The rapid evolution of echocardiography during the past 25 years

Transcription

1 Evaluation of Myocardial Mechanics in the Fetus by Velocity Vector Imaging Adel K. Younoszai, MD, David E. Saudek, MD, Stephen P. Emery, MD, and James D. Thomas, MD, Denver, Colorado; Cleveland, Ohio; and Pittsburgh, Pennsylvania There are limited data on the properties of fetal myocardium with only a small number of Doppler tissue imaging based studies published. We evaluated the feasibility of using velocity vector imaging, a novel technique for analyzing 2-dimensional images offline, to study myocardial mechanics in the normal fetal heart at different of gestational ages. A single 2-dimensional 4-chamber image of the heart was interrogated offline using velocity vector imaging software. Longitudinal velocity, strain, and strain rate were measured in the right ventricular free wall, ventricular septum, and left ventricular free wall. Images from 24 of 27 (89%) were successfully analyzed. The systolic and diastolic longitudinal velocities increased with gestational age in all myocardial segments analyzed (r , P.05). Systolic strain and strain rate were not found to have significant correlation with gestational age in any of the wall segments. This implies that increasing velocities during normal gestation are the result of somatic growth rather than changes in myocardial contractility. The rapid evolution of echocardiography during the past 25 years hasledtoagreaterunderstandingoffetalcardiacphysiology.cardiac functionhastraditionallybeenmeasuredbysubjectiveassessmentof contractility and observation for hydrops fetalis. Newer indices such as the myocardial performance index, which measures the proportion of the time in isovolumic contraction and relaxation compared with ejection, are under evaluation as surrogate measures of global cardiac function; however, aprecise understanding of cardiac mechanics and function in the developing fetus is lacking. 1 The investigation of cardiac mechanics has become an area of increasing interest. Doppler tissue imaging (DTI) uses Doppler shifts ofthemyocardiumtoevaluatetissuedisplacementandvelocity.this techniquehasimprovedourunderstandingoftissuemotion,segmental deformation, and ventricular twisting. The use of longitudinal myocardial velocity profiles has significantly improved the ability to diagnose diastolic heart failure in the adult. It has also been used in combination with Doppler flow velocities across the mitral valve to predict left ventricular (LV) filling pressures. From displacement and velocity data, cardiac strain and strain rate can be calculated. Strain is definedasrelativedeformationofthemyocardiumandisthechange inlength(l)ofamyocardialsegmentreferencedtoitsoriginallength (L o ),asgivenbythefollowingequation:strain L L o /L o,whereas strain rate describes the rate of deformation, or how quickly atissue shortens or lengthens. Strain rate will be decreased when the tissue From Pediatric Cardiology, The University of Colorado (A.K.Y.), Denver, Colorado; Pediatric Cardiology (D.E.S.) and Cardiology (J.D.T.), The Cleveland Clinic Foundation, Cleveland, Ohio; and Obstetrics and Gynecology, The University of Pittsburgh (S.P.E.), Pittsburgh, Pennsylvania. Disclosure: Siemens has provided funding for imaging research to the Cleveland Clinic, Department of Cardiology. Reprint requests: Adel K. Younoszai, MD, Cardiac Imaging, The Children s Hospital, 1056 E 19th Ave, Cardiology/B100, Denver, CO ( Younoszai.Adel@ tchden.org) /$34.00 Copyright 2008 by the American Society of Echocardiography. doi: /j.echo changes aspecific length more slowly or has adecreased change in length in the same time span. Cardiac strain is not affected by translational motion of the heart and is theoretically, therefore, more accurate than simple velocity measurement. The fetal myocardium has unique properties. It is made up of primarily noncontractile elements with fewer myocytes and is stiffer compared with the mature heart. 2 The presence of fluid-filled lungs that externally limit ventricular relaxation may also contribute to decreased compliance during fetal diastole. 3 Myocardial tissue motion and velocities have now been investigated in the fetal heart with the first published report indicating its feasibility appearing in 1999 by Harada et al. 4 Since that time a number of preliminary studies have been performed validating the techniqueandreportingnormallongitudinalvelocities. 4-8 Inarecent report, Di Salvo et al 9 demonstrated that evaluation of strain and strain rate can also be measured in the fetus with limited variability. These Doppler-based techniques are subject to some technical limitations and are dependent on obtaining an optimal angle of interrogation. Velocity vector imaging (VVI) is anew offline analysis software package that allows evaluation of myocardial tissue motion and velocitywithoutthelimitationsofdopplerechocardiography.ituses acombination of speckle tracking with complexgeometricanalysisto follow the myocardium through the cardiac cycle. It offers an intuitive analysisofmyocardialmechanics,however,ithasnotbeenwellstudied andtherearenoreportsofitsuseinthefetus.theaimofthisstudywas toevaluatethefeasibilityofperformingvviinthenormalfetalheartat avarietyofgestationalagesandevaluatethevariationinvelocity,strain, and strain rate with gestational development and heart rate. METHODS Image Acquisition Data were collected from normal obstetric ultrasounds performed on an ultrasound system (Sequoia, Siemens Medical Systems, Malvern, PA) in the department of perinatology. Cardiac imaging, including a4-chamber view, was obtained as part of the standard examination by an experienced perinatologist. Any fetus in which

