Archives of Cardiovascular Disease (2010) 103, 603—614
REVIEW
Developments in echocardiographic techniques forthe evaluation of ventricular function in children
L’évaluation de la fonction ventriculaire chez l’enfant — nouvelles techniqueséchocardiographiques
Andreea Dragulescu, Luc L. Mertens ∗
Department of Cardiology, Hospital for Sick Children, 555 University Avenue,Toronto, ON, M5G 1X8 Canada
Received 9 August 2010; received in revised form 7 September 2010; accepted 9 September2010Available online 25 November 2010
KEYWORDSEchocardiography;Tissue Doppler;Speckle tracking;Ventricular function
Summary Echocardiography is a very important tool for the diagnosis and follow-up ofchildren with congenital and acquired heart disease. One of the challenges that remains inpaediatric heart disease is the assessment of systolic and diastolic function in children, asthis is influenced by growth, morphology and loading conditions. New echocardiographic tech-niques, such as tissue Doppler, deformation imaging and three-dimensional echocardiography,have great potential application in this field. They may provide new insights into the influence of
growth, morphology and loading on cardiac mechanics, and could become useful clinical tools.In this review, we discuss the potential use and limitations of these new echocardiographictechniques in paediatric and congenital heart disease.© 2010 Elsevier Masson SAS. All rights reserved.Abbreviations: 2D, two-dimensional; 3D, three-dimensional; ASD, atrial septal defect; EF, ejection fraction; IVA, myocardial accelerationduring isovolumic contraction; LV, left ventricle/ventricular; MRI, magnetic resonance imaging; RV, right ventricle/ventricular; TD, tissueDoppler.
∗ Corresponding author. Fax: +416 813 5857.E-mail addresses: [email protected], [email protected] (L.L. Mertens).
1875-2136/$ — see front matter © 2010 Elsevier Masson SAS. All rights reserved.doi:10.1016/j.acvd.2010.09.004
604 A. Dragulescu, L.L. Mertens
MOTS CLÉSÉchocardiographie ;Doppler tissulaire ;Speckle tracking ;Fonctionventriculaire
Résumé L’échocardiographie est un outil très important dans le diagnostic et le suivi despatients avec cardiopathies congénitales et acquises. L’évaluation de la fonction systolique etdiastolique chez l’enfant reste un problème due aux influences liées à la croissance, la mor-phologie ventriculaire et les conditions de charge. Les nouvelles techniques échographiquescomme le Doppler tissulaire, l’imagerie de déformation et l’échographie 3D ont des applica-tions potentielles importantes dans ce cadre. Elles peuvent apporter des nouvelles informationsconcernant la mécanique cardiaque dans différentes conditions pendant la croissance. Danscette revue, on présente les indications potentielles et les limitations de ces techniqueséchographiques chez l’enfant.© 2010 Elsevier Masson SAS. Tous droits réservés.
I
Evdraapmoert
m(ehhacadithvoloRm
tdhtpsmtiaev
cor
T
Th1tdwiiiaTib
atvrwavbwdahtscese
i
ntroduction
chocardiography has become the most important nonin-asive technique for the diagnosis and follow-up of heartisease in children. Cross-sectional Doppler echocardiog-aphy allows a detailed description of cardiac anatomynd haemodynamics. Currently, the majority of childrenre referred for cardiac surgery based on echocardiogra-hy only. The diagnostic accuracy for describing cardiacorphology is extremely high, with a reported incidence
f diagnostic errors of only 87 errors in more than 50,660chocardiograms in an established paediatric echocardiog-aphy laboratory [1], demonstrating the level of accuracyhat can be reached.
Cardiac MRI and cardiac computed tomography are theain imaging techniques used for extracardiac anatomy
pulmonary arteries and aortic arch) if not visualized prop-rly by echocardiography. Diagnostic cardiac catheterizationas become more obsolete and is restricted mainly toaemodynamic assessment in more complex lesions and thessessment of pulmonary vascular resistance. One of thehallenges remaining for paediatric echocardiography is thevailability of good techniques for assessing systolic andiastolic ventricular function. Most functional variables usedn echocardiography were developed for the assessment ofhe morphologically normal LV. The diversity of congenitaleart defects complicates the interpretation of functionalariables, because of the anatomical variability, the effectf growth on myocardial function and the differences inoading conditions. For the LV, adult techniques are extrap-lated to paediatrics often without good validation. For theV and the single ventricle, qualitative subjective assess-ent is the technique used routinely in most laboratories.During the past decade, different echocardiographic
echniques have been developed that allow a moreetailed analysis of cardiac function. These techniquesave potential application in and substantial benefit forhe assessment of ventricular dysfunction in paediatricatients. TD echocardiography and speckle-tracking-basedtrain imaging provide direct quantitative information aboutyocardial motion and deformation, which is more geome-
ry independent than measurement of EF; they give more
nsight into myocardial mechanics and could provide guid-nce in treatment and response to therapies. Current 3Dchocardiographic techniques enable the acquisition of fullolumetric datasets, which can be analysed offline for themcems
alculation of ventricular volumes, mass and EF. The impactf these technologies on paediatric functional echocardiog-aphy will be discussed in the current review.
issue Doppler velocities
he analysis of pulsed TD signals to interpret cardiac motionas been used since the early 1960s [2]. It took until the990s for it to be recognized as a potentially useful clinicalechnique for the assessment of global and regional myocar-ial function [3]. In a pig ischaemic model, tissue velocitiesere shown to change very quickly and consistently after the
nduction of ischaemia [4]. Several subsequent clinical stud-es investigated the use of regional myocardial velocitiesn various adult diseases, such as ischaemic heart disease,ortic insufficiency and hypertrophic cardiomyopathy [4,5].he advantage for paediatric and congenital heart disease
s that these techniques are geometry independent and cane applied to any chamber morphology.
