Prediction of outcome of fetal congenital heart ... - Wiley Online Library

Ultrasound Obstet Gynecol 2008; 31: 284–288 Published online 5 February 2008 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.5177

Prediction of outcome of fetal congenital heart disease using a cardiovascular profile score A. WIECZOREK*, J. HERNANDEZ-ROBLES*, L. EWING†, J. LESHKO†, S. LUTHER‡ and J. HUHTA* *Department of Pediatrics, University of South Florida College of Medicine, †All Children’s Hospital, St. Petersburg and ‡Department of Health Policy and Management, University of South Florida College of Public Health, Tampa, FL, USA

K E Y W O R D S: cardiomegaly; cardiovascular profile score; congenital heart disease; congestive heart failure; Doppler echocardiography; fetal echocardiography; hydrops fetalis

ABSTRACT

INTRODUCTION

Objectives Congestive heart failure in fetuses with congenital heart defects (CHD) is associated with high perinatal mortality. The clinical condition can be characterized by five ultrasound markers that comprise the 10-point cardiovascular profile (CVP) score. Our aim was to assess the value of the CVP score in evaluating the condition and in maintaining surveillance of fetuses with CHD.

Outcome is poor in a proportion of neonates diagnosed prenatally with congenital heart defects (CHD). Although heart anatomy plays the most important role in fetal and neonatal survival, coexisting congestive heart failure (CHF) has a potential impact on the outcome of CHD, obstetric and perinatal management, and timing of delivery1,2 . CHF is characterized by inadequate cardiac output and compromised tissue perfusion, which leads to a series of complex reflexes and adaptations to improve forward blood flow to vital organs. It is this functional rather than the morphological variable that could be related to perinatal mortality. Fetal CHF may be assessed by the cardiovascular profile (CVP) score1 , a concept similar to the biophysical profile score of Manning et al.4 . It comprises five categories of ultrasound marker in a 10-point scoring system that combines variables known to be associated with perinatal mortality (Table 1): hydrops5 , cardiomegaly6,7 , abnormal myocardial function8,9 , redistribution of cardiac output10 – 12 and abnormal venous Doppler13 – 17 . Any one of these markers may become abnormal prior to the onset of hydrops fetalis and all have been associated with poor fetal outcome. The maximum CVP score is 10 points and the minimum is 0. Deductions from the maximum of 1 or 2 points for each of the five categories are made for abnormal findings. Monitoring the severity of the fetal condition and grading the severity of fetal CHF are vital for determining suitable treatment plans as well as optimal timing and environment for delivery. Awareness of such factors is of great importance to multidisciplinary perinatal care teams, which include perinatologists/obstetricians, pediatric cardiologists, neonatologists, pediatric cardiac surgeons and

Methods We evaluated retrospectively 131 singleton pregnancies with a diagnosis of fetal CHD, which had been assessed by serial echocardiographic examinations, during which the CVP score was obtained. Fetal and neonatal outcomes, including perinatal mortality and Apgar scores, were assessed. Results Fetuses with a final CVP score ≤ 7 were statistically significantly more likely to suffer mortality than were fetuses with a final CVP score ≥ 8 (87.5% vs. 15.2% mortality; P < 0.0001, chi square = 24.5). Significance was maintained after controlling for birth weight, lag time between the final examination and delivery and the dichotomized 5-min Apgar score (odds ratio, 22.3; P = 0.024). For low Apgar score and mortality, the CVP score had low sensitivity (0.25 and 0.27, respectively) but high specificity (0.98 and 0.99, respectively). The presence of hydrops and severe cardiomegaly were statistically significantly associated with mortality (P < 0.05). Conclusions Fetuses with CHD and a CVP score below 8 are at risk of perinatal death. The CVP score may be used to assess the severity of fetal CHD and to plan perinatal management. Copyright  2008 ISUOG. Published by John Wiley & Sons, Ltd.

Correspondence to: Dr J. Huhta, Department of Pediatrics, University of South Florida, 140 7th Ave. S., CRI-4007, St Petersburg, FL 33701, USA (e-mail: [email protected]) Accepted: 24 April 2007

Copyright  2008 ISUOG. Published by John Wiley & Sons, Ltd.

