In hypertrophic cardiomyopathy (HC), electrocardiographic (ECG) changes have been postulated to be an early marker of disease, detectable in sarcomere mutation carriers when left ventricular (LV) wall thickness is still normal. However, the ECG features of mutation carriers have not been fully characterized. Therefore, we systematically analyzed ECGs in a genotyped HC population to characterize ECG findings in mutation carriers (G+) with and without echocardiographic LV hypertrophy (LVH), and to evaluate the accuracy of ECG findings to differentiate at-risk mutation carriers from genetically unaffected relatives during family screening. The ECG and echocardiographic findings were analyzed from 213 genotyped subjects (76 G+/LVH−, 57 G+/LVH+ overt HC, 80 genetically unaffected controls). Cardiac magnetic resonance imaging was available on a subset. Q waves and repolarization abnormalities (QST) were highly specific (98% specificity) markers for LVH− mutation carriers, present in 25% of G+/LVH− subjects, and 3% of controls (p <0.001). QST ECG abnormalities remained independently predictive of carrying a sarcomere mutation after adjusting for age and impaired relaxation, another distinguishing feature of G+/LVH− subjects (odds ratio 8.4, p = 0.007). Myocardial scar or perfusion abnormalities were not detected on cardiac magnetic resonance imaging in G+/LVH− subjects, irrespective of the ECG features. In overt HC, 75% had Q waves and/or repolarization changes, but <25% demonstrated common isolated voltage criteria for LVH. In conclusion, Q waves and repolarization abnormalities are the most discriminating ECG features of sarcomere mutation carriers with and without LVH. However, owing to the limited sensitivity of ECG and echocardiographic screening, genetic testing is required to definitively identify at-risk family members.
Hypertrophic cardiomyopathy (HC) is the most common monogenic cardiovascular disorder. Sarcomere mutations can be identified in approximately 60% of affected patients. Conventionally, a clinical diagnosis of HC is made by identifying unexplained left ventricular hypertrophy (LVH) by cardiac imaging. However, LVH typically does not develop until adolescence or later in life. The delayed penetrance of this key diagnostic feature makes it difficult to identify family members who are at risk of developing HC early in life, when left ventricular (LV) wall thickness is normal. In contrast, genetic testing allows definitive identification of such individuals relatives who have inherited a pathogenic sarcomere mutation (G+) but do not yet have the diagnostic clinical features of HC (LVH−; denoted G+/LVH−). However, genetic testing may not always be feasible or the results might be ambiguous. As such, more precise characterization of the clinical features of G+/LVH− subjects is of great value.
Electrocardiographic (ECG) abnormalities are common in patients with overt HC and have been postulated to be an earlier and more sensitive manifestation of sarcomere mutations than increased LV wall thickness. Previous systematic investigation of electrocardiograms in genotyped HC populations has been limited. Therefore, we analyzed the electrocardiograms of well-characterized G+/LVH−, G+/LVH+ overt HC, and healthy G− relatives (as normal controls) to both better define the ECG features associated with sarcomere mutations and to assess the ability of such ECG findings to predict the genetic status of apparently healthy family members with normal LV wall thickness. By studying genotyped populations, the early manifestations of sarcomere mutations can be better characterized, improving our understanding of how mutations cause disease.
Methods
Genotyped HC probands and relatives identified by research protocols or through clinical evaluation were studied. Research subjects were recruited from 3 institutions: Brigham and Women’s Hospital (n = 106), the Minneapolis Heart Institute Foundation (n = 57), and Copenhagen University Hospital (n = 50). Genetic status was previously determined in the family proband by direct DNA sequencing of 8 sarcomere genes, including myosin-binding protein C ( MYBPC3 ), β-myosin heavy chain ( MYH7 ), cardiac troponin T ( TNNT2 ), cardiac troponin I ( TNNI3 ), α-tropomyosin ( TPM1 ); actin ( ACTC1 ), myosin regulatory light chain ( MYL2 ), and myosin essential light chain ( MYL3 ) using standard methods. Mutation confirmation testing was performed on family members to determine whether the family-specific mutation was present or absent.
The subjects were assigned to 3 different status groups according to genotype and echocardiographic wall thickness: G+/LVH− subjects, overt HC (G+/LVH+), and related normal controls (G−/LVH−). The G+/LVH− group consisted of mutation carriers without echocardiographic LVH (maximal LV wall thickness <12 mm). This strict definition for G+/LVH− was chosen to avoid including subjects with borderline LVH and emerging or mild clinical disease as G+/LVH− subjects. The overt HC group consisted of patients with a sarcomere gene mutation and diagnostic clinical features of HC, defined by a maximal echocardiographic LV wall thickness of ≥12 mm in adults or z score ≥2 in children. The control subjects were younger, healthy family members who did not carry the family-specific sarcomere mutation.
All subjects were assessed by history, physical examination, electrocardiography, and echocardiography. Exclusion criteria included poor echocardiographic images, systemic hypertension (taking medication or blood pressure ≥140/90 mm Hg), coronary artery disease, valvular heart disease, previous septal myectomy or alcohol septal ablation, or ventricular pacing. No subjects were taking cardiac glycosides. All participants provided informed consent using protocols approved by the institutional review board of Brigham and Women’s Hospital and the Local Science Ethics Committee, Copenhagen, Denmark.
Each subject underwent echocardiography with M-mode, 2-dimensional, and Doppler imaging according to standard clinical practice. The images were stored digitally and analyzed off-line by 2 investigators blinded to clinical and genetic status. The standard measures of cardiac dimensions were made using the mean of 3 cardiac cycles in accordance with the guidelines of the American Society of Echocardiography. Early myocardial tissue Doppler relaxation velocities (Ea) were measured at the lateral, septal, anterior, and inferior aspects of the mitral annulus. The mean Ea velocity of these 4 measurements was used to represent the global Ea velocity. Statistical analyses were performed using global Ea.
Standard 12-lead electrocardiograms were obtained at the time of echocardiographic examination with subjects resting quietly in the supine position. All electrocardiograms were analyzed by a single investigator who blinded to clinical, genetic, and echocardiographic information. The heart rate, QRS axis, intrinsicoid deflection, and QRS and QT intervals were measured according to the standard criteria. The QT interval was corrected for the heart rate according to Bazett’s formula (QTc = QT/√RR). The criteria for defining ECG abnormalities are summarized in Table 1 . Q waves in ≥2 contiguous leads were considered abnormal, if they were of sufficient amplitude (>1/3 the height of the R wave) or duration (>30 ms). A more stringent definition of pathologic Q waves required both sufficient duration and amplitude.
| Variable | Description |
|---|---|
| Q waves |
|
| Pathologic Q waves |
|
| Repolarization abnormalities | |
| T-wave inversion | ≥0.1 mV in ≥2 contiguous leads |
| ST-segment depression | Upsloping: ≥0.1 mV in ≥2 contiguous leads; downsloping or flat: ≥0.05 mV in ≥2 contiguous leads |
| Left ventricular hypertrophy | |
| Total voltage | Sum of greatest positive and negative QRS deflection in 12 standard leads >17.5 mV |
| Romhilt Estes ⁎ | Score ≥4 |
| Sokolow-Lyon | S V1 + R V5 or R V6 ≥3.5 mV |
| Cornell | S V3 + R aVL ≥2.1 mV (women) or 2.9 mV (men) |
| Composite abnormalities | |
| Major abnormality | Q waves or repolarization abnormality (excluding nonspecific abnormalities), or any LVH criteria |
| Minor abnormality † | Left atrial abnormality ‡ or prolonged interventricular conduction § or nonspecific ST-segment or T-wave abnormalities |
| QST | Q waves or T wave inversion or ST-segment depression |
⁎ Romhilt-Estes score assigns points for presence of increased voltage (3 points), repolarization abnormalities (3 points), intrinsicoid deflection >50 ms (1 point), QRS duration >90 ms (1 point), QRS axis left of −15° (2 points), and left atrial abnormality (3 points).
† In absence of any major abnormality.
‡ P-wave inversion in lead V 1 ≥40 ms in duration and 1 mV in depth.
T-wave inversion was considered if present in leads with R greater than S wave amplitude. T wave inversions in leads V 1 to V 3 were not considered pathologic in subjects <18 years old. ST-segment depression was considered pathologic if present in ≥2 contiguous leads with depth >0.1 mV if upsloping or >0.05 mV if horizontal or downsloping. Nonspecific ST or T-wave abnormalities were considered present if there were abnormalities in the ST-segment or T waves that did not meet the criteria for T-wave inversion or ST-segment depression. LVH was defined by standard criteria, including Romhilt-Estes, Cornell, Sokolow-Lyon, and total voltage, as described in Table 1 . Major ECG abnormalities were considered present if diagnostic ST-segment depression, T-wave inversion, Q waves, or any of the LVH criteria were identified. The minor ECG abnormalities included NSSTW, left atrial abnormality or delayed interventricular conduction (QRS >110 ms), in the absence of major abnormalities.
Cardiac magnetic resonance (CMR) imaging was available in an unselected subset of subjects with sarcomere mutations who had studies performed at Brigham and Women’s Hospital in the context of clinical evaluation or participation in other research studies examining G+ subjects. The images were analyzed by 2 investigators who were unaware of the ECG characteristics and genotype. All patients were studied in either a 1.5-T (General Electric) or 3.0-T (Siemens) CMR system. Myocardial perfusion at rest was assessed by the standard first-pass saturation-prepared T 1 -weighted fast gradient echo during injection of a gadolinium contrast bolus (Magnevist, Bayer, Wayne, New Jersey). The CMR protocol included late gadolinium enhancement imaging to evaluate for myocardial scar. The segmented inversion-recovery pulse sequence for late gadolinium enhancement imaging was used, starting 15 minutes after a cumulative 0.15-mmol/kg dose of gadolinium diethylenetriaminepentaacetic acid, as previously described. The images were analyzed off-line using QMass (Medis, Leiden, The Netherlands), and late gadolinium enhancement imaging was defined by any region of signal intensity >2 SDs above the mean signal intensity of the normal myocardium. On statistical analysis, late gadolinium enhancement imaging was treated as a dichotomous variable, present or absent, for each subject.
The clinical, echocardiographic, ECG, and CMR imaging characteristics are presented for each of the 3 status groups as the frequency and percentage or mean values and SDs. Statistical analysis was performed using SAS, version 9.1 (SAS Institute, Cary, North Carolina). To test for differences among the 3 groups and adjust for the influence of relations between family members, the GenMod procedure was used to perform analysis of variance with clustering, assuming an exchangeable correlation structure. A p value <0.017 was considered statistically significant to apply a post hoc Bonferroni correction for multiple comparisons across the 3 status groups. Logistic regression analysis using the GenMod procedure was performed to evaluate multiple predictors simultaneously, adjust for confounders, and adjust for the influence of age on the penetrance of LVH and effect on Ea velocity. p Values <0.05 were considered statistically significant in comparing the G+/LVH− and control cohorts.
Results
A total of 213 subjects from 60 different families were evaluated, including 76 G+/LVH− subjects, 57 overt (G+/LVH+) HC, and 80 genotype-negative (G−) healthy relatives who served as the normal controls. The clinical and echocardiographic characteristics are summarized in Table 2 . Reflecting the age-dependent penetrance of LVH, the overt HC cohort was nearly 2 decades older than the G+/LVH− cohort. Pediatric subjects (<18 years) constituted 42% of the G+/LVH− cohort compared to 33% and 4% of controls and patients with overt HC, respectively. Fifty-eight different mutations were represented (a full listing of all sarcomere mutations present in this cohort are provided in the Supplementary Table ). As typically seen, mutations in MYH7 (n = 69; 52%) and MYBPC3 (n = 48; 36%) were most prevalent. Although minor differences in heart rate and blood pressure were noted among the status groups, all were within the normal ranges.
| Variable | Related Controls (n = 80) | p Value (G/+LVH− vs controls) | G/+LVH− (n = 76) | p Value (HC vs G+/LVH−) | Overt HC (n = 57) | p Value (HC vs controls) |
|---|---|---|---|---|---|---|
| Age (years) | ||||||
| Mean ± SD | 27.7 ± 11.8 | 0.004 | 22.8 ± 11.9 | <0.001 | 40.0 ± 12.8 | <0.001 |
| Range | 3–56 | 4–47 | 14–75 | |||
| Women | 42 (53%) | 0.27 | 47 (62%) | 0.03 | 21 (37%) | 0.11 |
| Genotype (n) | ||||||
| MYH7 | 40 | 29 | ||||
| MYBPC3 | 28 | 19 | ||||
| TNNT2 | 5 | 6 | ||||
| TNNI3 | 3 | 2 | ||||
| TPM1 | — | 1 | ||||
| Body surface area (m 2 ) | 1.79 ± 0.35 | 0.09 | 1.68 ± 0.38 | <0.001 | 1.95 ± 0.22 | 0.003 |
| Systolic blood pressure (mm Hg) | 121 ± 12 | <0.001 | 111 ± 13 | 0.01 | 118 ± 11 | 0.27 |
| Diastolic blood pressure (mm Hg) | 72 ± 10 | <0.001 | 66 ± 8 | <0.001 | 74 ± 9 | 0.50 |
| Heart rate (beats/min) | 68 ± 11 | 0.91 | 68 ± 13 | 0.06 | 63 ± 12 | 0.01 |
| Interventricular septal thickness (cm) | 0.9 ± 0.1 | 0.94 | 0.9 ± 0.2 | <0.001 | 1.7 ± 0.5 | <0.001 |
| Posterior wall thickness (cm) | 0.8 ± 0.2 | 0.35 | 0.8 ± 0.1 | <0.001 | 1.1 ± 0.3 | <0.001 |
| Global Ea ⁎ (cm/s) ± SE | 14.4 ± 0.3 | <0.001 | 12.7 ± 0.3 | <0.001 | 10.0 ± 0.4 | <0.001 |
| Left ventricular ejection fraction (%) | 65 ± 5 | <0.001 | 70 ± 6 | 0.92 | 69 ± 8 | 0.001 |
The ECG features of the 3 study cohorts are summarized in Table 3 . Comparing G+ and G− family members with normal LV wall thickness, Q waves and repolarization abnormalities (T-wave inversions and ST-segment depression) were significantly more prevalent in mutation carriers (p = 0.01 and p = 0.03, respectively) than in G− normal controls. Representative tracings are shown in Figure 1 . Other isolated and composite ECG abnormalities did not discriminate G+/LVH− subjects from controls. Collectively, Q waves and repolarization changes (the presence of either is denoted QST) were present in 25% of G+/LVH− subjects compared to 2% of controls ( Figure 2 ) and most commonly seen in the inferior leads. Fourteen (18%) G+/LVH− subjects had Q waves, including 10 with pathologic Q waves, in contrast to only 2 (3%) and 1 (1%) of the control subjects, respectively. Repolarization abnormalities were completely absent in the control subjects. Only 1 G+/LVH− subject had both repolarization abnormalities and Q waves. Fifty-four percent of G+/LVH− subjects had entirely normal ECG findings compared to 64% of controls (p = 0.19).
| Variable | Related controls (n = 80) | p Value (G/+LVH− vs Controls) | G/+LVH− (n = 76) | p Value (HC vs G+/LVH−) | Overt HC (n = 57) | p Value (HC vs Controls) |
|---|---|---|---|---|---|---|
| Rhythm | ||||||
| Sinus | 78 ⁎ (98%) | 1.00 | 75 ⁎ (99%) | 0.19 | 55 (96%) | 0.18 |
| Atrial fibrillation | 0 | 0 | 2 (4%) | |||
| QRS axis (°) | 55 ± 32 | 0.68 | 52 ± 31 | 0.12 | 38 ± 48 | 0.05 |
| Conduction disturbance | 1.00 | 0.002 | 0.004 | |||
| Interventricular conduction delay | 3 (4%) | 2 (3%) | 9 (16%) | |||
| Left bundle branch block | — | — | 1 (2%) | |||
| Right bundle branch block | — | — | 1 (2%) | |||
| Q wave | 2 (3%) | 0.028 | 14 (18%) | <0.001 | 24 (42%) | <0.001 |
| Q-wave location | 1.00 | 0.41 | 0.78 | |||
| Inferior | 2 (2%) | 8 (10%) | 7 (12%) | |||
| Anterior | 0 | 1 (1%) | 4 (7%) | |||
| Lateral | 0 | 2 (3%) | 8 (14%) | |||
| Multiple | 0 | 3 (4%) | 5 (9%) | |||
| Repolarization abnormalities | ||||||
| ST-segment depression | 0 (0) | 0.12 | 3 (4) | <0.001 | 26 (46) | <0.001 |
| T-wave inversion | 0 (0) | 0.12 | 3 (4) | <0.001 | 31 (54) | <0.001 |
| Electrocardiographic criteria for left ventricular hypertrophy | ||||||
| Sokolow-Lyon | 6 (8%) | 0.52 | 7 (9%) | 0.03 | 14 (25%) | 0.007 |
| Cornell | 3 (4%) | 0.43 | 5 (6%) | 0.06 | 10 (18%) | 0.015 |
| Total 12-lead QRS voltage >17.5 mV | 17 (21%) | 0.45 | 20 (26%) | 0.006 | 29 (49%) | <0.001 |
| Total 12-lead QRS voltage (mV) | 14.7 ± 4.0 | 0.45 | 15.1 ± 4.4 | 0.006 | 18.6 ± 5.9 | <0.001 |
| Romhilt-Estes | 5 (6%) | 0.43 | 7 (9%) | <0.001 | 28 (48%) | <0.001 |
| Composite electrocardiographic findings | ||||||
| Any electrocardiographic criteria for left ventricular hypertrophy | 23 (29%) | 0.70 | 24 (32%) | <0.001 | 37 (65%) | <0.001 |
| Any major or minor electrocardiographic finding | 29 (36%) | 0.19 | 35 (46%) | <0.001 | 49 (86%) | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
