prospectively analyzed in 28 (40%) and retrospectively analyzed in 42 (60%) patients. Echocardiographic data were interpreted by 2 different observers blinded to the outcome data and to each other. Clinical information regarding demographics and symptoms was obtained from hospital records at the time of baseline TTE. This study was approved by the institutional review board of each participating center, and in accordance with Swiss law, patients gave written informed consent for prospective inclusion into this study.

From left parasternal, apical, and subcostal windows, conventional M-mode, 2D, and color Doppler echocardiography was performed by an experienced investigator trained in the analysis of the complex anatomy of the right ventricle in all patients using ultrasound transducers with a frequency range between 1 and 5 MHz (Acuson Sequoia C 512; Acuson Corporation, Siemens, Mountain View, CA; Vivid 7 or Vivid E9; GE Vingmed Ultrasound AS, Horten, Norway; iE33; Philips Healthcare, Best, the Netherlands) and according to a standard clinical protocol that did not differ among the 3 centers. Dimensions and function of both ventricles were assessed according to established guidelines for 2D TTE.²⁵ For FAC, final values were obtained after averaging 2 measurements over 1 cardiac cycle, whereas for all other parameters, 1 measurement was taken. The degree of tricuspid regurgitation was determined as none, mild, moderate, or severe.²⁶

Echocardiography-derived 2D RV area measurements were taken from the apical 4-chamber view (4CV) at end-diastole (area D) and end-systole (area S), and FAC was defined as the ratio between the difference of the end-diastolic and end-systolic RV areas and the end-diastolic area (FAC=[area D−area S]/area D) as previously reported (Figure 2).^9,23 TAPSE was determined by placing the M-mode cursor perpendicularly through the lateral tricuspid annulus in the 4CV and measuring the amount of longitudinal motion at peak systole (Figure 2).^17,23 A FAC ≥33%^14,24 and a TAPSE ≥17 mm²⁷ were considered normal. Echocardiographic exams and reports were also reviewed for regional wall motion abnormalities and left ventricular (LV) involvement. LV involvement was considered when regional wall motion abnormalities or a reduced EF (<50% by Simpson biplane method) were present and other causes were excluded.^9,14

To determine intra- and interobserver variability, measurements were reanalyzed by the same observer (at least 1 week after the first measurement) and by an independent observer in 31 randomly selected cases.

Follow-up for survival data was performed by review of hospital charts, including implantable cardioverter defibrillator (ICD) interrogations with stored electrograms, Holter ECGs, clinical visits, and telephonic interviews of patients, their relatives, or treating physicians. The role of FAC, TAPSE, and other parameters on long-term clinical outcome was compared in 2 patient groups: (1) patients with major adverse cardiovascular events (MACE)—adverse outcome (composite of cardiac death, heart transplantation, survived SCD, ventricular fibrillation [VF], sustained ventricular tachycardia [VT], and arrhythmogenic syncope), and (2) all remaining patients—favorable outcome. For Kaplan–Meier estimates and Cox regression analyses, time from baseline to MACE was the event of interest. Survived SCD, VF, sustained VT, and arrhythmogenic syncope were defined as previously reported.²⁸

Continuous variables are presented as mean±SD or median (interquartile range [ICR]). Categorical variables are reported as frequency (percentage). Comparisons between groups with a favorable outcome and MACE (Tables 1 and 2) were performed by the 2-sided unpaired Student t-test (normal distribution) or the Mann–Whitney U-test (skewed distribution) for continuous variables, and by Fisher exact test for categorical variables. For comparison of echocardiographic data between baseline and follow-up, we used the paired Student t-test (normal distribution) or the Wilcoxon signed-rank test (skewed distribution) for continuous variables and the McNemar test (without continuity correction) for categorical variables. Spearman correlation was performed to analyze correlations between TTE and CMR parameters. Cumulative probabilities of survival free of MACE were determined by the Kaplan–Meier method. We applied the Grambsch and Therneau method for testing the proportional hazards assumption for univariable Cox models. There was no evidence that proportionality assumption was not fulfilled. Accordingly, baseline variables that were significantly associated with MACE were identified by univariable Cox regression analysis, and differences in survival between groups were calculated with the log-rank test. The performance of TAPSE and FAC for classifying an end point was measured using Harrell c-statistic, which accounts for censored measurements. Additionally, we calculated receiver operating characteristic curves and derived sensitivity and specificity for the determined cut-off values. The cut-off values for FAC and TAPSE were determined as follows: the log-rank test of all cut-off points between the 20th and 80th percentiles was calculated, and the cut-off with the minimal adjusted P value was chosen according to the formula by Altman.²⁹ For the purpose of this study, we were mainly interested in echocardiographic variables at initial assessment that could be important to predict MACE in patients with ARVC/D. Based on the univariable Cox regression analysis, we performed separate bivariable analyses with the only 2 RV functional echocardiographic parameters (FAC and TAPSE) and with the 2 right-sided dimensional echocardiographic parameters that yielded the lowest P values in univariable analysis (RV end-diastolic area [RVEDA] and right atrial [RA] short-axis diameter). These bivariable analyses were performed to better delineate the prognostic properties of each RV functional and right-sided dimensional parameter on adverse outcome. The initial echocardiographic assessment will guide the physician to choose an appropriate therapy for each individual patient based on the risk for MACE. As in most clinical cases, it is not appropriate to wait for an assessment of changes of functional and structural echocardiographic parameters to evaluate the patient’s risk for MACE; we have not used time-updated echocardiographic parameters to predict MACE. Intraclass correlation (ICC) was used to quantify intra- and interobserver variability, with ICC values >0.75 representing good reliability of measurements. A 2-sided P value <0.05 was considered significant. Statistical analysis was performed using R programming language (R Development Core Team 2009) and GraphPad Prism 5 (GraphPad Software Inc, La Jolla, CA).

Of 131 patients screened in the 3 centers, 70 had a complete echocardiographic examination at baseline and were included in the study (Figure 1). All of these were referred as index patients with a definite or borderline diagnosis of ARVC/D according to the 2010 TFC. Borderline cases fulfilled 1 major and 1 minor (n=13) or 3 minor (n=4) criteria. Baseline characteristics are provided in Table 1.

Echocardiographic data are presented in Tables 2 and 3. Ten (14%) patients had echocardiographic evidence of LV involvement at baseline. When the reference values were indexed to BSA, 15 (21%) patients had a reduced TAPSE <7 mm/m², 44 (63%) patients a dilated RV >12.3 cm²/m², 18 (26%) patients a dilated RA (short axis >25 and long axis >30 mm/m²), and 10 (14%) patients a dilated LA >23 mm/m².³⁰ FAC correlated better with RVEF as determined by CMR than TAPSE (r=0.480; P=0.05 for FAC versus RVEF; r=0.34; P=0.19 for TAPSE versus RVEF).

From the whole cohort, a follow-up TTE was available in 58 patients. Of those, 33 patients belonged to the MACE group and 25 patients to the favorable group (P=0.21). The median time between baseline TTE and follow-up TTE was 2244 (IQR, 1099–3591; range, 95–9744) days. More patients had LV involvement compared with baseline (29% versus 14%; P=0.02). Averaged for all patients in whom a follow-up measurement was available, median FAC, mean TAPSE, and mean TAPSE indexed to BSA decreased over time (P=0.03 for FAC; P=0.03 for TAPSE; P=0.01 for TAPSE/BSA, each versus baseline). In contrast, median RVEDA increased (P=0.001 versus baseline). TAPSE decreased in 47% of patients, whereas there was no change in 24% and an increase in 29% (range of change, −10 to +8 mm). TAPSE adjusted for BSA decreased in 45% of patients, whereas there was no change in 29% and an increase in 26% (range of change, −5 to +4 mm/m²). FAC decreased in 57% of patients, whereas there was no change in 4% and an increase in 39% (range of change, −33% to +18%). RVEDA increased in 63% of patients, no change in 11%, and a decrease in 26% (range of change, −15 to +25 cm²). Figure 3 depicts the data for FAC and TAPSE at baseline and their changes during long-term follow-up in patients from whom measurements of both parameters were available. A significant positive correlation between baseline FAC and TAPSE (r=0.5; P<0.001) and a negative correlation between baseline FAC and baseline right chamber dimensions (r=−0.57; P<0.001 for FAC versus RVEDA; r=−0.44; P<0.001 for FAC versus RA short axis/BSA) were observed. There was no evidence for an association between an impaired TAPSE and subtricuspidal regional wall motion abnormalities (P=1.0).

Over a median follow-up period of 5.3 (IQR 1.8–9.8) years, 37 (53%) patients experienced MACE (Table 4). Median follow-up was not significantly different between patients with favorable versus those with an adverse outcome (4.4 years; IQR 1.6–8.1 versus 6.3 years; IQR 2.1–10.5 years, respectively; P=0.48). No patient died. Heart transplantation was performed in 5 patients (due to incessant VT/VF combined with heart failure in 3 patients and heart failure only in 2 patients). The most frequent MACE were survived sustained VT (67%) and VF (19%). Thirty-one (63%) of 49 patients with an ICD experienced an appropriate ICD intervention. This intervention was because of VT in 24 patients and VF in 7 patients. Table 2 shows baseline echocardiographic parameters in patients with favorable outcome and those with MACE. Variables analyzed as potential predictors of MACE are presented in Table 5. A lower FAC and TAPSE as well as an increased RVEDA and RA short-axis diameter were identified as univariable predictors (Table 5). In bivariable analyses stratified into functional and dimensional echocardiographic parameters, FAC (functional parameter), RVEDA, and RA short-axis diameter (dimensional parameters) remained independent predictors of MACE. A reduced FAC constituted the strongest predictor of MACE (hazard ratio, 1.08 per 1% decrease; 95% confidence interval [CI], 1.04–1.12; P<0.001) on bivariable analysis. Consistent with this observation, Harrell c-statistic for FAC equaled 0.67 and for TAPSE 0.61, demonstrating that the measurement of baseline FAC had a better predictive value compared with baseline TAPSE. Figure 4 illustrates the results of the receiver operating characteristic analysis for distinguishing patients with MACE from those without. For baseline FAC, a sensitivity >80% for predicting MACE was obtained at a cut-off <23% (sensitivity, 88%; 95% CI, 73–95%) and for baseline TAPSE at a cut-off <17 mm (sensitivity, 85%; 95% CI, 69–93%), yet both parameters displayed a limited specificity for these cut-off values (FAC, 46%; 95% CI, 31–62%; TAPSE, 41%; 95% CI, 26–57%). Patients with a FAC <23% or a TAPSE <17 mm had a cumulative probability for MACE ≈75% after 2 years (Figure 5), which was higher than for patients with a FAC ≥23% (P log-rank unadjusted and adjusted <0.001) and a TAPSE ≥17 mm (P log-rank unadjusted=0.02, adjusted=0.19).

For interobserver variability, ICC values were 0.50 (95% CI, 0.18–0.72) for FAC and 0.83 (95% CI, 0.67–0.91) for TAPSE, whereas for intraobserver variability, ICC values were 0.59 (95% CI, 0.30–0.86) for FAC and 0.86 (95% CI, 0.72–0.93) for TAPSE.

Previous echocardiographic studies have demonstrated that RV dilation and reduced regional or global RV function are important features of ARVC/D.^31–33 Our study underscores that ARVC/D is a progressive disease.^8,31,34 During long-term follow-up, right chamber dimensions significantly increased, whereas both FAC and TAPSE decreased. Moreover, LV involvement increased by 2.3-fold compared with baseline, with almost one third of patients demonstrating biventricular disease at follow-up. In line with previous findings from a smaller ARVC/D cohort from Scandinavia, FAC and TAPSE did not decrease, and RVEDA did not increase in a substantial proportion of patients. This reflects some heterogeneity in the individual disease course but may also be related to variability in the measurement of those echocardiographic parameters.³⁴ Our results differ from those of a smaller Italian ARVC/D cohort in which RVEF and RV end-diastolic volume as determined by 2D TTE did not change significantly between baseline and long-term follow-up.³⁵ A likely explanation for this discrepancy may be related to differences in the assessment of RV function and dimension, as well as a lack of power to detect significant differences occurring over time due to small sample size.

Different ARVC/D cohorts independently confirmed that spontaneous ventricular tachyarrhythmias, syncope, heart failure, LV involvement, and inducibility of sustained ventricular arrhythmias constitute the risk factors for MACE.^{8,9,12–14,28,36} Yet previous studies have yielded conflicting evidence regarding the usefulness of RV function and dimension for risk prediction in ARVC/D.^7–15,37 Controversial findings by others may be related to different study methodologies, divergent imaging modalities, and heterogeneous patient cohorts. Neither occurrence nor severity of RV disease was defined homogenously throughout these studies. A large international multicenter study, Multidisciplinary Study of Right Ventricular Dysplasia, has defined diffuse RV involvement as severe regional involvement in ≥2 RV regions or an RVEF ≤45%.¹² Yet imaging modalities to measure these parameters have not been standardized across different centers.¹⁹ Data about the predictive role of FAC reflecting RVEF in a single 2D plane acquired in the 4CV are scarce in patients with ARVC/D, and 3 previous studies analyzing the predictive role of FAC in ARVC/D came to different conclusions.^9,14,15 Echocardiographic assessment of RV dimension and function is challenging because of the complex RV geometry, and both FAC and TAPSE do not always correlate well with the more accurate 3-dimensional volumes and EF obtained by CMR.^16–20 This is particularly true for FAC, which depends on an optimal image quality. FAC also does not fully reflect the regional function of the RV outflow tract and subtricuspid area that are often involved in ARVC/D.¹⁴ Furthermore, FAC consistently displayed higher intra- and interobserver variability compared with TAPSE in previous studies as well as the present study, which may also be related to the fact that we did not exclude any patient due to difficulties in measuring FAC. Yet both parameters showed good ICC in our analysis.^23,38 Despite these limitations, the current study demonstrates that even small reductions in FAC were associated with MACE, and that a reduced FAC was the strongest predictor of MACE on univariable and bivariable Cox regression analysis. Of note, an FAC >33%, which is proposed as the lower normal range in the current guidelines,¹⁹ still conferred a 3-fold increased risk for adverse events and supports the 2010 TFC, in which the combination of regional wall motion abnormalities and an FAC between 33% and 40% are considered minor criteria for ARVC/D diagnosis.¹²

Although angle-dependent, TAPSE can easily be determined by M-mode echocardiography, is less dependent on image quality, and constitutes a robust and reproducible parameter for the echocardiographic assessment of RV function in ARVC/D. It also correlates well with RVEF determined by CMR or angiography.^{16,20–23,39,40} The low intra- and interobserver variability for TAPSE in the current study is consistent with those previous findings. TAPSE has the potential to discriminate well between patients with ARVC/D and healthy probands.^18,33 Although TAPSE has evolved as an important prognostic parameter in other cardiac pathologies such as pulmonary hypertension, myocarditis, and dilated cardiomyopathy,^16,17,41 it has not been studied as a prognostic marker in ARVC/D. To our knowledge, our study is the first to demonstrate that TAPSE may be valuable for predicting MACE in ARVC/D. A value of TAPSE <17 mm best identified patients with MACE. However, TAPSE was inferior to FAC with regard to its predictive value and also correlated less well with RVEF as determined by CMR, which may well be related to the fact that TAPSE only represents the longitudinal systolic movement of the RV lateral free wall and hence does not reflect the entire RV systolic function. This property may constitute a particular drawback in patients with ARVC/D, which can affect multiple regions of the RV. Such a regional heterogeneity in systolic function is certainly better reflected by FAC. Another new finding of our study is the predictive value of RA dilation as previously shown in Eisenmenger Syndrome.^30,42 This finding is in line with a recent study showing a prognostic impact of significant tricuspid regurgitation in ARVC/D.¹⁴

This study analyzing patient data from 3 Swiss tertiary care centers may have limitations in patient selection and risk stratification by its observational nature. Of note, we studied a rather high-risk cohort of patients with ARVC/D, given the fact that 57% of our patients had a sustained episode of VT/VF before baseline echocardiography and given the observed high MACE rates. Definite disease was more common in the MACE group at baseline, although definite disease was not associated with MACE in Cox regression analysis. Thus, our data may not be generalizable to other, potentially lower-risk ARVC/D populations. Yet we analyzed whether the patients who were excluded from the final analysis constituted a lower-risk population than those included. In patients with a borderline and definite diagnosis of ARVC/D, MACE were observed in 17 of 34 (50%) excluded patients compared with 37 of 70 (53%) included patients, minimizing the risk of selection bias toward a higher-risk population within our ARVC/D cohort. Although the studied cohort was not small for patients with ARVC/D, the sample size limits the ability to control for all confounders in multivariable analysis. Although median follow-up was not significantly different between patients with a favorable outcome and those with MACE, from a clinical perspective, we cannot fully exclude that the median difference of 1.9 years may have contributed to more adverse events observed in the MACE group and that MACE may also have occurred in the favorable group if follow-up period would have been equal. When RV function is regionally impaired in areas not imaged in the 4CV, such as the RV outflow tract, FAC may be an imperfect measure of RV systolic function. Therefore, FAC may perform less well as a predictor of MACE in those patients. Furthermore, the assessment of RV function with FAC and TAPSE may be limited by the acoustic shadowing of ICD leads. The advent of tissue Doppler imaging and recently strain imaging and 3-dimensional tools may improve both diagnosis and risk stratification in ARVC/D,^15,18,33,43 although their use is still hampered by wide variations, noise artifacts, limited availability, and intervendor variations.^18,43–45

This long-term observational study indicates that TAPSE and dilation of right-sided cardiac chambers are associated with an increased risk for MACE in patients with ARVC/D with advanced disease and a high risk for adverse events. However, FAC is the strongest echocardiographic predictor of adverse outcome in these patients. Our data advocate a role for TTE in risk stratification in patients with ARVC/D, although our results may not be generalizable to lower-risk ARVC/D cohorts.