2 Journal of the American Society of Echocardiography Younoszai et al 471 Volume 21 Number 5 Figure 1 Velocity vector profile. Example of processed velocity vector image. Pictures represent still frames of left ventricle (LV) with velocity vectors in diastole and systole. Length of each arrow represents magnitude of velocity in that direction of representative myocardium. LA, Left atrium; RA, right atrium; RV, right ventricle. Color figure online. Figure 2 Example of velocity, strain, and strain rate profiles of left ventricle superimposed on background M-mode. Green tracing represents sample at midseptal wall and red represents sample of mid-left ventricular free wall. Velocities are measured as moving toward (positive) or away (negative) from icon of transducer placed at apex of heart as point of reference (top left). Color figure online. there was a significant fetal anomaly, known chromosomal abnormality, abnormal cardiac rhythm, or inappropriate size for gestational age was excluded. All data were stored as standard Digital Imaging and Communication in Medicine (DICOM) standard images with a frame rate of 30 Hz. Offline Analysis The images were interrogated offline using software (Axius VVI 2, Siemens Medical Solutions). The subendocardium of both ventricles was traced at end systole and a total of 2 to 3 cardiac cycles were averaged to obtain a velocity vector profile of myocardial motion (Figure 1). Longitudinal velocity measurements (peak systolic and peak diastolic) were performed at the right ventricular (RV) free wall, septum, and LV free wall at the level of the atrioventricular valves. A peak diastolic velocity was reported as the velocity profiles did not separate out early and late diastolic waveforms. Longitudinal strain and strain rate were measured in the midportion of the myocardium of the same 3 walls (Figure 2).

3 472 Younoszai et al Journal of the American Society of Echocardiography May 2008 Statistics Interobserver variation, defined as the difference of the values obtained between reviewers over the mean of those values, was calculated from a randomly chosen sample of 10 data sets. All variables of myocardial function were correlated with increasing gestational age and heart rate. RESULTS Feasibility Images from 27 fetal examinations were presented for offline analysis. Of these, 24 (89%) were successfully analyzed. The median gestational age was 25 and 3/7 weeks and ranged from 18 and 2/7 weeks to 39 weeks. The median fetal heart rate, calculated from reconstructed M-mode peak-to-peak cardiac intervals, was 146/min and ranged from 126 to 168/min. Interobserver Variability The interobserver variabilities for systolic velocity were 0.20,.018, and 0.22 in the RV free wall, septum, and LV free wall, respectively. Similar values for the diastolic velocity were 0.18, 0.24, and For strain they were 0.24, 0.17, and 0.15 and for strain rate they were 0.15, 0.30, and Changes with Gestational Age Systolic and Diastolic Annular Velocities The systolic and diastolic longitudinal velocities increased with gestational age in all myocardial segments measured (Figure 3). The strongest correlation was in the septum for both systolic and diastolic velocities (septal systolic r 0.78, P.0001, septal diastolic r 0.73, P.0001). The lowest correlation was in the RV free wall (RV systolic r 0.56, P.004, RV diastolic r 0.42, P.040) and the LV free wall was intermediate between the two (LV systolic r 0.65, P.0009, LV diastolic r 0.67, P.0006). Changes with Gestational Age Systolic Strain and Strain Rate Systolic strain and strain rate were not found to have significant correlation with gestational age in any of the wall segments (Figure 3). Changes with Heart Rate Systolic and diastolic velocities were not found to have a significant correlation with heart rate in any of the myocardial walls measured. Strain did not vary with heart rate either. Strain rate did appear to increase with an increase in heart rate in the RV free wall (r 0.55, P.01) and the septum (r 0.51, P.01). A similar trend was seen in the LV free wall (r P.12) (Figure 4). There was no significant change in heart rate with advancing gestational age. DISCUSSION VVI is a new technique for analyzing echocardiographic B-mode images offline that allows for assessment of myocardial mechanics. VVI uses complex mathematic modeling to track myocardial motion and then is able to calculate myocardial velocity, strain, and strain rate for these data. The software tracks echocardiographically generated speckles and factors annulus motion, periodicity, and border tracking to track the myocardium. Advantages of this technique include the ability to measure motion independent of the angle of image acquisition and the ability to measure motion in multiple directions (eg, longitudinal and radial motion) from a single image. These factors are Figure 3 Myocardial velocity, strain, and strain rate plotted versus gestational age. Plots of systolic and diastolic velocity, strain, and strain rate against advancing gestational age (weeks). Displayed range (y-axis) for systolic and diastolic velocity 0 to 5 cm/s, strain 0% to 40%, and strain rate 0 to 4/s. Note that there is statistically significant increase in velocity profiles with advancing gestational age (P.05). No significant variation in strain or strain rate is detected. LV, Left ventricle; RV, right ventricle. Color figure online. particularly important in fetal echocardiography when fetal positioning limits the angle of insonation. Feasibility Evaluation of myocardial dynamics in the fetus by VVI is feasible and was accomplished in the majority of fetal scans reviewed. The main reason for failing to obtain a reliable processed image for analysis was poor image quality. This is an ongoing challenge in imaging the fetus, especially in late gestation with increased shadowing from calcifying ribs and the spine. There was a learning curve to tracing the endocardium and it may be that future studies will yield a higher success rate. Interobserver variability was relatively high and likely also reflects a learning curve. As experience with this new technique increases, the reproducibility may improve as well. Changes with Gestational Age The systolic and diastolic velocities of all wall segments increased with gestational development; however, no significant change in strain or strain rate was observed. As the heart grows in fetal development, the distance that the myocardium travels in the longitudinal plane increases. Given that the time interval of the cardiac cycle remains constant, velocity increases proportionally. Strain is time independent and, therefore, a more reliable predictor of myocyte deformation. The fact that strain was not found to correlate with advancing gestational age and growth suggests that myocardial function is constant through mid and late gestation. This finding is supported by the variation of strain rate with heart rate. As the cardiac cycle length decreases the strain rate increases. This shows that within the normal

4 Journal of the American Society of Echocardiography Younoszai et al 473 Volume 21 Number 5 Figure 4 Strain rate versus heart rate. Strain rate (1/s) plotted against heart rate (beats/min). As heart rate increases, strain rate also increases in both right ventricular (RV) free wall and septum (P.05). There appears to be similar trend in left ventricular (LV) free wall (P.12). Color figure online. heart rate range of the developing fetus, myocardial strain appears to remain stable requiring the rate of deformation to compensate for the change in cycle length. If this finding is generalized to the entire myocardium, it would support a maintained stroke volume within the normal heart rate range, allowing an increase in heart rate to directly increase cardiac output. These data are consistent with the limited data available from chick embryos using microspheres to track strain. Taber et al 10 found that longitudinal and circumferential strain did not change during the early stages of ventricular looping. However, few other data are available. Data in adults (age years) using DTI have shown that strain does not change significantly with aging whereas myocardial velocities and strain rate decrease with advancing age. 11 Comparison with DTI Studies Several reports have used DTI to measure myocardial velocity, strain, and strain rate. Harada et al 4 published the first report using DTI to measure myocardial velocities. They found that RV and LV systolic velocities and early diastolic velocities at the RV, LV, and septum each correlated with gestational age. They did not show a correlation with systolic velocity at the septum or late diastolic velocity. Paladini et al 7 measured transverse myocardial velocities of the LV and RV at the subendocardium and subepicardium at various gestational ages. They found a significant correlation with gestational age at each point of measurement except late diastolic velocity of the subepicardium. More recently, Chan et al 5 looked at longitudinal myocardial velocities in 302 normal fetuses and found an increase in systolic and diastolic velocities in all segments. These previous reports measured peak myocardial velocities, whereas VVI measures mean modal velocity. Similar to findings in these studies that used DTI, we found an increase in longitudinal velocities through mid to late gestation. Little has been published regarding the evaluation of myocardial strain and strain rate in normal gestation. Using DTI, Di Salvo et al 9 found that strain and strain rate of the LV wall increased with gestational age. They did not report any data on changes in strain and strain rate for the RV free wall or septum. Therefore, it is unclear whether the correlation they found was consistent throughout all segments of the myocardium. In contrast, we found that strain and strain rate did not correlate with gestational age at any measured segment. Because VVI analysis is performed offline, measurements are made independent of imaging angle. This is particularly important in fetal echocardiography, where fetal positioning can limit the angle of insonation. Success in obtaining measurements using DTI ranged from 63% to 88% in studies requiring an angle of less than 30 degrees from the angle of measurement. 