Hypertrophic cardiomyopathy (HCM) and aortic stenosis (AS) may influence left ventricular (LV) systolic function, despite preservation of LV ejection fraction. The aim of this study was to determine the relative importance of cardiac afterload and myocardial hypertrophy in the potential dysfunction of myocardial deformation, at rest and during standardized exercise.
Patients with moderate to severe (≤1.5 cm 2 ) asymptomatic AS and patients with HCM in sinus rhythm were prospectively studied using resting and exercise echocardiography during submaximal exercise. Myocardial deformations were assessed using two-dimensional strain. Exclusion criteria were altered LV ejection fraction (<50%), coronary artery disease, intra-LV obstruction > 30 mm Hg at rest, diastolic LV thickness ≥ 30 mm, and New York Heart Association class > II. Thus, 50 patients (25 with AS, 25 with HCM) were selected and matched for age, sex, rest and exercise blood pressure, degree of LV hypertrophy (defined by maximal wall thickness), and LV ejection fraction.
Mean resting global longitudinal strain (GLS) was −14.9 ± 4.7% in patients with AS and −16.1 ± 3.9% in those with HCM ( P = .30). During exercise (mean heart rate, 110 ± 10 beats/min), mean GLS was −13.9 ± 4.2% in patients with AS and −18.1 ± 5.4% in those with HCM ( P = .004). GLS decreased in patients with AS but increased in those with HCM (ΔGLS, 0.9 ± 3.1% and −1.9 ± 3.2%, respectively, P = .003). The same results were observed for global circumferential strain. Mean resting global circumferential strain was −16.4 ± 5.8% in patients with AS and −17.9 ± 4.5% in those with HCM ( P = .36). During exercise, mean global circumferential strain was −13.8 ± 4.1% in patients with AS and −18.6 ± 5.3% in those with HCM ( P = .011). Afterload was higher, particularly during exercise, in patients with AS than in those with HCM.
Longitudinal and circumferential LV deformation during exercise was lower in patients with AS compared with those with HCM, despite similar resting characteristics. The greater afterload observed in patients with AS led to reduced contractile reserve.
Two-dimensional (2D) strain has been validated as a new, easy, and fast method to analyze myocardial deformation. Abnormalities of 2D strain have been validated as a prognostic factor in a large number of cardiac diseases, including systolic heart failure, myocardial infarction, and aortic stenosis (AS). Two-dimensional strain has also been proposed as an additional tool for differentiation between physiologic and pathologic left ventricular (LV) hypertrophy (LVH).
However, the determinants of LV longitudinal dysfunction remain controversial. In patients with AS, LV longitudinal function studied by echocardiography correlates with the degree of LVH. Fibrosis and myocardial fiber disarray have been reported to alter longitudinal function in patients with LVH. We previously reported that LV longitudinal function is blunted in patients with AS and that it is emphasized by an increase in afterload generated by standardized exercise.
To describe the relative effects of afterload and of myocardial hypertrophy on the three components of the myocardial deformation (longitudinal, radial, and circumferential), we compared 2D strain parameters in patients with AS and those with hypertrophic cardiomyopathy (HCM) with the same echocardiographic degree of LVH. These parameters were recorded both at rest and during standardized submaximal exercise, performed to emphasize an increase in afterload. We expected that afterload would be more elevated and would thus alter significantly the myocardial deformation in patients with AS compared with those with HCM.
From February 2007 to January 2011, we prospectively and consecutively enrolled 150 patients with moderate to severe AS, along with 40 patients with HCM.
All patients were in New York Heart Association class I or II, were in sinus rhythm, and had normal LV ejection fractions (≥50%) as calculated by 2D echocardiography. No patient had a history of coronary artery disease. All presented with good echocardiographic windows and were able to perform exercise tests.
In the AS group, all patients demonstrated moderate to severe AS, defined by an aortic valve area ≤ 1.5 cm 2 . In all cases, no more than mild associated other valve lesions were noted. In the HCM group, all patients had HCM, defined by recent guidelines. In this group, exclusion criteria were resting intra-LV pressure gradient > 30 mm Hg or diastolic LV thickness ≥ 30 mm.
All patients included underwent clinical examinations, 12-lead electrocardiography, resting echocardiography, and submaximal exercise echocardiography. From these two populations, 25 patients with AS and 25 with HCM were selected and matched according to age, gender, resting and exercise blood pressure, level of LV hypertrophy (defined by maximal wall thickness), and LV ejection fraction.
This study was approved by the hospital ethics committee and was conducted in accordance with the Declaration of Helsinki. All patients gave informed consent.
All patients underwent clinical examinations, resting arterial blood pressure measurements (Dinamap Procare Auscultatory 100; GE Healthcare, Milwaukee, WI), resting 12-lead electrocardiography, and resting and exercise transthoracic echocardiography (Vivid 7; GE Vingmed Ultrasound AS, Horten, Norway).
Patients underwent standard exercise echocardiography using an integrated electromagnetic cycle ergometer tilt table (Ergometrics, Lynnwood, WA). The initial workload was 30 W, with 20-W increments every 2 min until a stable heart rate of 100 to 120 beats/min was obtain. The pedaling rate was fixed at 60 rpm. As previously validated, echocardiographic exercise data were recorded at a submaximal stage with a stable heart rate of 100 to 120 beats/min. Exercise tests were interrupted in case of typical chest pain, severe limiting breathlessness, dizziness, muscular exhaustion, severe hypertension (systolic blood pressure ≥ 250 mm Hg), or significant ventricular arrhythmia. Exercise was followed by an active (2 min, 30 W, 60 rpm) and then a passive (4 min) recovery phase. An electrocardiogram was recorded continuously, and blood pressure was measured every 2 min during both exercise and recovery.
Echocardiography was performed using standard acquisitions in the parasternal, apical, and subcostal views. Recordings were made both at rest and during the exercise test on a Vivid 7 machine. All resting and exercise data were stored on a workstation for offline analysis (EchoPAC; GE Vingmed Ultrasound AS). Analysis was performed offline on the EchoPAC workstation by two experienced echocardiographers (F.S., E.D.). For each patient, conventional analysis of the echocardiogram preceded the 2D strain analysis.
For each measurement, at least two cardiac cycles were averaged. LV diameters and wall thicknesses were assessed using M-mode imaging; concentric LV hypertrophy was defined as (interventricular septal thickness in diastole + posterior wall thickness in diastole)/LV end-diastolic diameter ratio > 0.42. Aortic valve area was calculated using the continuity equation; mean and maximum transaortic gradients were obtained using continuous-wave Doppler. LV end-diastolic volume, LV end-systolic volume (LVESV), and LV ejection fraction were measured using the biplane method of disks. Peak E-wave and A-wave velocities of mitral inflow were measured using pulsed-wave Doppler. Doppler tissue imaging was recorded at the level of the septal and lateral mitral annulus. Peak velocities during systole and early (e′) diastole were calculated first separately and then averaged. The E/e′ ratio and Pr/Vol ([E/e′ ratio]/LV end-diastolic volume) were subsequently calculated.
LV end-systolic meridional wall stress (LVESMWS) was estimated as 0.334 × LV pressure × LV diameter in systole/posterior wall thickness in systole [1 + (posterior wall thickness in systole/LV diameter in systole)], where LV pressure is estimated LV pressure (systolic blood pressure + mean aortic pressure gradient).
Strain measurement was based on the speckle-tracking approach: to complete the analysis of LV systolic function, global longitudinal, circumferential, and radial myocardial deformation was evaluated from standard 2D images (frame rates ≥ 70 frames/sec) using 2D strain software (EchoPAC). In brief, by tracing the endocardial borders on an end-systolic frame, the software automatically tracked the contour on the subsequent frames. Adequate tracking was verified in real time and was manually corrected, if necessary.
Global longitudinal strain (GLS) was the average of the segment strains from the apical four-chamber view ( Figures 1 and 2 ). Global radial strain and global circumferential strain (GCS) were the averages of the segment strains in the mid parasternal short-axis view. The image acquisition frame rate was 60 to 90 Hz (mean, 75 Hz).
The readers of the echocardiograms were, obviously, able to recognize patients with and without AS. But they were blinded to the results of the exercise stress tests. They did not have any clinical, electrocardiographic, or blood pressure data when reading the echocardiograms.
Strain values were indexed to LVESV to take into account LV geometric modifications induced by both pathologic conditions. Resting and exercise values and the Δ value between rest and exercise were calculated.
Data are expressed as mean ± SD or as percentages unless otherwise specified. Data were analyzed using parametric statistics after mathematical confirmation of normal distribution using Shapiro-Wilk tests. Group (AS vs HCM) comparisons for categorical variables were obtained using χ 2 tests and for continuous variables using one-way analysis of variance with Bonferroni’s test when necessary. Correlations were determined between deformation and afterload parameters Then, a multivariate linear regression was used to identify the independent predictors of alteration in exercise longitudinal and circumferential strain. We used a forward elimination procedure with the other variables. P values < .05 were considered statistically significant. Statistical analysis was performed using SPSS version 15 (SPSS, Inc., Chicago, IL).
Demographic and Hemodynamic Parameters
As expected, there was no significant difference in age, gender, resting and exercise systolic blood pressure, and exercise load between patients with AS and those with HCM.
