The treatment of fetal anemia secondary to maternal red cell isoimmunization has been one of the most successful fetal therapies to date. First described in 1963 by Sir William Liley,1 intrauterine fetal transfusion has dramatically improved the outcome for severe isoimmunization with fetal survival rates now in excess of 90%. During the past five decades, there have been improvements in the management of red cell isoimmunization, including the use of prophylactic anti-D, the universal use of intravascular rather than intraperitoneal blood transfusion, and the introduction of ultrasonography to noninvasively recognize fetal anemia. Although these medical interventions have been highly successful in decreasing fetal loss, there have been only a few reports assessing the long-term outcomes of surviving children after therapies for severe isoimmunization.2–6 The available long-term data have centered principally on neurodevelopmental outcomes and have consistently demonstrated no excess of neurologic handicap rates when adjusted for gestational age at birth. However, there is a paucity of long-term outcome data on other organ systems. Since the pioneering epidemiologic work of David Barker in 19897 it has been recognized that the intrauterine environment defines the lifelong health of the human heart. It has been hypothesized that chronic hypoxia may predispose a fetus to lifelong cardiac disease.8 Although there are human data to assess the impact of anemia on the fetal cardiac status,9,10 there are no long-term studies to evaluate the impact of these profound changes on survivors.
The aim of this study is to estimate by echocardiography the cardiac structure and function of children who received intrauterine blood transfusions as therapy for red cell isoimmunization to investigate the potential effects of anemia and hypoxia on the developing fetal heart.
MATERIALS AND METHODS
All surviving children from pregnancies of women with severe red cell isoimmunization who received at least one intravascular blood transfusion during their pregnancy between 1992 and 2003 at King Edward Memorial Hospital were identified from the institutional fetal medicine service database. Children were required to be at least 4 years of age for study eligibility to ensure cooperation during echocardiography. Age- (±1.9 years) and sex-matched children were chosen for the control group based on “best friend” selection. Each child in the case group was asked to invite a best friend of similar age and sex to participate in the study to assist in controlling for similar socioeconomic group and background. The study protocol was approved prospectively by the institutional ethics committees of King Edward Memorial Hospital and Princess Margaret Hospital for Children.
The maternal and neonatal medical record charts of the recruited children were reviewed and data on the perinatal course obtained. The study design consisted of two parts: a clinical assessment and an echocardiography examination. The clinical evaluation consisted of a standard history and physical examination by a pediatric cardiologist (L.D.). Height, weight, and blood pressure were measured in a standardized manner. A transthoracic echocardiography examination was performed to evaluate cardiac structure and function with an M-mode and two dimensions of the heart. Three-dimensional full-volume acquisition was attempted on all children, but owing to technical difficulties with participant cooperation for adequate imaging and accurate analysis, especially for left ventricular mass, these data were not used. Echocardiographic standards as described by the American Society of Echocardiography guidelines, the American Society of Echocardiography Guidelines for Clinical Trials Task Force for Echocardiography,11–13 were abided by. All studies were performed on iE33 ultrasound machines (Philips Medical Systems, Bothell, WA) in the Children's Cardiac Centre at Princess Margaret Hospital using high-resolution multifrequency transducers S8-3 (8-3 MHz) and S5-1 (5-1 MHz) by one senior cardiac ultrasonographer (J.S.). Data analysis was performed on cart after the acquisition of all images. Children in the case and control groups presented in a random order, and the cardiac ultrasonographer was blinded to participant identification at the time of image acquisition and data analysis. Images were downloaded to the Children's Cardiac Centre digital archiving unit at Princess Margaret Hospital. A pediatric cardiologist (L.D.) reviewed and reported all studies and data analysis and was blinded to participant identification. All measurements were recorded at least five times each, with the best five consecutive measurements averaged and selected by a single observer (J.S.).
The following dimensions, time intervals, and velocities were measured: 1) fractional shortening was calculated using M-mode imaging as an indicator of systolic function; 2) left ventricular ejection fraction Simpson's biplane method was calculated using two-dimensional apical four- and two- chamber views at end diastole and systole as an indicator of systolic function; 3) M-mode or two-dimensional measurements, whichever was deemed most accurate: right ventricular end diastolic dimensions, interventricular septum at end diastole, left ventricular internal diameter at end diastole, left ventricular posterior wall at end diastole, and left ventricular internal diameter at end systole from parasternal long or short axis views; 4) left ventricular mass was calculated using M-mode according to the recommendations of the American Society of Echocardiography: LV mass=(1.04)(0.8) [(LVIDd+PWT+IVS)3−LVIDd3]+0.6; 5) left atrial area from the apical four-chamber view using two dimensions; 6) the myocardial performance index using tissue Doppler time intervals and the formula (a-b)/b was calculated to assess global left ventricular function; 7) mitral annular planar systolic excursion and tricuspid annular planar systolic excursion using M-mode as indicators of systolic function; 8) isovolumic relaxation time from the apical four-chamber view using Doppler tissue imaging at the mitral annulus; 9) pulsed-wave Doppler: mitral valve peak E, mitral valve peak A, deceleration time, and mitral valve atrial duration from the apical four-chamber view; 10) pulsed-wave Doppler: pulmonary venous S velocity, pulmonary venous D velocity, pulmonary venous atrial reversal velocity, and pulmonary venous atrial reversal duration from the apical four-chamber view and sample volume in the right upper pulmonary vein; 11) the A delta score was calculated as the difference between the duration of the transmitral and pulmonary atrial waves (ie, delta=MVAdur−PVARdur); 12) Doppler tissue imaging both septal (medial) and lateral annulus: mitral annulus medial S velocity, mitral annulus medial E velocity Doppler tissue imaging), and mitral annulus medial A velocity from the apical four-chamber view, with the cursor placed in the myocardium immediately below the septal annulus as well as mitral lateral S′, lateral E′, and lateral A′. Diastolic function was evaluated using mitral valve peak E, mitral valve peak A, mitral valve E velocity/mitral valve A velocity ratio, deceleration time, mitral valve atrial duration, isovolumic relaxation time pulmonary venous S velocity, pulmonary venous D velocity, pulmonary venous atrial reversal velocity, pulmonary venous atrial reversal duration, Delta, ratio, E′, A′, and E/E′ ratio. When mitral valve E–mitral valve A fusion occurred, deceleration time and mitral valve atrial duration were not measured; and 13) coronary artery dimension by two dimensions: left main coronary artery, left anterior descending coronary artery, right coronary artery, and circumflex coronary artery.
To detect a difference of 0.6 between the group means of standardized continuous outcomes with 80% power at a 5% significance level, a sample size of 25 case–control pairs was required.
Clinical and echocardiography measurements were summarized using medians and ranges (R) for continuous data and frequency distributions for categorical data. Categorical comparisons between children in the case and control groups were made using McNemar's test and comparisons of echocardiography outcomes were made using the Wilcoxon signed-rank test to account for the pair-matching of the participants. Echocardiography outcomes with univariable P<.15 were selected for further assessment. Multivariable comparisons of selected echocardiography outcomes between children in the case and control groups were conducted using linear mixed models with matched pairs modeled as random effects. Standard residual diagnostic techniques for linear models were used to assess normality. Data transformations to the natural logarithm were performed when necessary. All models included an adjustment for body surface area (BSA) and severity of red cell isoimmunization. Body surface area was calculated using the formula: BSA (m2)=0.024265×height(cm)0.3964× weight(kg)0.5378.14 Severity of disease was a composite variable based on the following characteristics: initial hemoglobin less than 50 g/L (1 point), presence of fetal hydrops at initial transfusion (1 point), one to three total transfusions (1 point) or four or more transfusions (2 points), more than 28 weeks of gestation at initial transfusion (1 point) or less than 28 weeks (2 points), and mean hemoglobin before transfusions less than 50 g/L (1 point). The points were summed and categorized into two groups: none to mild (0–3 points) and severe (4 points or more). All hypothesis tests were two-tailed and P<.05 was considered statistically significant. SPSS 15.0 statistical software was used for data analyses.
During the time period of the study, intravascular transfusions were performed on 40 fetuses in pregnancies complicated by red cell isoimmunization. We were able to contact and recruit 25 of 38 survivors for the study and these children form the study population. The most frequent red cell antibody was anti-D (23 of 25 children in the case group) either in isolation (nine of 23) or in association with other red cell antibodies (anti-c in 14 children in the case group, anti-E in five, and anti-Kidd in two). Anti-Kell isoimmunization complicated two children in the case group (both in combination with other red cell antibodies). Four women (16%) had received intravascular transfusions in prior pregnancies. The perinatal data for these 25 pregnancies are presented in Table 1. The median hemoglobin at delivery was 129 g/L (range=63–171 g/L) with a median cord bilirubin of 85 mmol/L (range=52–134 mmol/L) and median maximum bilirubin 202 mmol/L (range=62–345 mmol/L). Four neonates (16%) received an exchange transfusion. The median duration of hospital stay was 9 days (range=3–170 days).
The baseline clinical data for the children in the case and control groups are presented in Table 2. There were no significant differences between children in the case and control groups for any cardiovascular parameters, although systolic blood pressure tended to be lower in children in the case group than in those in the control group. Children in the case group had on average 5.1 mm Hg lower systolic blood pressure (95% confidence interval [CI] 0.12–10.4 mm Hg, P=.055) than did those in the control group.
The echocardiographic data are shown in Tables 3 and 4. Children in the case group had on average an 11-ms lower mitral valve duration than did those in the control group (95% CI 3–19 ms less, P=.009; Table 5). In addition, the left atrial area was 9% less in children in the case group (95% CI 2–16% less, P=.020) and the left ventricular mass was 10% less (95% CI 1–19% less, P=.039) than in those in the control group. The left ventricular end diastolic dimension was on average 0.21 cm less in children in the case group (95% CI 0–0.41 cm less, P=.053).
Severity of disease (score 4+ compared with score less than 4) was not associated with left ventricular mass either in a model including group (children in the case group compared with those in the control group) and adjusted for BSA (P=.787) or in a model with group excluded and adjusted for BSA (P=.293) (Table 5; Fig. 1). