The incidence of systemic thromboembolism is high in patients with hypertrophic cardiomyopathy (HCM). The authors hypothesized that vulnerability to such vascular events could be caused by depressed left atrial appendage (LAA) function during normal sinus rhythm (SR). The aim of this cross-sectional study was to investigate LAA contractile function during SR in patients with HCM.
LAA function was assessed in 62 patients with HCM in SR and compared with that in 53 age-matched and sex-matched controls. Patients with histories of atrial fibrillation and documented episodes of paroxysmal atrial fibrillation on 24-hour Holter monitoring and depressed left ventricular ejection fractions (<50%) were excluded. Multiplane transesophageal echocardiography was performed for determination of the morphology and function of the LAA.
LAA thrombi were present in five patients (8%) with HCM. LAA emptying and filling Doppler velocities were significantly depressed in the HCM group. LAA emptying and filling velocities were negatively correlated with age in controls ( r = −0.4, P = .005), but these velocities were not associated with age in the HCM group. Moreover, LAA velocities were not associated with left ventricular mass index, left ventricular outflow tract gradient, or the degree of diastolic dysfunction in the HCM group. All Doppler tissue imaging velocities obtained from LAA walls were also significantly depressed in the HCM group.
LAA thrombus formation was not rare in this patient population. The significantly depressed LAA filling and emptying velocities in SR may predispose patients with HCM to thromboembolic events. The depressed Doppler tissue imaging LAA parameters in patients with HCM may indicate the presence of a possible intrinsic atrial myopathy. Thromboembolic risk should be taken into account, and the evaluation of LAA morphology and function by transesophageal echocardiography might become a component of routine workup in patients with HCM in the future.
The incidence of stroke and systemic arterial embolism in patients with hypertrophic cardiomyopathy (HCM) is higher than expected and is significantly associated with increased morbidity and mortality. Such vascular events are more common in patients with atrial fibrillation (AF). However, there are reports that such embolism could occur even during sinus rhythm in patients with HCM. The detection of patients with HCM who are predisposed to thromboembolism might be crucial to prevent its devastating consequences. Impaired left atrial (LA) appendage (LAA) function detected by transesophageal echocardiography (TEE) have been reported as an important predictor of thromboembolism both in patients with AF and in patients in sinus rhythm.
In the presence of a high incidence of systemic thromboembolism in patients with HCM, we hypothesized that vulnerability to systemic thromboembolism could be caused by depressed LAA function during normal sinus rhythm. The aim of this cross-sectional study was to investigate LAA contractile function during sinus rhythm in patients with HCM.
Eighty consecutive unrelated patients with HCM were prospectively included in a case-control study conducted in the Cardiology Department of Ege University Medical School. These patients with HCM were either newly diagnosed or were previously known to have HCM, and most of them were referred from six affiliated cardiology centers (December 2007 to June 2009). Eighteen patients were excluded because of the following cardiac and/or systemic diseases: coronary artery disease (≥70% stenosis of any major epicardial vessel; n = 2), presence of AF and/or atrial tachycardia episodes on 24-hour Holter monitoring ( n = 7), presence of subaortic membrane ( n = 1), history of septal myocardial ablation or surgical myectomy ( n = 2), presence of mitral stenosis and/or moderate to severe mitral regurgitation in the absence of systolic anterior motion of the mitral leaflet ( n = 4), valvular aortic stenosis ( n = 1), and end-stage renal disease ( n = 1). The remaining 62 patients (34 men, 28 women) formed the patient population. The patient population was compared with 53 age-matched and sex-matched asymptomatic, healthy controls. Ten of the controls were healthy volunteers from the medical staff, and the remainder were subjects in sinus rhythm who had been referred to our echocardiography laboratory for the investigation of patent foramen ovale, atrial septal defects, and interatrial aneurysms. All control subjects had normal results on TEE. The study protocol was approved by the institutional review board, and written informed consent was obtained from all participants.
