http://informahealthcare.com/jmf ISSN: 1476-7058 (print), 1476-4954 (electronic) J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437 ! 2014 Informa UK Ltd. DOI: 10.3109/14767058.2013.878695
ORIGINAL ARTICLE
Evaluation of prenatal risk factors for prediction of outcome in right heart lesions: CVP Score in fetal right heart defects
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Ana Luisa Neves1,2, Leigh Mathias3, Marilyn Wilhm4, Jennifer Leshko5, Kersti K. Linask6, Tiago Henriques-Coelho1,2, Jose´ C. Areias1,2, and James C. Huhta7,8 1
Hospital S. Joa˜o, Porto, Portugal, 2Faculty of Medicine, University of Porto, Porto, Portugal, 3College of Public Health, University of South Florida, Tampa, FL, USA, 4All Children’s Hospital, Heart Institute, St. Petersburg, FL, USA, 5All Children’s Hospital, St. Petersburg, FL, USA, 6Children’s Research Institute, University of South Florida, Tampa, FL, USA, 7Department of Perinatal Cardiology, Pediatrix Medical Group, University of Florida, St. Petersburg, FL, USA, and 8All Children’s Hospital Outpatient Care Center, St. Petersburg, FL, USA
Abstract
Keywords
Objective: To determine the prenatal variables predicting the risk of perinatal death in congenital right heart defects. Methods: Retrospective analysis of 28 fetuses with right heart defects was performed. Logistic regression analyses were performed to obtain odds ratios (OR) for the relationship between the risk of death and echocardiographic parameters. The parameters that correlated with the outcome were incorporated in an attempt to devise a disease-specific cardiovascular profile score. Results: Fetal echocardiograms (143) from 28 patients were analyzed. The cardiovascular profile score predicted the risk of death. A lower right ventricle (RV) pressure was associated with mortality (OR 0.959; 95% confidence intervals (CI) 0.940–0.978). Higher peak aortic velocity through the aortic valve (OR 0.104; 95% CI 0.020–0.529) was associated with a better outcome. These cardiac function parameters were incorporated in a modified disease-specific CVP Score. Patients with a mean modified cardiovascular profile score of 6 were over 3.7 times more likely to die than those with scores of 7–10. Conclusions: The original Cardiovascular Profile Score predicted the risk of death in right heart defects. The modified score was not validated as a good prediction tool by this study. Fetal RV pressure estimate and peak aortic velocity can be used as independent prognostic predictors.
Congenital heart defects, fetal echocardiography, outcomes
Introduction Congestive heart failure (CHF) is a clinical syndrome resulting from a variety of disorders that impair the ability of the ventricles to fill with or to eject blood to ensure adequate tissue perfusion to meet metabolic demands of the body [1]. CHF in fetuses with congenital heart diseases is associated with high perinatal mortality [2]. The clinical use of echocardiography for prenatal imaging has the major advantage that it is non-invasive, easily accessible and can be used in pregnancy to evaluate the fetal pathophysiological state. The Cardiovascular Profile Score (CVP), an echocardiography-based multivariate approach, for the assessment of fetal cardiac hemodynamics. It evaluates the degree of hydrops, the venous flow velocity pattern (umbilical vein and ductus venosus), the heart size Address for correspondence: James C. Huhta, Professor of Pediatrics, Department of Perinatal Cardiology, Pediatrix Medical Group, University of Florida, United States of America and All Children’s Hospital Outpatient Care Center, 601 5th Street South, Dept. #7125, St. Petersburg, FL 33701, USA. Tel: +1 727 767 4782. Fax: +1 727 767 8111. E-mail: [email protected]
History Received 16 May 2013 Revised 28 November 2013 Accepted 20 December 2013 Published online 29 January 2014
(cardiac to thoracic area ratio), cardiac function (tricuspid or mitral regurgitation), and arterial flow (umbilical artery). The CVP Score has been shown in our previous studies, which is to be useful in predicting the outcome of fetuses with hydrops [3], growth restriction [4], and congenital heart disease (CHD) [5]. It has been shown that growth-restricted fetuses with adverse neonatal outcome had lower CVP Scores than did fetuses with favorable neonatal outcome. The strongest predictors for adverse neonatal outcome in the CVP Score were cardiomegaly, abnormal cardiac function with monophasic atrioventricular filling or holosystolic tricuspid regurgitation (TR), and abnormal flow patterns suggestive of abnormal right atrium/venous pressure [4]. It has been shown that fetuses with congenital heart defects and a CVP Score below 8 were at risk of perinatal death and that the CVP Score may be used to assess the severity of fetal CHD and to plan perinatal management [5]. It is also useful for appropriate timing and monitoring of the success of intrauterine therapy for CHF [6]. There is significant variability in the hemodynamics in different types of CHD, so a specific CVP Score for different heart defects may be useful.
1432
A. L. Neves et al.
Our hypothesis is that the outcome of fetuses with right heart defects may be related to prenatal condition and left ventricular (LV) function. Accordingly, our aim was to determine the variables predicting the risk of perinatal death in right heart defects in order to achieve a diseasespecific CVP Score that could be used in clinical perinatal medicine.
Methods
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Study design and patient population Medical records and fetal echocardiograms from fetuses with right heart defects, evaluated at the All Children’s Hospital Perinatal Care Center, St. Petersburg, Florida, in five consecutive years, were analyzed retrospectively. The reviewing was blinded to outcomes. This study was approved by the All Children’s Hospital Institutional Review Board and all patients provided informed consent. Right heart defects included pulmonary valve atresia with intact septum (PVAIS), severe pulmonary stenosis (PS), and Ebstein’s anomaly (EbA). Two patients were excluded, one for pregnancy termination and the other for lack of follow-up information. Outcome was defined as either survival or perinatal death (intrauterine death or death within the first month of postnatal life). The degree of fetal CHF was quantified, as previously published [3–5], by the CVP Score, an assessment that evaluates the degree of hydrops, the venous flow velocity pattern (umbilical vein and ductus venosus), the heart size (cardiac to thoracic area ratio), the cardiac function (tricuspid or mitral regurgitation), and the arterial flow (umbilical artery). The presence or absence of hydrops was noted. One point was deducted if ascites, pleural effusion, or pericardial effusion was present. Two points were deducted if skin edema was present. Venous Doppler flow patterns were described as either normal or abnormal. The normal umbilical vein is non-pulsatile. An abnormal flow pattern was umbilical venous pulsations in a free loop of the umbilical cord (two points deducted). In the ductus, venosus Doppler is normally pulsatile, but the abnormal Doppler flow pattern was defined as atrial systolic flow reversal (one point deducted), suggesting either right heart diastolic dysfunction or elevated central venous pressure. The cardiothoracic area ratio (CTR) was obtained as the simple ratio of areas within the ellipses drawn around the pericardium of the heart and around the thorax. This transverse section was taken to include opposing ribs and the spine in the transthoracic four-chamber view. The normal range values are 0.25–0.35. The presence or absence of atrioventricular valve regurgitation was assessed by color flow and pulsed Doppler. Holosystolic TR or depressed shortening fraction demanded a one point deduction while with mitral regurgitation or monophasic filling, two points were deducted About arterial Doppler, the mid-segment of the cord (free loop) was assessed. Umbilical artery absent diastolic flow resulted in a one-point deduction, while reverse diastolic flow deducted two points. A CVP of 10 was indicative of no CHF. Points were deducted for abnormalities of each category, so the lower the CVP Score, the more the severe the CHF [3]. All echocardiographic studies were performed with
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
Table 1. Echocardiographic parameters correlated with mortality. Parameter
OR
Right ventricle hemodynamic parameters RV gradient 0.384 RV pressure 0.959 RV/LV pressure 0.395 Left ventricle hemodynamic parameters LV peak aortic velocity 0.104 LV MPI 0.065 MV E/A 0.775 MV E/E’ 1.021 MV E’/A’ 2.149 Morphometric parameters TV/MV size 0.956 RA/LA width 1.166 RV/LV width 0.369
95% CI
p
0.228–0.646 0.940–0.978 0.237–0.659
0.0003* 50.0001* 0.0004*
0.020–0.529 0.002–1.792 0.366–1.641 0.869–1.200 0.977–4.729
0.0064* 0.1063 0.5047 0.7977 0.5233
0.573–1.596 0.783–1.735 0.190–0.715
0.8638 0.4500 0.0032*
RV, right ventricle; LV, left ventricle; MPI, myocardial performance index; MV, mitral valve; RA, right atrium; LA, left atrium. *Statistically significant (p50.05).
a Siemens Sequoia 512 (Siemens Healthcare, Malvern, PA) with curvilinear transducers. Specific study parameters for right heart defects The estimate of the degree to which the right ventricular pressure was greater than the systemic pressure was obtained by converting the peak velocity of the TR jet to a gradient using the four times velocity squared modified Bernoulli equation [7]. The peak pressure gradient of the TR jet and the right ventricle (RV) estimated pressure (obtained by adding the right atrium (RA) estimated pressure of 5 to the peak pressure gradient of the TR jet) were assessed. The peak aortic velocity through the aortic valve is a sensitive index of global left ventricular performance [8,9]. Peak aortic velocity through the aortic valve was used as a surrogate for cardiac output (stroke volume) of the LV. The myocardial performance index (MPI) or Tei index [10] is the sum of the isovolumic times divided by the ejection time. The isovolumic time is the sum of the two intervals – the isovolumic relaxation time and the isovolumic contraction time. This term was measured by subtracting the ejection time from the time between two inflows in the LV [7]. Peak velocities of the inflow E and A waves at the annulus of the mitral valve (MV E/A) were analyzed, as well as the ratio between maximum inflow blood velocity during early diastole and annular myocardial lengthening velocity (by tissue Doppler) during early ventricular relaxation (E/E’) which reflects ventricular end-diastolic and atrial pressures [11,12]. The use of the ratios obviated the need for agespecific normal data [13]. The cardiac function parameters that correlated with the outcome (Table 1) were incorporated in a specific modified CVP Score (Figure 1), with a one point deducted for TR jet velocity 52 m/s, and two points deducted for LV ejection velocity 51 m/s. This specific Cardiovascular Profile Score was applied and statistical analyses were performed. Statistical analysis Observed clinical data and cardiac function category values for all subjects were analyzed using SAS statistical software,
CVP score in fetal right heart defects
1433
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
DOI: 10.3109/14767058.2013.878695
Figure 1. Modified Cardiovascular Profile Score specific for fetal right heart defects. UV, umbilical vein; DV, ductus venosus; UA, umbilical artery; white arrow, pericardial effusion; TV, tricuspid valve; TR, tricuspid regurgitation.
version 9.1 (SAS, Cary, NC). Chi-squared analysis was used to identify statistically significant differences between survivors and non-survivors with respect to all the clinical cardiac indicators mentioned above. Crude odds ratio (OR) with corresponding 95% confidence intervals (CI) were obtained through separate unconditional logistic regression analyses for the relationship between risk of death and the following parameters: CVP Score, modified CVP score, RV pressure, and parameters of LV performance (peak systolic velocity through the aortic valve, LV myocardial performance index, Doppler E/A, tissue Doppler E’/A’, and E/E’).
Results Fetal echocardiogram exams (143) from 28 patients were analyzed, ranging from 1 to 13 echocardiograms (median of 4) per patient. The mean ± SD gestational age (GA) at the delivery was 37 ± 2 weeks; the mean ± SD birth weight was 2705 ± 677 g; the median Apgar scores at 1 and 5 min ranged between 1–9 (median 8) and 5–9 (median 8), respectively. Seventeen patients had pulmonary atresia with intact ventricular septum (PAIVS) (TV/MV51 in 12,41 in 5), seven PS, and four EbA. From patients with PAIVS, eight had
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
1434
A. L. Neves et al.
amniocentesis, with normal karyotype and were negative for 22q11.2 deletion. Three patients with critical pulmonary stenosis had abnormal karyotypes (one had trisomy 18, one had terminal deletion of chromosome 3, and one had trisomy 21). From the 28 fetuses, 25 were born alive. The total mortality was 14/28 (50%) by 30 d of postnatal life (Table 2). Twelve newborns had surgery. Four died after surgery. Seven died before intervention. From these, three had PAIVS with TV/MV51 (one had Staph aureus endocarditis, two had RVdependent coronary circulation and were listed for heart transplant). One had PAIVS with TV/MV41 and died of arrhythmia. Two patients with critical pulmonary stenosis died before intervention, one had trisomy 18 and the other had terminal deletion of one chromosome 3, were a twin pregnancy with a birth weight of 800 g and did not survive. One patient with Ebstein had sepsis and brain parenchymal hemorrhage and was too critically ill to attempt intervention. The original CVP Score appeared to be a predictor of the risk of death in fetuses with right heart defects. The observed OR of 0.9252 (95% CI 0.866–0.988) indicated that a higher CVP Score was associated with a decreased risk of death. The lower the CVP Score, the greater the risk of death, meaning that for each one-point score reduction of the CVP, there is an additional 8% risk of death. The five CVP categories were analyzed separately and correlated with mortality (Table 3). Pericardial effusion was found in at least one echocardiogram of 16 patients, two of them additionally had skin edema. Of the patients with points deducted in the hydrops category, 50% died. For venous Doppler, 16 had abnormal results, 12 had ductus venosus reversal while three also had umbilical vein pulsations. Mortality associated with umbilical vein pulsations was 67%. The CTR was not correlated with death with an OR of 0.770 (95% CI 0.050–11.796). Holosystolic TR was found in 12 patients and monophasic filling in 8. Mortality occurred in three of the patients with holosystolic TR (25%) and five of the patients with monophasic filling (63%). The velocity of the TR jet varied between 2.66 m/s and 4.35 m/s. Umbilical artery Doppler abnormalities were found rarely (two patients had absent diastolic flow) and no patient had reversed diastolic flow. Observed values of RV pressure correlated with an increased risk of mortality, with corresponding OR of 0.959 (95% CI 0.940–0.978). This finding indicated that an increase in the RV pressure is associated with a decrease in risk of death among subjects (Table 1). Higher observed peak aortic velocity through the aortic valve was found to be associated with decreased risk of death, with an OR of 0.104 (95% CI 0.020–0.529) among subjects. This suggests that as the peak aortic velocity through the aortic valve increases, the risk of death is reduced (Table 1). No statistically significant differences were found between the LV myocardial performance index and survival. While not statistically significant (95% CI 0.977–4.729), the observed association between the ratio of the mitral tissue Doppler E’/ A’ in the LV and risk of death (OR 2.149) suggests that each 1-unit increase in the MV E’/A’ ratio equates to a two-fold increase in the risk of death. No differences were found in the ratio E/E’.
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
The cardiac function parameters that correlated with the outcome (Table 1) were incorporated in a modified diseasespecific CVP adjusted to the right heart defects (Figure 1). The parameters were peak aortic velocity through the aortic valve and peak velocity of the TR jet. In comparison with the original Cardiovascular Profile Score, changes were made at the cardiac function category, with a one-point deducted for TR jet velocity 52 m/s, and two points deducted for peak aortic velocity through the aortic valve 51 m/s (maximum deduction in this category is two points in any category). The specific CVP Score was applied to all the echocardiograms, and a statistical analysis was performed using the mean score for each patient (Table 3). With the mean CVP Score for each patient, we were able to observe an OR of 0.292. The corresponding 95%CI (0.074– 1.160) was not significant. When the analysis was performed with a categorical variable of good (46)/bad (6) modified CVP Score, we also observed some strong results. Those patients with a mean modified CVP Score of 6 or less were over 3.7 times as likely to die in comparison to those patients with mean CVP Scores of seven or more.
Discussion The clinical presentation of CHF in the fetus can be characterized by findings in at least five categories. Abnormalities in the CVP Score may occur prior to the clinical state of fetal hydrops [14]. The final common pathway of fetal death in CHF is poor tissue perfusion and acidosis [14]. The CVP Score has been shown to be useful in predicting the outcome of fetus with hydrops [3], twin–twin transfusion [15], intrauterine growth restriction [4], and CHD [5]. A disease-specific cardiovascular Score could be a way of improving the interpretation of hemodynamic data [1]. The original CVP Score combines five areas of physiology assessment: hydrops – a measure of capillary permeability and/or elevated capillary venous pressure and/or hypoproteinemia; venous Doppler – a measure of flow patterns suggestive of abnormal right atrium/venous pressure, and/or RV diastolic function; heart size – a measure of remodeling of the heart in response to increased preload, afterload or anemia; arterial Doppler – a measure of placental resistance and/or combined cardiac stroke volume at falling outputs; and cardiac function – a heterogeneous measure of afterload and/ or annular dilation (tricuspid valve regurgitation), ventricular shortening, extreme diastolic filling abnormality (biphasic versus monophasic filling), and ventricular contractility (dP/dt estimate). In right heart defects, we observed that the LV-dependent circulation should manifest signs of dysfunction that could affect prognosis. Knowing that mitral regurgitation is rare, we searched for markers of LV function that were easy to obtain in the fetus, namely E/A, E’/A’, E’/E, and aortic peak velocity. Not surprisingly, univariate analysis showed that the aortic velocity was important as an incremental variable in prognosis and LV diastolic parameters approached significance. Lower RV pressure was associated with the risk of death. The RV pressure correlated with mortality such that when the RV pressure is higher, the risk of death is lower.
Range of CVP Scores/ change in CVP Score
2/1 12/10 1 1
Ebstein anomaly (n ¼ 4) 25 39 3515 26 37 2685 27 32 1828 28 N/A N/A 7 5–7/decrease 3 6
6 5–7/increase 8 5–8/decrease 7–8/decrease 3–8/decrease 8–9/decrease increase No Yes (8) No No
No Yes (9) No Yes (3) No Yes (4) No
Yes (11) Yes (6) No No Yes (1)
No No Yes (5) No No No No No No
No No
Digoxin/length of treatment (w)
Tricuspid valve plasty No BT shunt No
No PV balloon valvuloplasty PV balloon valvuloplasty No PV balloon valvuloplasty PV balloon valvuloplasty Balloon valvuloplasty
BT shunt, RA and RV outflow tract plasty, tricuspid valve annuloplasty Waterson shunt, RA and RV plasty No No No
BT shunt, Glenn Open pulmonary valvuloplasty BT shunt
Atrial septostomy, BT shunt BT shunt Atrial septostomy, BT shunt No Atrial septostomy PV balloon valvuloplasty BT shunt, Bi-Glenn Heart transplant, bilateral PA repair
Intervention
Survival Pre-intervention Post-intervention Prenatal
Pre-intervention Survival Survival Pre-intervention Survival Survival Survival
Survival Survival Prenatal Prenatal Pre-intervention
Post-intervention Post-intervention Survival Pre-intervention Pre-intervention Survival Survival Post-intervention Pre-intervention Survival Survival Survival
Outcome mortality
Intervention was defined as heart surgery or catheter intervention. Outcome was defined at 30 d of postnatal life, as prenatal mortality, postnatal mortality pre-intervention, postnatal mortality post-intervention and survival. Pt, patient; GA, gestational age at birth or intra-uterine death, in weeks (w); BT shunt, Blalock–Taussig modified shunt; PV, pulmonary valve.
1 6/9 4/13 8/8 4/16 13/12 4/10
(n ¼ 7)
2325 2975 800 3860 2030 3565
Pulmonary stenosis 18 35 19 37 20 38 21 33 22 38 23 34 24 39
6–8/decrease 6–8/decrease increase 3–9/decrease 6 6
No. echocardiograms/ weeks studied
Pulmonary valve atresia with TV/MV41 (n ¼ 5) 13 38 2690 13/14 14 39 3004 11/17 15 N/A N/A 4/6 16 N/A N/A 1 17 36 2200 7/1
Birth weight (g) 6–9/decrease 9 8–9/decrease 8–10/decrease 8–10 6–8/decrease 7–10/increase 10 10 8–9/decrease 6–7/decrease 9–10/decrease
GA (w)
Pulmonary valve atresia with TV/MV51 (n ¼ 12) 1 38 2900 3/6 2 35 2875 1 3 38 2845 3/5 4 40 2590 6/21 5 37 3285 7/19 6 38 2082 5/8 7 37 3035 6/12 8 38 3300 5/14 9 38 3085 4/16 10 38 2000 3/13 11 39 3200 4/9 12 38 2250 4/14
Pt
Table 2. Characteristics of 28 patients by diagnostic subgroups.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
DOI: 10.3109/14767058.2013.878695
CVP score in fetal right heart defects 1435
1436
A. L. Neves et al.
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
Table 3. Cardiovascular Profile Score and subcategories correlated with mortality.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Parameter Original CVP Hydrops Pericardial effusion Skin edema Venous Doppler UV pulsations DV reversals Heart size (CTR) 0.35–0.5 40.5 Cardiac function Holosystolic TR Monophasic filling Arterial Doppler Modified CVP Score Mean CVP Categorical CVP CVP 6 CVP 46 (null)
OR
95% CI
0.9252 1.015 1.984 3.344 1.103 2.345 0.9101 0.770 0.652 1.018 0.9510 1.001 2.264 1.835
0.866–0.988* 0.687–2.435 0.915–3.051 0.079–9.085 0.642–3.888 0.048–5.622 0.756–1.002 0.050–11.796 0.015–23.548 0.289–18.552 0.7987–0.9715* 0.927–1.213 0.983–3.566 0.881–4.532
0.292
0.074–1.160
3.666 –
0.771–17.426 –
CVP, Cardiovascular Profile Score; UV, umbilical vein; DV, ductus venosus; CTR, cardiothoracic ratio; TR, tricuspid regurgitation. *Statistically significant (p50.05).
Preservation of RV systolic pressure generation suggests that myopathic changes in the RV function have not occurred. Considered another way, a low RV pressure leads to atrophy of RV myocardium that is not suited for post-natal pressure work. Cardiac function can be assessed measuring the cardiac output, although this is often technically challenging [16]. Cardiac output is dependent on fetal heart rate and stroke volume, on preload, afterload, and myocardial contractility [16,17]. Peak aortic velocity was found to be a discriminator for mortality in fetal right heart defects. The higher the peak aortic velocity through the aortic valve is, the lower the risk of death is. Our use of the peak aortic velocity through the aortic valve makes physiologic sense and this parameter could be used serially to assess changes in forward flow from the LV. Specifically, a falling peak aortic velocity could be useful in predicting impending fetal acidosis or hydrops. The abnormal RV influences the LV function in systole and diastole. There is compromise of septal motion by the hypertensive or dilated RV and compromise of LV filling in diastole. This so-called cross-talk phenomenon is the most likely explanation of these findings. Later, in fetal congestive heart failure, systemic resistance rises because of redistribution of cardiac output with a concomitant fall in peak aortic velocity. Our study population consists of very complex fetuses, with heart failure and high mortality risk. Patients with Ebstein’s disease have an even higher mortality risk, and that was confirmed in our study; however, even the critical pulmonary stenosis and pulmonary atresia groups had high mortality either because of genetic syndromes or the patients were too critically ill. Limitations to our study include its retrospective design. Another limitation is the small sample size of patients observed in this investigation that could be overcome by a multicenter approach. Nevertheless, the presented results give
important directions for future investigations. For the association between CVP Score and death, all clinical assessments were included in the analysis to increase the statistical power. This could have biased the observed OR. However, we can be sure of the direction of the association between lower CVP Scores and the risk of death. This is an important topic as fetal cardiologists are responsible for not only providing an accurate diagnosis of CHD but also counseling the family appropriately in regard to prognosis and guiding prenatal decision making. The more we know about key echocardiographic parameters which are predictive of outcome the better and more accurate our counseling and planning will be. In conclusion, the original Cardiovascular Profile Score predicted the risk of fetal or neonatal death in right heart defects. The modified score was not validated as a good prediction tool by this study. Fetal RV pressure estimate and peak aortic valve velocity can be used as independent prognostic predictors. Higher peak aortic velocity and peak pressure gradient of the TR jet are associated with lower risk of death. With these findings, we suggest the use of peak aortic velocity and RV pressure estimate combined with the original CVP Score.
Acknowledgements We wish to thank the members of the Perinatal Department of All Children’s Hospital.
Declaration of interest The authors report no conflicts of interest.
References 1. Acharya G, Archer N, Huhta JC. Functional assessment of the evolution of congenital heart disease in utero. Curr Opin Pediatr 2007;19:533–7. 2. Holzgreve W, Curry CJ, Golbus MS, et al. Investigation of nonimmune hydrops fetalis. Am J Obstet Gynecol 1984;150: 805–12. 3. Hofstaetter C, Hansmann M, Eik-Nes SH, et al. A Cardiovascular Profile Score in the surveillance of fetal hydrops. J Matern Fetal Neonatal Med 2006;19:407–13. 4. Makikallio K, Rasanen J, Makikallio T, et al. Human fetal Cardiovascular Profile Score and neonatal outcome in intrauterine growth restriction. Ultrasound Obstet Gynecol 2008;31: 48–54. 5. Wieczorek A, Hernandez-Robles J, Ewing L, et al. Prediction of outcome of fetal congenital heart disease using a Cardiovascular Profile Score. Ultrasound Obstet Gynecol 2008;31:284–8. 6. Patel D, Cuneo B, Viesca R, et al. Digoxin for the treatment of fetal congestive heart failure with sinus rhythm assessed by Cardiovascular Profile Score. J Matern Fetal Neonatal Med 2008; 21:477–82. 7. Huhta JC. Right ventricular function in the human fetus. J Perinat Med 2001;29:381–9. 8. Sabbah HN, Khaja F, Brymer JF, et al. Non-nvasive evaluation of left ventricular performance based on peak aortic blood acceleration measured with a continuous-wave doppler velocity meter. Circulation 1986;74:323–9. 9. Clur SA, Oude Rengerink K, Ottenkamp J, Bilardo CM. Cardiac function in trisomy 21 fetuses. Ultrasound Obstet Gynecol 2011;37: 163–71.
CVP score in fetal right heart defects
DOI: 10.3109/14767058.2013.878695
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
10. Tei C. New non-invasive index for combined systolic and diastolic ventricular function. J Cardiol 1995;26: 135–6. 11. Nagueh SF, Middleton KJ, Kopelen HA, et al. Doppler tissue imaging: a non-invasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 1997;30:1527–33. 12. Ommen SR, Nishimura RA, Appleton CP, et al. Clinical utility of doppler echocardiography and tissue doppler imaging in the estimation of left ventricular filling pressures: a comparative simultaneous doppler-catheterization study. Circulation 2000;102: 1788–94. 13. Iacobelli R, Pasquini L, Toscano A, et al. Role of tricuspid regurgitation in fetal echocardiographic diagnosis of pulmonary
14. 15.
16. 17.
1437
atresia with intact ventricular septum. Ultrasound Obstet Gynecol 2008;32:31–5. Huhta JC. Fetal congestive heart failure. Semin Fetal Neonatal Med 2005;10:542–52. Shah AD, Border WL, Crombleholme TM, Michelfelder EC. Initial fetal Cardiovascular Profile Score predicts recipient twin outcome in twin–twin transfusion syndrome. J Am Soc Echocardiogr 2008; 21:1105–8. Tutschek B, Schmidt KG. Techniques for assessing cardiac output and fetal cardiac function. Semin Fetal Neonatal Med 2011;16:13–21. Van Mieghem T, DeKoninck P, Steenhaut P, Deprest J. Methods for prenatal assessment of fetal cardiac function. Prenat Diagn 2009;29:1193–203.
ORIGINAL ARTICLE
Evaluation of prenatal risk factors for prediction of outcome in right heart lesions: CVP Score in fetal right heart defects
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Ana Luisa Neves1,2, Leigh Mathias3, Marilyn Wilhm4, Jennifer Leshko5, Kersti K. Linask6, Tiago Henriques-Coelho1,2, Jose´ C. Areias1,2, and James C. Huhta7,8 1
Hospital S. Joa˜o, Porto, Portugal, 2Faculty of Medicine, University of Porto, Porto, Portugal, 3College of Public Health, University of South Florida, Tampa, FL, USA, 4All Children’s Hospital, Heart Institute, St. Petersburg, FL, USA, 5All Children’s Hospital, St. Petersburg, FL, USA, 6Children’s Research Institute, University of South Florida, Tampa, FL, USA, 7Department of Perinatal Cardiology, Pediatrix Medical Group, University of Florida, St. Petersburg, FL, USA, and 8All Children’s Hospital Outpatient Care Center, St. Petersburg, FL, USA
Abstract
Keywords
Objective: To determine the prenatal variables predicting the risk of perinatal death in congenital right heart defects. Methods: Retrospective analysis of 28 fetuses with right heart defects was performed. Logistic regression analyses were performed to obtain odds ratios (OR) for the relationship between the risk of death and echocardiographic parameters. The parameters that correlated with the outcome were incorporated in an attempt to devise a disease-specific cardiovascular profile score. Results: Fetal echocardiograms (143) from 28 patients were analyzed. The cardiovascular profile score predicted the risk of death. A lower right ventricle (RV) pressure was associated with mortality (OR 0.959; 95% confidence intervals (CI) 0.940–0.978). Higher peak aortic velocity through the aortic valve (OR 0.104; 95% CI 0.020–0.529) was associated with a better outcome. These cardiac function parameters were incorporated in a modified disease-specific CVP Score. Patients with a mean modified cardiovascular profile score of 6 were over 3.7 times more likely to die than those with scores of 7–10. Conclusions: The original Cardiovascular Profile Score predicted the risk of death in right heart defects. The modified score was not validated as a good prediction tool by this study. Fetal RV pressure estimate and peak aortic velocity can be used as independent prognostic predictors.
Congenital heart defects, fetal echocardiography, outcomes
Introduction Congestive heart failure (CHF) is a clinical syndrome resulting from a variety of disorders that impair the ability of the ventricles to fill with or to eject blood to ensure adequate tissue perfusion to meet metabolic demands of the body [1]. CHF in fetuses with congenital heart diseases is associated with high perinatal mortality [2]. The clinical use of echocardiography for prenatal imaging has the major advantage that it is non-invasive, easily accessible and can be used in pregnancy to evaluate the fetal pathophysiological state. The Cardiovascular Profile Score (CVP), an echocardiography-based multivariate approach, for the assessment of fetal cardiac hemodynamics. It evaluates the degree of hydrops, the venous flow velocity pattern (umbilical vein and ductus venosus), the heart size Address for correspondence: James C. Huhta, Professor of Pediatrics, Department of Perinatal Cardiology, Pediatrix Medical Group, University of Florida, United States of America and All Children’s Hospital Outpatient Care Center, 601 5th Street South, Dept. #7125, St. Petersburg, FL 33701, USA. Tel: +1 727 767 4782. Fax: +1 727 767 8111. E-mail: [email protected]
History Received 16 May 2013 Revised 28 November 2013 Accepted 20 December 2013 Published online 29 January 2014
(cardiac to thoracic area ratio), cardiac function (tricuspid or mitral regurgitation), and arterial flow (umbilical artery). The CVP Score has been shown in our previous studies, which is to be useful in predicting the outcome of fetuses with hydrops [3], growth restriction [4], and congenital heart disease (CHD) [5]. It has been shown that growth-restricted fetuses with adverse neonatal outcome had lower CVP Scores than did fetuses with favorable neonatal outcome. The strongest predictors for adverse neonatal outcome in the CVP Score were cardiomegaly, abnormal cardiac function with monophasic atrioventricular filling or holosystolic tricuspid regurgitation (TR), and abnormal flow patterns suggestive of abnormal right atrium/venous pressure [4]. It has been shown that fetuses with congenital heart defects and a CVP Score below 8 were at risk of perinatal death and that the CVP Score may be used to assess the severity of fetal CHD and to plan perinatal management [5]. It is also useful for appropriate timing and monitoring of the success of intrauterine therapy for CHF [6]. There is significant variability in the hemodynamics in different types of CHD, so a specific CVP Score for different heart defects may be useful.
1432
A. L. Neves et al.
Our hypothesis is that the outcome of fetuses with right heart defects may be related to prenatal condition and left ventricular (LV) function. Accordingly, our aim was to determine the variables predicting the risk of perinatal death in right heart defects in order to achieve a diseasespecific CVP Score that could be used in clinical perinatal medicine.
Methods
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Study design and patient population Medical records and fetal echocardiograms from fetuses with right heart defects, evaluated at the All Children’s Hospital Perinatal Care Center, St. Petersburg, Florida, in five consecutive years, were analyzed retrospectively. The reviewing was blinded to outcomes. This study was approved by the All Children’s Hospital Institutional Review Board and all patients provided informed consent. Right heart defects included pulmonary valve atresia with intact septum (PVAIS), severe pulmonary stenosis (PS), and Ebstein’s anomaly (EbA). Two patients were excluded, one for pregnancy termination and the other for lack of follow-up information. Outcome was defined as either survival or perinatal death (intrauterine death or death within the first month of postnatal life). The degree of fetal CHF was quantified, as previously published [3–5], by the CVP Score, an assessment that evaluates the degree of hydrops, the venous flow velocity pattern (umbilical vein and ductus venosus), the heart size (cardiac to thoracic area ratio), the cardiac function (tricuspid or mitral regurgitation), and the arterial flow (umbilical artery). The presence or absence of hydrops was noted. One point was deducted if ascites, pleural effusion, or pericardial effusion was present. Two points were deducted if skin edema was present. Venous Doppler flow patterns were described as either normal or abnormal. The normal umbilical vein is non-pulsatile. An abnormal flow pattern was umbilical venous pulsations in a free loop of the umbilical cord (two points deducted). In the ductus, venosus Doppler is normally pulsatile, but the abnormal Doppler flow pattern was defined as atrial systolic flow reversal (one point deducted), suggesting either right heart diastolic dysfunction or elevated central venous pressure. The cardiothoracic area ratio (CTR) was obtained as the simple ratio of areas within the ellipses drawn around the pericardium of the heart and around the thorax. This transverse section was taken to include opposing ribs and the spine in the transthoracic four-chamber view. The normal range values are 0.25–0.35. The presence or absence of atrioventricular valve regurgitation was assessed by color flow and pulsed Doppler. Holosystolic TR or depressed shortening fraction demanded a one point deduction while with mitral regurgitation or monophasic filling, two points were deducted About arterial Doppler, the mid-segment of the cord (free loop) was assessed. Umbilical artery absent diastolic flow resulted in a one-point deduction, while reverse diastolic flow deducted two points. A CVP of 10 was indicative of no CHF. Points were deducted for abnormalities of each category, so the lower the CVP Score, the more the severe the CHF [3]. All echocardiographic studies were performed with
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
Table 1. Echocardiographic parameters correlated with mortality. Parameter
OR
Right ventricle hemodynamic parameters RV gradient 0.384 RV pressure 0.959 RV/LV pressure 0.395 Left ventricle hemodynamic parameters LV peak aortic velocity 0.104 LV MPI 0.065 MV E/A 0.775 MV E/E’ 1.021 MV E’/A’ 2.149 Morphometric parameters TV/MV size 0.956 RA/LA width 1.166 RV/LV width 0.369
95% CI
p
0.228–0.646 0.940–0.978 0.237–0.659
0.0003* 50.0001* 0.0004*
0.020–0.529 0.002–1.792 0.366–1.641 0.869–1.200 0.977–4.729
0.0064* 0.1063 0.5047 0.7977 0.5233
0.573–1.596 0.783–1.735 0.190–0.715
0.8638 0.4500 0.0032*
RV, right ventricle; LV, left ventricle; MPI, myocardial performance index; MV, mitral valve; RA, right atrium; LA, left atrium. *Statistically significant (p50.05).
a Siemens Sequoia 512 (Siemens Healthcare, Malvern, PA) with curvilinear transducers. Specific study parameters for right heart defects The estimate of the degree to which the right ventricular pressure was greater than the systemic pressure was obtained by converting the peak velocity of the TR jet to a gradient using the four times velocity squared modified Bernoulli equation [7]. The peak pressure gradient of the TR jet and the right ventricle (RV) estimated pressure (obtained by adding the right atrium (RA) estimated pressure of 5 to the peak pressure gradient of the TR jet) were assessed. The peak aortic velocity through the aortic valve is a sensitive index of global left ventricular performance [8,9]. Peak aortic velocity through the aortic valve was used as a surrogate for cardiac output (stroke volume) of the LV. The myocardial performance index (MPI) or Tei index [10] is the sum of the isovolumic times divided by the ejection time. The isovolumic time is the sum of the two intervals – the isovolumic relaxation time and the isovolumic contraction time. This term was measured by subtracting the ejection time from the time between two inflows in the LV [7]. Peak velocities of the inflow E and A waves at the annulus of the mitral valve (MV E/A) were analyzed, as well as the ratio between maximum inflow blood velocity during early diastole and annular myocardial lengthening velocity (by tissue Doppler) during early ventricular relaxation (E/E’) which reflects ventricular end-diastolic and atrial pressures [11,12]. The use of the ratios obviated the need for agespecific normal data [13]. The cardiac function parameters that correlated with the outcome (Table 1) were incorporated in a specific modified CVP Score (Figure 1), with a one point deducted for TR jet velocity 52 m/s, and two points deducted for LV ejection velocity 51 m/s. This specific Cardiovascular Profile Score was applied and statistical analyses were performed. Statistical analysis Observed clinical data and cardiac function category values for all subjects were analyzed using SAS statistical software,
CVP score in fetal right heart defects
1433
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
DOI: 10.3109/14767058.2013.878695
Figure 1. Modified Cardiovascular Profile Score specific for fetal right heart defects. UV, umbilical vein; DV, ductus venosus; UA, umbilical artery; white arrow, pericardial effusion; TV, tricuspid valve; TR, tricuspid regurgitation.
version 9.1 (SAS, Cary, NC). Chi-squared analysis was used to identify statistically significant differences between survivors and non-survivors with respect to all the clinical cardiac indicators mentioned above. Crude odds ratio (OR) with corresponding 95% confidence intervals (CI) were obtained through separate unconditional logistic regression analyses for the relationship between risk of death and the following parameters: CVP Score, modified CVP score, RV pressure, and parameters of LV performance (peak systolic velocity through the aortic valve, LV myocardial performance index, Doppler E/A, tissue Doppler E’/A’, and E/E’).
Results Fetal echocardiogram exams (143) from 28 patients were analyzed, ranging from 1 to 13 echocardiograms (median of 4) per patient. The mean ± SD gestational age (GA) at the delivery was 37 ± 2 weeks; the mean ± SD birth weight was 2705 ± 677 g; the median Apgar scores at 1 and 5 min ranged between 1–9 (median 8) and 5–9 (median 8), respectively. Seventeen patients had pulmonary atresia with intact ventricular septum (PAIVS) (TV/MV51 in 12,41 in 5), seven PS, and four EbA. From patients with PAIVS, eight had
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
1434
A. L. Neves et al.
amniocentesis, with normal karyotype and were negative for 22q11.2 deletion. Three patients with critical pulmonary stenosis had abnormal karyotypes (one had trisomy 18, one had terminal deletion of chromosome 3, and one had trisomy 21). From the 28 fetuses, 25 were born alive. The total mortality was 14/28 (50%) by 30 d of postnatal life (Table 2). Twelve newborns had surgery. Four died after surgery. Seven died before intervention. From these, three had PAIVS with TV/MV51 (one had Staph aureus endocarditis, two had RVdependent coronary circulation and were listed for heart transplant). One had PAIVS with TV/MV41 and died of arrhythmia. Two patients with critical pulmonary stenosis died before intervention, one had trisomy 18 and the other had terminal deletion of one chromosome 3, were a twin pregnancy with a birth weight of 800 g and did not survive. One patient with Ebstein had sepsis and brain parenchymal hemorrhage and was too critically ill to attempt intervention. The original CVP Score appeared to be a predictor of the risk of death in fetuses with right heart defects. The observed OR of 0.9252 (95% CI 0.866–0.988) indicated that a higher CVP Score was associated with a decreased risk of death. The lower the CVP Score, the greater the risk of death, meaning that for each one-point score reduction of the CVP, there is an additional 8% risk of death. The five CVP categories were analyzed separately and correlated with mortality (Table 3). Pericardial effusion was found in at least one echocardiogram of 16 patients, two of them additionally had skin edema. Of the patients with points deducted in the hydrops category, 50% died. For venous Doppler, 16 had abnormal results, 12 had ductus venosus reversal while three also had umbilical vein pulsations. Mortality associated with umbilical vein pulsations was 67%. The CTR was not correlated with death with an OR of 0.770 (95% CI 0.050–11.796). Holosystolic TR was found in 12 patients and monophasic filling in 8. Mortality occurred in three of the patients with holosystolic TR (25%) and five of the patients with monophasic filling (63%). The velocity of the TR jet varied between 2.66 m/s and 4.35 m/s. Umbilical artery Doppler abnormalities were found rarely (two patients had absent diastolic flow) and no patient had reversed diastolic flow. Observed values of RV pressure correlated with an increased risk of mortality, with corresponding OR of 0.959 (95% CI 0.940–0.978). This finding indicated that an increase in the RV pressure is associated with a decrease in risk of death among subjects (Table 1). Higher observed peak aortic velocity through the aortic valve was found to be associated with decreased risk of death, with an OR of 0.104 (95% CI 0.020–0.529) among subjects. This suggests that as the peak aortic velocity through the aortic valve increases, the risk of death is reduced (Table 1). No statistically significant differences were found between the LV myocardial performance index and survival. While not statistically significant (95% CI 0.977–4.729), the observed association between the ratio of the mitral tissue Doppler E’/ A’ in the LV and risk of death (OR 2.149) suggests that each 1-unit increase in the MV E’/A’ ratio equates to a two-fold increase in the risk of death. No differences were found in the ratio E/E’.
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
The cardiac function parameters that correlated with the outcome (Table 1) were incorporated in a modified diseasespecific CVP adjusted to the right heart defects (Figure 1). The parameters were peak aortic velocity through the aortic valve and peak velocity of the TR jet. In comparison with the original Cardiovascular Profile Score, changes were made at the cardiac function category, with a one-point deducted for TR jet velocity 52 m/s, and two points deducted for peak aortic velocity through the aortic valve 51 m/s (maximum deduction in this category is two points in any category). The specific CVP Score was applied to all the echocardiograms, and a statistical analysis was performed using the mean score for each patient (Table 3). With the mean CVP Score for each patient, we were able to observe an OR of 0.292. The corresponding 95%CI (0.074– 1.160) was not significant. When the analysis was performed with a categorical variable of good (46)/bad (6) modified CVP Score, we also observed some strong results. Those patients with a mean modified CVP Score of 6 or less were over 3.7 times as likely to die in comparison to those patients with mean CVP Scores of seven or more.
Discussion The clinical presentation of CHF in the fetus can be characterized by findings in at least five categories. Abnormalities in the CVP Score may occur prior to the clinical state of fetal hydrops [14]. The final common pathway of fetal death in CHF is poor tissue perfusion and acidosis [14]. The CVP Score has been shown to be useful in predicting the outcome of fetus with hydrops [3], twin–twin transfusion [15], intrauterine growth restriction [4], and CHD [5]. A disease-specific cardiovascular Score could be a way of improving the interpretation of hemodynamic data [1]. The original CVP Score combines five areas of physiology assessment: hydrops – a measure of capillary permeability and/or elevated capillary venous pressure and/or hypoproteinemia; venous Doppler – a measure of flow patterns suggestive of abnormal right atrium/venous pressure, and/or RV diastolic function; heart size – a measure of remodeling of the heart in response to increased preload, afterload or anemia; arterial Doppler – a measure of placental resistance and/or combined cardiac stroke volume at falling outputs; and cardiac function – a heterogeneous measure of afterload and/ or annular dilation (tricuspid valve regurgitation), ventricular shortening, extreme diastolic filling abnormality (biphasic versus monophasic filling), and ventricular contractility (dP/dt estimate). In right heart defects, we observed that the LV-dependent circulation should manifest signs of dysfunction that could affect prognosis. Knowing that mitral regurgitation is rare, we searched for markers of LV function that were easy to obtain in the fetus, namely E/A, E’/A’, E’/E, and aortic peak velocity. Not surprisingly, univariate analysis showed that the aortic velocity was important as an incremental variable in prognosis and LV diastolic parameters approached significance. Lower RV pressure was associated with the risk of death. The RV pressure correlated with mortality such that when the RV pressure is higher, the risk of death is lower.
Range of CVP Scores/ change in CVP Score
2/1 12/10 1 1
Ebstein anomaly (n ¼ 4) 25 39 3515 26 37 2685 27 32 1828 28 N/A N/A 7 5–7/decrease 3 6
6 5–7/increase 8 5–8/decrease 7–8/decrease 3–8/decrease 8–9/decrease increase No Yes (8) No No
No Yes (9) No Yes (3) No Yes (4) No
Yes (11) Yes (6) No No Yes (1)
No No Yes (5) No No No No No No
No No
Digoxin/length of treatment (w)
Tricuspid valve plasty No BT shunt No
No PV balloon valvuloplasty PV balloon valvuloplasty No PV balloon valvuloplasty PV balloon valvuloplasty Balloon valvuloplasty
BT shunt, RA and RV outflow tract plasty, tricuspid valve annuloplasty Waterson shunt, RA and RV plasty No No No
BT shunt, Glenn Open pulmonary valvuloplasty BT shunt
Atrial septostomy, BT shunt BT shunt Atrial septostomy, BT shunt No Atrial septostomy PV balloon valvuloplasty BT shunt, Bi-Glenn Heart transplant, bilateral PA repair
Intervention
Survival Pre-intervention Post-intervention Prenatal
Pre-intervention Survival Survival Pre-intervention Survival Survival Survival
Survival Survival Prenatal Prenatal Pre-intervention
Post-intervention Post-intervention Survival Pre-intervention Pre-intervention Survival Survival Post-intervention Pre-intervention Survival Survival Survival
Outcome mortality
Intervention was defined as heart surgery or catheter intervention. Outcome was defined at 30 d of postnatal life, as prenatal mortality, postnatal mortality pre-intervention, postnatal mortality post-intervention and survival. Pt, patient; GA, gestational age at birth or intra-uterine death, in weeks (w); BT shunt, Blalock–Taussig modified shunt; PV, pulmonary valve.
1 6/9 4/13 8/8 4/16 13/12 4/10
(n ¼ 7)
2325 2975 800 3860 2030 3565
Pulmonary stenosis 18 35 19 37 20 38 21 33 22 38 23 34 24 39
6–8/decrease 6–8/decrease increase 3–9/decrease 6 6
No. echocardiograms/ weeks studied
Pulmonary valve atresia with TV/MV41 (n ¼ 5) 13 38 2690 13/14 14 39 3004 11/17 15 N/A N/A 4/6 16 N/A N/A 1 17 36 2200 7/1
Birth weight (g) 6–9/decrease 9 8–9/decrease 8–10/decrease 8–10 6–8/decrease 7–10/increase 10 10 8–9/decrease 6–7/decrease 9–10/decrease
GA (w)
Pulmonary valve atresia with TV/MV51 (n ¼ 12) 1 38 2900 3/6 2 35 2875 1 3 38 2845 3/5 4 40 2590 6/21 5 37 3285 7/19 6 38 2082 5/8 7 37 3035 6/12 8 38 3300 5/14 9 38 3085 4/16 10 38 2000 3/13 11 39 3200 4/9 12 38 2250 4/14
Pt
Table 2. Characteristics of 28 patients by diagnostic subgroups.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
DOI: 10.3109/14767058.2013.878695
CVP score in fetal right heart defects 1435
1436
A. L. Neves et al.
J Matern Fetal Neonatal Med, 2014; 27(14): 1431–1437
Table 3. Cardiovascular Profile Score and subcategories correlated with mortality.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
Parameter Original CVP Hydrops Pericardial effusion Skin edema Venous Doppler UV pulsations DV reversals Heart size (CTR) 0.35–0.5 40.5 Cardiac function Holosystolic TR Monophasic filling Arterial Doppler Modified CVP Score Mean CVP Categorical CVP CVP 6 CVP 46 (null)
OR
95% CI
0.9252 1.015 1.984 3.344 1.103 2.345 0.9101 0.770 0.652 1.018 0.9510 1.001 2.264 1.835
0.866–0.988* 0.687–2.435 0.915–3.051 0.079–9.085 0.642–3.888 0.048–5.622 0.756–1.002 0.050–11.796 0.015–23.548 0.289–18.552 0.7987–0.9715* 0.927–1.213 0.983–3.566 0.881–4.532
0.292
0.074–1.160
3.666 –
0.771–17.426 –
CVP, Cardiovascular Profile Score; UV, umbilical vein; DV, ductus venosus; CTR, cardiothoracic ratio; TR, tricuspid regurgitation. *Statistically significant (p50.05).
Preservation of RV systolic pressure generation suggests that myopathic changes in the RV function have not occurred. Considered another way, a low RV pressure leads to atrophy of RV myocardium that is not suited for post-natal pressure work. Cardiac function can be assessed measuring the cardiac output, although this is often technically challenging [16]. Cardiac output is dependent on fetal heart rate and stroke volume, on preload, afterload, and myocardial contractility [16,17]. Peak aortic velocity was found to be a discriminator for mortality in fetal right heart defects. The higher the peak aortic velocity through the aortic valve is, the lower the risk of death is. Our use of the peak aortic velocity through the aortic valve makes physiologic sense and this parameter could be used serially to assess changes in forward flow from the LV. Specifically, a falling peak aortic velocity could be useful in predicting impending fetal acidosis or hydrops. The abnormal RV influences the LV function in systole and diastole. There is compromise of septal motion by the hypertensive or dilated RV and compromise of LV filling in diastole. This so-called cross-talk phenomenon is the most likely explanation of these findings. Later, in fetal congestive heart failure, systemic resistance rises because of redistribution of cardiac output with a concomitant fall in peak aortic velocity. Our study population consists of very complex fetuses, with heart failure and high mortality risk. Patients with Ebstein’s disease have an even higher mortality risk, and that was confirmed in our study; however, even the critical pulmonary stenosis and pulmonary atresia groups had high mortality either because of genetic syndromes or the patients were too critically ill. Limitations to our study include its retrospective design. Another limitation is the small sample size of patients observed in this investigation that could be overcome by a multicenter approach. Nevertheless, the presented results give
important directions for future investigations. For the association between CVP Score and death, all clinical assessments were included in the analysis to increase the statistical power. This could have biased the observed OR. However, we can be sure of the direction of the association between lower CVP Scores and the risk of death. This is an important topic as fetal cardiologists are responsible for not only providing an accurate diagnosis of CHD but also counseling the family appropriately in regard to prognosis and guiding prenatal decision making. The more we know about key echocardiographic parameters which are predictive of outcome the better and more accurate our counseling and planning will be. In conclusion, the original Cardiovascular Profile Score predicted the risk of fetal or neonatal death in right heart defects. The modified score was not validated as a good prediction tool by this study. Fetal RV pressure estimate and peak aortic valve velocity can be used as independent prognostic predictors. Higher peak aortic velocity and peak pressure gradient of the TR jet are associated with lower risk of death. With these findings, we suggest the use of peak aortic velocity and RV pressure estimate combined with the original CVP Score.
Acknowledgements We wish to thank the members of the Perinatal Department of All Children’s Hospital.
Declaration of interest The authors report no conflicts of interest.
References 1. Acharya G, Archer N, Huhta JC. Functional assessment of the evolution of congenital heart disease in utero. Curr Opin Pediatr 2007;19:533–7. 2. Holzgreve W, Curry CJ, Golbus MS, et al. Investigation of nonimmune hydrops fetalis. Am J Obstet Gynecol 1984;150: 805–12. 3. Hofstaetter C, Hansmann M, Eik-Nes SH, et al. A Cardiovascular Profile Score in the surveillance of fetal hydrops. J Matern Fetal Neonatal Med 2006;19:407–13. 4. Makikallio K, Rasanen J, Makikallio T, et al. Human fetal Cardiovascular Profile Score and neonatal outcome in intrauterine growth restriction. Ultrasound Obstet Gynecol 2008;31: 48–54. 5. Wieczorek A, Hernandez-Robles J, Ewing L, et al. Prediction of outcome of fetal congenital heart disease using a Cardiovascular Profile Score. Ultrasound Obstet Gynecol 2008;31:284–8. 6. Patel D, Cuneo B, Viesca R, et al. Digoxin for the treatment of fetal congestive heart failure with sinus rhythm assessed by Cardiovascular Profile Score. J Matern Fetal Neonatal Med 2008; 21:477–82. 7. Huhta JC. Right ventricular function in the human fetus. J Perinat Med 2001;29:381–9. 8. Sabbah HN, Khaja F, Brymer JF, et al. Non-nvasive evaluation of left ventricular performance based on peak aortic blood acceleration measured with a continuous-wave doppler velocity meter. Circulation 1986;74:323–9. 9. Clur SA, Oude Rengerink K, Ottenkamp J, Bilardo CM. Cardiac function in trisomy 21 fetuses. Ultrasound Obstet Gynecol 2011;37: 163–71.
CVP score in fetal right heart defects
DOI: 10.3109/14767058.2013.878695
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 10/08/14 For personal use only.
10. Tei C. New non-invasive index for combined systolic and diastolic ventricular function. J Cardiol 1995;26: 135–6. 11. Nagueh SF, Middleton KJ, Kopelen HA, et al. Doppler tissue imaging: a non-invasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 1997;30:1527–33. 12. Ommen SR, Nishimura RA, Appleton CP, et al. Clinical utility of doppler echocardiography and tissue doppler imaging in the estimation of left ventricular filling pressures: a comparative simultaneous doppler-catheterization study. Circulation 2000;102: 1788–94. 13. Iacobelli R, Pasquini L, Toscano A, et al. Role of tricuspid regurgitation in fetal echocardiographic diagnosis of pulmonary
14. 15.
16. 17.
1437
atresia with intact ventricular septum. Ultrasound Obstet Gynecol 2008;32:31–5. Huhta JC. Fetal congestive heart failure. Semin Fetal Neonatal Med 2005;10:542–52. Shah AD, Border WL, Crombleholme TM, Michelfelder EC. Initial fetal Cardiovascular Profile Score predicts recipient twin outcome in twin–twin transfusion syndrome. J Am Soc Echocardiogr 2008; 21:1105–8. Tutschek B, Schmidt KG. Techniques for assessing cardiac output and fetal cardiac function. Semin Fetal Neonatal Med 2011;16:13–21. Van Mieghem T, DeKoninck P, Steenhaut P, Deprest J. Methods for prenatal assessment of fetal cardiac function. Prenat Diagn 2009;29:1193–203.
