14
Key to abbreviations
2D-STE Two-dimensional speckle tracking echocardiography AAT Aortic outflow acceleration time
AGp Aortic outflow peak velocity pressure gradient AVp Aortic outflow peak velocity
BW Body weight
CS Peak circumferential strain
CSg Global peak circumferential strain CSR Peak circumferential systolic strain rate CSR-A Peak circumferential late diastolic strain rate
CSR-Ag Global peak circumferential late diastolic strain rate CSR-E Peak circumferential early diastolic strain rate
CSR-Eg Global peak circumferential early diastolic strain rate CSRg Global peak circumferential systolic strain rate EFM Ejection fraction derived from m-mode
EFste Ejection fraction derived from speckle tracking echocardiography EPSS E point to septal separation
ESVI End systolic volume index FAC Fractional area change FS Fractional shortening
HR Heart rate
IVRT Isovolumic relaxation time
IVS% Percentage thickening of the interventricular septum IVSd Interventricular septal thickness in end diastole LA/Ao The ratio of left atrium and aortic diameter LS Peak longitudinal strain
LSg Global peak longitudinal strain LSR Peak longitudinal systolic strain rate LSR-A Peak longitudinal late diastolic strain rate
LSR-Ag Global peak longitudinal late diastolic strain rate LSR-E Peak longitudinal early diastolic strain rate
LSR-Eg Global peak longitudinal early diastolic strain rate LSRg Global peak longitudinal systolic strain rate LVDd Left ventricular dimension in end-diastole LVDs Left ventricular dimension in end-systole
LVFW% Percentage thickening of the left ventricular free wall LVFWd Left ventricular free wall thickness in end-diastole
15
LVFWs Left ventricular free wall thickness in end-systole LVMI Left ventricular mass index
MAvp Peak late mitral inflow velocity ME/A Early and late mitral inflow ratio MEVp Peak early mitral inflow velocity
PEP:ET Pre-ejection time period and ejection time ratio PVp Pulmonary outflow peak velocity
QP-Ao Difference between pulmonary pre-ejection time and aortic pre-ejection time
RS Peak radial strain
RSg Global peak radial strain RSR Peak radial systolic strain rate RSR-A Peak radial late diastolic strain rate
RSR-Ag Global peak radial late diastolic strain rate RSR-E Peak radial early diastolic strain rate
RSR-Eg Global peak radial early diastolic strain rate RSRg Global peak radial systolic strain rate
RT-cε Range of the 6 segment time to peak circumferential strain RT-lε Range of the 6 segment time to peak longitudinal strain RT-rε Range of the 6 segment time to peak radial strain RT-ε Range of the 6 segment time to peak strain
SDT-cε Standard deviation of the 6 segment time to peak circumferential strain SDT-lε Standard deviation of the 6 segment time to peak longitudinal strain SDT-rε Standard deviation of the 6 segment time to peak radial strain SDT-ε Standard deviation of the 6 segment time to peak strain Sg Global peak strain
SH-cε Circumferential segmental heterogeneity SH-lε Longitudinal segmental heterogeneity SH-rε Radial segmental heterogeneity
SH-ε Segmental heterogeneity SR Peak systolic strain rate SR-A Peak late diastolic strain rate
SR-Ag Global peak late diastolic strain rate SR-E Peak early diastolic strain rate
SR-Eg Global peak early diastolic strain rate SRg Global peak strain rate
SV Stroke volume
16
TAVp Peak late tricuspid inflow velocity TE/A Early and late tricuspid inflow ratio Tei index Myocardial performance index TEVp Peak early tricuspid inflow velocity TH-ε Transmural heterogeneity
Vcf Velocity of circumferential fiber shortening
ε Langarian strain
17
Two-Dimensional Speckle Tracking Echocardiographic Assessment of
Mechanical Ventricular Synchrony and Heterogeneity in Clinically Normal Cats Yueh-Lun Hsu, Hui-Pi Huang
Department of Veterinary Medicine, National Taiwan University, Taipei, Taiwan Abstract
Two-dimensional speckle tracking echocardiography (2D-STE) makes the regional cardiac deformation and function can be evaluated. Regional myocardial synchrony and heterogeneity in the human cardiac disease can be featured as pathogenesis or the consequence of specific diseases.
