Background
The aim of this study was to investigate the impact of acute right ventricular pressure overload (RVPO) on left ventricular (LV) function and regional uniformity using speckle-tracking displacement and strain analyses in patients with acute pulmonary embolism (PE).
Methods
Twenty-five patients with acute PE (mean age, 59 ± 16 years) and 25 normal subjects were enrolled. Radial, longitudinal, and circumferential LV wall motion and myocardial deformation were analyzed using speckle-tracking displacement and strain imaging echocardiography, respectively, from the mid-LV short-axis and apical four-chamber views. The standard deviation of the heart rate–corrected intervals from QRS onset to peak systolic displacement (PSD) and peak systolic strain for the six segments was used to quantify LV systolic dyssynchrony. The standard deviation of regional PSD and peak systolic strain divided by their global values was used to quantify LV systolic heterogeneity. Mechanical discoordination of LV regional wall motion and myocardial deformation was assessed by averaging the frame-by-frame percentage discordance between segmental and global signal changes in the six segments.
Results
Patients with acute PE had reduced radial PSD and peak systolic strain and a large extent of displacement-derived nonuniformities (PSD dyssynchrony, 74 ± 32 vs 40 ± 20 m sec; PSD heterogeneity, 0.39 ± 0.13 vs 0.17 ± 0.08; and PSD discoordination, 23 ± 2% vs 15 ± 3%; P < .05 vs normal subjects for all comparisons) associated with a leftward shift of the interventricular septum. In contrast, all indices of strain-derived radial LV nonuniformities were not augmented by acute RVPO in patients with acute PE. Patients with acute PE also had impaired LV systolic function and regional uniformities in the longitudinal and circumferential directions. After the amelioration of acute RVPO by primary treatment, most of the indices of LV function and regional uniformity were restored to normal values. Multiple regression analysis indicated that only radial LV wall motion discoordination was a significant determinant of cardiac index.
Conclusions
Acute RVPO induces reversal LV regional uniformities, which are closely associated with reduced LV function and abnormal geometry of the left ventricle, and radial LV wall motion coordination plays a key role in the short-term regulation of cardiac output in patients with acute PE.
Acute pulmonary embolism (PE) represents an example of the etiology of acute right ventricular (RV) pressure overload (RVPO), which alters ventricular geometry and function. We previously reported that acute PE impaired RV function and regional uniformity assessed using speckle-tracking strain echocardiography. More important, the left ventricle is impeded by the leftward movement of the septum and has a paradoxical septal motion, which can alter global left ventricular (LV) performance and coordination of LV regional kinetics, even if intrinsic LV function is normal. Speckle-tracking displacement and strain echocardiography can measure global and regional LV myocardial kinetics independent of echo angle and chamber translation. Accordingly, we aimed to extend our previous work to elucidate the impact of acute RVPO on LV systolic function and regional uniformity using speckle-tracking displacement and strain echocardiography in patients with acute PE.
Methods
Study Population
The study group consisted of 25 consecutive patients with massive or submassive acute PE referred to Mie University Hospital between May 2005 and June 2009 (mean age, 59 ± 16 years; range, 19–82 years; 5 men). Twenty-three of 25 patients in this study had been included in the previously published study. All patients were diagnosed as having acute PE using personal histories and the results of physical examinations, laboratory tests, echocardiography, and thoracic computed tomography. The clinical definition of massive PE was established in the presence of cardiogenic shock or hypotension, the latter defined as systemic systolic blood pressure <90 mm Hg or a pressure drop >40 mm Hg for >15 minutes not caused by arrhythmia, hypovolemia, or sepsis. Normotensive patients with PE and evidence of RV dysfunction were defined as having submassive PE. The criteria applied for the diagnosis of RV dysfunction were a diastolic diameter of the right ventricle >30 mm, an RV diastolic diameter/LV diastolic diameter ratio >1, paradoxical septal movement, hypokinesia of the RV free wall, loss of inspiratory collapse of the inferior vena cava, and tricuspid regurgitation (TR) at a velocity >2.5 m/sec in the absence of inspiratory collapse of the inferior vena cava or >2.8 m/sec. In all patients, RV dysfunction was confirmed by transthoracic echocardiography on hospital admission. We also studied 25 age-matched and gender-matched normal subjects (the control group; mean age, 57 ± 16 years; range, 19–82 years; 8 men) who had no histories of cardiopulmonary disease, no known coronary risk factors, normal electrocardiographic results, and normal echocardiographic results. Twenty-three of 25 control subjects in this study had also been included in the previously published study. Written informed consent was obtained from all subjects, and the protocol was approved for use by the Human Studies Subcommittee of Mie University Graduate School of Medicine.