Colour TD imaging was introduced in the early 1990ss an alternative technique for measuring tissue veloci-ies. In contrast to pulsed Doppler, which measures peakelocities, it uses autocorrelation techniques to measureegional mean velocities. This technical difference explainshy colour TD-derived myocardial velocities are on aver-ge 15—20% lower than pulsed wave-derived myocardialelocities [6] (Fig. 1A and B). Very high frame rates cane obtained by image optimization (> 250 frames/second),hich is very useful for the analysis of short-lived myocar-ial movements, such as during the isovolumetric periods,nd is important for adequate temporal resolution at highereart rates. An advantage of colour TD is that tissue veloci-ies can be recorded simultaneously in different myocardialegments during the same cardiac cycle. This allows theomparison of regional wall motion and timing of cardiacvents between different myocardial segments during theame cardiac cycle, which is important for dyssynchronyvaluation.
The limitations of TD velocity imaging are related tots angle dependency (Doppler technique) and the unidi-
ensional assessment of myocardial motion (longitudinal,ircumferential or radial). Global cardiac translation of thentire heart during the cardiac cycle will also affect theeasurement and tethering effects between myocardial
egments; if a dysfunctional segment is moved by a healthy
New echocardiographic techniques in children 605
Figure 1. Tissue Doppler imaging in a patient after Fontan completion for hypoplastic left heart syndrome. (A) Pulse Doppler in theic vellowebnor
twSadwlaipgswstwLrwr
basal segment of the lateral wall with reduced systolic and diastolof the systemic ventricle. The absolute value for each segment isDoppler-derived strain and strain rate for the same ventricle with a
segment, this can also contribute to regional motion maskingregional dysfunction.
Clinical application of tissue velocities inchildren
One of the challenges when introducing new techniquesinto paediatric echocardiography is that the methodologymust be validated first for paediatric use, especially theestablishment of normal values for different paediatric agegroups. Normal paediatric TD data have been publishedby different groups [7—9]. In these studies, it was shownthat tissue velocities vary with age and heart rate. Eidemet al. included 325 children and showed that pulsed-waveTD velocities also correlate with cardiac growth variables,especially the LV end-diastolic dimension and LV mass [7],indicating that tissue velocities are not entirely geometryindependent. This has important implications when applying
this methodology to children with congenital heart disease,where there is a large variability in ventricular geometry.Apart from the influence of geometry, changes in load-ing conditions also affect TD velocity measurements. Acutepreload changes clearly affect tissue velocities [10], whileMff
d
ocities. (B) Simultaneous tissue Doppler velocities in six segmentsr than the corresponding pulse Doppler velocity. (C and D) Tissuemal curves for the septal segments (not involved in ejection).
his is less clear for the effect of chronic volume loading,here the ventricle has adjusted to the chronic load [11].tudies in children with chronic LV volume loading related toventricular septal defect or patent ductus arteriosus haveocumented only minimal changes in TD velocities comparedith normal paediatric controls [11]. Acute increase in after-
oad results in decreased TD velocities [12], while chronicdaptation by the concentric hypertrophic response resultsn decreased longitudinal velocities during the remodellingrocess. In children with aortic valve stenosis, systolic lon-itudinal velocities were shown to be reduced in the basalegments of the interventricular septum and the LV lateralall. This could be an effect of hypertrophy and associated
ubendocardial dysfunction. In an adult population with aor-ic stenosis, the degree of reduction in longitudinal functionas shown to be related to the degree of fibrosis in the
V and was also predictive of outcome after aortic valveeplacement [13]. Kiraly et al. [14] showed that in childrenith aortic valve stenosis, longitudinal TD velocities were
educed more significantly compared with radial velocities.
ore research is warranted in children, as the mechanismsor adaptation to increased pressure loading could be dif-erent in the paediatric population.
As congenital heart disease frequently affects the RV,ata on right ventricular TD velocities have been published
6
idqtsl[tntoapafwiflHu
hshutwHtcttcclcmawr
iTwaTdasrd
cdppdb
lwT
sIpgambd[Adhcstit
mptpasipIstfbbdHsdwsrppir
ptiitmrpspcsp
06
n different conditions that affect the RV. With the pre-ominantly longitudinal orientation of the RV myofibres, theuantification of longitudinal function is especially impor-ant when assessing RV function. Good correlations betweenystolic velocities and RV EF were found in an adult popu-ation that included patients with congenital heart disease15]. Several studies have shown elevated RV systolic veloci-ies in patients with ASDs before closure of the defect, whichormalized within 24 hours after closure [16,17]. Quantita-ive assessment of right ventricular performance after repairf tetralogy of Fallot has also been the subject of consider-ble investigation. TD velocities are decreased in tetralogyatients after repair, with some RV regional wall motionbnormalities [18]. However, in these patients, regionalunctional variables obtained at the base of the RV freeall correlate less well with RV EF assessed by MRI, as this
ncludes the patched and often dilated right ventricular out-ow tract, influencing the assessment of global RV function.ence, the use of regional variables in this condition is stillnproven and requires further study.
The use of TD velocities in functionally univentricularearts has also been studied. Frommelt et al. [19] usederial measurements for the evaluation of patients withypoplastic left heart syndrome from the neonatal periodntil after the second stage palliation, and showed a trendowards a decrease in systolic and diastolic tissue velocities,ith no difference regarding the type of initial palliation.owever, the interpretation of these data is difficult, ashere were significant changes in loading conditions and alsohanges in ventricular growth that could have influencedhe observed changes. Vitarelli et al. [20] evaluated ven-ricular function at midterm (7.4 ± 2.8 years) after Fontanompletion, using myocardial velocities derived from theomputed tomography dose index. They compared morpho-ogically LVs (tricuspid atresia, double inlet LV) with normalontrols and analysed the anterior and inferior wall (groupsore comparable). There were significantly lower systolic
nd early diastolic tissue velocities in the Fontan group,hich correlated well with the EF and the mass/volume
atio, respectively.Tissue velocities are used in the evaluation of mechan-
cal dyssynchrony. Mechanical dyssynchrony, as assessed byD imaging and speckle tracking, is often present in patientsith cardiomyopathy unrelated to electrical dyssynchrony,nd correlates with the severity of LV dysfunction [21,22].his probably reflects regional differences in myocardialysfunction. Mechanical dyssynchrony by TD imaging wasnalysed in only 64% of patients in a multicentre Europeantudy evaluating the current practice and results of cardiacesynchronization therapy in paediatric and congenital heartisease [23].
