The purpose of the present study was to evaluate the role of left ventricular global afterload and various echocardiographic parameters of systolic function in a prospective cohort of 52 asymptomatic patients with severe aortic stenosis (indexed aortic valve area 0.4 ± 0.1 cm 2 /m 2 ) and normal left ventricular ejection fraction (61 ± 5%). Using 2-dimensional speckle tracking echocardiography, myocardial strain, rotation, and twist were evaluated. The valvuloarterial impedance (Zva) was calculated as a measure of left ventricular global afterload. The predefined end points were the occurrence of symptoms (dyspnea, angina, syncope), aortic valve replacement, and death. At study entry, all patients had decreased longitudinal strain (LS) (−15 ± 4%) and increased circumferential strain (−22 ± 5%), twist (24 ± 7°), and Zva (5.8 ± 2 mm Hg/ml/m 2 ). Increased Zva was closely associated with the circumferential strain increase (r = 0.59, p = 0.02) and LS decrease (r = −0.56, p = 0.016). In contrast, no relation was found between myocardial function and transaortic gradients. During follow-up (11 ± 7.5 months), on univariate Cox regression analysis, the predictors of events were the left ventricular ejection fraction (p = 0.02), mass index (p = 0.01), LS (p <0.0001), radial strain (p = 0.04), and Zva (p = 0.0002). On multivariate Cox regression analysis, only the global LS (p = 0.03) and Zva (p = 0.03) were independently associated with the combined end point. Using receiver operating characteristic curve analysis, a LS of ≤−18% (sensitivity 96%, specificity 73%) and a Zva of ≥4.7 mm Hg/ml/m 2 (sensitivity 100%, specificity 91%) were identified as the best cutoff values to be associated with events. In conclusion, in asymptomatic patients with severe aortic stenosis, the degree of global afterload and its consequences on longitudinal function might play a role in clinical practice.
In the complex relation between the pathophysiologic and clinical changes that occur in aortic stenosis (AS), it is known that the severity of valvular load and its rate of progression, the extent of valve calcification, and left ventricular (LV) hypertrophy might affect the clinical outcome of asymptomatic patients. In contrast, the role of pathophysiologic changes in arterial load (i.e., increased systemic arterial resistance) and their influence on both myocardial function and patient prognosis remain to be fully elucidated. We hypothesized that the assessment of global LV afterload, including both valvular and arterial impedance rather than the mere valvular component evaluation, could provide more information for stratifying the relative risk in asymptomatic patients. On the basis of this assumption, we investigated whether (1) the global afterload, noninvasively evaluated using valvuloarterial impedance (Zva), might influence LV myocardial function more extensively than the severity of the valve disease, and (2) whether the Zva and/or other myocardial functional parameters (strain, rotation, and twist), evaluated using speckle tracking echocardiography, might represent a marker of clinical outcome. We studied a prospective cohort of consecutive patients with asymptomatic severe AS and preserved LV ejection fraction (EF) who did not fulfill the current criteria for surgical treatment.
Methods
A total of 52 asymptomatic patients (17 men, mean age 72 ± 11 years) with severe AS were prospectively enrolled in the present study. All subjects met the following enrollment criteria: severe AS defined as an indexed aortic valve area <0.6 cm 2 /m 2 , no symptoms according to a careful history, normal LVEF (≥50%) as calculated by 2-dimensional echocardiography, no more than mild associated heart valve disease, sinus rhythm, no renal failure, no previous myocardial infarction, and optimal quality for speckle tracking echocardiography analysis. At study entry, the following clinical data were collected: age, gender, history of hypercholesterolemia (total cholesterol >190 mg/dl or patients receiving lipid-lowering therapy), diabetes mellitus (fasting blood glucose >126 mg/dl on 2 occasions or patients currently receiving oral hypoglycemic medication or insulin), systemic arterial hypertension (blood pressure ≥140/90 mm Hg or patients receiving antihypertensive treatment), and overweight (body mass index >25 kg/m 2 ). The regional ethics committee approved the protocol, and all patients gave written informed consent.
Echocardiographic examinations were performed using a VIVID-7 ultrasound machine (GE Vingmed Ultrasound, Horten, Norway), equipped with a phased-array transducer. M-mode, 2-dimensional, color Doppler, pulsed-wave, and continuous-wave Doppler data were stored on a dedicated workstation (EchoPAC, version 8.0.0; GE Medical Systems, Horten, Norway), for off-line analysis. The measurements were made for 3 cardiac cycles, and the average value was calculated. The LV diameters, wall thickness, and outflow tract diameter were measured according to the recommendations of the American Society of Echocardiography. The transvalvular aortic velocity time integral, mean pressure gradient, and peak aortic velocity were obtained using continuous wave Doppler ultrasonography. The right parasternal view was used whenever possible.
The aortic valve area was determined using the continuity equation method and indexed to the body surface area, using the Du Bois and Du Bois formula. The stroke volume was calculated using the Doppler method as follows: 0.785 × (LV outflow tract diameter) × LV outflow tract velocity time integral. The LVEF was derived using the biplane Simpson disk method. The LV mass was determined with the area − length method, and the mass index was calculated as the LV mass/body surface area (g/m 2 ) ratio. The diagnosis of LV hypertrophy was determined using a LV mass index >102 g/m 2 in men and >81 g/m 2 in women. A cutoff value of ≥0.45 for the relative wall thickness was considered to define a concentric remodeling. The mitral flow peak velocities (E and A) and E/A ratio were measured using pulsed wave Doppler. Furthermore, from stored color tissue Doppler imaging loops, the value of E′ was obtained by averaging the peak early-diastolic velocities calculated at the level of the septal, lateral, anterior, and inferior corner of the mitral annulus. The E/E′ ratio was also included as an estimate of LV filling pressure.
To complete the analysis of LV systolic function, myocardial deformation was assessed by speckle tracking echocardiography and automated function imaging for the evaluation of longitudinal strain (LS). Automated function imaging was performed on apical long-axis, 4-chamber, and 2-chamber views, following an on-screen guided workflow. The results were presented as a bull’s-eye display showing color-coded and numeric values for peak systolic LS. For assessing global circumferential strain and radial strain, the standard parasternal short-axis views at the basal, mid, and apical levels with a frame rate of >70 frames per second were acquired. The relative global strains were obtained by calculating the average strain derived by myocardial tracking from each short-axis view.
Finally, the LV rotations were derived from the apical and basal short-axis images as the average angular displacement of the 6 standard segments referring to the ventricular centroid, frame by frame. LV twist was defined as the net difference (in degrees) of apical and basal rotations at the isochronal time points. The opposite rotation after LV twist was defined as LV untwist and the time derivative of LV untwist was defined as the LV untwisting rate (°/s).
Systemic arterial pressure was measured using an arm cuff sphygmomanometer at the Doppler echocardiographic examination. To estimate the global LV afterload, we calculated the Zva as the sum of the systemic arterial pressure and the mean transvalvular pressure gradient divided by the stroke volume index, as suggested.
Follow-up information was obtained from interviews with patients, their relatives, or their physicians every 6 months, according to the current guidelines. The predefined end points for assessing the outcome were the occurrence of symptoms (dyspnea, angina, syncope), aortic valve replacement, or death during follow-up.
The data are expressed as the mean ± SD. Statistical analysis was performed using the Statistical Package for Social Sciences statistical software, version 17 for Windows (SPSS, Chicago, Illinois). To compare each parameter between groups, the chi-square test and Fisher’s exact test, when appropriate, or the Mann-Whitney U test, were used for categorical and continuous variables, respectively. Linear regression analysis, with Pearson’s and Spearman’s coefficients, were used to estimate correlation between variables. P values of ≤0.05 were considered significant.
The combined end point of the study included the development of significant symptoms (angina, dyspnea, syncope), cardiac death, and the clinical need for aortic valve replacement. Multivariate Cox proportional hazards regression analysis was used to identify independent predictors of events. Clinically relevant variables with p <0.05 on univariate analysis were incorporated into the multivariate model. The Kaplan-Meier curve was used for cumulative survival analysis at the end of follow-up. Receiver operating characteristic curves were generated to determine the performance of independent variables for the prediction of cardiac events.
Results
The mean patient age was 72 ± 11 years, and 33% were men. The body surface area and heart rate was 1.7 ± 0.1 m 2 and 72 ± 10 beats/min, respectively. Every enrolled subject was classified as having New York Heart Association class I. A history of hypertension was reported in 15 (29%), 13 (25%) had hypercholesterolemia, and 4 (8%) of 52 patients had diabetes, and none were obese. On the basis of the case history and echocardiographic study findings, the pathophysiologic mechanism of AS was presumptively degenerative (calcified aortic valve) in all patients. None of the patients was diagnosed with a bicuspid aortic valve.
The ventricle geometry and function analysis, assessed using standard echocardiography, showed a pattern of concentric hypertrophy with a normal EF in most patients. In contrast, the analysis of myocardial systolic function using speckle tracking echocardiography and automated function imaging reveled a decreased global LS, no alteration of global radial strain, mildly increased global circumferential strain, and increased rotations and twist of the left ventricle ( Table 1 ). All the speckle tracking echocardiography and automated function imaging results were interpreted according to previous findings from various studies on myocardial function in patients with severe AS and preserved LVEF. The mean Zva in asymptomatic patients with AS at enrollment was also increased ( Table 1 ).
| Variable | Patients (n = 52) |
|---|---|
| Left ventricular mass index (g/m 2 ) | 127 ± 38 |
| Ejection fraction (%) | 61 ± 5 |
| Aortic valve area (cm 2 ) | 0.6 ± 0.2 |
| Indexed aortic valve area (cm 2 /m 2 ) | 0.4 ± 0.1 |
| Aortic peak gradient (mm Hg) | 90 ± 25 |
| Aortic mean gradient (mm Hg) | 60 ± 16 |
| Stroke volume index (ml/m 2 ) | 35 ± 10 |
| Systolic arterial pressure (mm Hg) | 134 ± 22 |
| E/A | 0.8 ± 0.2 |
| E/E′ | 18 ± 13 |
| Global longitudinal strain (%) | −15 ± 4 |
| Global circumferential strain (%) | −22 ± 5 |
| Global radial strain (%) | 42 ± 12 |
| Basal rotation (°) | −7.5 ± 3 |
| Apical rotation (°) | 16.5 ± 7 |
| Twist (°) | 24 ± 7 |
| Untwisting rate (°/s) | −148 ± 44 |
| Valvuloarterial impedance (mm Hg/ml/m 2 ) | 5.8 ± 2 |
Increased Zva was closely associated with decreased LS (r = −0.56, p = 0.016) and increased circumferential strain (r = 0.59, p = 0.02). In addition, decreased LS correlated with LVEF (r = 0.66, p <0.001), mass index (r = −0.46, p = 0.015), twist (r = −0.39, p = 0.036), radial strain (r = 0.46, p = 0.04), circumferential strain (r = −0.45, p = 0.009), and aortic valve area/m 2 (r = 0.37, p = 0.04). In contrast, we did not find any correlation between longitudinal function and the severity of AS, expressed as transvalvular pressure gradients (peak gradient, r = −0.21, p = 0.23; mean gradient, r = −0.30, p = 0.23). Finally, the circumferential strain and radial strain were closely related to each other (r = −0.60, p = 0.004).