ORIGINAL PAPER

CVP score in CHD

285

Table 1 The cardiovascular profile score Category

2 points

1 point

0 points

Hydrops

None

Skin edema

Heart size (HA/CA) Cardiac function

> 0.2 and ≤ 0.35 Normal MV and TV, biphasic diastolic filling, LV or RV SFs > 0.28 Normal Normal

Ascites or pericardial or pleural effusion 0.35–0.5 Holosystolic TR or LV or RV SFs < 0.28

Arterial umbilical Doppler Venous Doppler

AEDV DV atrial reversal

< 0.2 or > 0.5 Holosystolic MR or TR dP/dt < 400, monophasic diastolic filling REDV UV pulsations

AEDV, absent end-diastolic velocity; DV, ductus venosus; HA/CA, heart to chest area ratio; LV, left ventricle; MR, mitral valve regurgitation; MV, mitral valve; REDV, reversed end-diastolic velocity; RV, right ventricle; SF, ventricular shortening fraction; TR dP/dt, change in pressure over time of TR jet; TR, tricuspid valve regurgitation; TV, tricuspid valve; UV, umbilical vein.

case management nursing personnel. Such preparation also helps parents to cope with the birth of a sick child3 . The aim of this study was to assess the role of the CVP score in evaluating the condition, and in maintaining the surveillance, of fetuses with CHD. We addressed two questions: 1) Is the CVP score associated with the ensuing perinatal outcome? (2) Is there any dominant category in the CVP score that determines the likelihood of the outcome in neonates affected by CHD?

METHODS This was a retrospective study of 131 consecutive pregnancies that were affected by fetal CHD and for which the CVP score was evaluated in the first and last perinatal cardiology examinations. The study was internal review board-approved. Gestational age of fetal echocardiography ranged from 16 + 0 to 40 + 1 weeks and was based on the final obstetric estimated date of delivery as determined by last menstrual period and serial ultrasound biometry (both were compared and no discrepancies were found). Cases were referred for fetal echocardiography for a variety of indications (Table 2) over a 2-year period. Reasons for exclusions were: fetal arrhythmia (supraventricular tachycardia, complete heart block, bradycardia), extracardiac malformations, cardiomyopathy and biventricular hypertrophy due to maternal diabetes, chromosomal abnormalities, as well as termination and multiple pregnancy. All ultrasound examinations were carried out by both a sonographer and a physician using an Acuson Sequoia 512 system (Siemens) equipped with 3.5- and 5.0-MHz transabdominal transducers. Recordings were made on digital media and were analyzed retrospectively. Pulsed, color and continuous wave Doppler velocimetry were used to assess cardiovascular structural anatomy and function. Doppler intensity was maintained at less than 100 mW/cm2 spatial peak temporal average. The highfrequency filter was set at 125 Hz to remove signals from slow-moving tissues during cardiac examination and it was set at 50 Hz for venous and arterial Doppler examinations. The sonographic anatomical findings were segregated into six groups (Table 3): single ventricle (n = 56, 43%),


Table 2 Indications for fetal echocardiographic examination Indication

n

Family history Sibling with CHD Sibling with chromosomal abnormalities Sibling with cardiomyopathy Sibling death at birth Maternal Medications Age Drug abuse Hypertension Diabetes mellitus History of preterm labor Fetal CHD on routine screening CHF Arrhythmia

35 4 1 1 9 7 6 2 2 1 58 3 2*

*Benign premature atrial contractions. CHD, congenital heart defect; CHF, congestive heart failure.

conotruncal malformations (n = 37, 28%), shunts (n = 13, 10%), right and left heart defects (n = 8, 6% and n = 13, 10%, respectively) and complex heart defects (n = 4, 3%). Two fetuses underwent balloon valvuloplasty in utero. Forty-eight neonates had cardiac surgery after birth. All diagnoses were confirmed postnatally and results for the last examination prior to either delivery or fetal demise were compared with the actual outcome of pregnancy. Fetal outcomes were: perinatal mortality, as defined by demise from 22 weeks’ gestation up to 28 days after delivery (‘28-day mortality’)18 , gestational age at time of delivery, birth weight and 1- and 5-min Apgar scores19 . Outcomes were available for all neonates. Ninety-nine of the 131 fetuses were examined longitudinally and the CVP scores from the first and last prenatal echocardiographic examinations were compared. The five categories comprising the CVP score were analyzed.

Statistical analysis To facilitate analysis we dichotomized the CVP scores into those ≤ 7 and those ≥ 8, and 5-min Apgar scores into those

Ultrasound Obstet Gynecol 2008; 31: 284–288.

Wieczorek et al.

286 Table 3 Groups of congenital heart defects Single ventricle (n = 56) Hypoplastic left heart syndrome (28) Hypoplastic right heart syndrome (10) Tricuspid atresia (4) Double inlet left ventricle (4) Double outlet right ventricle, mitral atresia (4) Single inlet left ventricle (2) Aortic atresia (1) Aortic stenosis, pulmonary stenosis (1) Atrioventricular canal defect, pulmonary atresia (2) Conotruncal malformations (n = 37) Tetralogy of Fallot (19) Transposition of great arteries (11) Double outlet right ventricle (7) Shunts (n = 13) Atrioventricular canal defect (5) Ventricular septal defect (5) Secundum atrial septal defect (2) Total anomalous pulmonary venous return (1) Right heart defects (n = 8) Ebstein’s anomaly (3) Pulmonary stenosis (4) Tricuspid valve dysplasia (1) Left heart defects (n = 13) Coarctation of aorta (7) Aortic stenosis (5) Interrupted aortic arch (1) Complex heart defects (n = 4) Right atrial isomerism (4) Numbers of fetuses are given in parentheses.

≤ 6 and those ≥ 7. To determine the CVP score’s ability to predict patient outcome, we calculated its sensitivity and specificity based on its ability to predict both 5-min Apgar score and 28-day mortality. We further explored the relationship between CVP score and outcome through a series of multivariate (logistic) regression models that controlled for the effect of potential covariates. All covariates were included in the initial models, after which non-significant covariates (P > 0.05) were eliminated to obtain the models reported here20 . In the first model, Apgar score acted as the dependent variable and the potential covariates included type of CHD, gestational age, birth weight and lag time between last examination and delivery. This time variable was included in the analysis because follow-up examinations more frequent than 4-weekly were only indicated in CHD fetuses with abnormalities of cardiac function. In the second model, mortality was the dependent variable and Apgar score was included as a potential covariate.

RESULTS The 131 fetuses were diagnosed with CHD in utero at an average gestational age of 26 (range, 16–37) weeks. The mean ± SD maternal age was 29 ± 7 (range, 15–47) years. The mean ± SD gestational age at delivery was 38 ± 2 (range, 32–42) weeks, the mean ± SD birth weight was 2908 ± 650 g and the median Apgar scores at 1 and 5 min


131 fetuses

7/131 (5%) died in utero

124/131 (95%) liveborn

9/131 (7%) died before surgery

48/131 (37%) underwent surgery

10/131 (8%) died during postsurgical period

26/131 (20%) died

67/131 (51%) survived without surgery

38/131 (29%) survived after surgery

105/131 (80%) survived

Figure 1 Outcome of fetuses with congenital heart defects up to 1 month of age. Surgery refers to a cardiac operation in the first month after delivery.

were 8 and 9, respectively. The lag time ranged from 1 to 153 (median, 52) days. Of the 131 fetuses, seven (5%) died in utero and 124 (95%) were live born. A further nine (7%) died before any operation and 10 (8%) died during the postoperative period, giving an overall perinatal mortality of 26/131 (20%) (Figure 1). There were 22 (17%) preterm births at 32–36 weeks of gestation, and five (23%) of these died during the first month of postnatal life. The CVP score data were highly skewed (skewness = −2.17) with a median score of 10 (interquartile range = 1). Therefore, we dichotomized the patients into two groups for analysis: those with CVP scores of ≥ 8 in one group and those with scores of ≤ 7 in the other. The perinatal mortality rate was 16% for affected fetuses with a CVP score of 10 on the last examination before delivery or intrauterine death, it was 12% for those with a score of 9, 17% for those with a score of 8, 67% for those with a score of 7, and 100% for those with a score of ≤ 6. Fetuses with a final CVP score of ≤ 7 were found to be statistically significantly more likely to have died than were patients with a final CVP score of ≥ 8 (mortality 87.5% vs. 15.2%; P < 0.0001, chi square = 24.5). The CVP scores obtained on the first echocardiographic examinations were not statistically significantly associated with mortality. For the sensitivity and specificity calculations, the dichotomized CVP score proved to have a relatively low sensitivity to predict both poor Apgar score (sensitivity = 0.25) and mortality (sensitivity = 0.27). However, the specificity was 0.99 for mortality and 0.98 for 5-min Apgar score. Two regression models were developed to investigate the relationship between dichotomized CVP score and the outcome variables, after controlling for potential confounding variables. In the first model, gestational age and lag time were found to be statistically significantly


CVP score in CHD

287

Table 4 Postnatal follow-up in 131 fetuses with congenital heart defects

Type of defect All congenital heart defects Single ventricle Conotruncal malformations Shunts Right heart defects Left heart defects Complex heart defects

Birth weight (g)

GA (weeks)

1-min Apgar score*

5-min Apgar score*

Mortality† (%)

Last CVP score*

2908 2903 2974 2781 2917 3221 2099

37.5 37.6 37.6 37.4 37.5 37.9 33.2

8 8 7 7 3 8 5

9 9 8 8 8 9 7

20 29 10 9 38 8 25

10 10 10 10 7 10 9

Note that, compared with the other fetuses with congenital heart defects, the right heart defect group had lower 1-min Apgar score and higher mortality. *All scores given as median, rounded to two significant figures. †The differences in mortality in the congenital heart defects groups were not statistically significant. CVP, cardiovascular profile; GA, gestational age at delivery.

associated with dichotomized 5-min Apgar score and were retained in the final model. After controlling for these variables, patients with CVP scores of ≤ 7 were found to be 14 times more likely than were those with a score of ≥ 8 to have poor Apgar scores (odds ratio (OR) = 14.1, P = 0.013). In the second model, birth weight, lag time and the dichotomized 5-min Apgar score were found to be statistically significantly associated with mortality and were retained in the final model. After controlling for these variables, patients with CVP scores of ≤ 7 were more than 22 times more likely to die than were those with a score or ≥ 8 (OR = 22.3, P = 0.024). Table 4 gives the outcome measures of the study according to CHD group (see Table 3). The lowest CVP scores on the first and last examinations and the highest perinatal mortality (38%) were in the group with right heart defects (median final CVP score, 7). In the complex CHD group, which had a median final CVP score of 9, mortality was 25%. Fetuses with single ventricle, shunts, conotruncal malformations and left heart defects had median CVPs of 10 and no change with GA in the longitudinal analysis. In the five CVP categories of ultrasound marker, two findings were statistically significantly associated with increased mortality (P < 0.05): hydrops (OR, 12.4; P = 0.03) and cardiomegaly with a heart to chest area ratio > 0.5, (OR, 11.4; P = 0.04) (Table 5).

DISCUSSION Neonates with CHD diagnosed prenatally and delivered under controlled circumstances in a pediatric cardiology center have a better prognosis then do patients with CHD unsuspected prior to birth21 . Our results showed that a CVP score of ≥ 8 was associated with a good perinatal outcome. Later time in gestation enhanced the predictive accuracy of the CVP score. The high positive predictive value and specificity found in this study could facilitate selection of patients at high risk who need increased surveillance. The majority of fetuses with CHD did not develop CHF in utero. Nevertheless, in a considerable proportion of cases (47/131; 36%), some signs of heart failure appeared and were diagnosed prenatally by fetal echocardiography.


Table 5 Mortality for different cardiovascular profile (CVP) score categories No. of cases CVP category Hydrops Arterial Doppler Heart size (HA/CA)* < 0.20 0.35–0.50 > 0.50 Venous Doppler Cardiac function

Mortality

Total

Alive

Dead

(%)

6 3 22 0 15 7 6 19

1 1 15 0 12 2 4 14

5 2 7 0 3 5 2 5

83† 67 32 0‡ 20 71† 33 26

*HA/CA ratio for a normal heart is 0.20–0.35. †Based on logistic regression on all CVP score categories simultaneously, only the presence of hydrops (odds ratio (OR), 12.4, P = 0.03) and HA/CA > 0.50 (OR, 11.4, P = 0.04) were found to be statistically significantly associated with mortality (P < 0.05). ‡No patients had a very small heart size. HA/CA, heart to chest area ratio.

Putting these signs into perspective regarding the impact on outcome was the goal of this study. It is clear that one or two abnormal findings yielding a score of 8 or more out of 10 did not significantly affect outcome. In our study there was a lack of fetuses with an extremely low CVP score. During a 5-year period at the same center, there was only one patient with a CVP score of < 5, and the pregnancy ended in intrauterine death. Therefore, our analysis was in a relatively fit population. The CHD subgroup with the highest perinatal mortality was hydrops fetalis (83% mortality in patients with this finding). The outcome of fetuses with hydrops without CHD is poor, with mortality varying between 30% and 80%, and this is predicted by the CVP score22 . Holosystolic regurgitation is always a pathological finding in CHF and may be the first sign of its development, although trivial tricuspid regurgitation (non-holosystolic and with maximum velocity < 2 m/s) may be found in 7% of normal fetuses23 . A four-degree grading scale for the severity of tricuspid regurgitation (trace, mild, moderate or severe) may be used by the perinatal cardiologist in addition to the CVP score. Since regurgitation may or may not progress in gestation, serial evaluation is indicated. Tricuspid valve regurgitation


288

would cause a change in the systolic wave of the ductus venosus blood velocity waveform but does not cause A-wave reversal. The time of appearance of tricuspid regurgiation may give information, depending on whether it occurs as a manifestation of increased afterload or as a sign of myocardial dysfunction. Valvular regurgitation associated with a congenitally abnormal atrioventricular valve is included in the CVP score because it presents a functional manifestation of the CHD. For example, mitral regurgitation may occur in fetal aortic stenosis. In the future, the CVP score may need to be weighted for specific cardiac conditions. Umbilical artery Doppler sonography changes occurred rarely in the fetuses with isolated CHD. Increased afterload is the most likely cause of decreased cardiac output late in fetal morbidity. In the study of Meise et al.12 , arterial Doppler was abnormal due to diminished cardiac output. Cases with significant extracardiac malformations were excluded from our analysis. The prognostic importance of venous Doppler abnormalities has been confirmed in several studies1,14 – 17 . Baez et al.24 found an important link between venous Doppler changes in the ductus venosus and hydrops in fetuses with CHD. In our study only six fetuses had venous Doppler abnormalities (only one had pulsations in the umbilical vein) and there was a 33% mortality rate in this subgroup. In conclusion, the CVP score is useful in the assessment of the fetus and may aid in predicting perinatal outcome when there is CHD. A CVP score of ≤ 7/10 may warrant closer fetal surveillance. More intensive cardiology and obstetric evaluation may be indicated in such cases to determine whether CHF is associated with CHD. In terms of timing, the last CVP score before delivery has the best predictive value. A decrease in CVP score with the development of hydrops, severe cardiomegaly or arterial Doppler abnormalities suggests impending fetal compromise in the fetus with CHD.

REFERENCES 1. Falkensammer CB, Paul J, Huhta JC. Fetal congestive heart failure: correlation of Tei-index and Cardiovascular-score. J Perinat Med 2001; 29: 390–398. 2. Hecher K, Campbell S, Doyle P, Harrington K, Nicolaides K. Assessment of fetal compromise by Doppler ultrasound investigation of the fetal circulation. Arterial, intracardiac, and venous blood flow velocity studies. Circulation 1995; 91: 129–138. 3. Sklansky M, Tang A, Levy D, Grossfeld P, Kashani I, Shaughnessy R, Rothman A. Maternal psychological impact of fetal echocardiography. J Am Soc Echocardiogr 2002; 15: 159–166. 4. Manning FA, Harman CR, Morrison I, Menticoglou SM, Lange IR, Johnson JM. Fetal assessment based on fetal biophysical profile scoring. IV. An analysis of perinatal morbidity and mortality. Am J Obstet Gynecol 1990; 162: 703–709. 5. Tulzer G, Gudmundsson S, Wood DC, Cohen AW, Weiner S, Huhta JC. Doppler in non-immune hydrops fetalis. Ultrasound Obstet Gynecol 1994; 4: 279–283.


Wieczorek et al. 6. Respondek M, Respondek A, Huhta JC, Wilczynski J. 2D echocardiographic assessment of the fetal heart size in the 2nd and 3rd trimester of uncomplicated pregnancy. Eur J Obstet Gynecol Reprod Biol 1992; 44: 185–188. 7. Paladini D, Chita SK, Allan LD. Prenatal measurement of cardiothoracic ratio in evaluation of heart disease. Arch Dis Child 1990; 65: 20–23. 8. Tulzer G, Gudmundsson S, Huhta JC, Wood DC, Tews G. The value of Doppler in evaluation and prognosis of fetuses with non-immunologic hydrops fetalis. Gynakol Rundsch 1991; 31: 152–153. 9. Rizzo G, Arduini D, Romanini C. Doppler echocardiographic assessment of fetal cardiac function. Ultrasound Obstet Gynecol 1992; 2: 434–445. 10. Rasanen J, Debbs RH, Huhta JC. Echocardiography in intrauterine growth restriction. Clin Obstet Gynecol 1997; 40: 796–803. 11. Dubiel M, Breborowicz GH, Gudmundsson S. Evaluation of fetal circulation redistribution in pregnancies with absent or reversed diastolic flow in the umbilical artery. Early Hum Dev 2003; 71: 149–156. 12. Meise C, Germer U, Gembruch U. Arterial Doppler ultrasound in 115 second- and third-trimester fetuses with congenital heart disease. Ultrasound Obstet Gynecol 2001; 17: 398–402. 13. Kiserud T, Eik-Nes SH, Blaas H-GK, Hellevik LR. Ultrasonographic velocimetry of the fetal ductus venosus. Lancet 1991; 388: 1412–1414. 14. Gudmundsson S, Huhta JC, Wood DC, Tulzer G, Cohen AW, Weiner S. Venous Doppler ultrasonography in the fetus with nonimmune hydrops. Am J Obstet Gynecol 1991; 164: 33–37. 15. Kiserud T, Eik-Nes SH, Hellevik RL, Blaas H-G. Ductus venosus blood velocity changes in fetal cardiac diseases. J Matern Fetal Invest 1993; 3: 15–20. 16. Hecher K, Snijders R, Campbell S, Nicolaides K. Fetal venous, intracardiac, and arterial blood flow measurements in intrauterine growth retardation: relationship with fetal blood gases. Am J Obstet Gynecol 1995; 173: 10–15. 17. Gudmundsson S, Tulzer G, Huhta JC, Marsal K. Venous Doppler in the fetus with absent end-diastolic flow in the umbilical artery. Ultrasound Obstet Gynecol 1996; 7: 262–267. 18. World Health Organization. Neonatal and Perinatal Mortality: Country, Regional and Global Estimates. WHO, Department of Making Pregnancy Safer: Geneva, 2006. 19. Thorngren-Jerneck K, Herbst A. Low 5-minute Apgar score: a population-based register study of 1 million term births. Obstet Gynecol 2001; 98: 65–70. 20. Hosmer DW, Lemeshow S. Applied Logistic Regression (2nd edn). John Wiley and Sons, Inc.: New York, 2000; 150–156. 21. Hsu DT, Gersony WM. Medical management of the neonate with congenital heart disease. In Intensive Care of the Fetus and Neonate, Spitzer AR (ed). Mosby-Year Book: St. Louis, 1996; 787–796. 22. Hofstaetter C, Hansmann M, Eik-Nes SH, Huhta JC, Luther SL. A cardiovascular profile score in the surveillance of fetal hydrops. J Matern Fetal Neonatal Med 2006; 19: 407–413. 23. Respondek ML, Kammermeier M, Ludomirsky A, Weil SR, Huhta JC. The prevalence and clinical significance of fetal tricuspid valve regurgitation with normal heart anatomy. Am J Obstet Gynecol 1994; 171: 1265–1270. 24. Baez E, Steinhard J, Huber A, Vetter M, Hackeloer BJ, Hecher K. Ductus venosus blood flow velocity waveforms as a predictor for fetal outcome in isolated congenital heart disease. Fetal Diagn Ther 2005; 20: 383–389.
