Correlation analysis of ductus venosus velocity indices and fetal cardiac function

Transcription

1 Ultrasound Obstet Gynecol 2014; 43: Published online 3 April 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: /uog Correlation analysis of ductus venosus velocity indices and fetal cardiac function L. SANAPO, O. M. TURAN, S. TURAN, J. TON, M. ATLAS and A. A. BASCHAT Center for Advanced Fetal Care, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Maryland School of Medicine, Baltimore, MD, USA KEYWORDS: atrioventricular valve; Doppler; E/A ratio; myocardial performance index; velocity ratios ABSTRACT INTRODUCTION Objective To examine the relationships between the ductus venosus (DV) pulsatility index for veins (PIV), individual DV velocity ratios and diastolic and global myocardial cardiac function. Methods Doppler measurements of the DV, atrioventricular (AV) valves and ventricular in- and outflow were analyzed. The DV-PIV and velocity ratios for individual phases (systole (S), end-systolic relaxation (v), early diastole (D), atrial systole (a), and S/v, S/D, S/a, v/d, v/a and D/a ratios) were calculated. The ratio of early and late diastolic peak velocities across AV valves was calculated (E/A ratio). Left modified myocardial performance index (MPI) was calculated from time intervals between valve clicks defining isovolumetric contraction/relaxation and ejection times. All values were transformed to Z-scores. The distributions of DV velocity ratios and DV-PIV were correlated with cardiac Doppler parameters. Results A total of 1163 examinations from 213 fetuses, most of which were at risk for cardiac dysfunction, were included in the study. In 742 the PIV was normal and in 421 PIV was elevated > 2 SD above the normal mean. The DV-PIV correlated with velocity ratios (P < ) but not with E/A ratios and the MPI. S/v and v/d ratios were related to tricuspid and mitral E/A ratios and left ventricular MPI. The S/D ratio was only related to both E/A ratios. There was no relationship between a-waverelated velocity ratios and cardiac function. Conclusions Velocity ratios of the DV show relationships with cardiac function that are not reflected by the PIV alone. In cases of suspected fetal cardiac dysfunction based on elevated DV-PIV, analysis of velocity ratios or direct cardiac evaluation is suggested to determine the underlying pathophysiology. Copyright 2013 ISUOG. Published by John Wiley & Sons Ltd. When a ductus venosus (DV) Doppler examination is performed, the flow-velocity waveform is typically analyzed using semiquantitative indices such as the pulsatility index for veins (PIV) 1. The phases of the DV flow-velocity waveform are related in timing to the phases of the cardiac cycle and concurrent volume and pressure changes in the cardiac chambers 2. As a result, changes of the venous flow-velocity profile may be interpreted to reflect cardiac function. However, although widely considered a parameter of cardiac preload, DV Doppler indices do not show reproducible relationships with parameters of cardiac function in normal and high-risk pregnancies 3 9. One of the possible explanations is that traditional DV Doppler indices are not directly related to cardiac function 9,10. Understanding the relationship between venous flow dynamics and cardiac function is of critical importance to reach an accurate interpretation of Doppler findings 8. The events during the cardiac cycle coincide with four distinct phases of the venous flow-velocity waveform. The S -wave corresponds to ventricular systole followed by the v -wave during end-systolic ventricular relaxation and the ascent of the atrioventricular (AV) valves before the onset of diastole. With the opening of the AV valves, the D and a -waves correspond to early passive and late active diastolic filling, respectively. We have recently derived reference ranges for velocity ratios for all four phases of the DV waveform from pregnancies with documented normal outcome 11. Three of these ratios relate consecutive cardiac events: the S/v ratio quantifies relative forward flow into the atria as the ventricle relaxes before the opening