Identification of hypertrophic cardiomyopathy (HC) in young athletes is challenging when left ventricular (LV) wall thickness is between 13 and 15 mm. The aim of this study was to revise the ability of simple echocardiographic and clinical variables for the differential diagnosis of HC versus athlete’s heart. Twenty-eight athletes free of cardiovascular disease were compared with 25 untrained patients with HC, matched for LV wall thickness (13 to 15 mm), age, and gender. Clinical, electrocardiographic, and echocardiographic variables were compared. Athletes had larger LV cavities (60 ± 3 vs 45 ± 5 mm, p <0.001), aortic roots (34 ± 3 vs 30 ± 3 mm, p <0.001), and left atria (42 ± 4 vs 33 ± 5 mm, p <0.001) than patients with HC. LV cavity <54 mm distinguished HC from athlete’s heart with the highest sensitivity and specificity (both 100%, p <0.001). Left atrium >40 mm excluded HC with sensitivity of 92% and specificity of 71% (p <0.001). Athletes showed higher e′ velocity by tissue Doppler imaging than patients with HC (12.5 ± 1.9 vs 9.3 ± 2.3 cm/second, p <0.001), with values <11.5 cm/second yielding sensitivity of 81% and specificity of 61% for the diagnosis of HC (p <0.001). Absence of diffuse T-wave inversion on electrocardiography (specificity 92%) and negative family history for HC (specificity 100%) also proved useful for excluding HC. In conclusion, in athletes with LV hypertrophy in the “gray zone” with HC, LV cavity size appears the most reliable criterion to help in diagnosis, with a cut-off value of <54 mm useful for differentiation from athlete’s heart. Other criteria, including LV diastolic dysfunction, absence of T-wave inversion on electrocardiography, and negative family history, further aid in the differential diagnosis.
Highlights
- •
We studied athletes and patients with HC with borderline LV hypertrophy (13 to 15 mm).
- •
Clinical and echocardiographic parameters were compared.
- •
LV cavity <54 mm distinguished HC from athlete’s heart with the highest sensitivity and specificity.
- •
Athletes were characterized by larger left atria compared with patients with HC.
- •
Also, diastolic functional parameters, electrocardiography, and family history proved useful in the differential diagnosis.
Intensive athletic training is associated with a spectrum of morphologic and functional cardiac changes (i.e., athlete’s heart), considered to be physiologic adaptations to increased hemodynamic load and neurohormonal changes. In most athletes, morphologic cardiac changes are mild and do not raise clinical concern, but in some highly trained athletes, left ventricular (LV) remodeling may be substantial, prompting differential diagnosis with structural heart disease, most commonly hypertrophic cardiomyopathy (HC). Indeed, accurate identification of HC in athletes has relevant clinical implications, because this disease is one of the most common causes of athletic field deaths and usually represents the basis for disqualification from competitive or professional sports. Distinguishing athlete’s heart from HC may prove particularly difficult when absolute LV wall thickness is in the range of 13 to 15 mm, which identifies the “gray zone” of overlap between these 2 clinical entities. Although previous studies have proved useful in aiding this differential diagnosis, at present, reliable identification of HC continues to be challenging in athletes with such ambiguous morphology. Therefore, the aim of the present study was to revise the ability of simple echocardiographic and clinical variables for the differential diagnosis of HC versus athlete’s heart.
Methods
From January 2008 to June 2009, 1,191 consecutive highly trained athletes were evaluated at the Institute of Sport Medicine and Science in Rome, as potential participants in the 2008 Beijing Olympic Games and/or the 2009 Pescara Pan-Mediterranean Games. Of these, 28 athletes (2.3%) were selected for the present study, on the basis of age 18 to 40 years and the echocardiographic finding of absolute LV wall thickness of 13 to 15 mm, which has been defined as the gray zone of overlap of physiologic LV hypertrophy and HC. Diagnosis of physiologic LV hypertrophy was based on the presence of mild LV wall thickening, judged consistent with the intensity and duration of sport discipline participated (as previously described), associated with normal systolic and diastolic function and in the absence of systolic anterior motion of the mitral valve and LV outflow tract obstruction, and negative family history for the disease. The athletes were asymptomatic male Caucasians aged 26 ± 4 years; each had been competing at the national or international level for ≥3 years, participating in rowing or canoeing (n = 11), cycling (n = 6), swimming (n = 4), water polo (n = 2), judo (n = 2), basketball (n = 1), wrestling (n = 1), and hammer throwing (n = 1). According to our medical program, cardiovascular evaluation included personal and family history, physical examination, 12-lead and exercise electrocardiography, and 2-dimensional Doppler echocardiography.
For comparison, a group of 25 patients with nonobstructive HC, matched for age (range 18 to 40 years) and gender (all men) from the HC center at Tufts University Medical Center (Boston, Massachusetts) was selected for this analysis, on the basis of the presence of mild LV hypertrophy (i.e., wall thickness 13 to 15 mm), in the absence of systolic anterior motion of the mitral valve and LV outflow tract obstruction. They were Caucasian (n = 23 [92%]) or African American (n = 2 [8%]), and diagnosis of HC was based on the 2-dimensional echocardiographic (and/or cardiac magnetic resonance) demonstration of a hypertrophied, nondilated left ventricle in the absence of any cardiac or systemic disease that should lead to LV hypertrophy of the extent evident. None of the patients with HC had evidence of systemic hypertension (i.e., blood pressure >140/90 mm Hg), and none was engaged in competitive sports or systematic exercise training programs (i.e., <3 hours of exercise per week). Diagnosis of HC was supported, indeed, by a positive family history of HC (n = 11 [44%]) and/or identification of pathogenic sarcomere protein mutation (n = 4 [16%]).
Athletes and patients with HC were followed up for the subsequent 4 years after baseline evaluation, to assess the incidence of cardiac events, symptoms, or echocardiographic evidence of HC. The requirement for written informed consent was waived for all subjects, and the study design was approved by the local ethics committees.
Echocardiographic examinations were obtained at the 2 institutions by using Philips iE33 machines (Philips Medical Systems, Andover, Massachusetts) equipped with S3 probes (2 to 4 MHz). All acquisitions and measurements were performed by expert cardiologists (S.C., A.P., M.S.M., and N.G.P.), who were aware of patients’ histories and clinical findings. The imaging protocol was defined at Tufts University Medical Center, where the Italian investigators (S.C. and A.P.) were trained in echocardiography. Specifically, 2-dimensional assessment of LV cavity diameters, wall thickness, the left atrium, and the aortic root was performed according to European Association of Cardiovascular Imaging and American Society of Echocardiography criteria. The LV ejection fraction was calculated by the biplane Simpson’s rule.
To assess LV diastolic function, early (E) and late (A) pulsed-wave Doppler diastolic peak-flow velocities were measured in the apical 4-chamber view, with the sample volume placed at the tip of the mitral leaflets; E-wave deceleration time and isovolumic relaxation time were measured in a standard fashion. Tissue Doppler imaging (TDI) measurements of mitral annular motion were achieved in the apical 4-chamber view, with a 10-ml sample volume placed at the septal border of the mitral annulus. Early (e′) and late (a′) diastolic peak velocities and their ratio were recorded.
Standard 12-lead electrocardiography was performed with the subject in the supine position and recorded at 10 mV and 25 mm/second. Analysis of the electrocardiographic tracings was performed according to widely used criteria. Standard treadmill or bicycle exercise testing was performed in athletes and patients with HC under continuous 12-lead electrocardiographic and blood pressure monitoring. Twenty-four-hour Holter monitoring was selectively performed in athletes and patients in whom ≥3 premature ventricular beats were present at baseline or during exercise electrocardiography.
Continuous data are expressed as mean ± SD. Categorical data are expressed as frequencies. Statistical significance was set for a 2-tailed p value <0.05. Differences between groups in terms of continuous variables were calculated by means of unpaired-samples Student’s t tests. Differences between proportions were calculated by chi-square tests. Receiver-operating characteristic curve analysis was used to test the sensitivity and specificity of those variables that showed significant differences on unpaired Student’s t tests. Sensitivity and specificity were reported when the p value was <0.05. Data were analyzed by using PASW Statistics version 18 (SPSS, Inc., Chicago, Illinois).
Results
Comparative echocardiographic and Doppler LV findings in athletes and patients with HC are listed in Table 1 . The distribution of LV hypertrophy was different in the 2 groups, in that anterior ventricular septum was thicker in patients with HC compared with athletes, whereas the posterior ventricular septum, posterior free wall, and anterolateral wall were thicker in athletes.
| Variable | Athletes (n = 28) | HC (n = 25) | p-Value |
|---|---|---|---|
| Anterior septum (mm) | 12.5 ± 0.6 | 13.8 ± 1.5 | <0.001 |
| Posterior septum (mm) | 13.1 ± 0.4 | 12.0 ± 1.7 | 0.002 |
| LV Posterior free wall (mm) | 11.7 ± 0.7 | 9.3 ± 1.5 | <0.001 |
| LV Lateral free wall (mm) | 11.3 ± 0.8 | 8.7 ± 1.1 | <0.001 |
| LV end-diastolic diameter (mm) | 60 ± 3 | 45 ± 4 | <0.001 |
| LV end-systolic diameter (mm) | 37 ± 5 | 24 ± 4 | <0.001 |
| Relative wall thickness ∗ | 0.42 ± 0.03 | 0.62 ± 0.11 | <0.001 |
| Left atrium diameter (mm) | 42 ± 4 | 34 ± 5 | <0.001 |
| Aortic root (mm) | 34 ± 3 | 30 ± 3 | <0.001 |
| Ejection fraction (%) | 63 ± 5 | 64 ± 6 | 0.488 |
| E wave (cm/s) | 82 ± 18 | 80 ± 21 | 0.767 |
| A wave (cm/s) | 44 ± 8 | 57 ± 18 | 0.001 |
| E/A ratio | 1.9 ± 0.5 | 1.6 ± 0.6 | 0.032 |
| Deceleration time (ms) | 207 ± 53 | 175 ± 30 | 0.011 |
| IVRT (ms) | 97 ± 16 | 88 ± 13 | 0.135 |
| TDI e′ wave (cm/s) | 12.5 ± 1.9 | 9.1 ± 2.4 | <0.001 |
| TDI a′ wave (cm/s) | 7.3 ± 1.6 | 6.7 ± 2.5 | 0.452 |
| TDI e′/a′ ratio | 1.77 ± 0.40 | 1.55 ± 0.70 | 0.170 |
| TDI s wave (cm/s) | 9.1 ± 1.8 | 8.2 ± 1.4 | 0.087 |
| E/e′ ratio | 6.6 ± 1.2 | 9.2 ± 2.5 | <0.001 |
∗ Relative wall thickness = ratio of the septal and free wall thickness normalized to LV cavity.
The LV cavity (end-diastolic and end-systolic) was substantially larger in athletes compared with patients with HC; therefore, relative wall thickness (i.e., the ratio between LV wall thickness and cavity size) was lower in athletes. On individual analysis, none of the athletes showed an absolute LV end-diastolic diameter <55 mm, while most athletes (n = 16 [67%]) exceeded 60 mm. None of the patients with HC exceeded an absolute LV cavity dimension of 53 mm, and most (n = 20 [88%]) were <50 mm ( Figures 1 and 2 ).
Left atrial (LA) transverse diameter was larger in athletes than in patients with HC ( Figure 3 ). Specifically, 7 athletes (25%), but none of the patients with HC, showed LA dimensions >45 mm. The aortic root was larger in athletes, although absolute values remained within normal limits (i.e., <40 mm) in all subjects. Ejection fractions were similar in athletes and patients with HC (63 ± 5% vs 64 ± 6%, p = 0.48) and none had values <50% or showed segmental wall motion abnormalities.
Pulsed Doppler-derived parameter are reported in Table 1 . Athletes showed similar E-wave velocity but lower transmitral A-wave velocity compared with patients with HC; consequently, the E/A ratio was higher in athletes. Additionally, E-wave deceleration time was shorter in HC. Doppler waveform patterns were normal in all athletes, but abnormal filling was found in 7 patients with HC (28%, p = 0.003). With TDI, athletes showed increased e′ velocity but similar a′ velocity compared with patients with HC, with greater e′/a′ ratio; finally, the E/e′ ratio was lower in athletes compared with patients with HC.