Overall, the effect of growth and variability in loadingonditions on myocardial velocities limits their use in pae-iatric heart disease, except for serial follow-up in the sameatient. However, only limited data are available. For exam-le, after heart transplantation in children, a significantecrease in systolic and diastolic RV tissue velocities has
een noted 3—6 months before terminal graft failure [24].Tissue velocity tracings have also been used to calcu-ate myocardial performance index, with good correlationsith standard pulse Doppler measurements [25]. The use ofD imaging tracings is especially useful for the RV, allowing
waaTd
A. Dragulescu, L.L. Mertens
imultaneous measurement of systolic and diastolic events.t has been described as being useful in the assessment ofatients after repair of tetralogy of Fallot and other con-enital heart diseases [26]. An animal study by Cheung etl., using invasive measurements, demonstrated that theyocardial performance index was affected significantlyy acute changes in loading conditions and was unable toetect acute changes in contractile function consistently27], making its interpretation difficult in clinical settings.dditionally, we believe that if you combine systolic andiastolic time intervals in a single index, you do not knowow to interpret the index when it is abnormal or when ithanges; when it is abnormal, you can only conclude thatomething is wrong with either the systolic or diastolic func-ion or that the loading of the heart changed. We do not uset routinely in our laboratory because of these considera-ions.
To overcome the effect of loading conditions on theeasurements, myocardial acceleration during IVA was pro-osed as an index of myocardial contractility. IVA calculateshe average rate of myocardial acceleration during the IVAeriod, which makes it insensitive to afterload changes andlso, to some extent, to preload changes. In experimentaltudies, IVA has been validated as a sensitive noninvasivendex of RV and LV contractility, which is unaffected byreload and afterload within a physiological range [28].t requires imaging at very high frame rates, as IVA is ahort-lived event (30—40 ms). The disadvantage of IVA ishat it is extremely sensitive to heart rate due to theorce—frequency relationship. The heart rate sensitivity haseen used to its advantage, to assess contractile reservey studying the force—frequency relationship during pacing,obutamine stress echocardiography or exercise [29—31].owever, intra- and interobserver variabilities have beenhown to be problematic. The reproducibility of LV IVA wasemonstrated to be relatively poor, especially in patientsith impaired LV function [32]. Moreover, Lyseggen et al.
howed that LV IVA is not a good variable for assessingegional myocardial function and, with non-physiologicalreload changes, is also preload sensitive. In paediatricatients, LV IVA was used after heart transplantation, wheret was shown to be a useful noninvasive marker of allograftejection [33].
The clinical use of RV IVA seems more promising. Inatients with repaired tetralogy of Fallot, RV IVA was showno be reduced and was related to the degree of pulmonarynsufficiency, suggesting reduced contractile RV functionn patients with pulmonary regurgitation or, alternatively,hat the increased loading also affects the RV IVA measure-ent. For patients with a systemic RV, systolic functional
eserve was assessed by dobutamine stress echocardiogra-hy, and the change in IVA during dobutamine exposure washown to correlate well with the change in the end-systolicressure—volume relationship assessed by conductanceatheter measurements [34]. The force—frequency relation-hip was studied noninvasively using IVA in the perioperativeeriod in patients after congenital heart surgery [30]. There
as a large variability in the postoperative response, withpreserved force—frequency relationship after ASD closurend significant reduction after the arterial switch operation.he same principle could be used during stress echocar-iography, as demonstrated by Pauliks et al. [31]. In single
utcrptctalttTtctpa
twtiisr[priwmu[taaaacpco[t
qpdtpEtrbf
New echocardiographic techniques in children
ventricles, the technique can be used to assess contractilereserve, and initial data suggest a preserved systolic reservein children after the Fontan operation [29].
Deformation imaging: strain, strain rate
Myocardial velocities are influenced by global cardiac trans-lational motion and myocardial tethering, which limits theiruse in the assessment of regional myocardial function. Thislimitation can be overcome by using regional myocardialdeformation or strain imaging. Two different technologiesare currently available for studying regional myocardialdeformation. The first technique is based on TD imagingand is based on calculating myocardial velocity gradients.The second is based on tracking speckles on the grey-scaleimages in different frames throughout the cardiac cycle, andcalculates the displacement of the speckles throughout thecardiac cycle. Two variables for cardiac deformation can becalculated (Fig. 1C and D). Regional strain rate is the rate ofdeformation (per second) and is calculated as the velocitydifference between two segments of myocardium dividedby the distance between them. In younger children, we usecomputational distances of 4—5 mm in the radial directionand 8—9 mm in the longitudinal direction. Regional strainrepresents the amount of deformation (%) or the fractionalchange in length, caused by an applied force, and is calcu-lated by integrating the strain rate curve over time duringthe cardiac cycle.
Experimental studies suggest that deformation variablescan be used as noninvasive indices of ventricular function.End-systolic strain measurements correlate well with EFmeasurements, while peak systolic strain rate correlatesrelatively well with dP/dt and end-systolic elastance [35].
Tissue Doppler-derived deformationimaging
The underlying principle of TD-derived strain imaging isthat instantaneous differences in tissue velocity betweentwo adjacent segments reflect either expansion or compres-sion of the tissue in between. The disadvantage of tissuevelocity-based strain imaging is that it is angle dependentand is a unidimensional technique; it also requires extensivepostprocessing, influencing intra- and interobserver vari-abilities [36]. Despite this limitation, if well standardized,radial strain evaluation in the LV posterior wall remains oneof the more reproducible techniques and performs betterthan speckle-tracking techniques [36]. The high frame rates(> 200 frames/second) that can be obtained provide a majoradvantage, especially in younger children with fast heartrates. This was the first technique available for measuringstrain and strain rate and the first clinical data in childrenwere obtained using this method. In particular, longitudi-nal strain measurements in the interventricular septum and
LV and RV lateral walls were used to quantify regional andglobal myocardial function. The first paper on normal strainrate and strain data in children was published by Weidemannet al. [37]. Recently, Pena et al. published normal values inneonates (Table 1) [38].olc[t
607
Quantification of regional myocardial function can beseful for different indications in children. Firstly, quan-ification of regional myocardial function can be used inhildren where coronary perfusion can be an issue andegional myocardial ischaemia and dysfunction can beresent. This includes patients after coronary reimplanta-ion, i.e. arterial switch operation, repair of abnormal leftoronary artery from the pulmonary artery, the Ross opera-ion and other similar interventions. Long-term follow-upfter coronary reimplantation for abnormal origin of theeft coronary artery from the pulmonary artery showedhat, despite normalization of the EF, longitudinal deforma-ion remained abnormal in the septal and lateral wall [39].his is probably related to residual subendocardial dysfunc-ion resulting in decreased longitudinal function. Althoughoronary-related myocardial ischaemia is rare in children,he analysis of regional function by deformation imaging canrovide additional information in children with congenital orcquired heart disease.
Secondly, strain imaging is also useful for the early detec-ion of regional myocardial function. While in most patientsith hypertrophic cardiomyopathy, global LV systolic func-
ion is generally considered to be preserved, deformationmaging has demonstrated significant regional differencesn systolic deformation variables. Peak systolic strain washown to be reduced significantly in the more hypertrophicegions, and was independent of the underlying aetiology40,41]. In young patients with Duchenne muscular dystro-hy, deformation analysis showed a significant decrease inadial and longitudinal peak systolic strain and strain raten the LV inferolateral and anterolateral walls in patientsith normal EF. This corresponds to the regions in theyocardium where early fibrotic changes can be observed
sing gadolinium late-enhancement cardiac MRI techniques42]. In Duchenne patients, these changes progressed withime, suggesting that early deformation changes could ben early marker for cardiac dysfunction. Also, in patientsfter anthracycline exposure, acute changes were observedfter low-to-moderate doses of anthracycline, with alter-tion of systolic and diastolic TD-derived variables. Thesehanges worsened after subsequent infusions. Moreover, inatients studied during long-term follow-up after anthra-ycline exposure, a decrease in deformation variables wasbserved, despite normal EF and fractional shortening43,44]. The significance of these early changes for long-erm outcomes needs to be further defined.
Thirdly, deformation imaging is particularly useful for theuantitative assessment of RV function (Fig. 1C and D). Inostoperative tetralogy of Fallot patients, Weidemann et al.emonstrated that in the basal, mid and apical segments ofhe right ventricular free wall and interventricular septum,eak systolic strain and strain rate values were reduced [45].yskens et al. demonstrated that these changes were relatedo the degree of pulmonary regurgitation [46]. Reducedegional peak systolic strain and strain rate values in theasal, mid and apical segments of the RV free wall were alsoound in patients with a systemic RV, such as after a Senning
r Mustard operation [47]. Similarly to these data, reducedongitudinal RV deformation was measured in patients withongenitally corrected transposition of the great arteries48]. Another study compared RV ventricular function afterhe Sano operation vs the classical Norwood initial pallia-
608 A. Dragulescu, L.L. Mertens
Table 1 Normal strain and strain rate values for longitudinal and radial components in the basal segments.
Pena et al., 2009 [38]a Weidemann et al., 2002 [37]a Lorch et al., 2008 [51]b
(n = 55) (n = 33) (n = 284)
Age Neonates 4—6 years 0—18 years
Longitudinal strain basal segmentsSystolicLV lat −24.46 ± 3.82 −26 ± 11 −20.68 ± 8.08IVS −25.86 ± 4.83 −24 ± 6 −18.3 ± 6.67RV lat −36 ± 11
Longitudinal SR basal segmentsSystolicLV lat −1.83 ± 0.37 −2.2 ± 1.1 −1.34 ± 0.68IVS −1.89 ± 0.60 −1.8 ± 0.6 −1.13 ± 0.53RV lat −2.4 ± 0.6Early diastolicLV lat 3.15 ± 1.53 2.7 ± 1.1 1.68 ± 0.89IVS 3.19 ± 1.57 2.6 ± 0.9 1.32 ± 0.52RV lat 3.2 ± 1.1
Late diastolicLV lat 2.12 ± 1.29 1.9 ± 1.2 0.40 ± 0.26IVS 2.39 ± 0.90 1.4 ± 0.6 0.47 ± 0.29RV lat 2.1 ± 1
Radial strain basal PW (SAX) systolic 49.72 ± 12.86 58 ± 12
Radial SR basal PW (SAX)Systolic 2.98 ± 0.78 3.7 ± 1.1Early diastolic −5.53 ± 1.70 −8 ± 2.3Late diastolic −3.89 ± 1.73 −2.2 ± 1.3
IVS: interventricular septum; LV lat: left ventricular lateral wall; PW: posterior wall; RV lat: right ventricular lateral wall; SAX: shortaxis; SR: strain rate.a Tissue Doppler-derived strain and strain rate.b Two-dimensional-based deformation imaging.
tRpt
uin
Ti
GmdsiOtti
hatttws
terwab
w[s
ion for hypoplastic left heart syndrome, and showed thatV free wall TD-derived strain and strain rate after RV toulmonary artery conduit palliation was significantly betterhan after the classical Norwood intervention [49].
Finally, TD and deformation techniques can also besed for the detection of myocardial dyssynchrony and thedentification of patients who might benefit from resynchro-ization therapy [21].
wo-dimensional-based deformationmaging—speckle tracking
iven the various limitations of Doppler-derived defor-ation imaging, new echocardiographic techniques wereeveloped to analyse cardiac deformation, based on grey-cale imaging. Speckle tracking is based on 2D images and
s relatively easy to perform with a short analysis time.ur recent work in children has shown that speckle-trackingechniques can be used in children to reliably quantify longi-udinal and circumferential strain with reasonable inter- andntraobserver variabilities. Radial strain measurements areawaai
ighly variable and strain rate measurements are less reli-ble [36]. An example of LV longitudinal strain by speckleracking is shown in Fig. 2. Further optimization of theechniques is required in children. One of the problems ishe lower temporal resolution of the technique comparedith Doppler-based technology, which might be an issue in
maller children with higher heart rates.The major advantage of speckle-tracking technology is
hat it allows the study of radial, longitudinal and circumfer-ntial deformation as well as the assessment of ventricularotation and torsion [50]. It also is angle independent. Asith any strain technique, it is, however, load dependent,nd the influence of geometry on the measurements has noteen well defined.