The aims of this study were firstly to assess the reproducibility and reproducibility of 2D-STE in clinical healthy feline, secondly to determine the influence of physical characteristic (age, body weight, heart rate, sex and blood pressure) in strain, strain rate (SR), synchrony/heterogeneity of left ventricle, and thirdly to compare the indices derived from 2D-STE with hemodynamic analysis obtained by conventional echocardiography.
Thirty-three clinical healthy cats were included. Both standard conventional echocardiography and 2D-STE were performed. Indices of circumferential, radial and longitudinal strain and SR, synchrony and heterogeneity were collected.
Indices of 2D-STE were almost free from the influence of physiological factors, but there were presence of gender differences. The deformations of left ventricle in clinically healthy cats possess heterogeneity, which the longitudinal strain/SR was lower in apex and circumferential strain was decreasing from subendocardium to subepicardium. Indices derived from 2D-STE were not correlated to M-mode indices of conventional echocardiography, and weakly correlated to pulse-wave Doppler indices. It is implied that regional deformation alteration may compensate by each other and makes 2D-STE more sensitive in detection of left ventricular dysfunction.
Meanwhile, systolic strain or SR was correlated to diastolic indices in conventional echocardiography. It is indicated that the boundary of diastolic and systolic function in the regional deformation may not be aware. Need more diseases and health research or longitudinal study in order to further understand the significance of the
deformation.
Key words: two-dimensional speckle tracking echocardiography, heterogeneity, synchrony, feline
_____________________________________________________________________
Part of the study is going to present as an abstract in the 22nd European College of Veterinary Internal Medicine Congress, Maastricht, the Netherlands. Sep. 6-8, 2012.
18
Introduction
Ventricular contraction normally occurs in a highly coordinated process.
Diseased myocardium is usually resulted from impaired electrophysiological activity (electrical dyssynchrony), and induce electrical conduction delay [1, 2]. Non-uniform in timing of electromechanical activation in different areas of the myocardium can produce myocardial contraction dyssynchrony (mechanical dyssynchrony) [3, 4].
In recent years, ventricular dyssynchrony are found to be associated with causes and consequence both valvular insufficiency and myocardial disease in human patients [5-7]. Cardiac failure would amplify the regional stress disparities that result from discoordinate activation, and this may be an important interaction [8].
Dyssynchrony may induce asymmetric ventricular hypertrophy and alterations in regional myocardial blood flow [8, 9]. Mitral regurgitation, caused by delay in both the rise in left ventricular pressure and discoordinate papillary muscle contraction, can exacerbate this inefficiency further [8]. Presence of dyssnchronous ventricular
contraction is an independent predictor of worsening heart failure [10, 11].
Quantification of ventricular synchrony and heterogeneity in clinical healthy cats is an essential first step in assessment of ventricular synchrony in cats. The aims of this study are to investigate left ventricular synchrony and heterogeneity in clinically healthy cats using two dimensional speckle tracking echocardiography (2D-STE).
Material and Method
Animals
Thirty three client-owned cats admitted to the National Taiwan University Veterinary Hospital during 2010-2012 were included for this perspective study.
Without history of respiratory or cardiac disease, all cats were deemed to be clinically
19
healthy based on physical examination, blood pressure measurements, routine blood work (complete blood counts and biochemical profiles), chest radiographs and
conventional echocardiography. The mean age was 3.11 ± 2.17 years (range 6 months to 8 years old) and a mean body weight was 4.32 ± 1.32 kg, 18 were males, and 16 were females. The represented breeds were 19 Domestic Short-hairs, 6 American Short-hairs, 7 Persians, 1 Bengal cat and 1 British Blue.
Measurement reliability
All echocardiographic studies were performed by 1 examiner (YLH), using an ultrasound unit equipped with a 5.5-7.5 MHz phased-array transducer (MyLabTM50 XVision, Esaote, Genova, Italy).
STE examinations, including circumferential, radial and longitudinal orientation, of five clinical cats were acquired by one echocardiographer (YLH) thrice in one day to calculate intra-observer coefficient of variation. The resultant echocardiograms were analyzed by another investigator(BIY). Each variable was measured 3 times on 3 consecutive cardiac cycles using the same recorded loop. The intra-observer
coefficient of variation (CV) was expressed as a percent value, calculated as CV=Standard deviation/ mean X 100% in order to assess the repeatability of the measurements used in this study.
Blood pressure
Systemic blood pressure was measured by Doppler flow detector and an inflatable cuff attached to a sphygmomanometer which position between the carpal and metacarpal pad. Record the average result of 5 consecutive measurements.
Conventional echocardiography
Ultrasound examinations were performed without sedation in gently restrained cats in lateral recumbency [12].
20
Indices of left ventricle was including left ventricular diastolic dimension
(LVDd), left ventricular systolic dimension (LVDs), the ratio of Left atrium and aortic diameter (LA/Ao), interventricular septal thickness in diastole (IVSd), left ventricular free wall thickness in diastole (LVFWd), and left ventricular mass index (LVMI) were obtained from the standard views [13, 14].
Parameters of left ventricular systolic function were derived from M-mode, including end-posterior E point to septal separation (EPSS), percentage thickening of the interventricular septum (IVS%), percentage thickening of the left ventricular free wall(LVFW%), fractional shortening (FS), ejection fraction by M-mode (EFM);
parameters were derived from pulse-wave Doppler including velocity of
circumferential fiber shortening (Vcf), the ratio of pre-ejection period an ejection time (PEP:ET), Aortic outflow acceleration time (AAT), including aortic outflow peak velocity (AVp), aortic outflow peak velocity pressure gradient (AGp), pulmonary outflow peak velocity (PVp), stroke volume (SV), and derived from B mode, including end systolic volume index (ESVI) were obtained from the standard views [14-23].
Parameters of left ventricular diastolic function were including mitral early and late diastolic inflow velocity (MEVp, MAVp), and tricuspid early and late diastolic inflow velocity (TEVp, TAVp), the ratio between MEVp and MAVp (ME/A), the ratio between TEVp and TAVp (TE/A), isovolumic relaxation time (IVRT), and myocardial performance index (Tei index) were obtained from the standard views [16, 24, 25].
Interventricular synchrony, the difference between pulmonary pre-ejection time and aortic pre-ejection time (QP-Ao) was obtained from standard views [26].
Two-dimensional speckle tracking echocardiography Measurement regional strain and strain rate
21
Indices of left ventricular longitudinal and circumferential/radial/longitudinal strain and strain rates were obtained using right parasternal apical 4-chamber and short axis views, respectively. All images were acquired in cine loops of 3 cardiac cycles recorded at frame rate of 54-111 frames per seconds, saved in digital format, and analyzed by off-line software (XStrainTM software for MyLabTM50 X Vision).
In quantification of circumferential and radial strain and strain, 6 segments (cranio-septum, cranial, lateral, caudal, ventral, and septal) of endocardium were semi-automatically selected by Aided Heart Segmentation (AHS) for analysis. In left ventricular longitudinal strain and strain rate, 6 segments (basal-septal, mid-septal, apical-septal, apical-lateral, mid-lateral, basal-lateral) of endocardium were
semi-automatically automatically selected by the AHS tool for analysis. Left ventricular ejection volume (EFste) and fraction area change (FAC) were also calculated by modified Simpson’s rule.
The 2D-STE indices included in this study were
circumferential/radial/longitudinal peak systolic strain (CS/RS/LS),
circumferential/radial/longitudinal peak systolic strain rate (CRS/RSR/LSR), circumferential/radial/longitudinal peak early diastolic strain rate
(CSR-E/RSR-E/LSR-E) and circumferential/radial/longitudinal peak late diastolic strain rate (CSR-A/RSR-A/LSR-A). The 2D-STE off-line analysis was performed by 1
examiner (YLH).
The images without adequate visualization of one or more segments of the endocardium were excluded from this study.
Measurement of synchrony and heterogeneity
Interventricular synchrony, defined as time difference between mean pulmonary and aortic pre-ejection periods (QP-Ao), was calculated based on conventional pulse
22
wave-Doppler.
Intraventricular synchrony, defined as the standard deviation of the six segments time to reach peak strain (SDT-ε) at longitudinal (SDT-lε), circumferential (SDT-cε) and radial (SDT-rε) directions, and the range of the six segments time to peak strain (RT-ε) at longitudinal (RT-lε), circumferential (RT-cε) and radial (RT-rε) directions were also calculated.
Segmental heterogeneity (SH-ε), defined as the range of the peak strain of the six segments, and transmural heterogeneity (TH-cε), defined as the difference of peak circumferential strain between endocardium and epicardium were also calculated.
Statistical analysis
All statistical analyses were performed with commercial computer static software (SPSS 12.0 Inc., Chicago, Illinois, USA). Data are express as the mean ± standard deviation (SD).
Compare the strain, SR, SR-E and SR-A between six segments by one
way-ANOVA and pairwise comparison by Scheffe’s method as post-hoc analysis if variances agree to variance homogeneity test, if not, using Robust test
(Brown-Forysthe test substitute for ANOVA) and Games-Howell method as post-hoc analysis. A 2-tailed P<0.05 was considered significant.
Spearman correlation was used to examine the linear association between continuous variables. A 2-tailed P<0.05 was considered significant.
Results
The descriptive statistics of the indices derived from conventional echocardiography indices are listed in Table 1.
The intra-observer CV of 2D-STE variables are almost less than 25% except
23
RSR-Ag was 25.39% (Table 2).
Myocardial strain and strain rate
Based on the direction, values for circumferential and longitudinal peak systolic strain and strain rate were negative and values for radial peak systolic strain and peak systolic strain rate were positive for all cats (Table 3). Heterogeneity of strain and strain rate values across the 6 myocardial segments in the longitudinal analysis was found.
In general, circumferential, radial and longitudinal strain and strain rate were not affected by age, bodyweight, heart rate or blood pressure, except LSR-Eg was
negatively correlated with the heart rate (r=-0.591, P=0.002, Table 4). Global circumferential (P=0.015) and radial strain (P=0.018) were lower in male.
Indices of 2D-STE and conventional echocardiography
Either systolic or diastolic indices derive from B- or M-mode echocardiography was not correlated with indices of STE. The Vcf (FS/ET), derived from pw-Doppler and M-mode, was correlated with LSg (r=0.437), LSRg (r=0.518) and CSR-Ag
(r=0.524), and AVp, also an index derived from pw-Doppler, was correlated with CSR-Ag(r=0.478), RSg (r=-0.432), LSg (r=0.473) and LSRg (r=0.454, Table 5).
The TE/A was correlated to CSRg (r=-0.440), CSR-Ag (r=-0.695) and LSR-Ag
(r=-0.593). The ME/A was correlated to LSg (r=-0.482) and LSRg (r=-0.591). In addition, MAVp was correlated to LSRg (r=0.434, Table 6).
Segmental heterogeneity and synchrony
The segmental heterogeneity (SH-ε) of SH-cε, SH-rε and SH-lε was 13.1 ± 5.9%, 19.1 ± 10.3%, and 15.4 ± 6.8%, respectively. The TH-cε was -14.3 ± 4.6% in
circumferential direction (Table 7). The TH-cε was correlated to body weight (r=-0.471, P=0.027) (Table 8).
24
The synchrony of SDT-cε, SDT-rε, and SDT-lε was 11.7 ± 4.2 ms, 16.5 ± 13.4 ms, and 19.4 ± 8.5 ms, respectively; RT-cε, RT-rε, and RT-lε was 32.5 ± 9.3 ms, 40.2
± 28.7 ms, and 44.2 ± 22.6 ms, respectively (Table 7). The SDT-lε was correlated to heart rate (r=-0.433, P=0.044, Table 8).
Segmental heterogeneity/synchrony and indices of conventional echocardiography
Radial heterogeneity and synchrony were not correlated to systolic indices of conventional echocardiography. Circumferential transmural heterogeneity was well correlated to FAC (r=-0.841, P<0.001). Longitudinal synchrony (RT-lε, SDT-lε) were affected by AAT, AVp and AGp (Table 9).
Circumferential heterogeneity (SH-cε) was correlated to TE/A (r=0.434). Radial synchrony (RT-rε) was correlated to ME/A (r=-0.470). Longitudinal synchrony (RT-lε, SDT-lε) were correlated to IVRT (r=0.489, r=0.461) respectively (Table 10).
Segmental heterogeneity/synchrony and indices of 2D-STE
Radial/longitudinal strain and strain rate were almost not correlated to synchrony.
Heterogeneity (SH-cε, SH-lε) were correlated to CSg and LSg, respectively, and heterogeneity (TH-cε) was correlated to CSg (r=-0.817, P<0.001), CSRg (r=-0.816, P<0.001), and CSR-Eg (r=-0.747, P<0.001) significantly (Table 11).
Discussion
Left ventricular strain and strain rate have been reported to be significantly associated with various physical characteristics. In human patients left ventricular strain and strain rate are affected by heart rate, but also by age, gender and body weight [27-32]. The effects of physical status on Indices of strain and strain rate in dogs are less consistent. Myocardial strain and strain rate may not be affected age in
25
dogs [28], the effects of body weight and heart rate remains inconclusive [28, 33-35].
In this study, left ventricular strain and strain rate in cats were not affected by age, body weight, heart rate or blood pressure, except peak longitudinal early diastolic strain rate was affected by heart rate. A significant association between age and diastolic myocardial velocity was reported in cats [36]. However, the effect of age on myocardial motion remains controversial [36, 37]. In this study, gender appeared to affect myocardial strain. Based on these findings, the effects of physical
characteristics on myocardial strain and strain rate should be taken into consideration, strain and strain rate might appear correlation with physiological factors if age or body weight range is expanded.
In general, cardiac systolic function is evaluated as contractility by pressure formation (PEP: ET and AAT derived from pulse-wave Doppler) or by deformation (EFM and FS derived from M-mode) [38]. The mechanics of segmental myocardium (elasticity deformation) are complied wall stress and contractile force (Hooke's Law) [39]. Deformation of segmental myocardium cannot directly represent contractility, it is affected by wall stress (loading, overall cardiac volume changes), neighboring segments interaction (tension), cardiac geometry and elasticity (tissue properties).
The relationship between the left ventricular strain and the indices derived from conventional echocardiography (B- and M-mode) was weak and inconsistent in this study. It was not surprised that no significant association between the indices of 2D-STE and conventional echocardiography was found. Ventricular deformation is complemented in three directions. No single direction of the myocardial was assigned to a larger proportion of systolic function. Thus, conventional echocardiography is insensitive to detect regional dysfunction [40]. On the other hand, the EFste and FAC derived from 2D-STE are highly correlated to the estimated volume by MRI and
26
3D-STE in human patients [41]. In this study, correlation between EFste/FAC and myocardial strain were found in cats. Correlations between LS, LSR and ME/A, MAVp have also been reported in human patients [30]. Several systolic indices derived from pulse-wave Doppler were correlated to myocardial strain.
Based on these findings, both systolic and diastolic parameters derived from Doppler echocardiography might closely reflect systolic and diastolic strain rate.
More cases with different cardiac conditions are needed to clarify the accuracy of the pressure gradient measured from Doppler echocardiography in association of
myocardial strain rate.
Heterogeneity and synchrony
In this study, both transmural and segmental heterogeneity of ventricle were demonstrated in clinically healthy cats. The characteristic of transmural heterogeneity of cats showed similar pattern of human subjects. The difference of TH-cε between endocardium and epicardium in this study was approximately 64% (TH-cε/Cg X 100%) in contrast to 36% in human subjects [42]. And CS was decreasing from endocardium to epicardium, which also consistent with the pattern of human studies [42,43].
In this study, FAC was negatively correlated with TH-cε. The effect of transmural heterogeneity has to be taken consideration when FAC is changed.
Heterogeneity of longitudinal direction was also found, significantly lower in the apex.
Left ventricular heterogeneity in longitudinal direction has also been reported in human patients, however the pattern is inconsistent [27, 43-45].
In general, ventricular synchrony was not affected by age, body weight and strain, strain rate, heart rate, except SDT-lε was negatively associated with heart rate. SDT-lε and RT- lε were positively correlated to IVRT [46]. The intraventricular synchrony in
27
our studies was 32.5 ± 9.3 ms in RT-cε, 40.2 ± 8.7 ms in RT-rε and 44.2 ± 22.6 ms in RT-lε, in contrast to value derived from in dogs with wide range of mean RT-rε varied from 15 to 41.8 ms (SD varied from 2 to 17.9 ms) [28, 47, 48]. Although ventricular synchrony was not affected by strain in this study, the ranges for segmental
intraventricular synchrony in clinically healthy cats were wide and the clinical application could be limited. However, the normal range of synchrohy in cat was similar to dog and human, the possibility of species conservatives made application of synchrony with cut off value to diagnose or predict outcome of cardiac diseases implementable because human medicine have already used cut off value in resynchronization therapy field successfully.
Conclusion
Longitudinal strain and strain rate was low in the segment of apex, CS decreasing from endocardium to epicardium. Left ventricular segmental and transmural
heterogeneity was found in clinically healthy cats. Global myocardial strain and strain rate, heterogeneity and synchrony were nearly independent to age, body weight and heart rate.
28
Reference
1. Yu C, Lin H, Zhang Q, Sanderson J: High prevalence of left ventricular systolic and diastolic asynchrony in patients with congestive heart failure and normal QRS duration. Heart 2003, 89:54-60.
2. Hawkins NM, Petrie MC, MacDonald MR, Hogg KJ, McMurray JJ: Selecting patients for cardiac resynchronization therapy: electrical or mechanical dyssynchrony? Eur Heart J 2006, 27:1270-1281.
3. Leclercq C, Faris O, Tunin R, Johnson J, Kato R, Evans F, Spinelli J, Halperin H, McVeigh E, Kass DA: Systolic improvement and mechanical
resynchronization does not require electrical synchrony in the dilated failing heart with left bundle-branch block. Circulation 2002,
106:1760-1763.
4. Kass DA: Predicting cardiac resynchronization response by QRS duration:
the long and short of it. J Am Coll Cardiol 2003, 42:2125-2127.
5. Fauchier L, Marie O, Casset-Senon D, Babuty D, Cosnay P, Fauchier JP:
Interventricular and intraventricular dyssynchrony in idiopathic dilated cardiomyopathy: a prognostic study with Fourier phase analysis of radionuclide angioscintigraphy. J Am Coll Cardiol 2002, 40:2022-2030.
6. Dohi K, Onishi K, Gorcsan III J, López-Candales A, Takamura T, Ota S, Yamada N, Ito M: Role of radial strain and displacement imaging to quantify wall motion dyssynchrony in patients with left ventricular
6. Dohi K, Onishi K, Gorcsan III J, López-Candales A, Takamura T, Ota S, Yamada N, Ito M: Role of radial strain and displacement imaging to quantify wall motion dyssynchrony in patients with left ventricular