Background
The diagnosis and management of paradoxical low-flow (PLF) aortic stenosis (AS) is challenging in clinical practice. In addition, its pathophysiology has not been fully understood. The aim of this study was to test the hypothesis that left ventricular (LV) myocardial function is deteriorated in PLF AS and that it is closely related to global LV afterload.
Methods
Echocardiographic data from 103 patients with severe AS (aortic valve area < 1.0 cm 2 ) with normal LV ejection fractions were prospectively collected. Global longitudinal and circumferential myocardial strain was analyzed using two-dimensional speckle-tracking imaging. PLF AS was defined as a stroke volume index < 35 mL/m 2 .
Results
Sixteen patients were classified as having PLF AS. Compared with those with normal-flow AS, patients with PLF AS were more likely to have worse functional status (mean New York Heart Association functional class, 2.38 ± 0.70 vs 1.96 ± 0.62; P = .02), worse global longitudinal strain (GLS) (−12.6 ± 4.4% vs −16.4 ± 4.0%, P < .01), lower aortic valve area (0.53 ± 0.15 vs 0.78 ± 0.19 cm 2 , P < .01), and higher valvuloarterial impedance (5.62 ± 1.33 vs 3.65 ± 0.83 mm Hg · m 2 /mL, P < .01). GLS showed a significant negative linear relationship with stroke volume index ( r = −0.324, P = .001) and a positive relationship with E/E′ ratio ( r = 0.367, P < .001). Multivariate analysis showed that age (β = 0.08, P = .07) and valvuloarterial impedance (β = 1.54, P < .01) were significant predictors of GLS.
Conclusions
GLS is depressed in patients with PLF AS. This implies that subclinical myocardial dysfunction may be more prominent in PLF AS compared with normal-flow AS and suggests the possible diagnostic and prognostic value of two-dimensional global strain in identifying PLF AS. In addition, global LV afterload is an important determinant of myocardial dysfunction in patients with severe AS.
Current American College of Cardiology and American Heart Association guidelines recommend the following Doppler measurements as the diagnostic criteria for severe aortic stenosis (AS): aortic valve area (AVA) < 1.0 cm 2 and/or AVA index < 0.6 cm 2 /m 2 , transaortic mean pressure gradient (PG) > 40 mm Hg, and peak velocity > 4.0 m/sec. However, these criteria have been challenged with recent reports of paradoxical low-flow (PLF), low-gradient AS, which do not meet the aforementioned criteria for severe AS. To make matters more perplexing, patients with PLF severe AS carry a worse prognosis if treated medically, and yet, the proportion of patients receiving the right diagnosis and timely operative correction is significantly lower than among patients with normal-flow severe AS. Therefore, the correct diagnosis and identification of PLF AS are important to improve clinical outcomes. In this regard, the pathophysiology of PLF AS needs to be clarified.
It is well known that impaired left ventricular (LV) myocardial function is an important determinant of clinical outcome in AS. Accordingly, we hypothesized that LV myocardial function is deteriorated and may lead to poor clinical outcomes in patients with PLF AS. We also hypothesized that global LV afterload is an important determinant of LV dysfunction in these patients. We used two-dimensional (2D) global strain on the basis of speckle-tracking imaging because it is useful for assessing multidirectional myocardial mechanics, relatively independent of loading conditions, and is also a useful prognostic factor of various conditions of the heart.
Methods
Patient Population
A total of 103 patients with severe AS and preserved LV ejection fractions (LVEFs; >50%) per the definition of AVA < 1.0 cm 2 in the American College of Cardiology and American Heart Association guidelines were enrolled in the study irrespective of symptom status. Patients with significant concomitant valvular heart disease of grade ≥3 other than AS (i.e., concomitant aortic regurgitation or mitral, tricuspid, or pulmonic valve disease; significant regional wall motion abnormality; or history of myocardial infarction or previous cardiac surgery) were excluded. PLF AS was defined as stroke volume index (SVi) < 35 mL/m 2 , according to previous literature, and normal-flow (NF) aortic stenosis (AS) was designated otherwise ( Figure 1 ). All patients gave informed consent to participate in the study, the protocol of which was approved by the institutional review board of Seoul National University Hospital. Baseline laboratory tests, anthropometric measures, and medical histories were taken at the time of echocardiographic examination.
Two-Dimensional Echocardiography
All patients underwent comprehensive 2D echocardiographic examinations using a 3.5-MHz transducer and commercially available equipment (Vivid 7; GE Vingmed Ultrasound AS, Horten, Norway). LV end-diastolic and end-systolic diameters and wall thickness were measured using standard M-mode tracings at the parasternal short-axis view of papillary muscle level, and the LVEF was calculated using these M-mode measurements. Relative wall thickness was estimated as 2 × (diastolic LV posterior wall thickness)/LV internal diastolic dimension. The dimensions of the aortic root (i.e. LV outflow tract [LVOT], sinotubular junction, and ascending thoracic aortic diameter) were measured in the standard parasternal long-axis view. End-diastolic and end-systolic volumes were obtained using the modified biplane method in apical four-chamber and two-chamber views. LV mass index was calculated as {1.05 × [(LV internal diastolic dimension + diastolic LV posterior wall thickness + diastolic interventricular septal thickness) 3 − (LV internal diastolic dimension) 3 ] − 13.6}/body surface area, where 1.05 is the specific gravity of the myocardium. Peak early (E) and late (A) diastolic velocity of mitral inflow and peak early (E′) and late (A′) diastolic mitral annular velocity on the septal side were measured in the apical four-chamber view. All echocardiographic measurements were averaged over three beats for patients in sinus rhythm and five beats for those in atrial fibrillation with baseline heart rates < 100 beats/min.
The time-velocity integral (TVI) of the LVOT was measured in the apical five-chamber view, and stroke volume was calculated with LVOT area (0.785 × [LVOT diameter ] 2 ) multiplied by the TVI at the LVOT. SVi was calculated by dividing stroke volume by body surface area, which in turn was calculated using the Mostella formula. AVA was calculated using the continuity equation (aortic valve TVI/[LVOT TVI × LVOT area]) and indexed by AVA/body surface area. The mean transaortic PG was estimated by manual tracing of the flow velocity using continuous-wave Doppler recordings in the apical five-chamber view. Valvuloarterial impedance (Z va ) was calculated as (systolic blood pressure + transaortic mean PG)/SVi.
Two-Dimensional Speckle-Tracking Imaging Analysis
Digital loop images were obtained at the most optimal frame rate (50–100 frames/sec), sector width, and image depth at an end-expiratory breath hold from the parasternal short-axis view at the papillary muscle level and apical views. The LV endocardial border was traced at the end-systolic frame, and the region of interest was defined automatically by the offline analysis program (EchoPAC version 5.0.1; GE Medical Systems, Milwaukee, WI) between the endocardial and epicardial borders. Myocardial tracking was verified, and the width of the region of interest was adjusted to optimize the tracking, if needed. Care was taken to ensure that at least five segments were tracked adequately from each plane to obtain an adequate global strain measure. The strain curve was obtained by a single independent observer blinded to the objective of the study. Peak global circumferential strain (GCS) and global longitudinal strain (GLS) were defined as the peak negative values of the strain curve in a single beat cardiac cycle and calculated for the entire circular or U-shaped LV myocardium as follows: global strain = (L [end-systole] − L [end-diastole])/L (end-diastole) × 100%, where L is the entire LV myocardium as one segment, and global strain is the myocardial deformity of the myocardium as a whole, not an average of each segmental strain as in previous studies of average strain in patients with severe AS. Peak GCS was measured in the midventricular parasternal short-axis view, whereas GLS was measured in the apical two-chamber, four-chamber, and three-chamber views and averaged.
Statistical Analysis
Continuous and dichotomous variables were compared using Student’s t tests and χ 2 tests, respectively, and are presented as mean ± SD and as percentages. Bivariate correlation analysis was drawn between GLS/GCS, and SVi, Z va , indexed AVA, and E/E′ ratio. The strength of association are presented as Pearson’s correlation coefficients. Variables that were considered related to myocardial function were initially analyzed with univariate analysis, and then, variables with P values < .10 were incorporated into a stepwise selection multiple linear regression model to take multicollinearity into account for the identification of factors significantly predictive of GLS. All analysis was done with SPSS version 17.0 (SPSS, Inc., Chicago, IL), and two-tailed P values < .05 were considered statistically significant.
Results
Among 103 patients, 16 were classified as having PLF AS. Patients with PLF AS tended to have larger body surface areas, were less likely to have hypertension, and were more likely to have atrial fibrillation ( Table 1 ). Interestingly, patients with PLF AS also tended to have worse functional status compared with those with NF AS (mean New York Heart Association functional capacity, 1.96 ± 0.62 for NF AS vs 2.38 ± 0.70 for PLF AS; P = .02).
| Variable | Total ( n = 103) | NF AS ( n = 87) | PLF AS ( n = 16) | P ∗ |
|---|---|---|---|---|
| Age (y) | 67.5 ± 9.7 | 67.2 ± 10.2 | 69.2 ± 6.8 | .34 |
| Men | 62 (60.2%) | 51 (58.6%) | 11 (68.8%) | >.99 |
| BSA (m 2 ) | 1.66 ± 0.17 | 1.64 ± 0.17 | 1.77 ± 0.18 | .02 |
| Diabetes mellitus | 27 (26.2%) | 24 (27.6%) | 3 (18.8%) | .55 |
| Hypertension | 52 (50.5%) | 47 (54.0%) | 5 (31.2%) | .09 |
| Systolic BP (mm Hg) | 129 ± 18 | 130 ± 19 | 120 ± 9 | .01 |
| Diastolic BP (mm Hg) | 71 ± 10 | 71 ± 10 | 71 ± 10 | .86 |
| Current smokers | 10 (9.7%) | 8 (9.2%) | 2 (12.5%) | .65 |
| Hyperlipidemia | 12 (11.7%) | 11 (12.6%) | 1 (6.2%) | .69 |
| NYHA functional status | 2.03 ± 0.65 | 1.96 ± 0.62 | 2.38 ± 0.70 | .02 |
| Atrial fibrillation | 7 (6.8%) | 4 (4.6%) | 4 (25.0%) | .02 |
| Baseline creatinine (mg/dL) | 1.07 ± 0.43 | 1.07 ± 0.46 | 1.10 ± 0.29 | .82 |
| Bicuspid AV | 48 (46.6%) | 37 (42.5%) | 11 (68.8%) | .05 |
| Operation | 63 (61.2%) | 49 (56.3%) | 14 (87.5%) | .02 |
∗ Differences between the NF AS group and the PLF AS group were calculated using Student’s t tests or χ 2 tests as appropriate.
Compared with the NF AS group, LVEFs were significantly lower in the PLF AS group, which was partially explained by larger end-systolic dimensions and volumes. Of note, the PLF AS group also tended to have thicker myocardium, as shown by significantly thicker interventricular septum and posterior wall, and also a tendency toward concentric remodeling, shown by relative wall thickness, compared with the NF AS group. Although there was no difference in aortic root dimensions between the two groups, the AVA and indexed AVA were significantly smaller in the PLF AS group. These results are summarized in Table 2 . Intriguingly, significantly worse GLS was observed in the PLF AS group, although there was no difference in GCS. In addition, Z va , a parameter of global LV afterload, was significantly increased in the PLF AS group.
| Variable | Total ( n = 103) | NF AS ( n = 87) | PLF AS ( n = 16) | P ∗ |
|---|---|---|---|---|
| LVEDD (mm) | 49.4 ± 4.9 | 49.3 ± 5.0 | 49.6 ± 4.2 | 0.82 |
| LVESD (mm) | 30.2 ± 4.6 | 29.9 ± 4.7 | 32.1 ± 4.2 | 0.08 |
| LVEDV (mL) | 108.9 ± 31.5 | 107.5 ± 29.3 | 116.5 ± 41.6 | 0.30 |
| LVESV (mL) | 41.3 ± 16.9 | 39.6 ± 15.4 | 49.9 ± 22.0 | 0.09 |
| LVEF (%) | 62.6 ± 7.2 | 63.5 ± 6.7 | 57.5 ± 7.7 | <0.01 |
| IVS thickness (mm) | 12.1 ± 2.2 | 11.9 ± 2.0 | 13.4 ± 2.9 | 0.02 |
| PWT (mm) | 11.6 ± 2.3 | 11.4 ± 2.1 | 12.9 ± 2.8 | 0.01 |
| Aortic annular diameter (mm) | 21.2 ± 2.1 | 21.1 ± 1.9 | 21.8 ± 2.9 | 0.25 |
| Sinus of Valsalva diameter (mm) | 32.0 ± 3.9 | 32.0 ± 3.7 | 32.4 ± 4.7 | 0.71 |
| Sinotubular junction diameter (mm) | 27.1 ± 3.4 | 27.0 ± 3.2 | 27.8 ± 4.6 | 0.39 |
| E (m/sec) | 0.78 ± 0.29 | 0.78 ± 0.29 | 0.79 ± 0.31 | 0.91 |
| A (m/sec) | 0.90 ± 0.27 | 0.91 ± 0.26 | 0.85 ± 0.36 | 0.42 |
| DT (msec) | 245 ± 79 | 248 ± 80 | 224 ± 78 | 0.29 |
| E′ (cm/sec) | 4.7 ± 1.7 | 4.8 ± 1.7 | 4.1 ± 1.8 | 0.17 |
| A′ (cm/sec) | 7.0 ± 2.0 | 7.1 ± 2.1 | 6.2 ± 1.3 | 0.14 |
| E/E′ | 18.8 ± 10.0 | 18.4 ± 10.4 | 20.9 ± 7.6 | 0.36 |
| V max (m/sec) | 4.6 ± 0.8 | 4.6 ± 0.8 | 4.3 ± 0.8 | 0.19 |
| AVA (cm ) | 0.74 ± 0.20 | 0.78 ± 0.19 | 0.53 ± 0.15 | <0.01 |
| AVA index (cm /m 2 ) | 0.45 ± 0.12 | 0.47 ± 0.10 | 0.30 ± 0.10 | <0.01 |
| Mean PG (mm Hg) | 51.1 ± 20.5 | 51.9 ± 21.4 | 46.6 ± 14.2 | 0.34 |
| Z va (mm Hg · m 2 /mL) | 3.96 ± 1.17 | 3.65 ± 0.83 | 5.62 ± 1.33 | <0.01 |
| GLS (%) | −15.8 ± 4.3 | −16.4 ± 4.0 | −12.6 ± 4.4 | <0.01 |
| GCS (%) | −17.7 ± 4.6 | −18.0 ± 4.6 | −16.1 ± 4.9 | 0.17 |
| LVMI (g/m 2 ) | 174.6 ± 51.6 | 171.5 ± 49.6 | 191.0 ± 59.9 | 0.17 |
| RWT | 0.48 ± 0.10 | 0.47 ± 0.09 | 0.52 ± 0.12 | 0.03 |
∗ Differences between the NF AS group and the PLF AS group were calculated using Student’s t tests.
By bivariate correlation analysis, a significant negative relationship existed between SVi and GLS ( r = −0.324, P = .001; Figure 2 A) and also between SVi and GCS ( r = −0.244, P = .024; Figure 2 B). Z va showed a good positive relationship with GLS ( r = 0.437, P < .001; Figure 2 C) and also for GCS, albeit to a lesser degree ( r = 0.321, P = .003; Figure 2 D). GLS also showed a significant negative relationship with AVA index ( r = −0.377, P < .001). GCS was marginally correlated with AVA index ( r = −0.219, P = .043; Figures 2 E and 2 F).