Normal data on LV longitudinal strain and strain rate, asell as rotation and twist, have been published recently
51,52] (Table 1). Longitudinal strain does not changeignificantly with maturation and decreasing heart rate,nd LV torsion and rotation remain relatively constanthen normalized by LV length and cardiac cycle, withtendency towards faster deformation at a younger
ge. More data are required in smaller children andnfants.
New echocardiographic techniques in children 609
pathme se
ri[cptt[
iavignbaui
Figure 2. Longitudinal strain in a patient with dilated cardiomyothe presence of dyskinetic motion and postsystolic shortening of so
Clinical applications
The major clinical applications of speckle-tracking tech-niques should be the same as for TD-derived deformationimaging. Despite the fact that it is relatively easy to use,only a few studies have been published so far in children. Ourgroup validated the methodology on different ultrasoundsystems [36]. An example of abnormal rotation in a patientwith hypertrophic cardiomyopathy is shown in Fig. 3. Othergroups have looked at different applications. One grouplooked at regional deformation properties after successfulrepair of aortic coarctation and showed decreased deforma-tion in the LV anterior wall [53]. A recent paper also showedthat in patients with aortic stenosis and aortic coarctation,LV torsion was increased before and decreased after inter-ventional treatment [54]. This is related to subendocardialdysfunction with an effect on epicardial force development.Most of the studies in the paediatric age group have beenperformed in patients with RV disease. However, as the tech-
nology was developed for the LV, most of the research workhas focused on the effect of RV disease on LV function. Leftventricular torsion seems to be impaired in conditions asso-ciated with RV volume load, mostly due to reduced basalwicr
y and severely impaired systolic function. It is easy to appreciategments. The global longitudinal strain is severely reduced.
otation. In young adults with secundum ASD, acute unload-ng of the RV improves LV twist by increasing basal rotation55]. Further studies are needed to assess this mechanism inhildren with other conditions with RV volume load, such asatients after tetralogy of Fallot repair. It has been shownhat RV dilatation has a negative impact on LV circumferen-ial deformation but not longitudinal or radial deformation56].
Speckle-tracking analysis has been applied to the RVn patients after tetralogy of Fallot repair to evalu-te the changes in RV function after surgical pulmonaryalve replacement [57]. There was a significant increasen peak systolic and diastolic velocities, but not inlobal longitudinal strain, and all indices remained sig-ificantly lower compared with normal values. This coulde related to persistent RV dysfunction but also to thecute effect of volume unloading. A similar study eval-ated the acute effect of transcatheter pulmonary valvemplantation, showing an acute improvement in RV free
all and septal longitudinal function [58]. More stud-es are needed to evaluate whether deformation analysisan predict RV functional recovery after pulmonary valveeplacement.
610 A. Dragulescu, L.L. Mertens
F al funo
T
3wwancsfiocwrhuaiaiabwbr
Ce
Dppp
ierhqstsiussR
igure 3. Hypertrophic cardiomyopathy with preserved longitudinf both apex and base.
hree-dimensional anatomy/volumetry
D rendering of cardiac structures gained more importanceith the development of high-frequency paediatric probesith improved image quality, allowing multiplanar reviewsnd structural reconstructions [59]. The benefit of this tech-ique has been described in several groups of patients withongenital heart disease, mostly related to valvar and septaltructures [60—63], where good correlations with surgicalndings were reported and/or additional information wasbtained compared with 2D echocardiography. Volume cal-ulation from 3D echocardiographic datasets has becomeidely available with the newer generation of echocardiog-
aphy machines, with good correlations with MRI [64,65]. Itas recently been demonstrated that the different methodssed to quantify EF by 2D techniques in children using therea-length or biplane Simpson’s methods, have importantntra- and interobserver variabilities that affect the reli-bility of the measurements. This is especially importantn patients with ventricular dysfunction, where the avail-bility of accurate echocardiographic measurements might
e helpful for clinical management. 3D echocardiographyith semi-automated border detection methods seems toe useful potentially for improving the reproducibility andeliability of the measurements.c
gt
ction but abnormal rotation pattern with counterclockwise rotation
linical applications of three-dimensionalchocardiography
etailed analysis of LV regional myocardial function can beerformed with segmentation of 3D volumes [66]. An exam-le of volumetric assessment of the LV and left atrium isresented in Fig. 4.
Most of the validation work for RV volumes has been donen adolescents and young adults with congenital heart dis-ase. Obtaining full echocardiographic 3D datasets of the RVemains challenging, and a study in patients with congenitaleart disease and dilated RVs showed that, using differentuantification techniques, 3D measurements were only fea-ible in about 50% of all patients [67]. This was due mainlyo inadequate image quality and incomplete datasets. Theame study showed that while RV EF was reliable in patientsn whom measurements could be obtained, RV volumes werenderestimated by ∼20% compared with MRI. Accurate mea-urement of RV volumes is important for certain diseasesuch as postoperative tetralogy of Fallot patients, whereV end-diastolic volume index is an important variable in
linical decision-making.An alternative to 3D echocardiography is 2D echocardio-raphy with knowledge-based 3D reconstruction—a methodhat has been developed, validated and introduced recently
New echocardiographic techniques in children 611
hearttole.
tiptgecnt
Figure 4. Three-dimensional volumetric assessment of the leftSegmental analysis of left ventricular volumetric change during sys
on a commercial basis, specifically to quantify RV volumesin patients after tetralogy of Fallot repair [68]. This methodallows the 3D reconstruction of the RV without requiring bor-der tracing or image processing, by using a database of RVsin both healthy and diseased states. A 3D space-localizingmagnetic device allows placement of 2D images/planes ina 3D volume. The user identifies anatomical landmarks bypoint placement on 2D images, on the valves, apex, septumand free wall. These points are used to create a 3D surfacewith the aid of the database [68,69]. Currently, datasets are
available for postoperative tetralogy of Fallot patients withand without conduit insertion. Fig. 5 shows an example of RVvolume reconstruction in a patient after repair of tetralogyof Fallot.srv
Figure 5. Knowledge-based three-dimensional reconstruction of rightof tetralogy of Fallot; septal and anterior view of the right ventricle.
using one of the commercially available pieces of software. (A)(B) Left atrial volume.
The anatomical variety of single ventricle anatomy makeshe standard assessment of ventricular function difficultn these cases. The method of discs has been used inatients with single ventricles for the estimation of sys-olic and diastolic volumes, ventricular mass and EJ, withood correlations with MRI [70]. This method requiresxtensive manual offline tracing and postprocessing. Theurrent semi-automated border detection programmes areot designed to perform the semi-automated segmenta-ion on the 3D volumes of the single ventricles. The
oftware is based on models of normal RVs or LVs. Moreesearch is required to make 3D volumetrics easier for singleentricles.ventricular systolic and diastolic volumes in a patient after repair
6
T
Tduasetfsmwloth
C
N
R
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
12
he future
he future of TD and deformation imaging techniquesepends on further validation and demonstration of clinicaltility. Currently, 3D strain techniques are becoming avail-ble but they are limited by lower frame rates and lowerpatial resolution. It seems that the continuous technicalvolutions are a challenge for paediatric cardiology, as eachime a new technique becomes available there is the needor validation and establishment of normal values. Industryhould come up with standards for TD and strain measure-ents, which make the techniques vendor independent, asith the other echocardiographic techniques. There is still a
ong way to go before the first clinical decision is made basedn these new echocardiographic tools. In the meantime,hey offer great new insights into the effect of congenitaleart disease on cardiac function and mechanics.
onflict of interest statement
o conflicts of interest to declare.
eferences
[1] Benavidez OJ, Gauvreau K, Jenkins KJ, et al. Diagnostic errorsin pediatric echocardiography: development of taxonomy andidentification of risk factors. Circulation 2008;117:2995—3001.
[2] Yoshida T, Mori M, Nimura Y, et al. Analysis of heart motionwith ultrasonic Doppler method and its clinical application. AmHeart J 1961;61:61—75.
[3] Sutherland GR, Stewart MJ, Groundstroem KW, et al. ColorDoppler myocardial imaging: a new technique for theassessment of myocardial function. J Am Soc Echocardiogr1994;7:441—58.
[4] Derumeaux G, Ovize M, Loufoua J, et al. Doppler tissue imagingquantitates regional wall motion during myocardial ischemiaand reperfusion. Circulation 1998;97:1970—7.
[5] Vinereanu D, Ionescu AA, Fraser AG. Assessment of left ven-tricular long axis contraction can detect early myocardialdysfunction in asymptomatic patients with severe aortic regur-gitation. Heart 2001;85:30—6.
[6] Kukulski T, Voigt JU, Wilkenshoff UM, et al. A comparison ofregional myocardial velocity information derived by pulsed andcolor Doppler techniques: an in vitro and in vivo study. Echocar-diography 2000;17:639—51.
[7] Eidem BW, McMahon CJ, Cohen RR, et al. Impact of cardiacgrowth on Doppler tissue imaging velocities: a study in healthychildren. J Am Soc Echocardiogr 2004;17:212—21.
[8] Roberson DA, Cui W, Chen Z, et al. Annular and septal Dopplertissue imaging in children: normal z-score tables and effects ofage, heart rate, and body surface area. J Am Soc Echocardiogr2007;20:1276—84.
[9] Rychik J, Tian ZY. Quantitative assessment of myocardial tissuevelocities in normal children with Doppler tissue imaging. AmJ Cardiol 1996;77:1254—7.
10] Firstenberg MS, Greenberg NL, Main ML, et al. Determinantsof diastolic myocardial tissue Doppler velocities: influences of
relaxation and preload. J Appl Physiol 2001;90:299—307.11] Eidem BW, McMahon CJ, Ayres NA, et al. Impact of chronicleft ventricular preload and afterload on Doppler tissue imag-ing velocities: a study in congenital heart disease. J Am SocEchocardiogr 2005;18:830—8.
[
A. Dragulescu, L.L. Mertens
12] Oki T, Fukuda K, Tabata T, et al. Effect of an acute increasein afterload on left ventricular regional wall motion velocity inhealthy subjects. J Am Soc Echocardiogr 1999;12:476—83.
13] Weidemann F, Herrmann S, Stork S, et al. Impact of myocardialfibrosis in patients with symptomatic severe aortic stenosis.Circulation 2009;120:577—84.
14] Kiraly P, Kapusta L, Thijssen JM, et al. Left ventricular myocar-dial function in congenital valvar aortic stenosis assessed byultrasound tissue-velocity and strain-rate techniques. Ultra-sound Med Biol 2003;29:615—20.
15] Wahl A, Praz F, Schwerzmann M, et al. Assessment of rightventricular systolic function: comparison between cardiacmagnetic resonance derived ejection fraction and pulsed-wavetissue Doppler imaging of the tricuspid annulus. Int J Cardiol2010 [Epub ahead of print].
16] Eyskens B, Ganame J, Claus P, et al. Ultrasonic strain rate andstrain imaging of the right ventricle in children before andafter percutaneous closure of an atrial septal defect. J AmSoc Echocardiogr 2006;19:994—1000.
17] Pauliks LB, Chan KC, Chang D, et al. Regional myocardial veloc-ities and isovolumic contraction acceleration before and afterdevice closure of atrial septal defects: a color tissue Dopplerstudy. Am Heart J 2005;150:294—301.
18] Vogel M, Sponring J, Cullen S, et al. Regional wall motion andabnormalities of electrical depolarization and repolarization inpatients after surgical repair of tetralogy of Fallot. Circulation2001;103:1669—73.
19] Frommelt PC, Sheridan DC, Mussatto KA, et al. Effect of shunttype on echocardiographic indices after initial palliations forhypoplastic left heart syndrome: Blalock-Taussig shunt versusright ventricle-pulmonary artery conduit. J Am Soc Echocar-diogr 2007;20:1364—73.