Patients fulfilling the following three criteria were included in this cross-sectional study: (1) presence of normal sinus rhythm at the time of enrollment, (2) left ventricular (LV) ejection fraction ≥ 50% and absence of any wall motion abnormality on transthoracic echocardiography (TTE), and (3) findings of HCM on TTE. The diagnosis of HCM was based on the presence of a hypertrophic (maximal wall thickness ≥ 15 mm) and nondilated left ventricle on two-dimensional echocardiography in the absence of any cardiac or systemic disease capable of inducing such hypertrophy.
Complete physical examinations and detailed clinical histories, including family histories of HCM, syncope, or sudden death in first-degree relatives at a young age (<40 years), symptoms, New York Heart Association functional class, and currently used drugs, were obtained from all participants. All study subjects underwent TTE and TEE. All patients with HCM underwent 24-hour Holter monitoring to rule out episodes of AF and/or atrial tachycardia lasting ≥ 30 sec.
The presence of LV outflow tract obstruction was assessed by TTE. Patients with resting pressure gradients ≥ 30 mm Hg in the outflow tract or mid portion of the left ventricle were accepted to have the obstructive form of HCM. Patients with resting gradients < 30 mm Hg underwent exercise testing using the standard Bruce protocol in the upright position. Patients with HCM with resting LV outflow pressure gradients < 30 mm Hg that increased to ≥30 mm Hg with exercise provocation were accepted to have latent obstructions.
TTE and TEE (Sonos 7500, Philips Medical Systems, Andover, MA) were performed according to the recommendations of the American Society of Echocardiography. All echocardiographic examinations were recorded on VHS videotape and analyzed at study completion by two independent experienced physicians who were blinded with respect to patients’ clinical characteristics.
Standard TTE included M-mode, two-dimensional, Doppler flow assessments, and Doppler tissue imaging (DTI) measurements. LV internal systolic and diastolic dimensions, posterior wall and interventricular septal thickness, and LA diameter were determined. LV mass was calculated by the method of Devereux and Reichek and was indexed to body surface area. LA volume was estimated using the biplane area-length formula after determining end-systolic area and LA long-axis length in the apical four-chamber and two-chamber views. LA volume was then calculated as [(0.85 × A 1 × A 2 )/ L ] and also indexed to body surface area. LV systolic function was assessed by the calculation of LV ejection fraction and fractional shortening. LV diastolic performance was evaluated using pulsed-wave Doppler and DTI. All Doppler measurements are given as the average values of three consecutive cardiac cycles. LV diastolic inflow velocities were obtained from the apical four-chamber view by placing the sample volume at the level of the mitral valve tip. Peak early diastolic flow velocity (E), peak atrial filling velocity (A), mitral deceleration time (from peak E wave to baseline), and isovolumic relaxation time were measured. All Doppler measurements were repeated after a Valsalva maneuver. DTI velocities of longitudinal mitral annular motion were recorded at the septal and lateral mitral annular borders. Spectral pulsed-wave Doppler was used with the instrument settings adjusted to record high-amplitude, low-velocity myocardial signals. The peak systolic (S), early diastolic (E′), and late diastolic (A′) tissue Doppler velocities over the mitral annulus were measured, and the ratio of mitral inflow E velocity to tissue Doppler (E/E′) for both the septal and lateral walls was calculated. For the assessment and grading of global LV diastolic function, LA volume index, mitral E/A ratio, deceleration time of early diastolic flow, mitral E/A ratio during the Valsalva maneuver, isovolumic relaxation time, and the E/E′ ratio for the septal and lateral annulus were considered when necessary, as recommended. Diastolic dysfunction (as determined by DTI, mitral inflow, and pulmonary vein flow velocity characteristics) was classified as mild (grade I), moderate (grade II), or severe (grade III). The presence of systolic anterior motion of the mitral leaflets was evaluated by M-mode echocardiography in the parasternal long-axis view and confirmed in the apical view. Dynamic obstruction of LV outflow was appreciated on continuous-wave Doppler in the apical five-chamber view using the modified Bernoulli equation. Obstruction of the LV outflow tract was considered to be present when the peak instantaneous gradient was estimated to be ≥30 mm Hg under resting conditions or 30 mm Hg during the strain phase of the Valsalva maneuver. The presence of mitral regurgitation was assessed semiquantitatively by the use of color Doppler imaging.
