Preparticipation screening of athletes with 12-lead electrocardiography has been promoted for the detection of asymptomatic cardiovascular disease, particularly hypertrophic cardiomyopathy (HC). Although false-positive electrocardiographic (ECG) results for HC are well recognized in athlete screening, expected false-negative rates are unknown. The aim of this study was to characterize the rate of false-negative ECG findings in a cohort of young asymptomatic patients with phenotypically expressed HC, defined by cardiovascular magnetic resonance, using the 2010 European Society of Cardiology recommended ECG criteria for the identification of suspected heart disease in trained athletes. Cardiac magnetic resonance studies and 12-lead electrocardiography were performed in 114 consecutive asymptomatic patients with HC aged ≤35 years (mean age 22 ± 8 years; 77% male patients). Electrocardiograms were analyzed to distinguish pathologic ECG patterns from alterations considered nonpathologic and physiologic consequences of athletic training. Among the 114 patients with HC, 103 (90%) demonstrated ≥1 pathologic ECG abnormality, while the remaining 11 patients (10%) had normal or nonpathologic ECG patterns and therefore defined a subgroup in whom ECG screening would not be expected to raise suspicion of heart disease (i.e., false-negative results). In this false-negative ECG results group, maximal left ventricular wall thickness was 17 ± 2 mm (range 15 to 21), compared to patients with pathologic ECG patterns, in whom maximal left ventricular wall thickness was 22 ± 5 mm (p = 0.003). In conclusion, a substantial minority of young asymptomatic patients with HC with phenotypically expressed left ventricular hypertrophy have nonpathologic ECG findings on the basis of the 2010 European Society of Cardiology guidelines. In principle, this high false-negative rate of 10% represents an important limitation in applying 12-lead electrocardiography to large, apparently healthy athletic populations for the detection of HC.
Sudden death in trained athletes has generated public health debate regarding the most efficacious screening strategy for the detection of cardiovascular disease. The routine addition of 12-lead electrocardiography to history and physical examination has been proposed by the European Society of Cardiology (ESC), on the basis of a long-standing mandatory program for the systematic screening of competitive sports programs, in particular to detect or raise suspicion for hypertrophic cardiomyopathy (HC), the most common cause of sudden death in young United States athletes. Although abnormal results on screening electrocardiography can trigger an unsuspected diagnosis of HC, a number of competing issues have been raised with respect to applying electrocardiography as a diagnostic tool in the setting of large, apparently healthy athlete populations in the United States. Prominent among these is the variable but generally high proportion of false-positive test results. The potential of electrocardiographic (ECG) screening for HC to generate false-negative results has been underrecognized or ignored. Indeed, in a very recent 2-month period in 2012, 5 elite athletes have had sudden death events during training or competition, all apparently after preparticipation screening. Therefore, it is particularly timely to critically investigate the HC phenotype (using cardiovascular magnetic resonance [CMR] imaging) with respect to the ECG criteria established by the 2010 ESC recommendations for trained athletes to characterize the expected false-negative rate of this test in such a population.
The study population comprised 114 consecutive, asymptomatic, and unrelated patients with HC aged ≤35 years evaluated using CMR from January 2004 to December 2010. All patients underwent evaluation with echocardiography and CMR, although ECG findings were compared only to morphologic data derived by CMR. The diagnosis of HC was based on a hypertrophied and nondilated left ventricle with a wall thickness ≥15 mm, in the absence of another cardiac or systemic disease that could produce the magnitude of hypertrophy evident. This study was approved by the internal review boards of the participating institutions.
The 12-lead electrocardiogram obtained at or near the time of CMR (interval 20 ± 70 days) was acquired during quiet respiration and recorded at a paper speed of 25 mm/s and an amplitude of 10 mm/mV. All electrocardiograms were assessed by 2 observers (E.J.R. and M.S.M.), blinded to CMR results. Disagreements between observers were resolved by consensus. ECG results were compared to the most contemporary standards, the 2010 ESC guidelines for the interpretation of 12-lead electrocardiograms in athletes and divided into 2 groups as defined in Table 1 : (1) pathologic and (2) nonpathologic patterns attributed to systematic athletic conditioning or patterns regarded to be within normal limits. For the purpose of this analysis, these latter ECG results were considered “false negative” when seen in patients with CMR evidence of HC. CMR examinations were performed using commercially available scanners (Philips ACS-NT 1.5 T Gyroscan-Intera, Philips Medical Systems, Best, The Netherlands; or Siemens Sonata 1.5 T, Siemens Healthcare, Erlangen, Germany) with commercial surface cardiac coils. Left ventricular (LV) end-diastolic and end-systolic volumes, mass, and ejection fraction were determined as previously described by experienced investigators using commercial software (Easy Vision 5.0 and View Forum, Philips Medical Systems; or Argus, Siemens Healthcare), blinded to the ECG results. LV mass was normalized for body surface area to calculate the LV mass index. Maximum end-diastolic LV wall thickness was taken as the dimension of greatest magnitude at any site within the LV wall. Late gadolinium enhancement CMR was performed 15 minutes after the intravenous administration of 0.2 mmol/kg gadolinium diethylenetriamine penta-acetic acid (Magnevist; Schering AG, Berlin, Germany) with a breath-held segmented inversion-recovery sequence (inversion time 240 to 300 ms), acquired in the same views as the cine images. All tomographic short-axis LV slices from base to apex were inspected visually to identify an area of completely nulled myocardium. Mean signal intensity (and SD) of normal myocardium was calculated, and a threshold ≥6 SDs exceeding the mean was used to define areas of late gadolinium enhancement.
|Group 1: Nonpathologic Alterations||Group 2: Pathologic Abnormalities|
|Sinus bradycardia||Nonvoltage criteria for LV hypertrophy §|
|First-degree atrioventricular block||T-wave inversions ∥|
|Incomplete right bundle branch block ⁎||ST-segment depression ¶|
|Early repolarization †||Pathologic Q waves #|
|Isolated QRS voltage criteria for LV hypertrophy ‡||Left or right atrial enlargement ⁎⁎|
|Left-axis deviation/left anterior hemiblock|
|Right-axis deviation/left posterior hemiblock|
|Right ventricular hypertrophy ††|
|Intraventricular conduction abnormalities ‡‡|
|Long- or short-QT interval §§|
|Brugada-like early repolarization|
‡ Sokolow-Lyon (sum of S-wave voltage in lead V 1 and R-wave voltage in lead V 5 or V 6 ≥35 mm) and Cornell voltage score (sum of S-wave in lead V 3 and R wave in lead aVL ≥20 mm for female subjects and ≥28 mm for male subjects).
Data are expressed as mean ± SD. Categorical data are expressed as counts and percentages. Categorical and binomial data were compared using Fisher’s exact test. Continuous variables were tested for normality before analysis. Continuous, normally distributed data were compared using an unpaired Student’s t test or analysis of variance with Bonferroni’s correction for multiple comparisons. The relation between continuous, normally distributed data were evaluated using standard correlation and linear regression. Two-sided p values <0.05 were used to determine statistical significance. All analyses were performed using SAS for Windows version 9.1 (SAS Institute Inc., Cary, North Carolina).
Demographic, clinical, and morphologic characteristics of the 114 patients with HC are listed in Table 2 . The mean age at evaluation was 22 ± 8 years; 88 patients (77%) were male; and 107 (93.8%) were white, 3 (2.6%) African American, 2 (1.8%) Asian, and 2 (1.8%) Hispanic. Maximal LV wall thickness was 21 ± 5 mm (range 15 to 37), and 17 patients (15%) had LV outflow tract obstructions ≥30 mm Hg at rest. Male and female patients did not differ significantly with respect to age (p = 0.55), maximal LV wall thickness (p = 0.13), or the presence of late gadolinium enhancement (p = 0.98).
|Parameter||Patients With Nonpathologic ECG Alterations||Patients With Pathologic ECG Abnormalities||p Value|
|(n = 11)||(n = 103)|
|Age (years)||22 ± 8||22 ± 8||0.99|
|Male||10 (91%)||78 (76%)||0.26|
|Maximal LV wall thickness (mm)||17 ± 2||22 ± 5||0.003|
|Maximal LV wall thickness ≥30 mm||0||6 (6%)||0.34|
|Late gadolinium enhancement present ⁎||0 (0%)||46 (45%)||0.004|
|LV mass (g)||144 ± 46||183 ± 79||0.10|
|LV mass index (g/m 2 )||70 ± 17||97 ± 33||0.01|
|End-diastolic volume (ml)||163 ± 36||165 ± 47||0.90|
|Left atrial dimension (mm)||36 ± 5||38 ± 6||0.39|
|LV outflow tract obstruction at rest (≥30 mm)||2 (18%)||15 (15%)||0.81|
|Ejection fraction (%)||71 ± 7||70 ± 6||0.49|
|Sinus bradycardia||1 (9%)||—|
|Atrioventricular block||0 (0%)||—|
|Cornell voltage score †||1 (9%)||—|
|Sokolow-Lyon index ‡||1 (9%)||—|
|Incomplete right bundle branch block||0 (0%)||—|
|Normal ECG results||9 (82%)||—|
|Nonvoltage criteria for LV hypertrophy §||—||62 (60%)|
|Pathologic Q waves ∥||—||49 (48%)|
|ST-segment depression or elevation||—||28 (27%)|
|Right or left axis deviation||—||20 (19%)|
|Right atrial enlargement or right ventricular hypertrophy||—||21 (20%)|
|Left atrial enlargement||—||22 (21%)|
|T-wave inversions||—||64 (62%)|
|Intraventricular conduction abnormalities||—||13 (13%)|
|Ventricular preexcitation||—||5 (5%)|
|Long-QT interval ¶||—||6 (6%)|
|Short-QT interval #||—||4 (4%)|
|Brugada-like ECG abnormalities||—||1 (1%)|