Author + information
- Received February 11, 2015
- Revision received March 16, 2015
- Accepted April 20, 2015
- Published online June 30, 2015.
- Abbas Zaidi, BSc (Hons), MBBS, MD∗,
- Nabeel Sheikh, BSc (Hons), MBBS∗,
- Jesse K. Jongman, MD†,
- Sabiha Gati, BSc (Hons), MBBS∗,
- Vasileios F. Panoulas, MD, PhD‡,
- Gerald Carr-White, BSc (Hons), MBBS, PhD§,
- Michael Papadakis, MBBS, MD∗,
- Rajan Sharma, BSc (Hons), MBBS, MD∗,
- Elijah R. Behr, MBBS, MD∗ and
- Sanjay Sharma, BSc (Hons), MBChB, MD∗∗ ()
- ∗St. George’s University of London, London, United Kingdom
- †Isala Clinics, Zwolle, the Netherlands
- ‡Imperial College Healthcare National Health Service Trust, London, United Kingdom
- §Guy’s and St. Thomas’s Hospital, London, United Kingdom
- ↵∗Reprint requests and correspondence:
Prof. Sanjay Sharma, Division of Cardiovascular Sciences, St. George’s University of London, Cranmer Terrace, SW17 0RE, London, United Kingdom.
Background Physiological cardiac adaptation to regular exercise, including biventricular dilation and T-wave inversion (TWI), may create diagnostic overlap with arrhythmogenic right ventricular cardiomyopathy (ARVC).
Objectives The goal of this study was to assess the accuracy of diagnostic criteria for ARVC when applied to athletes exhibiting electrocardiographic TWI and to identify discriminators between physiology and disease.
Methods The study population consisted of athletes with TWI (n = 45), athletes without TWI (n = 35), and ARVC patients (n = 35). Subjects underwent electrocardiography (ECG), signal-averaged electrocardiography (SAECG), echocardiography, cardiac magnetic resonance imaging (CMRI), Holter monitoring, and exercise testing.
Results There were no electrical, structural, or functional cardiac differences between athletes exhibiting TWI and athletes without TWI. When athletes were compared with ARVC patients, markers of physiological remodeling included early repolarization, biphasic TWI, voltage criteria for right ventricular (RV) or left ventricular hypertrophy, and symmetrical cardiac enlargement. Indicators of RV pathology included the following: syncope; Q waves or precordial QRS amplitudes <1.8 mV; 3 abnormal SAECG parameters; delayed gadolinium enhancement, RV ejection fraction ≤45%, or wall motion abnormalities at CMRI; >1,000 ventricular extrasystoles (or >500 non-RV outflow tract) per 24 h; and symptoms, ventricular tachyarrhythmias, or attenuated blood pressure response during exercise. Nonspecific parameters included the following: prolonged QRS terminal activation; ≤2 abnormal SAECG parameters; RV dilation without wall motion abnormalities; RV outflow tract ectopy; and exercise-induced T-wave pseudonormalization.
Conclusions TWI and balanced biventricular dilation are likely to represent benign manifestations of training in asymptomatic athletes without relevant family history. Diagnostic criteria for ARVC are nonspecific in such individuals. Comprehensive testing using widely available techniques can effectively differentiate borderline cases.
Individuals engaging in regular, intensive sporting activity frequently demonstrate a constellation of electrical and structural cardiac alterations that are collectively described as the “athlete’s heart.” Although such training-induced changes are generally considered physiological and benign (1), they occasionally overlap with phenotypic features of inherited cardiomyopathies, in which vigorous exercise is associated with an increased risk of sudden cardiac death (SCD) (2,3). Physiological remodeling of the athlete’s right ventricle (RV) may mimic changes observed in arrhythmogenic right ventricular cardiomyopathy (ARVC) (4), which is responsible for as many as 22% of SCD in young athletes (2). Accurate differentiation between physiological and pathological RV remodeling is essential because failure to identify the disease could jeopardize a young life, whereas an inappropriate diagnosis of ARVC may lead to an unnecessary exclusion from sporting activity. Whereas diagnostic algorithms to facilitate the differentiation between physiological left ventricular (LV) hypertrophy and hypertrophic cardiomyopathy are established, similar data are lacking for the RV. Furthermore, diagnostic criteria for ARVC are derived from patients with established disease (5) and may therefore not be applicable to low-risk individuals, such as athletes. The objectives of the present study were to assess the accuracy of current diagnostic criteria for ARVC when applied to athletes exhibiting phenotypic overlap with the condition and to identify clinical discriminators between RV physiology and disease.