of the AV valves. The v/d ratio reflects the early diastolic filling immediately following this event. The D/a ratio is a diastolic parameter that relates the magnitude of forward flow during passive and active diastolic filling, analogous to the E/A ratio for the AV valves. Three ratios relate non-consecutive Correspondence to: Dr A. A. Baschat, Department of Obstetrics, Gynecology and Reproductive Sciences, University of Maryland School of Medicine, Baltimore, 22 South Greene Street, 6 th floor, room 6NE11, Baltimore, MD 21201, USA ( abaschat@umm.edu) Accepted: 21 October 2013 Copyright 2013 ISUOG. Published by John Wiley & Sons Ltd. ORIGINAL PAPER

2 516 Sanapo et al. cardiac events: the S/D ratio quantifies ventricular systolic to early passive diastolic filling 12, the S/a ratio quantifies ventricular systolic to active diastolic filling 13 and the v/a ratio quantifies end-systolic relaxation and active diastolic filling. We hypothesized that DV velocity ratios will allow more detailed assessment of the relationship between the DV flow profile and cardiac function. It was the aim of this study to correlate the DV-PIV and velocity ratios to the diastolic E/A ratio and the modified myocardial performance index (MPI) in order to demonstrate the aspects of the venous waveform that relate most closely to cardiac function. METHODS We evaluated waveforms of the DV, AV valves and the left and right ventricular outflow tracts that were acquired during the same examination based on a clinical indication. We chose examinations that were performed at our Fetal Medicine Unit between 2006 and During this period all examinations were performed transabdominally, according to a prospectively predefined technique, by registered medical diagnostic sonographers or fetal medicine specialists using 4 8-MHz probes (Voluson E8, GE Healthcare Wauwatosa, WI, USA or Siemens Sequoia 512, Siemens, Mountain View, CA, USA). The DV Doppler measurements were obtained in the absence of fetal movements in a midsagittal or transverse abdominal plane after color Doppler imaging using the smallest possible insonation angle (< 30 ). After five or six constant waveforms were sampled with high signal-to-noise ratio, the image was frozen. The waveform was traced from the beginning of ventricular systoletotheendofatrialsystole.wemeasuredthepeak velocity of the four phases of the DV (S, v, D and a) and the time averaged maximum velocity in cm/s (Figure 1). From these, we calculated the PIV. We also calculated velocity ratios for consecutive cardiac events (S/v, v/d and D/a ratios) and ratios that related systolic to diastolic events (S/D, S/a and v/a ratios). Figure 2 Typical flow velocity waveform obtained with placement of the Doppler gate beyond the atrioventricular valves. From these waveforms, peak systolic velocity during early passive (E) and late active (A) ventricular filling was measured to calculate the E/A ratio for tricuspid and mitral valves (MV). Figure 3 Typical flow velocity waveform obtained with placement of the Doppler gate over atrioventricular inflow and ventricular outflow. The valve clicks allow timing of closure of the atrioventricular valve (1) and opening of the semilunar valve (2), closure of the semilunar valve (3) and reopening of the atrioventricular valve (4). Measuring the time intervals between these events yields the isovolumetric contraction time (ICT), ejection time (ET) and isovolumetric relaxation time (IRT). These intervals were used to calculate the modified myocardial performance index (ICT + IRT)/ET for the left ventricle. Figure 1 Typical flow velocity waveform of the ductus venosus (DV). For the purpose of the study, peak velocities during ventricular systole (S), end-systolic ventricular relaxation (v), early diastole (D) and atrial systole (a) were measured for calculation of velocity ratios. Left and right ventricular diastolic function was evaluated by the ratio of early and atrial peak velocities across the AV valves (E/A ratio) (Figure 2). For this measurement the sample gate was placed immediately distal to the respective AV valve. Left ventricular global myocardial performance was evaluated using the MPI described by Hernandez-Andrade et al. 14. In a fourchamber view, the Doppler sample volume was placed immediately distal to the mitral valve encompassing the ascending aorta, as previously described 15. After five or six constant waveforms were obtained that clearly demonstrated the valve clicks, the cursors were placed to measure the isovolumetric contraction time (ICT), ejection time (ET) and isovolumetric relaxation time (IRT) in milliseconds (Figure 3). From these measurements

3 Ductus venosus waveform and fetal cardiac function 517 the MPI was calculated according to the formula: MPI = (ICT + IRT)/ET for the left ventricle. The DV-PIV and individual velocity ratios were converted into Z-scores using reference ranges derived from a population with documented normal outcome 11. For the E/A ratio we calculated Z-scores from the reference ranges of van Splunder et al. 1 between 11 and 16 weeks and the nomogram by Fernández Pineda et al. 16 for measurements obtained thereafter. The MPI of the left ventricle was transformed into Z-scores utilizing the nomogram of Cruz-Martinez et al. 17. Because we sought to evaluate the general relationship between the DV waveform and cardiac function, all examinations meeting inclusion criteria were analyzed. In order to account for the potential bias of multiple examinations from one patient, we used a mixed-model analysis method for statistical calculation. Goodness of fit of the models was assessed by Akaike s information criterion (AIC) which accounts for random effects that can occur from repeat measurements derived from the same individual. The DV-PIV and DV velocity ratios, expressed as Z-scores, were each used as the dependent variable and the cardiac indices were assigned as covariates in these analyses. Data were analyzed using Microsoft Excel and SPSS version 21.0 (SPSS Co., Chicago, IL, USA). The study was approved by the Institutional Review Board of the University of Maryland School of Medicine. RESULTS A total of 1163 examinations from 213 fetuses were included in the study. The median gestational age at ultrasound was 21.4 (range, ) weeks. The DV- PIV was normal in 742 (63.8%) examinations. The majority of examinations were performed in pregnancies complicated by twin twin transfusion syndrome and pregestational diabetes (Table 1). DV-PIV was associated with a significant difference in the distribution of all phasic DV velocity indices (P < for all), in the following order of decreasing strength of relationship: S/a, D/a, v/a, S/v, v/d and S/D (AIC was , , , , and , respectively). However, DV-PIV was not associated with the E/A or MPI Z-scores (Table 2). There was no statistical relationship between a-waverelated phasic velocity ratios and cardiac function. The ratio between early and late systolic forward velocity (S/v ratio) was positively correlated with tricuspid and mitral E/A ratios and inversely related to the left ventricular MPI (P = 0.004, Table 2). The ratio between peak velocities during ventricular relaxation and early diastole (v/d ratio Z-score) showed a significant, inverse correlation with the mitral and tricuspid E/A ratio Z-score and was inversely associated with left ventricular MPI (P = 0.002, Table 2). The ratio of forward velocity during ventricular systole and early passive diastolic filling (S/D ratio) was inversely related to both E/A ratio Z-scores (P < for both). Table 1 Characteristics of the study population Characteristic Value Gestational age at ultrasound (weeks) 21.4 (11.3 to 34.2) PIV Z-score 1.30 ( 2.13 to 23.9) TTTS recipient 510 (43.9) TTTS donor 478 (41.1) s-iugr in twin pregnancies 11 (0.9) TRAP 2 (0.2) Pregestational diabetes 107 (9.2) FGR 20 (1.7) Fetal hydrops 4 (0.4) Cardiac defects 9 (0.8) Values are given as median (range) or n (%). FGR, fetal growth restriction (estimated fetal weight < 10 th percentile and Doppler abnormalities suggestive of placental dysfunction); s-iugr, selective growth restriction of one monochorionic twin; TRAP, twin reversed arterial perfusion; TTTS, twin twin transfusion syndrome. DISCUSSION The assumption that DV flow patterns reflect cardiac function is inconsistent with studies that evaluated this relationship using Doppler indices such as the PIV 5 8. The aim of this study was to evaluate the relationship between DV Doppler waveforms and cardiac function, regardless of the underlying pathology. Because venous flow fluctuates in four phases that coincide with distinct cardiac events we extended our analysis beyond the PIV by using velocity ratios that quantify relative forward flow during these events. We have confirmed that a relative decrease in the a-wave correlates most closely with an increased DV-PIV 5. However, neither the a- wave-related ratios nor the PIV correlate with diastolic and global myocardial performance measurements. In contrast, v-wave-related ratios significantly correlate with left ventricular cardiac function. There are few studies that have focused on phases of the DV flow profile other than the visually most apparent a-wave. Among the first velocity indices described was the preload index or S/a ratio for the inferior vena cava 13. Smrcek et al. described the use of the S/D ratio as providing a superior measure of passive diastolic filling in fetuses with tricuspid regurgitation 12. Picconi et al. developed a DV Doppler index that incorporates a direct measurement of the v-wave and was able to demonstrate a better prediction of adverse cardiovascular outcome and survival in fetal growth restriction (FGR) 18. Unfortunately, these studies did not correlate these velocity ratios with direct measurement of cardiac function. Measurement of cardiac function has primarily been related to traditional semiquantitative Doppler indices, such as the PIV, demonstrating that this index correlates neither with the E/A ratio 2,5 nor with the MPI The continuity of venous forward flow fluctuates with the capacity of the heart to accommodate venous return. This capacity depends on venous volume (preload), cardiac function (relaxation, compliance and contractility) and downstream arterial blood-flow resistance (afterload) 8,23,24. However, establishing the primary

4 518 Sanapo et al. Table 2 Correlation between ductus venosus (DV) and cardiac Doppler parameters E/A Z-score MPI Z-score Tricuspid valve Mitral valve Mitral valve Z-score Estimate (95% CI) P Estimate (95% CI) P Estimate (95% CI) P DV-PIV 0.09 ( 0.16 to 0.34) ( 0.09 to 0.32) ( 0.17 to 0.05) 0.29 Velocity ratio S/v 0.17 (0.07 to 0.28) (0.07 to 0.24) < ( 0.11 to 0.02) S/D 0.21 ( 0.32 to 0.12) < ( 0.27 to 0.11) < ( 0.02 to 0.07) 0.32 v/d 0.30 ( 0.44 to 0.17) < ( 0.39 to 0.18) < (0.03 to 0.16) S/a 0.03 ( 0.18 to 0.11) ( 0.40 to 0.10) ( 0.09 to 0.04) 0.39 v/a 0.10 ( 0.26 to 0.06) ( 0.21 to 0.04) ( 0.08 to 0.05) 0.65 D/a 0.04 ( 1.07 to 0.10) ( 0.13 to 0.09) ( 0.09 to 0.03) 0.36 a, atrial systole; D, early diastole; DV-PIV, ductus venosus pulsatility index for veins; E/A, ratio of early and late diastolic peak velocities across atrioventricular valves; MPI, myocardial performance index; S, ventricular systole; v, end-systolic ventricular relaxation. determinant of venous forward flow can be complex because several factors may be involved. For example, increased afterload, decreased ventricular contractility and compliance all can lead to increased end-systolic ventricular volume or pressures resisting venous forward flow. An example of increased afterload is severe FGR with high placental blood-flow resistance. In placentabased FGR the fetus is also at risk for myocardial hypoxemia, which in turn can lead to contractile dysfunction as well as abnormal relaxation 10,24. Without concurrent quantification of all underlying variables it is impossible to identify the primary contributor to abnormal venous forward flow 1,5,9,10,24. In the DV the matter is further complicated by modulation in vessel diameter, potentially affecting retrograde transmission of cardiac pressure volume relationships and especially the antegrade velocities during atrial systole 23. This study only considered diastolic and global myocardial performance, and without knowledge of venous volume, additional parameters of cardiac function and afterload we are unable to identify the mechanisms that determine individual velocity ratios. In the context of our hypothesis the velocity ratios offer the advantage of relating venous flow to distinct phases of the cardiac cycle 8,24,25. With the help of these ratios we were able to demonstrate that end-systolic v-wave changes most closely relate to cardiac function. In the venous system, resistance to forward flow during the v-wave is mostly generated by ascent of the closed AV valves toward the filling atria 5,8,23,24. Abnormal cardiac compliance and ventricular relaxation, both related to myocardial oxygenation in the fetus 24,26, can lead to increased end-systolic intracardiac pressure or residual volume, both of which oppose forward flow during the v-wave 23. The closer relationship between the v-wave ratios and cardiac function can be explained by the absence of other atrial mechanical events during this part of the cardiac cycle. In contrast, the a-wave is produced by the retrograde pressure that is superimposed by atrial systole. While increased end-systolic pressure and volume contribute to the magnitude of this retrograde pressure, our results suggest that atrial systole obscures the contribution of these ventricular variables; accordingly, a-wave-related ratios do not closely reflect diastolic and myocardial function. This is further supported by the observation that D/a and E/A ratios, which reflect concurrent events in the cardiac cycle, are unrelated in our and previous studies 2,5,10, One could therefore hypothesize that the v-wave is primarily attributable to cardiac events, whilst the a-wave reflects a combination of extra- and intracardiac variables, limiting the ability to discern the cardiac contribution. The limitations of our study include the retrospective selection of examinations based on the availability of measurements of venous and cardiac waveforms performed during the same examination. Although we present a large number of examinations, ascertainment bias, particularly of fetuses with overt cardiac dysfunction, cannot be excluded. Finally, there are more sophisticated techniques of cardiac evaluation that could enable us to study the pathophysiology of venous flow waveforms in relationship with additional cardiac function parameters. However, the study design allowed us to address our hypothesis in a large sample size. In summary, our observations indicate that relative forward venous flow during end-systolic relaxation, and, to a lesser extent, early diastole, has the strongest relationship with diastolic and global myocardial performance. Relative forward flow during atrial systole is the major numeric contributor to the PIV but does not correlate with these cardiac performance parameters. Accordingly, elevation of the DV-PIV requires examination of cardiac performance parameters in order to determine the primary underlying contributor 8 10,17,24. Further research into the clinical validity of alternative Doppler indices that emphasize the v- and D-waves may widen the information that can be gained through the clinical application of venous Doppler. REFERENCES 1. van Splunder P, Stijen T, Wladimiroff JW. Fetal atrioventricular flow-velocity waveforms and their relation to arterial and venous flow-velocity waveforms at 8 to 20 weeks of gestation. Circulation 1996; 94:

5 Ductus venosus waveform and fetal cardiac function Baschat AA. Examination of the fetal cardiovascular system. Semin Fetal Neonatal Med 2011; 16: Kiserud T. Physiology of the fetal circulation. Semin Fetal Neonatal Med 2005; 10: Crispi F, Gratacós E. Fetal cardiac function: technical considerations and potential research and clinical applications. Fetal Diagn Ther 2012; 32: Hecher K, Campbell S, Snijders R, Nicolaides K. Reference ranges for fetal venous and atrioventricular blood flow parameters. Ultrasound Obstet Gynecol 1994; 4: Crispi F, Hernandez-Andrade E, Pelsers MM, Plasencia W, Benavides-Serralde JA, Eixarch E, Le Noble F, Ahmed A, Glatz JF, Nicolaides KH, Gratacos E. Cardiac dysfunction and cell damage across clinical stages of severity in growth-restricted fetuses. Am J