Original article
Impaired right and left ventricular mechanics in adults with pulmonary hypertension and congenital shunts Rocio Toroa, Maria Luisa Cabeza-Letra´nb, Maribel Quezadac, Maria Jose Rodriguez-Purasb and Alipio Mangasa Aims To assess left ventricle mechanics in Eisenmenger physiology patients with congenital shunts, and their relationship with the right ventricle, and to consider the clinical usefulness of this information. Methods The study involved 28 patients with pulmonary artery hypertension (PAH) and congenital shunt, matched with 28 healthy participants. Standard echocardiography and pulsed wave tissue Doppler imaging were employed to analyze systolic and diastolic ventricular function, the myocardial performance index (MPI) of ventricles, and the strain and strain rate along the left ventricle lateral wall, septum, and right ventricle free wall.
six-minute walking test results were correlated negatively with left ventricle-MPI (r U S0.69, P < 0.001), whereas the functional class was positively correlated (r U 0.36, P < 0.001). In conclusion, left ventricle mechanics and geometry are impaired in Eisenmenger syndrome patients, although conventional evaluation is in the normal range. Our results highlight the significance of ventricular interdependence in PAH and provide a useful tool for improving the clinical management of these patients. J Cardiovasc Med 2016, 17:209–216
Results The left ventricle ejection fraction was similar in the two groups. However, despite normal standard left ventricle measures, patients presented parameters of defective myocardial mechanics: mitral peak systolic velocity (S0 ) (cm/s) (8.6 (7.6–10.9) vs. 10.7 (8.6–12.5); P U 0.002) was higher, whereas left ventricle-MPI was lower (0.54 W 01 vs. 0.32 W 0.07, P < 0.001). Right ventricle-MPI and right ventricle global strain were correlated significantly with left ventricle-MPI and left ventricle global strain (r U 0.74, P < 0.001; r U 0.442, P < 0.001, respectively). Clinically, the
Keywords: congenital shunts, echocardiography, Eisenmenger syndrome, left ventricular function
Introduction
more sensitive measures of early left ventricle dysfunction are considered to be required, although newer echocardiographic techniques have helped to identify myocardial mechanical abnormalities. Assessment of myocardial deformation derived from tissue Doppler imaging (TDI), which assesses regional and global myocardial function, is less influenced by loading conditions, has excellent temporal resolution, and has proven to be a well tolerated and reproducible representation of global systolic and diastolic function.9 In many studies, TDI has been considered a helpful and accessible tool for evaluating the right ventricle with reliability in Eisenmenger physiology patients, demonstrating that the myocardial performance index (MPI) of the tissue is superior to other conventional right ventricle parameters, and highlighting the advantages such as its independence from the right ventricle geometry.10,11 Lammers et al.12 have confirmed that quantitative assessment by TDI of ventricular function and ventricular–ventricular interactions in children with pulmonary hypertension may facilitate the understanding of the mechanism involved in this disease.
Pulmonary artery hypertension (PAH) associated with the systemic-pulmonary shunt is one of the most severe complications of adult congenital heart disease.1 The early adaptation of the right ventricle to the nearsystemic pulmonary arterial pressures has been the subject of several previous studies and is considered one of the reasons why these patients evolve relatively better than those with other forms of PAH.2–4 In fact, echocardiographic parameters of right ventricle function and right atrial area have recently been reported to predict mortality in Eisenmenger physiology patients.5 Furthermore, the right ventricle does not work in isolation and, as a consequence of a leftward shift of the interventricular septum secondary to the PAH, the left ventricle geometry and size may be impaired.6 Its systolic function, which is associated with prognosis, may also be compromised.7 On the contrary, the measurement of the left ventricle ejection fraction (LVEF) presents a number of challenges related to image quality, assumptions of left ventricle geometry, load dependence, and expertise.8 Therefore, 1558-2027 ß 2016 Italian Federation of Cardiology. All rights reserved.
a
Department of Medicine, Cadiz University School of Medicine, Cadiz, bAdult Congenital Heart Disease Unit, Clinical Management Area of the Heart, University Hospital ‘Virgen del Rocio’, Seville and cCardiology Department, Hospital Carlos III, Madrid, Spain Correspondence to Rocio Toro, Associate Professor, Department of Medicine, Cadiz University School of Medicine, Cadiz 11009, Spain Tel: +34 956015324; e-mail: [email protected] Received 5 February 2014 Revised 20 May 2014 Accepted 21 May 2014
DOI:10.2459/JCM.0000000000000172
© 2016 Italian Federation of Cardiology. All rights reserved
210 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
In the study reported, we have sought to analyze the mechanics of the left ventricle and the relationship between the two ventricles, in adults with PAH and congenital shunts; based on our findings, we then assess the clinical applicability of the echocardiographic technique used.
Methods Study design
We proposed a prospective observational case–control study conducted from July 2009 to March 2012. Estimated sample size was 48 participants: 24 patients and 24 controls. In fact we recruited consecutively 28 patients with Eisenmenger syndrome and congenital shunts, from the Adult Congenital Heart Disease Unit of the Clinical Management Area of the Heart, Seville. As controls, 28 healthy volunteers were selected and matched for age, sex, and body surface area. Actual sample size was therefore 56 participants. All gave their informed consent to the study, which was approved by the Ethics Committee of the University Hospital ‘Virgen del Rocı´o’ (Seville, Spain). We included all patients attending who presented Eisenmenger physiology with a structurally biventricular heart and a morphological left ventricle supporting the systemic circulation. Four patients presented an atrial septal defect and 24 had a posttricuspid defect (ventricular septal defect, atrioventricular septal defect, persistent ductus arteriosus, truncus arteriosus, and aortopulmonary window). The exclusion criteria were as follows: patients with congenital heart defects affecting the left heart; patients with a single ventricle physiology or with transposition of the great arteries; patients with pulmonary hypertension in which the subaortic ventricle was the morphological right ventricle; patients operated for congenital heart disease; and patients with a poor echocardiographic window. Clinical evaluation
All our patients were graded following the New York Heart Association (NYHA) classification to determine their functional situation. The six-minute walking test (SMWT) was also performed according to the centerspecific protocols.13–15 Echocardiographic examinations
A comprehensive echocardiographic study, including M-mode, 2-D, and Doppler echocardiography, was performed using a unique commercially available cardiovascular ultrasound system (Model iE33 from Philips, Andover, Massachusetts, USA) by one cardiologist with expertise in congenital echocardiography. Intraobserver variability agreement was excellent, with a Cohen’s kappa coefficient higher than 0.88 in all the measurements. The estimated systolic pulmonary artery pressure (PAP) was calculated as the sum of the trans-tricuspid gradient and the estimated right atrial pressure. A systematic
search was performed using two-dimensional and colorflow Doppler to identify the most complete tricuspid regurgitant jet, followed by continuous-wave Doppler acquisitions of spectral envelopes of the greatest maximum velocity and density. The systolic trans-tricuspid pressure gradient was calculated using the modified Bernoulli equation: P ¼ 4 V2, where V represents the maximum regurgitant velocity in meters per second. To estimate right atrial pressure, measurements of inferior vena cava were made from the long-axis subxiphoid view. The right atrial pressure was estimated using the caval respiratory index as described by Kircher et al.16 Chamber dimensions and Doppler-derived hemodynamics were assessed according to the criteria of the American Society of Echocardiography.17 Left ventricular function was assessed from the apical four-chamber view using the modified biplane method of Simpson.18 The eccentricity index was defined as the ratio between the left ventricle anteroposterior and the septolateral dimension; when this ratio was higher than 1, it was considered abnormal and suggestive of right ventricle overload. Standard measures of global right ventricle function included the fractional area change (FAC) and the MPI, as previously described. The tricuspid annular plane systolic excursion (TAPSE) was also used to quantify the longitudinal function of the right ventricle.19 For assessment of the left ventricle diastolic function, left atrium volume was estimated according to the algorithm recommended by the American and European Societies of Echocardiography, based on mitral inflow E and A wave peak velocities, E wave deceleration time, early mitral annular velocities (E0 ), and left atrium volume index.20 Spectral pulsed TDI was performed using a high temporal resolution (>100 frames/s) with alignment of the sample volume of 3 mm parallel to the lateral mitral annulus and tricuspid annulus in the apical four-chamber view.21 Mitral peak systolic velocity S0 , early diastolic (E0 ), late diastolic (A0 ), and the E0 /A0 ratio were recorded. TDI in combination with pulsed Doppler analysis of mitral filling allowed quantification of the E/E0 ratio of the left ventricle. For off-line assessment of one-dimensional peak global longitudinal left ventricle lateral wall, septum, and right ventricle strain as well as strain rate, a commercially available software package was used (QLAB, Philips, The Netherlands). For strain and strain rate measurements, three operator-selected regions of interest were positioned (basal, mid, and apical segments) along the left ventricular free wall and two were positioned along the right ventricle free wall (mid-basal and mid-apical segments). Longitudinal left ventricle or right ventricle strain was defined as the peak negative value on the strain curve during the entire cardiac cycle. Peak systolic strain and strain rate were calculated as previously described.22
© 2016 Italian Federation of Cardiology. All rights reserved
Ventricular mechanics in congenital shunts Toro et al. 211
Statistical analysis
Continuous variables were expressed as mean standard deviation, using the median (interquartile range) in case of asymmetry. Categoric variables were expressed as percentage frequencies. Between-group comparisons of categoric data were performed with the Pearson chisquared test. Quantitative variables were compared using unpaired Student’s t tests and equivalent nonparametric Mann-Whitney U tests. The correlation between quantitative parameters was determined by Pearson’s correlation analysis, or by Spearman’s correlation coefficient in case of asymmetry. A linear regression analysis was performed to establish the correlation of left ventricle and right ventricle parameters vs. left ventricle eccentricity index. Multivariate logistic regression models using a forward stepwise approach were constructed. Statistical significance was P < 0.05. We used the statistical package SPSS (version 19.SPSS Inc., Chicago, Illinois, USA).
Results Case–control study
The mean age of the patients was 37 15 years, and that of the healthy participants 36 11 years (P ¼ 0.72). The body surface area was 1.68 0.29 and 1.78 0.27 m2 for the patients and controls, respectively (P ¼ 0.21). Of the 28 patients, 24 showed posttricuspid defects and four showed pretricuspid defects. Clinically, three patients (10.7%) were in NYHA functional class I, 22 (78.2%) were in class II, and three (10.7%) were classified as class III. The SMWT distances were 358 127 and 627 63 m for patients and controls, respectively. Overall, 18 patients Table 1
(40.9%) were in advanced therapy at the time of the echocardiographic examination: four (14.2%) were receiving an endothelin receptor antagonist such as bosentan; six (21.4%) were receiving a phosphodiesterase-5 inhibitor, sildenafil; seven (25%) were receiving combined treatment (bosentan and sildenafil); and one (3.5%) patient was receiving treatment with an inhaled prostacyclin analogue (iloprost). Echocardiographic examinations: standard echocardiography
According to Doppler echocardiographic examination, the patients had elevated PAP (systolic PAP 93.4 17 mmHg, diastolic PAP 36.4 14 mmHg, mean PAP 57 18 mmHg). Standard echocardiographic parameters are shown in Table 1. Comparing both groups, patients’ right ventricle size was preserved, although it was higher than in the control group, whereas left ventricle values were smaller. Furthermore, the right atrium was significantly dilated, unlike the left atrium volume, which was conserved. Although the systolic function of both ventricles was sustained, patients had worse right ventricle systolic function parameters than healthy subjects, such as TAPSE or right ventricle FAC. The systolic eccentricity index was higher than 1 in the patients, but not in the control group. Diastolic function
Considering left ventricle diastolic dysfunction in the patient group compared with the controls, the mitral E/A velocity ratio was reversed, showing an impaired relaxation pattern, but with relatively normal filling
Standard echocardiographic parameters in both groups
Right ventricle free wall thickness (mm) Right ventricle diastolic diameter (basal) (mm) Right ventricle diastolic diameter (longitudinal) (mm) Right ventricle end-diastolic area (cm2) Right ventricle end-systolic area (cm2) Right ventricle fractional area change (%) TAPSE (mm) Right atrial volume index (ml/m2) Tricuspid E velocity (cm/s) Tricuspid A velocity (cm/s) Tricuspid E/A ratio Left ventricle posterior wall thickness (mm) Interventricular septum thickness (mm) Left ventricle end-diastolic diameter (mm) Left ventricle end-systolic diameter (mm) Left ventricle end-diastolic volume index (ml/m2) Left ventricle end-systolic volume index (ml/m2) Left ventricle fractional shortening (%) Left ventricle ejection fraction (%) left atrium volume index (ml/m2) Mitral E velocity (cm/s) Mitral A velocity (cm/s) E deceleration time (ms) E/E0 Mitral E/A ratio Left ventricle eccentricity index
Cases, N ¼ 28
Controls, N ¼ 28
P
9.15 (8.57–11) 37.4 8.3 59.2 11 17.5 6.1 9.7 (6.67–15) 43 10 17 (15–19) 29.2 (21.2–37) 52.2 17.6 53 (44–49.5) 0.94 0.39 9.83 2.3 11.8 2.8 39.5 (34.2–49.5) 23 (19.7–28.7) 26.9 (23.4–38.3) 9.1 (7–14.3) 40.1 9.4 65.6 7 18.2 (11.7–22.5) 73 29 70 25 218 53 6.54 (4.55–8.3) 1.03 0.41 1.4 (1.28–1.61)
3.4 (3.15–3.8) 30.5 4.4 52 6.3 11.9 2.46 6.15 (5.2–7.1) 49 8 20.7 (20–24) 15.2 (13.2–18.9) 56.4 8.8 37.2 (32.1–40.5) 1.53 0.3 8.82 1.78 8.9 1.57 44.5 (40.8–47.7) 25 (23–28.7) 38.5 (32.8–48.6) 13.6 (11.3–14.8) 41.2 7.1 65.4 6.7 15.25 (12.3–20.4) 78 15 55.4 17.4 173 28 4.99 (4.17–6.18) 1.5 0.45 1.01 (1–1.02)
0.01 0.001 0.007 0.001 0.001 0.022 0.001 0.001 0.27 0.001 0.001 0.076 0.001 0.029 0.14 0.001 0.007 0.6 0.9 0.43 0.44 0.02 0.04 0.004 0.002 0.001
Quantitative variables were expressed as mean standard deviation or as median (interquartile range) in case of asymmetry. TAPSE, tricuspid annular plane systolic excursion.
© 2016 Italian Federation of Cardiology. All rights reserved
212 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
Table 2
Tissue Doppler imaging echocardiographic parameters in both groups
Right ventricle (lateral tricuspid annulus) S0 (cm/s) E0 (cm/s) A0 (cm/s) E0 /A0 E/E0 MPI Global right ventricle strain (%) Global right ventricle strain rate (s1) Left ventricle (lateral mitral annulus) S0 (cm/s) E0 (cm/s) A0 (cm/s) E0 /A0 MPI Global left ventricle strain (%) Global left ventricle strain rate (s1) Septum Global septum strain (%) Global septum strain rate (s1)
Cases
Controls
P
9.58 (7.8–11.2) 8.3 (6–11.2) 12.1 (11–13.5) 0.7 0.28 5.06 (4.42–7.82) 0.67 0.2 18 9 1.05 (0.78 to 1.28)
13.2 (11.8–15) 12.9 (9.65–25) 9.75 (7.8–12.4) 1.38 0.5 4.8 (3.89–4.98) 0.31 0.08 29.26 5.5 1.27 (1.08 to 1.76)
0.001 0.001 0.007 0.001 0.003 0.001 0.001 0.02
8.6 (7.6–10.9) 10.5 3.6 11.2 (8.39–13.8) 0.91 (0.63–1.22) 0.54 0.1 18.6 (13.5 to 21.2) 0.81 (0.65 to 1.19)
10.7 (8.6–12.5) 15.2 3 7.25 (6.17–9) 2.17 (1.56–2.68) 0.32 0.07 21.37 (19.8 to 24) 1.26 (1.03 to 1.64)
0.002 0.001 0.001 0.001 0.001 0.002 0.001
11.8 5.8 0.71 0.43
19.3 3.2 1.06 0.25
0.001 0.001
Quantitative variables were expressed as mean standard deviation or as median (interquartile range) in case of asymmetry. A0 , peak late-diastolic myocardial velocity derived by pulsed-wave tissue Doppler; E0 , peak early-diastolic myocardial velocity derived by pulsed-wave tissue Doppler; MPI, myocardial performance index obtained by tissue Doppler image; S0 , peak systolic myocardial velocity derived by pulsed-wave tissue Doppler.
pressure based on the left atrium volume. The E/E0 ratio was preserved, although the E0 velocity of the mitral annulus was reduced (Table 2). Systolic function and strain imaging
Table 2 shows the echocardiographic TDI results of both populations. Mitral peak systolic velocity S0 , strain, and strain rate of the left ventricular free wall and septum were significantly lower in patients than in controls. The left ventricle-MPI was also significantly higher in our patients. At the level of the tricuspid annulus, patients presented a lower peak systolic velocity (S0 ) and a higher MPI according to our standard echocardiographic results. The strain and strain rate of the lateral tricuspid annulus were also significantly lower, compared with controls. When we analyzed the association between clinical and echocardiographic parameters, the SMWT and the functional class showed the following results: subjects who walked 332 m or more presented significantly better echocardiographic measures and strain than those who did not (left ventricle free wall strain (%): 20 vs. 17, P < 0.05; lateral mitral S0 wave: 10.40 vs. 8.67 cm/s; P < 0.02). Considering the functional class, the left ventricle mechanics of those patients with a stage equal to or greater than II presented worse parameters of myocardial deformation and diastolic function (left ventricle free wall strain (%): 16.7 vs. 22.1, P < 0.001; lateral mitral S0 wave: 8.7 vs. 11.03 cm/s; P < 0.001) than the rest. In the study group, the right ventricle-MPI correlated with the left ventricle-MPI (r ¼ 0.746, P ¼ 0.001), and right ventricle global strain correlated with left ventricle global strain (r ¼ 0.442, P < 0.001). The septum global strain correlated with the left ventricle and right ventricle global strain (r ¼ 0.433, P < 0.001; r ¼ 0.416, P ¼ 0.002,
respectively). The correlations obtained with the index of eccentricity are shown in Table 3. In a bivariate correlation analysis, the SMWT showed a strong negative correlation with left ventricle-MPI (R2 ¼ 0.69, P < 0.0001), but correlated positively with the S0 lateral mitral annulus (R2 ¼ 0.28, P < 0.001), left ventricle global strain rate and strain (R2 ¼ 0.21, P < 0.001, respectively). Likewise, in the bivariate correlation performed by including the NYHA functional classification, the left ventricle-MPI was positively (R2 ¼ 0.34, P < 0.001) correlated and so was the S0 wave velocity (R2 ¼ 0.236, P < 0.001) and the left ventricle global strain (R2 ¼ 0.234, P < 0.001) but weakly. Subsequently, in order to prove the strength of the left ventricle and the right ventricle TDI measures vs. eccentricity index, a regression lineal analysis was performed (Table 4.) A multivariate analysis showed that abnormal left ventricle TDI measures persisted, even after adjusting for PAPs and left ventricle eccentricity index as covariates. Thus, left ventricle-MPI showed a b: 0.62 Table 3 Correlation of the strain imaging and systolic function and eccentricity index Eccentricity index
P
0.433 0.366 0.549 0.181 0.542 0.382 0.421 0.689 0.527 0.605
0.001 0.006 0.001 0.191 0.001 0.004 0.001 0.001 0.001 0.001
Global left ventricle strain Global left ventricle strain rate Global right ventricle strain Global right ventricle strain rate Global septum strain Global septum strain rate Left ventricle TDI S0 velocity Left ventricle-MPI Right ventricle TDI S0 velocity Right ventricle-MPI
MPI, myocardial performance index; TDI, tissue Doppler imaging.
© 2016 Italian Federation of Cardiology. All rights reserved
Ventricular mechanics in congenital shunts Toro et al. 213
(0.76–1.52; < 0001) and S0 wave lateral mitral annulus revealed a b: 0.37 (0.07 to 0.01; P ¼ 0005).
Discussion Right ventricle systolic function
Our results confirm that the deterioration of the left ventricle myocardial mechanics in patients with PAH and congenital shunts is not only because of the right heart dysfunction but also because of intrinsic left ventricle impairment, although standard echocardiography reveals a normal LVEF. Despite the LVEF being in normal range, the impairment of the septum and the lateral wall longitudinal deformation measures can be explained by the damaged myocardial fiber. The limitations of conventional echocardiography for assessing the right ventricle function are even more evident during the evaluation of the left ventricle in patients with Eisenmenger physiology. Recently, the right ventricle function in patients with Eisenmenger physiology has been evaluated by standard echocardiography, thus corroborating our results.23 Traditional measurements like TAPSE or FAC showed lower values, but without confirming global right ventricle dysfunction. In fact, both parameters have been associated with a worse outcome in this population. On this point, Kalogeropoulos et al.4 have observed that although right ventricle function in the short axis was preserved by anatomical M mode, when more sensitive techniques such as TDI and strain or strain rate were used, the global right ventricle was found to be depressed with increased right ventricle-MPI and decreased global strain and strain rate values, thus confirming our results. It is known that the right ventricle cardiac output, under normal circumstances, is mostly produced through longitudinal strain, in contrast to the left ventricle wherein the function in the short axis is the main driving force. In fact, experimental studies have reported that right ventricle longitudinal deformation correctly expresses its global function.23 Thus, adaptation in PAH is performed by circumferential fiber development that could affect more distal segments of the septum, resulting in an increased thickness of the septum and longitudinal mechanical Univariate logistic regression analyses performed for all potential right ventricle and left ventricle echocardiographic measure predictors of left ventricle systolic dysfunction
Table 4
Left ventricle-MPI Right ventricle-MPI S0 wave lateral mitral annulus Global right ventricle strain Global right ventricle strain rate Global septum strain Global septum strain rate Global left ventricle Strain Global left ventricle strain rate
OR
95% CI
P
0.56 0.70 0.38 0.57 0.18 0.54 0.7 0.57 0.21
0.71–1.75 0.66–1.21 0.09 to 0.02 0.01–0.03 0.04–0.22 0.01–0.042 0.00–0.36 0.01–0.026 0.36 to 0.28
Impaired right and left ventricular mechanics in adults with pulmonary hypertension and congenital shunts Rocio Toroa, Maria Luisa Cabeza-Letra´nb, Maribel Quezadac, Maria Jose Rodriguez-Purasb and Alipio Mangasa Aims To assess left ventricle mechanics in Eisenmenger physiology patients with congenital shunts, and their relationship with the right ventricle, and to consider the clinical usefulness of this information. Methods The study involved 28 patients with pulmonary artery hypertension (PAH) and congenital shunt, matched with 28 healthy participants. Standard echocardiography and pulsed wave tissue Doppler imaging were employed to analyze systolic and diastolic ventricular function, the myocardial performance index (MPI) of ventricles, and the strain and strain rate along the left ventricle lateral wall, septum, and right ventricle free wall.
six-minute walking test results were correlated negatively with left ventricle-MPI (r U S0.69, P < 0.001), whereas the functional class was positively correlated (r U 0.36, P < 0.001). In conclusion, left ventricle mechanics and geometry are impaired in Eisenmenger syndrome patients, although conventional evaluation is in the normal range. Our results highlight the significance of ventricular interdependence in PAH and provide a useful tool for improving the clinical management of these patients. J Cardiovasc Med 2016, 17:209–216
Results The left ventricle ejection fraction was similar in the two groups. However, despite normal standard left ventricle measures, patients presented parameters of defective myocardial mechanics: mitral peak systolic velocity (S0 ) (cm/s) (8.6 (7.6–10.9) vs. 10.7 (8.6–12.5); P U 0.002) was higher, whereas left ventricle-MPI was lower (0.54 W 01 vs. 0.32 W 0.07, P < 0.001). Right ventricle-MPI and right ventricle global strain were correlated significantly with left ventricle-MPI and left ventricle global strain (r U 0.74, P < 0.001; r U 0.442, P < 0.001, respectively). Clinically, the
Keywords: congenital shunts, echocardiography, Eisenmenger syndrome, left ventricular function
Introduction
more sensitive measures of early left ventricle dysfunction are considered to be required, although newer echocardiographic techniques have helped to identify myocardial mechanical abnormalities. Assessment of myocardial deformation derived from tissue Doppler imaging (TDI), which assesses regional and global myocardial function, is less influenced by loading conditions, has excellent temporal resolution, and has proven to be a well tolerated and reproducible representation of global systolic and diastolic function.9 In many studies, TDI has been considered a helpful and accessible tool for evaluating the right ventricle with reliability in Eisenmenger physiology patients, demonstrating that the myocardial performance index (MPI) of the tissue is superior to other conventional right ventricle parameters, and highlighting the advantages such as its independence from the right ventricle geometry.10,11 Lammers et al.12 have confirmed that quantitative assessment by TDI of ventricular function and ventricular–ventricular interactions in children with pulmonary hypertension may facilitate the understanding of the mechanism involved in this disease.
Pulmonary artery hypertension (PAH) associated with the systemic-pulmonary shunt is one of the most severe complications of adult congenital heart disease.1 The early adaptation of the right ventricle to the nearsystemic pulmonary arterial pressures has been the subject of several previous studies and is considered one of the reasons why these patients evolve relatively better than those with other forms of PAH.2–4 In fact, echocardiographic parameters of right ventricle function and right atrial area have recently been reported to predict mortality in Eisenmenger physiology patients.5 Furthermore, the right ventricle does not work in isolation and, as a consequence of a leftward shift of the interventricular septum secondary to the PAH, the left ventricle geometry and size may be impaired.6 Its systolic function, which is associated with prognosis, may also be compromised.7 On the contrary, the measurement of the left ventricle ejection fraction (LVEF) presents a number of challenges related to image quality, assumptions of left ventricle geometry, load dependence, and expertise.8 Therefore, 1558-2027 ß 2016 Italian Federation of Cardiology. All rights reserved.
a
Department of Medicine, Cadiz University School of Medicine, Cadiz, bAdult Congenital Heart Disease Unit, Clinical Management Area of the Heart, University Hospital ‘Virgen del Rocio’, Seville and cCardiology Department, Hospital Carlos III, Madrid, Spain Correspondence to Rocio Toro, Associate Professor, Department of Medicine, Cadiz University School of Medicine, Cadiz 11009, Spain Tel: +34 956015324; e-mail: [email protected] Received 5 February 2014 Revised 20 May 2014 Accepted 21 May 2014
DOI:10.2459/JCM.0000000000000172
© 2016 Italian Federation of Cardiology. All rights reserved
210 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
In the study reported, we have sought to analyze the mechanics of the left ventricle and the relationship between the two ventricles, in adults with PAH and congenital shunts; based on our findings, we then assess the clinical applicability of the echocardiographic technique used.
Methods Study design
We proposed a prospective observational case–control study conducted from July 2009 to March 2012. Estimated sample size was 48 participants: 24 patients and 24 controls. In fact we recruited consecutively 28 patients with Eisenmenger syndrome and congenital shunts, from the Adult Congenital Heart Disease Unit of the Clinical Management Area of the Heart, Seville. As controls, 28 healthy volunteers were selected and matched for age, sex, and body surface area. Actual sample size was therefore 56 participants. All gave their informed consent to the study, which was approved by the Ethics Committee of the University Hospital ‘Virgen del Rocı´o’ (Seville, Spain). We included all patients attending who presented Eisenmenger physiology with a structurally biventricular heart and a morphological left ventricle supporting the systemic circulation. Four patients presented an atrial septal defect and 24 had a posttricuspid defect (ventricular septal defect, atrioventricular septal defect, persistent ductus arteriosus, truncus arteriosus, and aortopulmonary window). The exclusion criteria were as follows: patients with congenital heart defects affecting the left heart; patients with a single ventricle physiology or with transposition of the great arteries; patients with pulmonary hypertension in which the subaortic ventricle was the morphological right ventricle; patients operated for congenital heart disease; and patients with a poor echocardiographic window. Clinical evaluation
All our patients were graded following the New York Heart Association (NYHA) classification to determine their functional situation. The six-minute walking test (SMWT) was also performed according to the centerspecific protocols.13–15 Echocardiographic examinations
A comprehensive echocardiographic study, including M-mode, 2-D, and Doppler echocardiography, was performed using a unique commercially available cardiovascular ultrasound system (Model iE33 from Philips, Andover, Massachusetts, USA) by one cardiologist with expertise in congenital echocardiography. Intraobserver variability agreement was excellent, with a Cohen’s kappa coefficient higher than 0.88 in all the measurements. The estimated systolic pulmonary artery pressure (PAP) was calculated as the sum of the trans-tricuspid gradient and the estimated right atrial pressure. A systematic
search was performed using two-dimensional and colorflow Doppler to identify the most complete tricuspid regurgitant jet, followed by continuous-wave Doppler acquisitions of spectral envelopes of the greatest maximum velocity and density. The systolic trans-tricuspid pressure gradient was calculated using the modified Bernoulli equation: P ¼ 4 V2, where V represents the maximum regurgitant velocity in meters per second. To estimate right atrial pressure, measurements of inferior vena cava were made from the long-axis subxiphoid view. The right atrial pressure was estimated using the caval respiratory index as described by Kircher et al.16 Chamber dimensions and Doppler-derived hemodynamics were assessed according to the criteria of the American Society of Echocardiography.17 Left ventricular function was assessed from the apical four-chamber view using the modified biplane method of Simpson.18 The eccentricity index was defined as the ratio between the left ventricle anteroposterior and the septolateral dimension; when this ratio was higher than 1, it was considered abnormal and suggestive of right ventricle overload. Standard measures of global right ventricle function included the fractional area change (FAC) and the MPI, as previously described. The tricuspid annular plane systolic excursion (TAPSE) was also used to quantify the longitudinal function of the right ventricle.19 For assessment of the left ventricle diastolic function, left atrium volume was estimated according to the algorithm recommended by the American and European Societies of Echocardiography, based on mitral inflow E and A wave peak velocities, E wave deceleration time, early mitral annular velocities (E0 ), and left atrium volume index.20 Spectral pulsed TDI was performed using a high temporal resolution (>100 frames/s) with alignment of the sample volume of 3 mm parallel to the lateral mitral annulus and tricuspid annulus in the apical four-chamber view.21 Mitral peak systolic velocity S0 , early diastolic (E0 ), late diastolic (A0 ), and the E0 /A0 ratio were recorded. TDI in combination with pulsed Doppler analysis of mitral filling allowed quantification of the E/E0 ratio of the left ventricle. For off-line assessment of one-dimensional peak global longitudinal left ventricle lateral wall, septum, and right ventricle strain as well as strain rate, a commercially available software package was used (QLAB, Philips, The Netherlands). For strain and strain rate measurements, three operator-selected regions of interest were positioned (basal, mid, and apical segments) along the left ventricular free wall and two were positioned along the right ventricle free wall (mid-basal and mid-apical segments). Longitudinal left ventricle or right ventricle strain was defined as the peak negative value on the strain curve during the entire cardiac cycle. Peak systolic strain and strain rate were calculated as previously described.22
© 2016 Italian Federation of Cardiology. All rights reserved
Ventricular mechanics in congenital shunts Toro et al. 211
Statistical analysis
Continuous variables were expressed as mean standard deviation, using the median (interquartile range) in case of asymmetry. Categoric variables were expressed as percentage frequencies. Between-group comparisons of categoric data were performed with the Pearson chisquared test. Quantitative variables were compared using unpaired Student’s t tests and equivalent nonparametric Mann-Whitney U tests. The correlation between quantitative parameters was determined by Pearson’s correlation analysis, or by Spearman’s correlation coefficient in case of asymmetry. A linear regression analysis was performed to establish the correlation of left ventricle and right ventricle parameters vs. left ventricle eccentricity index. Multivariate logistic regression models using a forward stepwise approach were constructed. Statistical significance was P < 0.05. We used the statistical package SPSS (version 19.SPSS Inc., Chicago, Illinois, USA).
Results Case–control study
The mean age of the patients was 37 15 years, and that of the healthy participants 36 11 years (P ¼ 0.72). The body surface area was 1.68 0.29 and 1.78 0.27 m2 for the patients and controls, respectively (P ¼ 0.21). Of the 28 patients, 24 showed posttricuspid defects and four showed pretricuspid defects. Clinically, three patients (10.7%) were in NYHA functional class I, 22 (78.2%) were in class II, and three (10.7%) were classified as class III. The SMWT distances were 358 127 and 627 63 m for patients and controls, respectively. Overall, 18 patients Table 1
(40.9%) were in advanced therapy at the time of the echocardiographic examination: four (14.2%) were receiving an endothelin receptor antagonist such as bosentan; six (21.4%) were receiving a phosphodiesterase-5 inhibitor, sildenafil; seven (25%) were receiving combined treatment (bosentan and sildenafil); and one (3.5%) patient was receiving treatment with an inhaled prostacyclin analogue (iloprost). Echocardiographic examinations: standard echocardiography
According to Doppler echocardiographic examination, the patients had elevated PAP (systolic PAP 93.4 17 mmHg, diastolic PAP 36.4 14 mmHg, mean PAP 57 18 mmHg). Standard echocardiographic parameters are shown in Table 1. Comparing both groups, patients’ right ventricle size was preserved, although it was higher than in the control group, whereas left ventricle values were smaller. Furthermore, the right atrium was significantly dilated, unlike the left atrium volume, which was conserved. Although the systolic function of both ventricles was sustained, patients had worse right ventricle systolic function parameters than healthy subjects, such as TAPSE or right ventricle FAC. The systolic eccentricity index was higher than 1 in the patients, but not in the control group. Diastolic function
Considering left ventricle diastolic dysfunction in the patient group compared with the controls, the mitral E/A velocity ratio was reversed, showing an impaired relaxation pattern, but with relatively normal filling
Standard echocardiographic parameters in both groups
Right ventricle free wall thickness (mm) Right ventricle diastolic diameter (basal) (mm) Right ventricle diastolic diameter (longitudinal) (mm) Right ventricle end-diastolic area (cm2) Right ventricle end-systolic area (cm2) Right ventricle fractional area change (%) TAPSE (mm) Right atrial volume index (ml/m2) Tricuspid E velocity (cm/s) Tricuspid A velocity (cm/s) Tricuspid E/A ratio Left ventricle posterior wall thickness (mm) Interventricular septum thickness (mm) Left ventricle end-diastolic diameter (mm) Left ventricle end-systolic diameter (mm) Left ventricle end-diastolic volume index (ml/m2) Left ventricle end-systolic volume index (ml/m2) Left ventricle fractional shortening (%) Left ventricle ejection fraction (%) left atrium volume index (ml/m2) Mitral E velocity (cm/s) Mitral A velocity (cm/s) E deceleration time (ms) E/E0 Mitral E/A ratio Left ventricle eccentricity index
Cases, N ¼ 28
Controls, N ¼ 28
P
9.15 (8.57–11) 37.4 8.3 59.2 11 17.5 6.1 9.7 (6.67–15) 43 10 17 (15–19) 29.2 (21.2–37) 52.2 17.6 53 (44–49.5) 0.94 0.39 9.83 2.3 11.8 2.8 39.5 (34.2–49.5) 23 (19.7–28.7) 26.9 (23.4–38.3) 9.1 (7–14.3) 40.1 9.4 65.6 7 18.2 (11.7–22.5) 73 29 70 25 218 53 6.54 (4.55–8.3) 1.03 0.41 1.4 (1.28–1.61)
3.4 (3.15–3.8) 30.5 4.4 52 6.3 11.9 2.46 6.15 (5.2–7.1) 49 8 20.7 (20–24) 15.2 (13.2–18.9) 56.4 8.8 37.2 (32.1–40.5) 1.53 0.3 8.82 1.78 8.9 1.57 44.5 (40.8–47.7) 25 (23–28.7) 38.5 (32.8–48.6) 13.6 (11.3–14.8) 41.2 7.1 65.4 6.7 15.25 (12.3–20.4) 78 15 55.4 17.4 173 28 4.99 (4.17–6.18) 1.5 0.45 1.01 (1–1.02)
0.01 0.001 0.007 0.001 0.001 0.022 0.001 0.001 0.27 0.001 0.001 0.076 0.001 0.029 0.14 0.001 0.007 0.6 0.9 0.43 0.44 0.02 0.04 0.004 0.002 0.001
Quantitative variables were expressed as mean standard deviation or as median (interquartile range) in case of asymmetry. TAPSE, tricuspid annular plane systolic excursion.
© 2016 Italian Federation of Cardiology. All rights reserved
212 Journal of Cardiovascular Medicine 2016, Vol 17 No 3
Table 2
Tissue Doppler imaging echocardiographic parameters in both groups
Right ventricle (lateral tricuspid annulus) S0 (cm/s) E0 (cm/s) A0 (cm/s) E0 /A0 E/E0 MPI Global right ventricle strain (%) Global right ventricle strain rate (s1) Left ventricle (lateral mitral annulus) S0 (cm/s) E0 (cm/s) A0 (cm/s) E0 /A0 MPI Global left ventricle strain (%) Global left ventricle strain rate (s1) Septum Global septum strain (%) Global septum strain rate (s1)
Cases
Controls
P
9.58 (7.8–11.2) 8.3 (6–11.2) 12.1 (11–13.5) 0.7 0.28 5.06 (4.42–7.82) 0.67 0.2 18 9 1.05 (0.78 to 1.28)
13.2 (11.8–15) 12.9 (9.65–25) 9.75 (7.8–12.4) 1.38 0.5 4.8 (3.89–4.98) 0.31 0.08 29.26 5.5 1.27 (1.08 to 1.76)
0.001 0.001 0.007 0.001 0.003 0.001 0.001 0.02
8.6 (7.6–10.9) 10.5 3.6 11.2 (8.39–13.8) 0.91 (0.63–1.22) 0.54 0.1 18.6 (13.5 to 21.2) 0.81 (0.65 to 1.19)
10.7 (8.6–12.5) 15.2 3 7.25 (6.17–9) 2.17 (1.56–2.68) 0.32 0.07 21.37 (19.8 to 24) 1.26 (1.03 to 1.64)
0.002 0.001 0.001 0.001 0.001 0.002 0.001
11.8 5.8 0.71 0.43
19.3 3.2 1.06 0.25
0.001 0.001
Quantitative variables were expressed as mean standard deviation or as median (interquartile range) in case of asymmetry. A0 , peak late-diastolic myocardial velocity derived by pulsed-wave tissue Doppler; E0 , peak early-diastolic myocardial velocity derived by pulsed-wave tissue Doppler; MPI, myocardial performance index obtained by tissue Doppler image; S0 , peak systolic myocardial velocity derived by pulsed-wave tissue Doppler.
pressure based on the left atrium volume. The E/E0 ratio was preserved, although the E0 velocity of the mitral annulus was reduced (Table 2). Systolic function and strain imaging
Table 2 shows the echocardiographic TDI results of both populations. Mitral peak systolic velocity S0 , strain, and strain rate of the left ventricular free wall and septum were significantly lower in patients than in controls. The left ventricle-MPI was also significantly higher in our patients. At the level of the tricuspid annulus, patients presented a lower peak systolic velocity (S0 ) and a higher MPI according to our standard echocardiographic results. The strain and strain rate of the lateral tricuspid annulus were also significantly lower, compared with controls. When we analyzed the association between clinical and echocardiographic parameters, the SMWT and the functional class showed the following results: subjects who walked 332 m or more presented significantly better echocardiographic measures and strain than those who did not (left ventricle free wall strain (%): 20 vs. 17, P < 0.05; lateral mitral S0 wave: 10.40 vs. 8.67 cm/s; P < 0.02). Considering the functional class, the left ventricle mechanics of those patients with a stage equal to or greater than II presented worse parameters of myocardial deformation and diastolic function (left ventricle free wall strain (%): 16.7 vs. 22.1, P < 0.001; lateral mitral S0 wave: 8.7 vs. 11.03 cm/s; P < 0.001) than the rest. In the study group, the right ventricle-MPI correlated with the left ventricle-MPI (r ¼ 0.746, P ¼ 0.001), and right ventricle global strain correlated with left ventricle global strain (r ¼ 0.442, P < 0.001). The septum global strain correlated with the left ventricle and right ventricle global strain (r ¼ 0.433, P < 0.001; r ¼ 0.416, P ¼ 0.002,
respectively). The correlations obtained with the index of eccentricity are shown in Table 3. In a bivariate correlation analysis, the SMWT showed a strong negative correlation with left ventricle-MPI (R2 ¼ 0.69, P < 0.0001), but correlated positively with the S0 lateral mitral annulus (R2 ¼ 0.28, P < 0.001), left ventricle global strain rate and strain (R2 ¼ 0.21, P < 0.001, respectively). Likewise, in the bivariate correlation performed by including the NYHA functional classification, the left ventricle-MPI was positively (R2 ¼ 0.34, P < 0.001) correlated and so was the S0 wave velocity (R2 ¼ 0.236, P < 0.001) and the left ventricle global strain (R2 ¼ 0.234, P < 0.001) but weakly. Subsequently, in order to prove the strength of the left ventricle and the right ventricle TDI measures vs. eccentricity index, a regression lineal analysis was performed (Table 4.) A multivariate analysis showed that abnormal left ventricle TDI measures persisted, even after adjusting for PAPs and left ventricle eccentricity index as covariates. Thus, left ventricle-MPI showed a b: 0.62 Table 3 Correlation of the strain imaging and systolic function and eccentricity index Eccentricity index
P
0.433 0.366 0.549 0.181 0.542 0.382 0.421 0.689 0.527 0.605
0.001 0.006 0.001 0.191 0.001 0.004 0.001 0.001 0.001 0.001
Global left ventricle strain Global left ventricle strain rate Global right ventricle strain Global right ventricle strain rate Global septum strain Global septum strain rate Left ventricle TDI S0 velocity Left ventricle-MPI Right ventricle TDI S0 velocity Right ventricle-MPI
MPI, myocardial performance index; TDI, tissue Doppler imaging.
© 2016 Italian Federation of Cardiology. All rights reserved
Ventricular mechanics in congenital shunts Toro et al. 213
(0.76–1.52; < 0001) and S0 wave lateral mitral annulus revealed a b: 0.37 (0.07 to 0.01; P ¼ 0005).
Discussion Right ventricle systolic function
Our results confirm that the deterioration of the left ventricle myocardial mechanics in patients with PAH and congenital shunts is not only because of the right heart dysfunction but also because of intrinsic left ventricle impairment, although standard echocardiography reveals a normal LVEF. Despite the LVEF being in normal range, the impairment of the septum and the lateral wall longitudinal deformation measures can be explained by the damaged myocardial fiber. The limitations of conventional echocardiography for assessing the right ventricle function are even more evident during the evaluation of the left ventricle in patients with Eisenmenger physiology. Recently, the right ventricle function in patients with Eisenmenger physiology has been evaluated by standard echocardiography, thus corroborating our results.23 Traditional measurements like TAPSE or FAC showed lower values, but without confirming global right ventricle dysfunction. In fact, both parameters have been associated with a worse outcome in this population. On this point, Kalogeropoulos et al.4 have observed that although right ventricle function in the short axis was preserved by anatomical M mode, when more sensitive techniques such as TDI and strain or strain rate were used, the global right ventricle was found to be depressed with increased right ventricle-MPI and decreased global strain and strain rate values, thus confirming our results. It is known that the right ventricle cardiac output, under normal circumstances, is mostly produced through longitudinal strain, in contrast to the left ventricle wherein the function in the short axis is the main driving force. In fact, experimental studies have reported that right ventricle longitudinal deformation correctly expresses its global function.23 Thus, adaptation in PAH is performed by circumferential fiber development that could affect more distal segments of the septum, resulting in an increased thickness of the septum and longitudinal mechanical Univariate logistic regression analyses performed for all potential right ventricle and left ventricle echocardiographic measure predictors of left ventricle systolic dysfunction
Table 4
Left ventricle-MPI Right ventricle-MPI S0 wave lateral mitral annulus Global right ventricle strain Global right ventricle strain rate Global septum strain Global septum strain rate Global left ventricle Strain Global left ventricle strain rate
OR
95% CI
P
0.56 0.70 0.38 0.57 0.18 0.54 0.7 0.57 0.21
0.71–1.75 0.66–1.21 0.09 to 0.02 0.01–0.03 0.04–0.22 0.01–0.042 0.00–0.36 0.01–0.026 0.36 to 0.28