All participants provided written consent, and ethical approval was obtained from the local research ethics committee in accordance with the Declaration of Helsinki. In the United Kingdom, the charity Cardiac Risk in the Young subsidizes cardiovascular evaluations for several elite sporting organizations that mandate pre-participation screening of all member athletes. The screening protocol consists of a health questionnaire, physical examination, and 12-lead electrocardiogram (ECG). In order to facilitate a study group exhibiting diagnostic overlap with ARVC, 45 athletes with ECG T-wave inversion (TWI) were recruited between 2011 and 2013 for further detailed assessment (TWI+ athletes). The TWI+ athletes were required to exhibit anterior or lateral TWI as a minimum inclusion criterion, as per the 2010 Task Force Criteria (TFC) for the diagnosis of ARVC (5). A cohort of athletes without TWI (TWI– athletes), matched for age, sex, ethnicity, and sporting category, was recruited to act as a control group. The athletic cohorts were between 14 and 35 years of age and competed at international, national, or regional levels. Sporting disciplines were categorized as predominantly endurance or strength, and as high-dynamic/high-static or non–high-dynamic/high-static disciplines, according to accepted criteria (6). Athletes with any previous history of cardiac or pulmonary disease, systemic hypertension, or diabetes mellitus were excluded. The ARVC cohort consisted of patients between 14 and 35 years of age presenting to 2 U.K. tertiary cardiac referral centers with a new diagnosis of “definite” ARVC by 2010 TFC (5).
All study participants underwent resting ECG, signal-averaged electrocardiography (SAECG), transthoracic echocardiography, cardiac magnetic resonance imaging (CMRI), and exercise testing, and they were assessed with reference to the 2010 TFC (5). Tissue characterization of the RV wall was not performed in any case. Genetic testing was offered only to the ARVC patients.
A standard 12-lead ECG was performed in the supine position using either a MAC 5000 or MAC 5500 digital resting ECG recorder (GE Medical Systems, Milwaukee, Wisconsin). Measurements were made using calipers. The normal frontal cardiac axis was considered to be >–30°, but <120°. Left ventricular hypertrophy (LVH) and right ventricular hypertrophy (RVH) were defined according to the Sokolow-Lyon voltage criteria (LVH = SV1 + RV5/6 >3.5 mV; RVH = RV1 + SV5/6 >1.05 mV). TWI ≥–0.1 mV in 2 or more contiguous leads was considered significant. Deep TWI was defined as ≥–0.2 mV. Leads V1 to V4 were subclassified as anterior precordial leads. Biphasic T waves were defined as those with components above as well as below the PR-segment. TWI in leads V1 to V3 or beyond, in the absence of complete right bundle branch block (RBBB), was considered a major diagnostic criterion for ARVC. TWI in leads V1 to V2, or V4, V5, or V6 was considered a minor diagnostic criterion in the absence of complete RBBB, or in leads V1 to V4 with complete RBBB. Partial right bundle branch block was defined as QRS duration >100 ms, but <120 ms, with rSR′ morphology in lead V1 and qRS in V6. Early repolarization was defined as J-point elevation ≥0.1 mV in 2 or more consecutive leads. A novel index of maximal QRS amplitude in the precordial leads (V-Ampmax) was formulated, measured from the peak of the R-wave to the nadir of the S-wave (greatest single value in leads V1 to V6). Additional ECG markers compatible with ARVC were sought, including QRS terminal activation duration ≥55 ms in leads V1, V2, or V3 and the epsilon wave (5).