HCM was revealed by an arrhythmic event (palpitation, documented atrial fibrillation, or ventricular tachycardia; 24%), dyspnea (32%), stroke (20%), or screening of an asymptomatic patient (competitive sport participation or family screening of relatives with HCM; 24%). Thirty-six percent patients had family histories of HCM. At inclusion, 56% of patients with HCM and 92% of those with AS were in New York Heart Association class I ( P = .0037; Table 1 ).
|Parameter||( n = 25)||( n = 25)||P|
|Age (y)||56 ± 12||50 ± 13||.09|
|Systolic BP (mm Hg)|
|Resting||139 ± 22||133 ± 15||.26|
|Exercise||171 ± 29||167 ± 31||.67|
|Resting||74 ± 12||60 ± 10||<.0001|
|Exercise||109 ± 7||108 ± 11||.66|
|Maximal workload (W)||110 ± 31||103 ± 48||.51|
There was no difference in cardiovascular risk factors (hypertension, diabetes, dyslipidemia, smoking). Treatments were not different between groups, except for a significantly more frequent use of β-blockers in the HCM group ( Table 1 ). Therefore, resting heart rates were lower in patients with HCM, but there was no difference during submaximal exercise ( Table 1 ). Although many patients were chronically treated for hypertension, all had normal rest blood pressures at the time of examination ( Table 1 ).
By definition, aortic valve area in patients with AS was reduced (1.01 ± 0.2 cm 2 ), with the mean LV-aortic pressure gradient increased (50.8 ± 19.9 mm Hg).
No adverse events occurred during submaximal exercise testing. In only two patients with HCM was exercise LV obstruction observed (peak LV gradient ≥ 30 mm Hg). In patients with HCM, the maximal intraventricular pressure gradient was 6.1 ± 3.9 mm Hg at rest and increased to 14.1 ± 10.9 mm Hg at exercise. There was an abnormal response in blood pressure (decrease or flat response with an increase of blood pressure < 20 mm Hg) in 28.5% of patients with HCM and in 21.7% of those with AS, with no significant difference between groups ( P = .73).
Cardiac Hypertrophy Patterns
LV end-diastolic diameter and LV volumes were greater in patients with AS than those with HCM ( Table 2 ). However, there was no significant difference in ventricular wall thicknesses between groups. Both groups demonstrated concentric LVH remodeling, but because of the larger LV end-diastolic diameter, the (interventricular septal thickness in diastole + posterior wall thickness in diastole)/LV end-diastolic diameter ratio was lower in the AS group ( P = .001).
|Parameter||( n = 25)||( n = 25)||P|
|IVSd (mm)||14.7 ± 1.9||15.8 ± 2.3||.07|
|PWd (mm)||13.9 ± 2.1||14.4 ± 1.8||.29|
|LVEDD (mm)||49.1 ± 6.7||41.8 ± 7.9||.0004|
|(IVSd + PWd)/LVDd||0.59 ± 0.09||0.76 ± 0.22||.001|
|LVEDV (mL)||129.8 ± 49||94.8 ± 28.3||.003|
|LVESV (mL)||51.7 ± 31.0||31.5 ± 12.5||.003|
Resting diastolic parameters were similar in both groups ( Table 3 ); however, with exercise, the E/e′ and Pr/Vol ratios were significantly higher in patients with HCM than in those with AS ( P = .04 and P = .0001, respectively).
|Parameter||( n = 25)||( n = 25)||P|
|E/A ratio||1.05 ± 0.64||1.43 ± 0.87||.08|
|e′ (cm/sec)||8.66 ± 1.99||8.76 ± 2.06||.86|
|E/e′ ratio||9.63 ± 3.11||9.63 ± 4.28||.99|
|Pr/Vol (mL −1 )||0.08 ± 0.03||0.12 ± 0.11||.07|
|E/A ratio||2.1 ± 1.57||1.44 ± 1.07||.15|
|e′ (cm/sec)||14.32 ± 4.33||10.46 ± 2.70||<.001|
|E/e′ ratio||9.71 ± 2.82||11.80 ± 2.93||.04|
|Pr/Vol (mL −1 )||0.08 ± 0.04||0.15 ± 0.07||.0001|
Resting systolic parameters were similar between groups ( Tables 4 and 5 ), with LV ejection fractions increasing similarly during exercise in both groups ( Table 4 ). The global radial component of strain decreased in both groups during exercise, without a significant difference, for both absolute and Δ values ( Table 4 ). Adaptations of the longitudinal and circumferential strain components were different between patients with AS and those with HCM. Indeed, the mean absolute value of GLS was lower in patients with AS than in those with HCM. Moreover, GLS decreased in patients with AS and increased in those with HCM (ΔGLS, 0.9 ± 3.1% vs −1.9 ± 3.2%, respectively, P = .003; Table 4 ). Last, absolute values of GCS decreased in patients with AS and increased in those with HCM, but no difference was observed for Δ values between the groups (ΔGCS, 2.2 ± 8.3% vs −1.0 ± 5.2%, respectively, P = .11; Table 4 ).