Follow-up information was available for 38 (73%) of 52 patients. The mean follow-up time was 11 ± 7.5 months (range 1 to 23; Figure 1 ) . During follow-up, the predefined end points were reached in 26 patients, including 2 deaths (1 sudden death and 1 death preceded by pulmonary edema), 2 patients who developed symptoms of heart failure, and 22 patients who required aortic valve replacement. Surgery was required because of the occurrence of symptoms within 8 ± 2 months after inclusion. The predominant symptoms were severe dyspnea (17 patients), angina (3 patients), or syncope (2 patients). Twelve patients remained free of clinical events during follow-up. The clinical and standard echocardiographic characteristics of the patients who remained asymptomatic and those who experienced an event are listed in Table 2 . In particular, among the clinical parameters examined (age, gender, heart rate, risk factors), male gender, and hypercholesterolemia had a significantly greater prevalence in the group of patients who developed events than in those remaining asymptomatic (p <0.001). However, the very interesting result is that subjects in the “event” group had hypertension than those in the “event-free” group (p = 0.02). In addition, the patients who developed events during follow-up had higher systolic blood pressure, lower stroke volume index, and higher LV mass index than those free from any event (p = 0.044, p <0.001, and p = 0.007, respectively). Furthermore, no parameter of AS severity (indexed aortic valve area and peak and mean gradients) and no index of LV systolic and diastolic function (LVEF, mitral E/A, and E/E′) allowed significant distinction between the 2 groups (p = NS). Moreover, neither radial strain (39 ± 10% vs 51 ± 15%, p = 0.117), circumferential strain (−20 ± 4% vs −23 ± 4%, p = 0.158), rotations (basal rotation = −7.5 ± 2° vs −8.8 ± 4°, p = 0.416; apical rotation = 16.5 ± 6° vs 14.7 ± 6°, p = 0.346), twist (23.2 ± 7° vs 23.8 ± 6°, p = 0.946) nor untwisting rate (−147 ± 50°/s vs −141 ± 52°/s, p = 0.123) showed a significant difference between the “event” and “event-free” group. In contrast, the LV longitudinal function and global afterload were the only parameters significantly impaired in patients with events compared to those remaining asymptomatic (p <0.001; Figures 2 and 3 ) .
| Variable | Events | No Events | p Value |
|---|---|---|---|
| Patients | 26 (68%) | 12 (32%) | <0.001 |
| Men | 11 (42%) | 2 (17%) | <0.001 ⁎ |
| Age (years) | 73 ± 10 | 72 ± 10 | 0.555 |
| Body surface area (m 2 ) | 1.6 ± 0.1 | 1.8 ± 0.2 | 0.632 |
| Heart rate (beats/min) | 70 ± 10 | 68 ± 8 | 0.105 |
| Hypertension | 7 (27%) | 2 (17%) | 0.02 ⁎ |
| Hypercholesterolemia | 7 (27%) | 0 | <0.001 ⁎ |
| Systolic arterial pressure (mm Hg) | 138 ± 21 | 121 ± 21 | 0.044 ⁎ |
| Stroke volume index (ml/m 2 ) | 30 ± 7 | 43 ± 7 | <0.001 ⁎ |
| Ejection fraction (%) | 59 ± 5 | 63 ± 4 | 0.139 |
| Left ventricular mass index (g/m 2 ) | 138 ± 32 | 91 ± 42 | 0.007 ⁎ |
| Indexed aortic valve area (cm 2 /m 2 ) | 0.35 ± 0.11 | 0.37 ± 0.10 | 0.611 |
| Aortic peak gradient (mm Hg) | 92 ± 23 | 87 ± 32 | 0.328 |
| Aortic mean gradient (mm Hg) | 60 ± 13 | 53 ± 16 | 0.115 |
| E/A | 0.89 ± 0.2 | 0.76 ± 0.1 | 0.052 |
| E/E′ | 19 ± 7 | 17 ± 7 | 0.081 |
On univariate Cox regression analysis, the predictors of adverse events during follow-up, among the preselected clinical and echocardiographic variables, were LVEF (p = 0.02), LV mass index (p = 0.01), global LS (p <0.0001), global radial strain (p = 0.04), and Zva (p = 0.0002; Table 3 ). With multivariate Cox regression analysis, only 2 parameters remained significant: global LS (hazard ratio 1.41, 95% confidence interval 1.01 to 1.95, p = 0.03) and Zva (hazard ratio 2.78, 95% confidence interval 1.09 to 7.08, p = 0.03), emerging as independently associated with the combined end point. Using receiver operating characteristic curve analysis, a global LS of ≤−18% (area under the curve 0.87, 95% confidence interval 0.722 to 0.963; p <0.0001; sensitivity 96%, specificity 73%) and a Zva ≥4.7 mm Hg/ml/m 2 (area under the curve 0.98, 95% confidence interval 0.88 to 1.0; p <0.0001, sensitivity 100%, specificity 91%) were identified as the best cutoff values associated with events ( Figures 4 and 5 ) .