Obstet Gynecol 2008; 199: 254.e Rychik J, Tian Z, Bebbington M, Xu F, McCann M, Mann S, Wilson RD, Johnson MP. The twin-twin transfusion syndrome: spectrum of cardiovascular abnormality and development of a cardiovascular score to assess severity of disease. Am J Obstet Gynecol 2007; 197: 392.e Baschat AA, Turan OM, Turan S. Ductus venosus blood flow patterns more than meets the eye? Ultrasound Obstet Gynecol 2012; 39: Bellotti M, Pennati G, De Gasperi C, Bozzo M, Battaglia FC, Ferrazzi E. Simultaneous measurements of umbilical venous, fetal hepatic, and ductus venosus blood flow in growthrestricted human fetuses. Am J Obstet Gynecol 2004; 190: Baschat AA. Relationship between placental blood flow resistance and precordial venous Doppler indices. Ultrasound Obstet Gynecol 2003; 22: Turan OM, Turan S, Sanapo L, Wilruth A, Berg C, Harman CR, Baschat AA. Reference ranges for ductus venosus velocity ratios in pregnancies with normal outcomes. J Ultrasound Med 2014; 33: Smrcek JM, Krapp M, Axt-Fliedner R, Kohl T, Geipel A, Diedrich K, Gembruch U, Berg C. Atypical ductus venosus blood flow pattern in fetuses with severe tricuspid valve regurgitation. Ultrasound Obstet Gynecol 2005; 26: Kanzaki T, Chiba Y. Evaluation of the preload condition of the fetus by inferior vena caval blood flow pattern. Fetal Diagn Ther 1990; 5: Hernandez-Andrade E, Lopez-Tenorio J, Figueroa-Diesel H, Sanin-Blair J, Carreras E, Cabero L, Gratacos E. A modified myocardial performance (Tei index based on the use of valve clicks improves reproducibility of fetal left cardiac function assessment. Ultrasound Obstet Gynecol 2005; 26: Turan S, Turan OM, Miller J, Harman C, Reece EA, Baschat AA. Decreased fetal cardiac performance in the first trimester correlates with hyperglycemia in pregestational maternal diabetes. Ultrasound Obstet Gynecol 2011; 38: Fernández Pineda L, Tamariz-Martel Moreno A, Maître Azcárate MJ, López Zea M, Rico Gómez F, Cazzaniga Bullón M, Quero Jiménez M. Contribution of Doppler atrioventricular flow waves to ventricular filling in the human fetus. Pediatr Cardiol 2000; 21: Cruz-Martinez R, Figueras F, Bennasar M, Garcia-Posadas R, Crispi F, Hernandez-Andrade E, Gratacos E. Normal reference ranges from 11 to 41 weeks gestation of fetal left modified myocardial performance index by conventional Doppler with the use of stringent criteria for delimitation of the time periods. Fetal Diagn Ther 2012; 32: Picconi JL, Kruger M, Mari G. Ductus venosus S- wave/isovolumetric A-wave (SIA) index and A-wave reversed flow in severely premature growth-restricted fetuses. JUltrasound Med 2008; 27: Girsen A, Ala-Kopsala M, Mäkikallio K, Vuolteenaho O, Räsänen J. Cardiovascular hemodynamics and umbilical artery N-terminal peptide of prob-type natriuretic peptide in human fetuses with growth restriction. Ultrasound Obstet Gynecol 2007; 29: Mäkikallio K, Vuolteenaho O, Jouppila P, Räsänen J. Ultrasonographic and biochemical markers of human fetal cardiac dysfunction in placental insufficiency. Circulation 2002; 105: Cruz-Martinez R, Figureras F, Benavides-Serralde A, Crispi F, Hernandez-Andrade F, Gratacos E. Sequence of changes in myocardial performance index in relation to aortic isthmus and ductus venosus Doppler in fetuses with early-onset intrauterine growth restriction. Ultrasound Obstet Gynecol 2011; 38: Rozmus W, Wloch A, Acharya G, Cnota W, Czuba B, Sodowski K, Skrzypulec V. Reference values for variables of fetal cardiocirculatory dynamics at weeks of gestation. Ultrasound Obstet Gynecol 2010; 35: Pappworth MW. The cardiovascular system. In APrimerof Medicine (5 th edn). Butterworths: London, 1984; Huhta JC. Deciphering the hieroglyphics of venous Doppler velocities. Ultrasound Obstet Gynecol 1997; 9: Rizzo G, Capponi A, Talone PE, Arduini D, Romanini C. Doppler indices from inferior vena cava and ductus venosus in predicting ph and oxygen tension in umbilical blood at cordocentesis in growth-retarded fetuses. Ultrasound Obstet Gynecol 1996; 7: Nakamura Y, Wiegner AW, Bing OH. Measurement of relaxation in isolated rat ventricular myocardium during hypoxia and reoxygenation. Cardiovasc Res 1986; 20: