The purpose of this study was to investigate whether global longitudinal strain measured by two-dimensional speckle tracking echocardiography could detect incipient myocardial dysfunction in patients with chronic aortic regurgitation (AR). Disclosing left ventricular (LV) dysfunction is of decisive importance for optimal timing of surgery but challenging because of the altered loading conditions.
Forty-seven patients referred for aortic valve replacement because of chronic AR were studied, along with 31 healthy controls. Myocardial deformation as determined by longitudinal, circumferential, and radial strain was calculated using two-dimensional speckle-tracking echocardiography technique, in addition to LV volumes, dimensions, and ejection fraction. Strain values were normalized to end-diastolic volume to correct for the volume dependency of deformation.
Global systolic longitudinal strain was significantly lower in patients with AR before surgery compared with the healthy controls (−17.5 ± 3.1% vs −22.1 ± 1.8%, P < .01), while global circumferential strain and LV ejection fraction did not differ (−21.7 ± 3.4% vs −22.6 ± 2.5%, P = .22 and 59 ± 5% vs 59 ± 6%, P = .59, respectively). However, differences between patients and controls were evident for both longitudinal and circumferential strain when normalized to end-diastolic volume (−0.09 ± 0.04 vs −0.23 ± 0.08, P < .01, and −0.11 ± 0.05 vs −0.24 ± 0.08, P < .01, respectively). In contrast to their absolute values, both normalized variables demonstrated improvement in myocardial shortening after valve replacement ( P < .01).
The study demonstrated reduced global longitudinal strain in patients with chronic AR with preserved LV ejection fractions. Global longitudinal strain might therefore disclose incipient myocardial dysfunction with a consequent potential for improved timing of aortic valve surgery.
Optimal timing of cardiac surgery for chronic aortic regurgitation (AR) has been a challenge for years. The development of systolic dysfunction precedes the onset of symptoms in more than one fourth of patients with this condition. Preoperative left ventricular (LV) ejection fraction (LVEF) and cavity dimensions are the most important determinants of survival and LV function after aortic valve replacement (AVR) for chronic AR. However, volume-derived measures of LV function have important limitations in assessing myocardial contractile function, whereby a series of compensatory mechanisms, including an increase in LV end-diastolic volume (LV EDV) and hypertrophy, can mask underlying changes in myocardial force development. Therefore, to date, there is no established specific diagnostic method to detect changes in LV systolic function before irreversible dysfunction occurs.
Myocardial strain echocardiography has been introduced as a clinical index of regional and global LV function. Speckle-tracking echocardiography measures strain by tracing tissue scatter in grayscale images and enables the angle-independent assessment of myocardial deformation indices. As it is noninvasive and reproducible, strain might be well suited for follow-up and to guide the timing of surgical intervention.
Therefore, the purpose of this study was to investigate whether global systolic strain measured by two-dimensional speckle-tracking echocardiography could detect early onset of myocardial dysfunction in patients with chronic AR and preserved LVEFs.
We prospectively included 78 individuals: 47 patients with chronic severe AR and 31 healthy age-matched subjects (15 women; mean age, 48 ± 15 years; P = .86). The majority of the patients were referred to the Department of Cardiology between October 2003 and October 2006 with diagnoses of AR and indications for valve replacement according to the American College of Cardiology and American Heart Association guidelines. None of the patients had atrial fibrillation or other concomitant valve disease. Patients with significant coronary artery disease, defined as ≥50% stenosis in any coronary artery, were excluded. We considered 214 referrals, but because of stringent inclusion criteria, the number of patients included in the study was modest. Therefore, an additional 10 patients were included between January 2009 and November 2009. The entire study was done prospectively. Only patients with LVEFs > 50% were included.
Recordings were performed with a Vivid 7 scanner (GE Vingmed Ultrasound AS, Horten, Norway). They were obtained in the three standard apical planes (four-chamber, two-chamber, and long-axis) and in the parasternal short-axis plane at the mitral tip by conventional two-dimensional grayscale imaging. The average frame rate was 58 ± 19 frames/sec. The digital loops were stored and subsequently analyzed using EchoPAC software (GE Vingmed Ultrasound AS).
M-mode measurements included LV end-diastolic and end-systolic short-axis diameter. LV EDV and end-systolic volume were analyzed using the biplane Simpson’s technique, with subsequent calculation of LVEF. End-diastolic and end-systolic LV lengths were measured from the mitral annulus to the LV apex in an apical four-chamber view. Changes in LV end-diastolic shape were determined by the ratio of end-systolic LV length to end-diastolic LV length and the principal radii of curvatures. The principal radii of curvatures were mathematically defined as r 1 = L 2 /2 D and r 2 = D /2, where r 1 is the radius of the lateral wall long-axis curvature, r 2 is the radius of the circumferential short-axis curvature, L is LV length, and D is LV short-axis diameter ( Figure 1 ).
Myocardial longitudinal, circumferential, and radial strain values were calculated using two-dimensional speckle-tracking echocardiography. Regions of interest were manually outlined by marking the endocardial borders at the mitral annular level and at the apex on each digital loop and were adjusted when the automatic tracking was considered suboptimal by visual or automated assessment. Examples of the technique are shown in Figure 2 . Conversion from 18 segments to a 16-segment model was performed by averaging longitudinal strain values in the corresponding apical segments in the apical long-axis and four-chamber views. Similarly, circumferential strain was averaged from six LV short-axis segments at the mitral tip level of the left ventricle. End-systole was defined as aortic valve closure as determined in the apical long-axis view. Analysis of strain in the septum and the lateral wall, separately, was performed in the apical four-chamber view. The longitudinal velocities and displacements by Doppler tissue imaging were calculated at the septal and lateral mitral annulus. Deformation parameters were normalized to LV EDV (strain/LV EDV) to correct for the volume dependency of deformation.
Furthermore, patients were evaluated using M-mode echocardiography to assess paradoxical septal motion after cardiac surgery, defined as motion of the interventricular septum toward the right ventricle in systole with normal or delayed systolic thickening.
The echocardiographic studies in patients with AR were performed 58 ± 77 days before and 229 ± 159 days after AVR.
Values are expressed as mean ± SD. Differences between groups were analyzed using independent-samples t tests. Individual differences were tested using paired-samples t tests. Values were compared for correlation by linear regression analysis.
We selected 10 random patients (160 segments) for a reproducibility test of global strain. The reproducibility of LVEF was tested in 20 randomly selected patients. Global strain and LVEF were assessed by two independent observers, and reproducibility was calculated by intraclass correlation (Cronbach’s α value). SPSS version 15.0 for Windows (SPSS, Inc., Chicago, IL) was used for all statistical analyses. For all statistical comparisons, P values < .05 were considered significant.
The study was approved by the Regional Committee for Medical Research Ethics in Norway, and all subjects gave written informed consent to participate.
In three patients, strain analysis was not feasible, because of poor echocardiographic image quality, and these patients were excluded from further analyses. Clinical characteristics of the remaining 44 patients with AR are shown in Table 1 . Approximately one third of the patients had no symptoms at the time of AVR. Because all the patients had normal systolic function as assessed by LVEF > 50%, the indication for surgery in the asymptomatic patients was severe LV dilatation according to the American College of Cardiology and American Heart Association guidelines. The most common etiologies of AR were dilated ascending aorta, bicuspid aortic valve, and previous endocarditis. Fifteen of the patients with AR (34%) underwent AVR as part of an ascending aortic composite graft.
|Age (y)||49 ± 14|
|Calcium channel blockers||7|
|Idiopathic dilatation of the aorta||17|
|Rheumatic heart disease||1|
|Aortic valve prolapse||2|
|NYHA functional class|
Echocardiographic findings are presented in Table 2 . Global systolic longitudinal strain was significantly lower in patients with AR before surgery compared with the healthy individuals, while global circumferential strain did not differ. However, differences between patients and controls were evident for both longitudinal and circumferential strain when normalized for EDV. The individual data demonstrated a large overlap between the absolute strain values in patients and controls that was partly overcome when normalized to LV EDV, in particular for global longitudinal strain ( Figure 3 ). LV dimensions and volumes, at both end-diastole and end-systole, were larger in patients with AR, while LVEF did not differ between patients and controls.
|Variable||Patients before surgery ( n = 44)||P||Patients after surgery ( n = 44)||Controls ( n = 31)||P ∗||P †|
|Global longitudinal strain (%)||−17.5 ± 3.1||.01||−16.1 ± 3.1||−22.1 ± 1.8||<.01||<.01|
|Normalized global longitudinal strain||−0.09 ± 0.04||<.01||−0.12 ± 0.04||−0.23 ± 0.08||<.01||<.01|
|Global circumferential strain (%)||−21.7 ± 3.4||.55||−21.1 ± 4.1||−22.6 ± 2.5||.22||.07|
|Normalized circumferential strain||−0.11 ± 0.05||<.01||−0.16 ± 0.06||−0.24 ± 0.08||<.01||<.01|
|Radial strain (%)||51.7 ± 18.5||.81||41.5 ± 16.6||59.4 ± 20.3||.14||<.01|
|Septal longitudinal velocity (cm/sec)||5.8 ± 1.4||<.01||4.8 ± 1.2||6.6 ± 0.8||.01||<.01|
|Lateral longitudinal velocity (cm/sec)||6.2 ± 1.6||<.01||7.2 ± 1.8||7.0 ± 1.6||.09||.479|
|Septal longitudinal displacement (cm)||11.4 ± 3.9||<.01||9.7 ± 3.0||13.7 ± 1.8||.03||<.01|
|Lateral longitudinal displacement (cm)||10.3 ± 3.3||<.01||12.5 ± 3.7||13.0 ± 2.4||<.01||.46|
|LVEF (%)||59 ± 5||<.01||54 ± 7||59 ± 6||.59||<.01|
|LV EDV (mL)||214 ± 71||<.01||141 ± 36||105 ± 30||<.01||<.01|
|LV ESV (mL)||93 ± 36||<.01||68 ± 24||43 ± 15||<.01||<.01|
|Left atrial ESV (mL)||71 ± 27||.47||70 ± 29||55 ± 17||.03||.03|
|SEDD (mm)||66 ± 8||<.01||54 ± 7||50 ± 5||<.01||.04|
|SESD (mm)||44 ± 8||<.01||36 ± 7||32 ± 4||<.01||.02|
|End-diastolic LV length (mm)||96 ± 11||<.01||89 ± 10||86 ± 9||<.01||<.01|
|End-systolic LV length (mm)||78 ± 9||.70||77 ± 10||51 ± 5||<.01||<.01|
|SV (mL)||163 ± 36||<.01||74 ± 19||62 ± 17||<.01||.01|
|Systolic blood pressure (mm Hg)||142 ± 25||<.01||131 ± 16||135 ± 21||.26||.42|
|Diastolic blood pressure (mm Hg)||66 ± 12||<.01||77 ± 12||77 ± 11||<.01||.90|
|Heart rate (beats/min)||68 ± 12||.24||66 ± 13||71 ± 12||.41||.12|