Echocardiography
All subjects underwent complete transthoracic echocardiography using a Vivid 7 system (GE Vingmed Ultrasound AS, Horten, Norway), as previously described. Patients with acute PE underwent repeat echocardiography on hospital admission (4 ± 5 days after initial symptoms) and after the elimination of RVPO (12 ± 8 days after initial echocardiography) by primary treatments such as thrombolysis, catheter-based pulmonary embolectomy, and/or anticoagulation therapy. Second complete echocardiographic studies were performed after identifying decreased peak systolic RV pressure of <35 mm Hg using repeat bedside echocardiographic examinations, as previously described. Peak systolic RV pressure was calculated from the sum of the estimated mean right atrial pressure and the maximal pressure difference between the right ventricle and the right atrium as calculated by continuous-wave Doppler flow velocity. RV end-diastolic area, end-systolic area, and fractional area change from the apical four-chamber view were measured. The severity of TR was assessed using color Doppler flow mapping of the spatial distribution of the regurgitant jet within the right atrium and was graded as trace, mild, moderate, or severe when the jet area occupied <10%, >10% to 20%, >20% to 33%, or ≥33% of the right atrial area, respectively. To determine the noninvasive estimation of pulmonary vascular resistance (PVR) by echocardiography, the peak TR velocity and the time-velocity integral of the RV outflow tract were measured. Estimation of PVR by echocardiography was calculated as PVR (Wood units) = TR velocity (m/sec)/time-velocity integral of the RV outflow tract (cm) × 10 + 0.16.
LV end-diastolic dimension, end-systolic dimension, and fractional shortening were assessed from the parasternal long-axis view. LV volume and ejection fraction were assessed using the biplane Simpson’s rule. Eccentricity index, defined as the ratio of the LV anterior-to-posterior dimension to the septal-to-lateral dimension at end-diastole from midventricular short-axis images, was used as an index of septal geometric abnormality caused by RV diastolic pressure overload. Cardiac index was calculated as the product of the Doppler-derived stroke volume index and heart rate. All echocardiographic measurements represent the average of three beats.
Speckle-Tracking Displacement and Strain Analysis
Speckle-tracking analysis was used to generate regional LV displacement and strain ( Figures 1-4 ). Radial and circumferential LV regional kinetics were investigated in the parasternal short-axis view at the mid-LV level, and longitudinal LV kinetics were investigated in apical four-chamber views, with a frame rate of 82 ± 19 Hz. Routine B-mode grayscale images were analyzed using commercially available software (EchoPAC; GE Vingmed Ultrasound AS) for frame-by-frame movement of stable patterns of natural acoustic markers present in ultrasound tissue images over the cardiac cycle. The location shift of these acoustic markers from frame to frame, which represents tissue movement, provides the spatial and temporal data used to calculate velocity vectors. Temporal alterations in these stable speckle patterns are identified as moving farther apart or closer together, and a series of regional strain vectors are calculated as change in length divided by initial length. Therefore, myocardial strain is expressed as the percentage change from the original dimension at end-diastole, and myocardial thickening or lengthening is represented as a positive value and myocardial thinning or shortening as a negative value. Strain is theoretically less susceptible to translational motion and tethering artifacts, and thus peak systolic strain (PSS), representing the magnitude of maximal deformation, can depict regional and global myocardial function. Myocardial displacement toward contractile center in the short-axis view or toward the apex in the longitudinal direction was represented as a positive value. The software automatically divided the short-axis and apical four-chamber images into six standard segments ( Figure 1 ). Assessment of circumferential displacement is not possible. Peak systolic displacement (PSD) and PSS obtained from time-displacement and time-strain curves were defined as the indices of LV systolic wall motion and myocardial systolic deformation, respectively. The standard deviation of the heart rate–corrected regional time to PSD and time to PSS was used to quantify LV systolic dyssynchrony (PSD dyssynchrony and PSS dyssynchrony, respectively). Heterogeneity of LV regional systolic wall motion and myocardial deformation was analyzed by calculating the standard deviation of segmental PSD and PSS and were divided by their averaged values (PSD heterogeneity and PSS heterogeneity, respectively). If there were multiple distinct peaks, the largest peak was taken as PSD or PSS. Although PSD or PSS occurs at or near aortic valve closure, the timing of these events may be shortened or prolonged, occurring well after the aortic valve closure in various cardiac diseases. Accordingly, time to PSD and time to PSS were measured throughout the whole cardiac cycle. Mechanical discoordination of LV regional wall motion and myocardial deformation (PSD discoordination and PSS discoordination, respectively) was assessed on the basis of identifying the frame-by-frame discordance between segmental and global signal changes (i.e., velocity for displacement and strain rate for strain, respectively) in the six segments. Segmental signals were defined as discordant if their changes were opposite to the simultaneous change in the global LV signal, which was reconstructed by averaging simultaneous six-site signals at each time point, and overall LV discoordination was calculated as the percentage of discordance in all segments within the specified time interval, from QRS onset to the end of isovolumic relaxation defined by mitral inflow Doppler imaging ( Figure 5 ).
Intraobserver and interobserver variability were analyzed in 10 randomly selected studies and are expressed as the mean percentage error (difference/mean).
Statistical Analysis
Group data (mean ± SD) were compared using two-tailed Student’s t tests or paired t tests, and multiple comparisons were corrected for by using Bonferroni’s method. Correlations were determined using Pearson’s product-moment correlation analysis. P values < .05 were considered statistically significant. Analyses were performed using SPSS for Windows version 11.5 (SPSS, Inc., Chicago, IL).
Results
Clinical and Echocardiographic Characteristics
Table 1 shows the clinical characteristics of the study subjects. Fifteen patients exhibited massive acute PE representing RV dysfunction and hemodynamic exacerbation, and 10 patients exhibited submassive acute PE representing RV dysfunction without hemodynamic instability. All patients received continuous intravenous heparin infusions. Thrombolytic agents such as urokinase or tissue plasminogen activator were administered in 7 patients. Catheter-based pulmonary embolectomies were performed in 4 patients. Although there were no statistical differences in systolic and diastolic blood pressures, heart rates were higher in patients with acute PE on admission compared with those after treatment. There were no statistical differences in QRS duration among the three groups.
| Variable | Normal subjects ( n = 25) | Acute PE on admission ( n = 25) | Acute PE after treatment ( n = 25) |
|---|---|---|---|
| Age (y) | 57 ± 16 | 59 ± 16 | |
| Men | 8 (32%) | 5 (22%) | |
| Massive/submassive PE | 15/10 | ||
| Height (cm) | 157 ± 9 | 158 ± 9 | |
| Weight (kg) | 54 ± 9 | 60 ± 11 ∗ | |
| Body mass index (kg/m 2 ) | 22 ± 3 | 24 ± 3 ∗ | |
| Systolic blood pressure (mm Hg) | 117 ± 12 | 113 ± 20 | 125 ± 21 |
| Diastolic blood pressure (mm Hg) | 68± 10 | 75 ± 19 | 75 ± 14 |
| Heart rate (beats/min) | 65 ± 10 | 95 ± 15 ∗ | 72 ± 10 ∗ † |
| QRS duration (m sec) | 85 ± 8 | 91 ± 14 | 87 ± 10 |
∗ P <.05 versus normal subjects.
† P < .05 versus patients with acute PE on admission (with Bonferroni’s correction).
Table 2 shows the echocardiographic data of the study subjects. Patients with acute PE on admission had elevated peak systolic RV pressures. All patients were successfully treated, and peak systolic RV pressure decreased to normal levels after treatment. Patients with acute PE had dilated RV chambers and reduced RV fractional area change on admission, and 21 of the 25 subjects (84%) had trace or mild TR; there no subjects with severe TR. Abnormally high PVR was observed in patients with acute PE on admission, and they had smaller LV chamber dimensions with higher eccentricity indices and reduced cardiac indices, but their LV ejection fractions were similar to those of normal subjects. Associated with the reduction of PVR after the elimination of the RVPO, RV chamber size and the degree of TR were reduced, and RV fractional area change was improved; however, these RV geometric and functional parameters were still impaired in patients with acute PE after treatment compared with normal subjects. Associated with the improvement of RV chamber dilatation, LV chamber size, eccentricity index, and cardiac index returned to normal values in patients with acute PE after treatment.
| Variable | Normal subjects ( n = 25) | Acute PE on admission ( n = 25) | Acute PE after treatment ( n = 25) |
|---|---|---|---|
| Peak RVP (mm Hg) | — | 52 ± 15 | 20 ± 9 † |
| RV EDA (cm 2 ) | 14 ± 3 | 23 ± 5 ∗ | 15 ± 3 † |
| RV ESA (cm 2 ) | 7 ± 2 | 17 ± 5 ∗ | 9 ± 3 ∗ † |
| RV FAC (%) | 50 ± 6 | 27 ± 6 ∗ | 45 ± 7 ∗ † |
| TR (%) | — | 13 ± 9 | 4 ± 3 † |
| PVR (Wood units) | — | 4.7± 2.0 | 1.5± 0.5 † |
| LV Dd (mm) | 44 ± 4 | 37 ± 4 ∗ | 44 ± 4 † |
| LV Ds (mm) | 27 ± 5 | 24 ± 3 ∗ | 28 ± 4 † |
| LV FS | 0.39 ± 0.06 | 0.36 ± 0.04 | 0.37 ± 0.06 |
| LV eccentricity index | 1.1 ± 0.1 | 1.5 ± 0.3 ∗ | 1.1 ± 0.1 † |
| EDV (mL) | 63 ± 17 | 45 ± 12 ∗ | 64 ± 16 † |
| ESV (mL) | 22 ± 10 | 16 ± 6 ∗ | 23 ± 7 † |
| LV EF (%) | 66 ± 6 | 65 ± 7 | 64 ± 4 |
| CI (L/min/m 2 ) | 2.6 ± 0.5 | 2.2 ± 0.5 ∗ | 2.5 ± 0.6 † |
∗ P < .05 versus normal subjects.
† P < .05 versus patients with acute PE on admission (with Bonferroni’s correction).
Displacement and Strain Measurements
Speckle tracking was possible in 100% and 96% of 450 attempted segments in the short-axis and apical four-chamber views from the 75 echocardiographic studies with technically adequate images, respectively. A normal subject had coordinated segmental displacement and strain through a cardiac cycle in all three directions ( Figure 2 , Videos 1 and 2 ; 
view video clips online). A distinctive paradoxical septal motion characterized as early systolic outward and postsystolic inward motion was observed associated with leftward shift of interventricular septum in the radial displacement imaging ( Figure 3 , top left ; Video 3 ; 
view video clip online), whereas radial strain imaging showed coordinated segmental myocardial thickening in a patient with massive acute PE ( Figure 3 , bottom left ). In the longitudinal direction, discoordinated LV regional wall motion ( Figure 3 , top middle ; Video 4 ; 
view video clip online) and myocardial shortening ( Figure 3 , bottom middle ) between the inferoseptum and the lateral wall were clearly observed in the same patient with massive acute PE. Discoordinated segmental strain was also observed in the same patient with massive acute PE in the circumferential direction ( Figure 3 , bottom right ). The abnormal segmental wall motion and myocardial deformation were restored after the elimination of acute RVPO ( Figure 4 ). Table 3 and Figure 6 show comparisons of PSD and PSS in normal subjects, patients with acute PE on admission, and after treatment. Global and regional radial PSD and PSS were reduced in patients with acute PE on admission compared with normal subjects, and PSD but not PSS recovered to normal values after treatment. Global and regional longitudinal PSD in the basal and mid segments were reduced in patients with acute PE on admission compared with normal subjects, and they recovered after treatment. Global and regional longitudinal PSS except in the basal lateral wall were reduced in patients with acute PE on admission. Longitudinal PSS in the apical segments and midlateral wall improved after treatment associated with restoration of global PSS; however, PSS in the basal and mid inferoseptum did not recover after the elimination of RVPO. Global and regional circumferential PSS were impaired in patients with acute PE on admission, and they recovered to normal values after treatment. Table 3 and Figure 7 show comparisons of LV dyssynchrony, heterogeneity, and discoordination assessed using displacement and strain imaging in the three groups. All three indices of displacement-derived radial LV nonuniformities were significantly greater in patients with acute PE on admission compared with normal subjects and improved after treatment; however, indices of strain-derived radial LV nonuniformities were similar among the three groups. In the longitudinal direction, all indices of LV regional nonuniformities but not PSD heterogeneity were significantly augmented in patients with acute PE on admission, and they significantly improved after treatment, except PSS heterogeneity. In the circumferential direction, all indices of strain-derived LV nonuniformities were significantly greater in patients with acute PE on admission compared with normal subjects and improved after treatment. Figure 8 shows the relationship between LV eccentricity index and indices of LV discoordination. Radial and longitudinal PSD discoordination and longitudinal and circumferential PSS discoordination were significantly correlated with eccentricity index in all subjects, but radial PSS discoordination was not. Multivariate linear regression analysis was used to identify echocardiographic variables (RV end-diastolic area, RV fractional area change, LV end-diastolic dimension, LV ejection fraction, eccentricity index, global radial strain, global circumferential strain, global longitudinal strain, and indices of LV regional nonuniformities in the three directions) that might determine cardiac index with a stepwise method. The results indicated that radial PSD discoordination was only an independent determinant of cardiac index, with a standardized coefficient of −0.378.