Echocardiographic examinations were performed using the following commercially available ultrasonography systems: Vivid-I (GE Healthcare, Milwaukee, Wisconsin); CX50 (Philips Medical, Bothell, Washington); or iE33 (Philips Medical). A complete echocardiographic study of the left and right heart was performed according to international guidelines (7,8). Echocardiographic studies were saved to compact disks as numeric files to generate anonymity, and cardiac measurements were repeated independently by an experienced cardiologist (A.Z.) blinded to the identity of the subject. All RV measurements were made from end-diastolic frames acquired with the breath held in end expiration. A Philips Achiever 3.0-T TX scanner (Amsterdam, the Netherlands) was used for CMRI examinations. Delayed gadolinium enhancement images were acquired 10 min after administration of 0.2 mmol/kg intravenous gadolinium-diethylenetriaminepentaacetate (Guerbet Dotarem, Obex Medical Limited, Auckland, New Zealand) using an inversion-recovery gradient echo sequence. Ventricular volumes and function were measured using standard techniques and analyzed using semiautomated software (Extended MR workspace, Philips, Amsterdam, the Netherlands) (9). All imaging measurements were recorded as absolute values and were also indexed to body surface area according to the DuBois-DuBois formula (10).
Holter monitor, signal-averaged ECG, exercise, and genetic testing
Twenty-four-hour ambulatory ECG recording (Lifecard 12 Holters, Spacelabs Healthcare, Hawthorne, California) was used to detect ventricular arrhythmias. Subjects were encouraged to continue day-to-day activities including exercise during monitoring. Upright treadmill stress testing was performed using a standard Bruce protocol (11). Subjects were exercised to volitional exhaustion. T-wave pseudonormalization was categorized as complete, partial (positive increase in T-wave axis but persistent negative component), or absent. The same machines used for standard ECG were used according to accepted methodology for SAECG, with the use of a 40-Hz high-pass bidirectional filter (12). Late potentials were defined as abnormal values in ≥1 of the following parameters (in accordance with ARVC TFC): duration of filtered QRS complex >114 ms (with QRS duration <110 ms on standard ECG); duration of terminal QRS (with amplitude <40 μV) >38 ms; and root mean square voltage of the terminal 40 ms of filtered QRS <20 μV (5). Genetic analysis was performed on consenting ARVC patients for 5 desmosomal gene mutations: desmocollin-2; desmoglein-2; desmoplakin; junctional plakoglobin; and plakophillin-2 (13).
The Kolmogorov-Smirnov test was used to assess normality of distributions. Group differences were tested using 1-way analysis of variance (with Sidak post-hoc test) or the Kruskall-Wallis test (with Dunn post-hoc test). The chi-square or Fisher exact tests were used to assess proportional differences between groups. Forward stepwise binary logistic regression was used to create a 5-variable model for differentiating physiological RV remodeling from ARVC. Variables included were the following: V-Ampmax (mV); presence of 3 abnormal SAECG parameters (1 = yes); presence of >500 ventricular extrasystoles (VE) per 24 h (1 = yes); total exercise test duration (min); and echocardiographic ratio of RV basal dimension in apical view/LV end-diastolic dimension in parasternal long axis view >0.9 (1 = yes). Goodness of fit was assessed using Nagelkerke R-square test and Hosmer-Lemeshow test. Receiver-operating characteristic curves were used to assess the discriminatory power of the model. Youden criterion was used to derive an optimal diagnostic cutoff value. All analyses were performed using SPSS software (version 20, Chicago, Illinois). Values are expressed as mean ± SD or percentages as appropriate. Two-tailed p values <0.05 were considered significant.