5,7,9 The 89% success rate using VVI in this study compares favorably with previous reports using DTI. Limitations This study had a limited number of patients. Further validation of this technique and its applicability in distinguishing normal from abnormal fetal cardiac function is warranted. Although VVI does not have the angle dependency of DTI, the ability to track the myocardium is reliant on adequate image quality and axial dropout should be minimized. Assessment of myocardial motion was further limited by the frame rate at which images were acquired. Although our imaging frame rates were higher (typically frames/s), the ultrasound system (Sequoia, Siemens Medical Systems) under standard settings saves DICOM images at a rate of 30 frames/s. Normal fetal heart rates vary from 120 to 160/min, meaning that a single cardiac cycle may have only 11 to 15 frames. Because temporal resolution was limited by the rate of image acquisition, the E wave and A wave in diastole were not distinct. As the frame rate of image acquisition is increased, VVI will be able to study myocardial mechanics with greater temporal resolution. Future Study The complexities of the mechanics of myocardial contraction are beginning to be understood in a more comprehensive way. The ventricles contract in the longitudinal, radial, and circumferential directions during systole. The heart also twists during systole, with clockwise rotation at the base and counterclockwise rotation at the apex in the healthy adult. Finally, the LV and RV normally contract in synchrony. Each of these components of ventricular contraction may be different during normal maturation or in pathologic states. New techniques, such as 2-dimensional image analysis through speckle tracking and related techniques and DTI, may allow for quantification of the different aspects of this complex motion. Previous work has shown that pressure overloading of embryonic chick hearts alters passive stress-strain relationships during development. 12 VVI may be a useful tool for quantification of changes in cardiac function in fetal distress and with developing congenital heart disease. Conclusions From mid to late gestation there is a progressive increase in myocardial longitudinal velocity in all measured cardiac segments with relatively stable myocardial deformation as measured by strain and

5 474 Younoszai et al Journal of the American Society of Echocardiography May 2008 strain rate. This suggests that the increase in velocity is the result of somatic growth of the heart rather than increasing contractility. This implies that the systolic contractile function of the myocyte is established as early as 18 weeks of gestation. Further investigation is required to optimize the temporal resolution; however, VVI holds significant promise for improved understanding of fetal myocardial function. REFERENCES 1. Falkensammer CB, Paul J, Huhta JC. Fetal congestive heart failure: correlation of Tei-index and Cardiovascular-score. J Perinat Med 2001; 29: Friedman WF. The intrinsic physiologic properties of the developing heart. Prog Cardiovasc Dis 1972;15: Grant DA. Ventricular constraint in the fetus and newborn. Can J Cardiol 1999;15: Harada K, Tsuda A, Orino T, Tanaka T, Takada G. Tissue Doppler imaging in the normal fetus. Int J Cardiol 1999;71: Chan LY, Fok WY, Wong JT, Yu CM, Leung TN, Lau TK. Reference charts of gestation-specific tissue Doppler imaging indices of systolic and diastolic functions in the normal fetal heart. Am Heart J 2005;150: Huhta JC, Kales E, Casbohm A. Fetal tissue Doppler a new technique for perinatal cardiology. Curr Opin Pediatr 2003;15: Paladini D, Lamberti A, Teodoro A, Arienzo M, Tartaglione A, Martinelli P. Tissue Doppler imaging of the fetal heart. Ultrasound Obstet Gynecol 2000;16: Tutschek B, Zimmermann T, Buck T, Bender HG. Fetal tissue Doppler echocardiography: detection rates of cardiac structures and quantitative assessment of the fetal heart. Ultrasound Obstet Gynecol 2003;21: Di Salvo G, Russo MG, Paladini D, et al. Quantification of regional left and right ventricular longitudinal function in 75 normal fetuses using ultrasound-based strain rate and strain imaging. Ultrasound Med Biol 2005;31: Taber LA, Sun H, Clark EB, Keller BB. Epicardial strains in embryonic chick ventricle at stages 16 through 24. Circ Res 1994;75: Sun JP, Popovic ZB, Greenberg NL, et al. Noninvasive quantification of regional myocardial function using Doppler-derived velocity, displacement, strain rate, and strain in healthy volunteers: effects of aging. J Am Soc Echocardiogr 2004;17: Tobita K, Schroder EA, Tinney JP, Garrison JB, Keller BB. Regional passive ventricular stress-strain relations during development of altered loads in chick embryo. Am J Physiol Heart Circ Physiol 2002; 282:H