Clinical and electrocardiographic findings are listed in Table 2 . None of the athletes (by selection criteria) had a positive family history of sudden cardiac death or cardiomyopathy; conversely, of the 11 patients with HC (44%) with positive family histories, diagnosis of HC (either nonobstructive or obstructive) was confirmed in ≥1 first-degree relative, and in 4 patients (16%), pathogenic sarcomere protein mutations were identified (i.e., MYBPC3 772 G>A p.Glu258 Lys, MYBPC3 927-9G>A, MYBPC3 E21V1, and MYH7 Arg 845 Gly).
| Variable | Athletes (n = 28) | HC (n = 25) | p-Value |
|---|---|---|---|
| Age (years) | 26 ± 4 | 28 ± 10 | 0.247 |
| Body Surface Area (m 2 ) | 2.12 ± 0.27 | 2.05 ± 0.21 | 0.335 |
| Systolic Blood Pressure (mmHg) | 128 ± 10 | 120 ± 13 | 0.015 |
| Diastolic Blood Pressure (mmHg) | 79 ± 7 | 74 ± 9 | 0.016 |
| Heart rate (bpm) | 52 ± 10 | 64 ± 10 | <0.001 |
| Family history for HC | 0 | 11 (44%) | <0.001 |
| Sokolow-Lyon score (mm) | 38 ± 13 | 31 ± 14 | 0.100 |
| Left axis deviation | 1 (4%) | 0 | 0.340 |
| Left atrial enlargement | 5 (18%) | 4 (16%) | 0.857 |
| Q waves | 0 | 3 (12%) | 0.059 |
| T-wave abnormalities | 2 (7%) | 13 (52%) | <0.001 |
| Ventricular premature complexes | 2 (7%) | 2 (8%) | 0.552 |
Athletes were allowed to continue training and competition, undergoing periodical cardiovascular evaluation. Periodical ambulatory evaluation was planned for patients with HC. During the subsequent 4-year follow-up, no athlete or patient with HC incurred cardiac symptoms or a clinical event.
No differences were identified between athletes and patients with HC in Sokolow-Lyon index ( Table 2 ). Diffusely inverted T waves were common in patients with HC (n = 13 [52%]) and unusual in athletes (n = 2 [7%], p <0.001); Q waves were absent in athletes and present in 3 patients with HC (p = 0.059). Prevalence of left-axis deviation (1 athlete, no patients with HC; p = 0.34) and LA enlargement (5 athletes, 4 patients with HC; p = 0.85) did not differ between the 2 groups. The 2 athletes with inverted T waves underwent cardiac magnetic resonance with gadolinium contrast that confirmed the absence of structural cardiac abnormalities. No arrhythmias were recorded on exercise testing and 24-hour electrocardiographic monitoring in these 2 athletes.
In addition, 2 other athletes presented either frequent (>10,000/24 hours) focal premature ventricular beats or a single burst of nonsustained ventricular tachycardia on 24-hour Holter monitoring. In these 2 athletes, cardiac magnetic resonance excluded an underlying cardiac pathologic substrate, and they underwent successful radiofrequency ablation for right ventricular outflow tract tachycardia, with no recurrence over the 4-year follow-up period.
In the HC group, 2 patients showed 1 short run of nonsustained ventricular tachycardia (bursts of 8 beats with an RR interval 360 ms and 6 beats with an RR interval of 440 ms, respectively), which was regarded as not requiring antiarrhythmic treatment, but continued periodical follow-up.
LV end-diastolic dimension was the strongest morphologic criterion differentiating athlete’s heart from HC in our cohort, with the threshold value of ≥55 mm the most reliable cut point ( Table 3 ). LA enlargement >40 mm also showed high sensitivity (92%) and specificity (71%). Of the diastolic parameters, TDI e′ velocity <11.5 cm/second identified patients with HC with the highest accuracy (sensitivity 81%, specificity 61%). Diffuse negative T waves on electrocardiographyCG had lower sensitivity for HC (52%), but the absence of T-wave inversion excluded HC with high specificity (92%). Figure 4 depicts the flowchart with morphologic, electrocardiographic, and clinical criteria described for differentiation of athlete’s heart from HC.