Demographics, symptoms, and family history of athletes and ARVC patients
Athletes competed in 12 sporting disciplines, predominantly soccer (35.0%), athletics (12.5%), cycling (11.3%), rugby (11.3%), and triathlon (11.3%). All athletes were either Caucasian (71.3%) or black (Table 1). Almost one-half of the ARVC cohort (48.6%) performed ≤2 h/week of exercise. Two ARVC patients (5.7%) performed ≥6 h/week of exercise, although none were competitive athletes. Previous episodes of vasovagal syncope were reported in 3 athletes (3.8%), occurring in the context of concurrent injury or acute illness. Two athletes (2.5%) reported previous episodes of nonexertional, pleuritic chest pain. None of the athletes were symptomatic during exertion. After independent review by 2 experienced cardiologists (S.S., A.Z.), none of the athletes were deemed to express symptoms suggestive of underlying cardiovascular pathology. None of the athletes revealed a family history of cardiomyopathy or premature SCD. The vast majority of the ARVC cohort had been assessed in the context of family screening (40.0%), unexplained syncope (25.7%), or out-of-hospital cardiac arrest (20.0%). Significant cardiovascular symptoms were reported by 71.4% of the ARVC patients, most frequently syncope (exertional: 28.6%; at rest: 20.0%), palpitations (exertional: 17.1%; at rest: 17.1%), and pre-syncope (exertional: 11.4%; at rest: 2.9%). Comprehensive evaluation of first-degree relatives of the ARVC patients confirmed familial disease in 57.1% of cases, with a history of SCD in 25.7% of families. More than one-half of the ARVC cohort (54.3%) was found to harbor a known disease-causing mutation (plakophillin-2 alone: 28.6%; plakophillin-2 and desmoplakin: 11.4%; desmocollin-2: 8.6%; desmoplakin alone: 2.9%; desmoglein-2: 2.9%; negative genetic testing: 20.0%; untested: 25.7%). Comprehensive phenotypic and genetic characterization of the ARVC cohort is demonstrated in Online Tables 1 and 2.
Clinical evaluation of athletes and ARVC patients
Voltage LVH and RVH were more common, and there was a trend toward a greater prevalence of left and right atrial enlargement in athletes compared with ARVC patients (Table 2). Pathological Q waves were observed in 8.6% of ARVC patients, but in none of the athletes. V-Ampmax ranged from 1.8 to 7.3 mV in male athletes, compared with 0.8 to 3.3 mV in male ARVC patients. In female athletes, V-Ampmax ranged from 1.2 to 3.5 mV, compared with 1.1 to 3.0 mV in female ARVC patients. Almost two-thirds of the ARVC cohort (62.9%) demonstrated TWI at presentation (Table 2). Nearly one-fourth of ARVC patients (22.9%) revealed inferior TWI, which does not constitute part of the 2010 TFC (5). The distribution of TWI in athletes and ARVC patients is demonstrated on a lead-by-lead basis in Online Table 3. T-wave characteristics were compared between TWI+ athletes (n = 45) and TWI+ ARVC patients (n = 22). There were no differences between these 2 groups with respect to the distribution and depth of TWI (anterior TWI: 77.8% vs. 90.9%, p = 0.31; lateral TWI: 40.0% vs. 18.2%, p = 0.10; inferior TWI: 33.3% vs. 36.4%, p = 1.00; deep TWI: 62.2% vs. 50.0%, p = 0.43). However, the majority of TWI+ athletes exhibited biphasic T-wave morphology with preceding convex ST-segment elevation, compared with a minority of the TWI+ ARVC patients (71.1% vs. 13.6%; p < 0.001), who revealed isoelectric ST-segments in the majority of cases. Anterior and lateral early repolarization was 3-fold to 4-fold more common, and inferior early repolarization was 9-fold more common in athletes than in ARVC patients. Epsilon waves were not observed in any study subject. The QRS terminal activation duration in V1 to V3 and the prevalence of partial RBBB did not differ between athletes and ARVC patients. One ARVC patient (2.9%) exhibited ventricular pre-excitation.