All study subjects underwent TEE (6-MHz multiplane transducer, Sonos 7500) for the determination of LAA morphology and function. All patients were studied in ≥4-hour fasting condition and received pharyngeal anesthesia with topical lidocaine 10% spray. Sedation with midazolam was used if needed. The LAA was visualized in the midesophageal view by rotation of the imaging sector from 0° to 180°. In the longitudinal view approximately at 90°, maximal LAA area and volume were measured by the planimetry method by tracing the LAA appendage endocardial border just before mitral valve opening by means of frame-by-frame analysis. Specific care was taken for the assessment of the presence of spontaneous echo contrast (SEC) and/or thrombus in the left atrium and/or LAA. SEC was diagnosed semiquantitatively as swirling dynamic smokelike echoes within the LAA cavity different from the white noise artifact. For accurate assessment, studies were performed in the appropriate gain settings, and if there was a suspicion, gain settings were decreased in a stepwise manner to avoid white noise artifact. The severity of SEC was graded from 0 to 4+, as previously described. The presence of LAA thrombus was considered when a well-circumscribed intracavitary echo-dense mass with different acoustic characteristics from the atrial endocardium or pectinate muscles with independent motion was detected. For the accurate evaluation of thrombus formation, the LAA was imaged from different views, and the visualization of intracavitary mass from every view was deemed necessary. The presence of thrombus was verified by two independent echocardiographers. In any disagreement between the observers, cardiac magnetic resonance imaging was performed to clarify the findings. LAA blood flow velocities were obtained with pulsed Doppler by positioning the sample volume at the proximal third of the LAA cavity after necessary gain and filter adjustments. During the assessment of LAA flow, three blood flow velocities were evaluated. First, early diastolic outflow wave observed before the electrocardiographic P wave was termed V 1 . Biphasic waves following V 1 subsequent to the P wave were termed V 2 (LAA contraction [emptying] flow velocity) and V 3 (filling flow velocity), respectively ( Figure 1 ). The peak flow velocities of V 1 , V 2 , and V 3 waves were measured. For DTI analysis, sample volume was positioned at five different points of the LAA walls: the proximal (basal) and mid portions of both the lateral and medial LAA walls and at the tip of the LAA (the intersection point of medial and lateral walls) ( Figure 2 A). Care was taken to keep the cursor as parallel as possible to the LAA walls. A triphasic flow pattern was recorded from each point of the LAA: D 1 , D 2 , and D 3 . The initial early diastolic positive velocity just before the electrocardiographic P wave was termed D 1 . The following biphasic positive emptying and negative filling velocities were termed D 2 and D 3 , respectively ( Figure 2 B). The D 2 wave was accepted identical to the LAA emptying velocity, and D 3 was accepted as identical to the LAA filling velocity. The peak velocities of D 1 , D 2 , and D 3 were measured and averaged for five consecutive cardiac cycles. All transesophageal echocardiographic examinations were performed without major complications.
All statistical analysis was performed using SPSS for Windows version 15.0 (SPSS, Inc., Chicago, IL). Data are presented as percentages for discrete variables and as mean ± SD for continuous variables. Two-sided P values < .05 were regarded as statistically significant. Comparisons between groups were made using t tests or Mann-Whitney U tests. Discrete variables were compared using Fisher’s exact test or χ 2 analysis, as appropriate. Correlation analyses were performed to identify factors associated with the presence of thrombi and SEC in the LAA and impaired LAA function using Pearson’s correlation analysis (or Spearman’s correlation analysis when the data were not normally distributed or had ordered categories; coefficient ρ). Receiver operating characteristic curve analysis was performed to identify the optimal cutoff values of the LAA peak emptying velocity (V 2 ) and LAA DTI septal and lateral wall D 2 velocities for the presence of LAA thrombi in patients with HCM.
The clinical and echocardiographic characteristics and types of medical treatment of patients with HCM are summarized in Table 1 . The average time interval from the initial diagnosis of HCM was 57 ± 67 months (range, 1–240 months). Obstructive HCM was present in 37 patients (59.6%) (latent type, n = 15). Baseline comparisons of demographic characteristics of the patient population and control subjects are shown in Table 2 . Comparisons of echocardiographic characteristics of the left ventricles and left atria of the study population are shown in Table 3 .
|Subtypes of HCM|
|Septal hypertrophy||51 (82.25)|
|Concentric hypertrophy||9 (14.51)|
|Apical hypertrophy||1 (1.61)|
|Midventricular hypertrophy||1 (1.61)|
|Sudden death||11 (17.7)|
|HCM and sudden death||5 (8.1)|
|HCM and syncope||2 (3.2)|
|NYHA functional class|
|History of stroke||5 (8.1)|
|History of syncope||7 (11.3)|
|Atrial tachycardia||1 (1.6)|
|Systolic anterior motion||26 (41.9)|
|Antiarrhythmic agents||3 (4.8)|
|ACE inhibitors/ARBs||16 (25.8)|
|Calcium channel blockers||9 (14.5)|
|HCM group||Control group|
|Variable||( n = 62)||( n = 53)||P|
|Age (years)||49 ± 13 (range, 18–78)||45 ± 16 (range, 18–76)||.175|
|Heart rate (beats/min)||72.40 ± 11.20||74.40 ± 9.30||.302|
|Systolic blood pressure (mm Hg)||121.41 ± 19.26||121.41 ± 16.56||.999|
|Diastolic blood pressure (mm Hg)||74.30 ± 12.57||73.18 ± 10.96||.615|
|HCM group||Control group|
|Variable||( n = 62)||( n = 53)||P|
|Interventricular septal thickness (cm)||2.22 ± 0.40||1.06 ± 0.16||.0001|
|Posterior wall thickness (cm)||1.30 ± 0.30||0.89 ± 0.16||.0001|
|LV end-systolic diameter (cm)||2.20 ± 0.32||2.60 ± 0.46||.0001|
|LV end-diastolic diameter (cm)||4.06 ± 0.39||4.40 ± 0.48||.0001|
|Ejection fraction (%)||65.27 ± 7.58||63.22 ± 5.70||.11|
|LV end-diastolic volume (mL)||80.89 ± 28.63||95.66 ± 36.02||.016|
|LV end-systolic volume (mL)||28.25 ± 13.62||34.84 ± 16.24||.02|
|Fractional shortening (%)||45.59 ± 7.71||39.30 ± 7.44||.0001|
|LV mass (g)||319 ± 87.83||148.04 ± 48.19||.0001|
|LV mass index (g/m 2 )||174 ± 47.80||79.49 ± 20.51||.0001|
|LA diameter (cm)||4.20 ± 0.60||3.46 ± 0.45||.0001|
|LA volume index (mL/m 2 )||36.30 ± 16.14||20.81 ± 7.91||.0001|
|Isovolumic relaxation time (sec)||109.27 ± 25.41||98.30 ± 21.18||.013|
|Deceleration time (sec)||247.50 ± 53.03||223.03 ± 46.95||.01|
|Lateral wall E/E′ ratio||11.53 ± 4.77||7.02 ± 2.63||.0001|
|LV diastolic dysfunction (%)||56 (90%)||24 (45.3%)||.0001|
Evaluation of the LAA
Patients with HCM had significantly higher LAA areas and volumes compared with control subjects ( Table 4 ). Early diastolic flow velocity (V 1 ), emptying (contraction) velocity (V 2 ), and filling velocity (V 3 ) obtained by pulsed-wave Doppler were all significantly depressed in patients with HCM compared with control subjects ( Table 4 , Figure 3 ). Similarly, all tissue Doppler velocities obtained from five different points of the LAA walls were significantly decreased in patients with HCM compared with control subjects ( Table 4 ).
|HCM group||Control group|
|Variable||( n = 62)||( n = 53)||P|
|LAA area (cm 2 )||4.61 ± 1.95||3.19 ± 0.97||.0001|
|LAA volume (mL)||5.12 ± 3.19||2.98 ± 1.40||.0001|
|LAA volume index (mL/m 2 )||2.85 ± 1.88||1.61 ± 0.71||.0001|
|Pulsed-wave Doppler flow velocities of LAA (cm/sec)|
|V 1||13.87 ± 4.68||19.43 ± 6.38||.0001|
|V 2||48.59 ± 19.29||71.61 ± 16.74||.0001|
|V 3||44.28 ± 18.20||51.88 ± 17.97||.029|
|Tissue Doppler velocities of LAA (cm/sec)|
|S 1 D 1||4.12 ± 1.98||6.08 ± 2.87||.0001|
|S 1 D 2||12.89 ± 4.05||16.07 ± 4.11||.0001|
|S 1 D 3||7.74 ± 3.29||12.36 ± 4.00||.0001|
|S 2 D 1||4.03 ± 1.77||6.31 ± 2.33||.0001|
|S 2 D 2||15.38 ± 4.44||17.84 ± 4.55||.006|
|S 2 D 3||9.03 ± 2.84||12.36 ± 4.00||.0001|
|AD 1||4.39 ± 1.93||6.74 ± 2.80||.0001|
|AD 2||16.75 ± 4.04||19.70 ± 6.81||.007|
|AD 3||9.92 ± 3.43||12.30 ± 4.67||.004|
|L 2 D 1||4.23 ± 1.99||6.15 ± 2.98||.0001|
|L 2 D 2||16.48 ± 4.24||19.80 ± 3.49||.0001|
|L 2 D 3||9.98 ± 2.98||14.02 ± 3.88||.0001|
|L 1 D 1||3.52 ± 1.57||5.90 ± 2.29||.0001|
|L 1 D 2||13.80 ± 4.25||17.86 ± 4.14||.0001|
|L 1 D 3||8.72 ± 3.64||13.33 ± 4.13||.0001|
Correlation analysis of the LAA flow velocities obtained by pulsed-wave Doppler revealed that all three velocities were negatively correlated with age in the control subjects (for V 1 , r = −0.340, P = .014; for V 2 , r = −0.335, P = .014; for V 3 , r = −0.388, P = .005). There were no associations between LAA velocities and age in the HCM group (for V 1 , r = −0.218, P = .088; for V 2 , r = 0.14, P = .259; for V 3 , r = −0.045, P = .732) ( Figure 4 ). Besides age, the peak velocity of V 2 of the control subjects was associated with systolic blood pressure, isovolumic relaxation time, septal and lateral annular E′ velocities, and the presence of diastolic dysfunction. In the HCM group, the peak velocity of V 2 was correlated with systolic and diastolic blood pressures, end-systolic and end-diastolic volumes, and LA volume index, but there were no associations between V 2 velocity and the time from the initial diagnosis of HCM, LV mass index, and LV outflow gradient ( Table 5 ). There were also no correlations between LAA emptying velocity and β-blocker treatment ( r = −0.11, P = .38) and calcium channel blocker therapy ( r = −0.10, P = 0.44) in the HCM group.
|HCM group ( n = 62)||Control group ( n = 53)|
|Time from initial diagnosis||−0.237||.064|
|Systolic blood pressure||0.264||.038||−0.297||.031|
|Diastolic blood pressure||0.252||.048||−0.144||.302|
|LV end-diastolic volume||0.362||.004||0.077||.583|
|LV end-systolic volume||0.317||.012||0.089||.525|
|Isovolumic relaxation time||0.043||.740||−0.400||.003|
|LA volume index||−0.390||.009||−0.143||.308|
|LAA volume index||−0.180||.166||0.181||.196|
|LV mass index||−0.010||.941||−0.139||.321|
|Resting LV outflow tract gradient||−0.109||.399|
|Presence of diastolic dysfunction||−0.122||.340||−0.270||.048|