20] Vitarelli A, Conde Y, Cimino E, et al. Quantitative assess-ment of systolic and diastolic ventricular function with tissueDoppler imaging after Fontan type of operation. Int J Cardiol2005;102:61—9.
21] Friedberg MK, Silverman NH, Dubin AM, et al. Mechanicaldyssynchrony in children with systolic dysfunction secondary tocardiomyopathy: a Doppler tissue and vector velocity imagingstudy. J Am Soc Echocardiogr 2007;20:756—63.
22] Labombarda F, Blanc J, Pellissier A, et al. Health-e-Childproject: mechanical dyssynchrony in children with dilated car-diomyopathy. J Am Soc Echocardiogr 2009;22:1289—95.
23] Janousek J, Gebauer RA, Abdul-Khaliq H, et al. Cardiacresynchronisation therapy in paediatric and congenital heartdisease: differential effects in various anatomical and func-tional substrates. Heart 2009;95:1165—71.
24] Fyfe DA, Ketchum D, Lewis R, et al. Tissue Doppler imagingdetects severely abnormal myocardial velocities that identifychildren with pre-terminal cardiac graft failure after hearttransplantation. J Heart Lung Transplant 2006;25:510—7.
25] Tekten T, Onbasili AO, Ceyhan C, et al. Novel approach tomeasure myocardial performance index: pulsed-wave tissueDoppler echocardiography. Echocardiography 2003;20:503—10.
26] Yasuoka K, Harada K, Toyono M, et al. Tei index determinedby tissue Doppler imaging in patients with pulmonary regur-gitation after repair of tetralogy of Fallot. Pediatr Cardiol2004;25:131—6.
27] Cheung MM, Smallhorn JF, Redington AN, et al. The effectsof changes in loading conditions and modulation of inotropicstate on the myocardial performance index: comparisonwith conductance catheter measurements. Eur Heart J
2004;25:2238—42.28] Vogel M, Cheung MM, Li J, et al. Noninvasive assessment of leftventricular force—frequency relationships using tissue Doppler-derived isovolumic acceleration: validation in an animal model.Circulation 2003;107:1647—52.
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
[
New echocardiographic techniques in children
[29] Cheung MM, Smallhorn JF, McCrindle BW, et al. Non-invasiveassessment of ventricular force-frequency relations in the uni-ventricular circulation by tissue Doppler echocardiography: anovel method of assessing myocardial performance in congen-ital heart disease. Heart 2005;91:1338—42.
[30] Cheung MM, Smallhorn JF, Vogel M, et al. Disruption of the ven-tricular myocardial force—frequency relationship after cardiacsurgery in children: noninvasive assessment by means of tissueDoppler imaging. J Thorac Cardiovasc Surg 2006;131:625—31.
[31] Pauliks LB, Vogel M, Madler CF, et al. Regional responseof myocardial acceleration during isovolumic contractionduring dobutamine stress echocardiography: a color tissueDoppler study and comparison with angiocardiographic find-ings. Echocardiography 2005;22:797—808.
[32] Margulescu AD, Thomas DE, Ingram TE, et al. Can isovolumicacceleration be used in clinical practice to estimate ventricularcontractile function? Reproducibility and regional variation of anew noninvasive index. J Am Soc Echocardiogr 2010;23:423—31(31, e1—6).
[33] Pauliks LB, Pietra BA, DeGroff CG, et al. Non-invasivedetection of acute allograft rejection in children by tissueDoppler imaging: myocardial velocities and myocardial accel-eration during isovolumic contraction. J Heart Lung Transplant2005;24:S239—48.
[34] Frigiola A, Redington AN, Cullen S, et al. Pulmonary regur-gitation is an important determinant of right ventricularcontractile dysfunction in patients with surgically repairedtetralogy of Fallot. Circulation 2004;110:II153—7.
[35] Greenberg NL, Firstenberg MS, Castro PL, et al. Doppler-derived myocardial systolic strain rate is a strong index of leftventricular contractility. Circulation 2002;105:99—105.
[36] Koopman LP, Slorach C, Hui W, et al. Comparison between dif-ferent speckle tracking and color tissue Doppler techniques tomeasure global and regional myocardial deformation in chil-dren. J Am Soc Echocardiogr 2010;23:919—28.
[37] Weidemann F, Eyskens B, Jamal F, et al. Quantification ofregional left and right ventricular radial and longitudinal func-tion in healthy children using ultrasound-based strain rate andstrain imaging. J Am Soc Echocardiogr 2002;15:20—8.
[38] Pena JL, da Silva MG, Faria SC, et al. Quantification ofregional left and right ventricular deformation indices inhealthy neonates by using strain rate and strain imaging. J AmSoc Echocardiogr 2009;22:369—75.
[39] Di Salvo G, Eyskens B, Claus P, et al. Late post-repair ventricularfunction in patients with origin of the left main coronary arteryfrom the pulmonary trunk. Am J Cardiol 2004;93:506—8.
[40] Ganame J, Mertens L, Eidem BW, et al. Regional myocardialdeformation in children with hypertrophic cardiomyopa-thy: morphological and clinical correlations. Eur Heart J2007;28:2886—94.
[41] Ho CY, Carlsen C, Thune JJ, et al. Echocardiographic strainimaging to assess early and late consequences of sarcom-ere mutations in hypertrophic cardiomyopathy. Circ CardiovascGenet 2009;2:314—21.
[42] Mertens L, Ganame J, Claus P, et al. Early regional myocardialdysfunction in young patients with Duchenne muscular dystro-phy. J Am Soc Echocardiogr 2008;21:1049—54.
[43] Ganame J, Claus P, Eyskens B, et al. Acute cardiac functionaland morphological changes after anthracycline infusions inchildren. Am J Cardiol 2007;99:974—7.
[44] Ganame J, Claus P, Uyttebroeck A, et al. Myocardial dys-function late after low-dose anthracycline treatment inasymptomatic pediatric patients. J Am Soc Echocardiogr
2007;20:1351—8.[45] Weidemann F, Eyskens B, Mertens L, et al. Quantification ofregional right and left ventricular function by ultrasonic strainrate and strain indexes after surgical repair of tetralogy ofFallot. Am J Cardiol 2002;90:133—8.