Image quality was sufficient for 96.5% of the echocardiographic indexes of the 2010 TFC to be successfully quantified. The TWI+ and TWI– athletes did not differ in any biventricular echocardiographic structural or functional parameters (Table 3). Athletes exhibited greater LV cavity dimensions and mass and superior indexes of LV diastolic function than did ARVC patients. None of the linear measurements of RV size differed between athletes and ARVC patients, even after indexing for body surface area. Mean RV wall thickness was almost 2 mm greater in athletes than in ARVC patients. Among measures of global systolic and diastolic function, RV fractional area change and tissue Doppler RV long-axis velocity were greater in athletes than in ARVC patients; a cutoff value of RV fractional area change ≤30% demonstrated high specificity (89.1%) for ARVC (Central Illustration). Regional RV wall motion abnormalities (WMA) were observed in over one-half of the ARVC patients (51.4%), compared with only 3 of the athletes (3.8%).
Cardiac magnetic resonance imaging
The TWI+ and TWI– athletes did not differ in any biventricular CMRI structural or functional parameters (Table 3). Athletes exhibited greater LV volumes and LV mass than ARVC patients did. RV volumes did not differ between any of the groups, even after indexing for body surface area. A ratio of RV to LV end-diastolic volume (EDV) (RVEDV/LVEDV) >1.2 was, however, highly specific for ARVC, as was RV ejection fraction ≤45% (Central Illustration). Regional RV WMA was observed in 51.4% of the ARVC patients. The 3 athletes with apparent echocardiographic RV WMA demonstrated normal RV wall motion at CMRI. Fourteen ARVC patients (40.0%) exhibited delayed gadolinium enhancement (RV only: n = 3; interventricular septum only: n = 4; LV only: n = 4; biventricular: n = 3), whereas none of the athletes did. Nine ARVC patients (25.7%) revealed LV abnormalities (regional WMA with or without delayed gadolinium enhancement), whereas none of the athletes did.
The TWI+ and TWI– athletes did not differ in any SAECG parameters (Table 4). The ARVC cohort exhibited 3 abnormal SAECG parameters more frequently than either of the 2 athlete groups did (TWI+ athletes: 6.7%; TWI– athletes: 0%; ARVC patients: 25.7%; p = 0.019).
Arrhythmias on 12-lead ECG and Holter monitor
Two athletes (2.5%, both TWI–) revealed VE of RV outflow tract (RVOT) morphology on the resting ECG. Five ARVC patients (14.3%) exhibited VE on the ECG (2 patients with a single RVOT VE; 1 patient with 3 RVOT VE; 1 patient with a single LV VE; and 1 patient with 6 biventricular VE). Both patients with LV-origin VE were subsequently diagnosed with biventricular ARVC. One asymptomatic patient with mild biventricular ARVC and LV fibrosis on CMRI exhibited a single RVOT VE on an otherwise normal ECG. The TWI+ and TWI– athletes did not differ in any Holter monitoring parameters (Table 4). The ARVC cohort revealed a greater burden of ventricular ectopic activity compared with TWI+ and TWI– athletes (1,642.0 ± 2,204.2 vs. 25.5 ± 128.6 vs. 45.1 ± 148.0 VE/24 h, respectively). Two athletes exhibited >500 VE/24 h (RVOT morphology). Seven ARVC patients (20.0%) demonstrated sustained or nonsustained ventricular tachycardia on Holter monitoring, but none of the athletes did.
Cardiovascular symptoms were present in 4 ARVC patients (11.4%) during exercise testing (limiting chest pain: n = 1; severe dyspnea: n = 1; pre-syncope: n = 1; palpitations: n = 1), but in none of the athletes (Table 4). Less than one-half of the ARVC patients (42.9%) were able to complete 12 min of the Bruce protocol, whereas all of the athletes were able to. The majority of TWI+ athletes (73.3%) displayed complete T-wave pseudonormalization during exercise, whereas 40.0% of ARVC patients did. A proportion of TWI+ athletes (13.3%) and TWI+ ARVC patients (40.0%) failed to exhibit any degree of T-wave pseudonormalization. Exercise-induced ST-segment depression was observed in 3 ARVC patients (8.6%) and in none of the athletes. Isolated VE during exercise were observed in 8 athletes (10.0%), and the ectopic burden did not increase from the resting state in any case. In contrast, ventricular ectopic activity increased during exercise in 11 ARVC patients (31.4%), with nonsustained ventricular tachycardia in 4 cases (11.4%). None of the athletes exhibited an attenuated systolic blood pressure response to exercise (<20 mm Hg increase), whereas 4 of the ARVC patients (11.4%) did. Exercise-induced hypotension was observed in 1 ARVC patient (2.9%).
Diagnostic overlap between physiological RV remodeling and ARVC
After application of the 2010 TFC to all study subjects, almost all TWI+ athletes fulfilled “possible” (51.1%) or “borderline” (44.5%) criteria. All ARVC patients met “definite” criteria, whereas the TWI– athletes were classified as “normal” (88.2%) or “possible” ARVC (11.8%). Of note, 3 athletes (3.75%) initially fulfilled echocardiographic criteria, although CMRI disproved the presence of RV WMA in these individuals. By definition, all TWI+ athletes fulfilled repolarization criteria (major: 55.6%; minor: 44.4%). The majority of athletes (81.3%) met minor depolarization criteria, primarily due to an abnormal SAECG. Two athletes (2.5%) fulfilled minor arrhythmic criteria due to the presence of >500 VE (of RVOT morphology) in 24 h. One TWI+ athlete was erroneously diagnosed with “definite” ARVC at initial presentation, due to anterior TWI, RV dilation, and apparent RV WMA at echocardiography, although subsequent CMRI revealed normal wall motion. An additional comparison was performed among all athletes (n = 80), phenotypically severe ARVC cases fulfilling “definite” criteria even without considering family history (n = 24), and milder ARVC cases meeting “definite” criteria due primarily to family history (n = 11). The results (Online Table 4) reveal considerable diagnostic overlap between athletic adaptation and ARVC, even after controlling for patients with mild disease. Finally, the following score for differentiating physiological remodeling from ARVC was derived: 1/(1 + e-z), where z = (27.349 + 13.331 [presence of 3 abnormal SAECG parameters] + 9.992 [presence of >500 VE per 24 h] –3.068 [V-Ampmax] + 5.537 [RV basal dimension in apical view/LV end-diastolic dimension in parasternal long axis view >0.9] – 1.768 [exercise duration]). Nagelkerke R square value for the model was 0.918 and Hosmer-Lemeshow test revealed a p value of 0.997. Receiver-operating characteristic analysis for accurate diagnosis of ARVC demonstrated an area under the curve of 0.993 using a cutoff value of >0.45 (95% confidence interval: 0.979 to 1.00 [p < 0.001], sensitivity: 96.3%, specificity: 100%, positive predictive value: 100%, negative predictive value: 98.6%). This model is the basis for the multivariable calculator for the differentiation between physiological remodeling and arrhythmogenic right ventricular cardiomyopathy provided in the Online Appendix.
Comparison between TWI+ and TWI– athletes
Comprehensive testing did not reveal any differences between TWI+ and TWI– athletes in any electrical, structural, or functional cardiac parameters. This finding supports the notion that TWI may represent a benign manifestation of intensive training in the majority of asymptomatic athletes without relevant family history. Furthermore, it serves as a reference point that any differences observed between TWI+ athletes and ARVC patients in the present study may be interpreted as being due to athletic training in the former or cardiac pathology in the latter.
Differentiation between physiological TWI and ARVC
Conventional pathological markers exhibiting high accuracy
The study confirms the diagnostic power of certain established pathological indicators, including syncope in the absence of circumstances (such as pain) clearly leading to reflex-mediated changes in vascular tone or heart rate, exercise-induced cardiovascular symptoms, a high burden of ventricular ectopic activity, and sustained or nonsustained ventricular arrhythmias. Markers of myocardial fibrosis and regional WMA at CMRI also demonstrate potent discriminatory ability and are likely to form the cornerstone of diagnosis in phenotypically borderline cases (Figure 1).
Conventional pathological markers exhibiting poor accuracy
Several conventional pathological markers appear to demonstrate poor accuracy in the athletic setting. Consensus guidelines for ECG interpretation in athletes suggest that TWI in ≥2 consecutive leads, epsilon waves, prolonged QRS terminal activation duration in V1 to V3, reduced limb lead voltages, and VE with left bundle branch block morphology and superior axis should prompt further investigation for ARVC (14). We have recently demonstrated that TWI and voltage RVH should be interpreted with caution in athletes, particularly those of African/Afro-Caribbean origin, due to considerable diagnostic overlap with physiological adaptation (4,15). In the present study, QRS terminal activation duration failed to differentiate between athletic remodeling and ARVC. This phenomenon probably reflects prolonged conduction through a physiologically enlarged RV, further reflected in a high prevalence of partial RBBB and SAECG abnormalities in athletes (4,16,17). The current minor TFC of >500 VE/24 h fails to specify VE morphology, attributing minor criteria to 2 healthy athletes with RVOT ectopy in this study. Exercise-induced T-wave pseudonormalization, or lack thereof, is also a poor discriminator, consistent with recent reports (18). The study also reinforces the nonspecific nature of RV dilation and/or low-normal systolic function without regional WMA (4,19). Finally, although advanced ARVC is usually associated with ECG abnormalities (20), it is noteworthy that one-third of our ARVC cohort exhibited a normal ECG, in some cases with a severe structural phenotype.
Additional pathological markers exhibiting high accuracy
A number of disease markers additional to current TFC were identified in the present study. Pathological Q waves, which are known to correlate with structural phenotype in ARVC (20), were observed exclusively in our disease cohort and should be considered abnormal unless proven otherwise. Almost one-fourth of the ARVC cohort exhibited inferior TWI, which does not feature in current diagnostic criteria. A novel marker assessing precordial rather than limb lead voltages was identified, such that values for V-Ampmax <1.8 mV strongly favor pathology in male subjects. Current recommendations advocate further testing if ≥2 VE are present on the resting ECG (14); however, we observed that a single VE might be the only ECG manifestation of mild RV disease. The current minor criterion of 1 abnormal SAECG parameter proved nonspecific. In contrast, 3 abnormal parameters revealed strong predictive accuracy for ARVC. Novel ratio indicators of asymmetric ventricular dilation (echocardiographic RV basal dimension in apical view/LV end-diastolic dimension in parasternal long axis view >0.9 or CMRI RVEDV/LVEDV >1.2) also demonstrated high specificity for disease. Attenuated blood pressure responses to exercise were seen exclusively in ARVC patients, consistent with findings in hypertrophic cardiomyopathy (21). Finally, it is noteworthy that one-fourth of the ARVC cohort exhibited regional WMA or delayed enhancement in the LV, which, although well reported in the literature, does not form part of current diagnostic criteria (5).
Additional markers suggestive of physiological remodeling
Electrocardiographic markers, such as early repolarization in any territory or biphasic TWI with preceding ST-segment elevation, appear strongly indicative of physiological remodeling. Voltage LVH and RVH were rarely observed in ARVC patients, but they were common in athletes, lending further support to training-induced adaptation. RV dilation with normal regional wall motion and concomitantly increased LV dimensions also suggests physiological remodeling.
Due to the rarity of the disease in youths, the ARVC cohort was small, was not matched to the athletes for ethnicity, and was slightly older than the athlete cohorts. Genetic testing was only performed in ARVC patients; hence, the TFC could not be applied comprehensively to athletes. The ARVC cohort comprised patients who in some cases had advanced disease, and none were professional athletes. Although the enrollment of asymptomatic athletes with ARVC, or athletes reporting cardiovascular symptoms, might yield greater insight into the diagnostic “gray zone,” the present study provides comprehensive phenotypic characterization, which should assist in the differentiation between physiological and pathological remodeling in athletes with incomplete disease expression. Whereas the multivariable score derived from our data demonstrates excellent diagnostic performance, it was derived from studying small numbers of subjects and requires validation in larger cohorts. Finally, the cross-sectional study design precludes the categorical exclusion of future RV pathology in athletes with marked repolarization anomalies, although the high prevalence of TWI in athletic individuals favors physiological remodeling.
Electrocardiographic TWI and balanced biventricular dilation are likely to represent benign manifestations of intensive training in the majority of asymptomatic athletes without relevant family history. Diagnostic criteria for ARVC are nonspecific in such cases and may lead to erroneous diagnoses. Low precordial ECG amplitudes, unbalanced RV dilation, wall motion abnormalities, a high burden of ventricular extrasystoles, and abnormal responses to exercise testing are key markers of pathological remodeling. Comprehensive testing using widely available techniques can effectively differentiate phenotypically borderline cases.
COMPETENCY IN MEDICAL KNOWLEDGE: Healthy, athletic individuals may exhibit electrocardiographic TWI, RV enlargement, and ventricular late potentials. The current TFC for ARVC may, therefore, erroneously suggest a diagnosis in such individuals.
TRANSLATIONAL OUTLOOK: Studies on the basis of CMRI, cardiopulmonary exercise testing, and molecular genetics may improve diagnostic accuracy and clarify the differentiation between athletic adaptation and mild cardiomyopathy.
The authors would like to thank Cardiac Risk in the Young for providing the portable electrocardiography and echocardiography equipment used in this study and Teofila Bueser for her invaluable help with data acquisition and recruitment of study participants.
For supplemental tables and a multivariable calculator, please see the online version of this article.
Drs. Zaidi, Sheikh, Gati, and Papadakis have received research grants from the charitable organization Cardiac Risk in the Young. Dr. Sharma has been a coapplicant on previous grants from Cardiac Risk in the Young to study athletes and nonathletes. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- arrhythmogenic right ventricular cardiomyopathy
- cardiac magnetic resonance imaging
- end-diastolic volume
- left ventricle
- left ventricular hypertrophy
- right bundle branch block
- right ventricle
- right ventricular hypertrophy
- right ventricular outflow tract
- signal-averaged electrocardiography
- sudden cardiac death
- Task Force Criteria
- T-wave inversion
- maximal QRS amplitude in the precordial leads
- ventricular extrasystole(s)
- wall motion abnormality
- Received February 11, 2015.
- Revision received March 16, 2015.
- Accepted April 20, 2015.
- American College of Cardiology Foundation
- Pelliccia A.,
- Maron B.J.,
- De Luca R.,
- et al.
- Corrado D.,
- Basso C.,
- Rizzoli G.,
- et al.
- Zaidi A.,
- Ghani S.,
- Sharma R.,
- et al.
- Marcus F.I.,
- McKenna W.J.,
- Sherrill D.,
- et al.
- Mitchell J.H.,
- Haskell W.,
- Snell P.,
- Van Camp S.P.
- Marcus F.I.,
- Edson S.,
- Towbin J.A.
- Drezner J.A.,
- Ashley E.,
- Baggish A.L.,
- et al.
- Zaidi A.,
- Ghani S.,
- Sheikh N.,
- et al.
- Luijkx T.,
- Velthuis B.K.,
- Prakken N.H.,
- et al.
- Frenneaux M.P.,
- Counihan P.J.,
- Caforio A.L.,
- Chikamori T.,
- McKenna W.J.