[
613
46] Eyskens B, Brown SC, Claus P, et al. The influence of pulmonaryregurgitation on regional right ventricular function in childrenafter surgical repair of tetralogy of Fallot. Eur J Echocardiogr2010;11:341—5.
47] Eyskens B, Weidemann F, Kowalski M, et al. Regional rightand left ventricular function after the Senning operation:an ultrasonic study of strain rate and strain. Cardiol Young2004;14:255—64.
48] Bos JM, Hagler DJ, Silvilairat S, et al. Right ventricular functionin asymptomatic individuals with a systemic right ventricle. JAm Soc Echocardiogr 2006;19:1033—7.
49] Hughes ML, Shekerdemian LS, Brizard CP, et al. Improved earlyventricular performance with a right ventricle to pulmonaryartery conduit in stage 1 palliation for hypoplastic left heartsyndrome: evidence from strain Doppler echocardiography.Heart 2004;90:191—4.
50] Leitman M, Lysyansky P, Sidenko S, et al. Two-dimensionalstrain-a novel software for real-time quantitative echocar-diographic assessment of myocardial function. J Am SocEchocardiogr 2004;17:1021—9.
51] Lorch SM, Ludomirsky A, Singh GK. Maturational and growth-related changes in left ventricular longitudinal strain andstrain rate measured by two-dimensional speckle trackingechocardiography in healthy pediatric population. J Am SocEchocardiogr 2008;21:1207—15.
52] Takahashi K, Al Naami G, Thompson R, et al. Normal rotational,torsion and untwisting data in children, adolescents and youngadults. J Am Soc Echocardiogr 2010;23:286—93.
53] Kowalski M, Kowalik E, Kotlinski K, et al. Regional left ventric-ular myocardial shortening in normotensive patients late afteraortic coarctation repair - normal or impaired? Ultrasound MedBiol 2009;35:1947—52.
54] Laser KT, Haas NA, Jansen N, et al. Is torsion a suitableechocardiographic parameter to detect acute changes in leftventricular afterload in children? J Am Soc Echocardiogr2009;22:1121—8.
55] Dong L, Zhang F, Shu X, et al. Left ventricular torsionaldeformation in patients undergoing transcatheter closure ofsecundum atrial septal defect. Int J Cardiovasc Imaging2009;25:479—86.
56] Cheung EW, Liang XC, Lam WW, et al. Impact of right ven-tricular dilation on left ventricular myocardial deformation inpatients after surgical repair of tetralogy of fallot. Am J Cardiol2009;104:1264—70.
57] Kutty S, Deatsman SL, Russell D, et al. Pulmonary valvereplacement improves but does not normalize right ventricularmechanics in repaired congenital heart disease: a comparativeassessment using velocity vector imaging. J Am Soc Echocar-diogr 2008;21:1216—21.
58] Moiduddin N, Asoh K, Slorach C, et al. Effect of transcatheterpulmonary valve implantation on short-term right ventricularfunction as determined by two-dimensional speckle track-ing strain and strain rate imaging. Am J Cardiol 2009;104:862—7.
59] Simpson JM. Real-time three-dimensional echocardiography ofcongenital heart disease using a high frequency paediatricmatrix transducer. Eur J Echocardiogr 2008;9:222—4.
60] Bharucha T, Ho SY, Vettukattil JJ. Multiplanar review anal-ysis of three-dimensional echocardiographic datasets givesnew insights into the morphology of subaortic stenosis. Eur JEchocardiogr 2008;9:614—20.
61] Takahashi K, Guerra V, Roman KS, et al. Three-dimensionalechocardiography improves the understanding of the mech-
anisms and site of left atrioventricular valve regurgitationin atrioventricular septal defect. J Am Soc Echocardiogr2006;19:1502—10.62] Takahashi K, Inage A, Rebeyka IM, et al. Real-time 3-dimensional echocardiography provides new insight into
6
[
[
[
[
[
[
[
14
mechanisms of tricuspid valve regurgitation in patients withhypoplastic left heart syndrome. Circulation 2009;120:1091—8.
63] Vettukattil JJ, Bharucha T, Anderson RH. Defining Ebstein’smalformation using three-dimensional echocardiography.Interact Cardiovasc Thorac Surg 2007;6:685—90.
64] Angelini ED, Homma S, Pearson G, et al. Segmentation ofreal-time three-dimensional ultrasound for quantification ofventricular function: a clinical study on right and left ventri-cles. Ultrasound Med Biol 2005;31:1143—58.
65] Niemann PS, Pinho L, Balbach T, et al. Anatomically orientedright ventricular volume measurements with dynamic three-dimensional echocardiography validated by 3-Tesla magnetic
resonance imaging. J Am Coll Cardiol 2007;50:1668—76.66] Corsi C, Lang RM, Veronesi F, et al. Volumetric quantificationof global and regional left ventricular function from real-time three-dimensional echocardiographic images. Circulation2005;112:1161—70.
[
A. Dragulescu, L.L. Mertens
67] Khoo NS, Young A, Occleshaw C, et al. Assessments ofright ventricular volume and function using three-dimensionalechocardiography in older children and adults with congeni-tal heart disease: comparison with cardiac magnetic resonanceimaging. J Am Soc Echocardiogr 2009;22:1279—88.
68] Sheehan FH, Ge S, Vick GW, 3rd, et al. Three-dimensional shapeanalysis of right ventricular remodeling in repaired tetralogy ofFallot. Am J Cardiol 2008;101:107—13.
69] Sheehan FH, Kilner PJ, Sahn DJ, et al. Accuracy of knowledge-based reconstruction for measurement of right ventricularvolume and function in patients with tetralogy of Fallot. AmJ Cardiol 2010;105:993—9.
70] Soriano BD, Hoch M, Ithuralde A, et al. Matrix-array 3-dimensional echocardiographic assessment of volumes, mass,and ejection fraction in young pediatric patients with afunctional single ventricle: a comparison study with cardiacmagnetic resonance. Circulation 2008;117:1842—8.
Recommended
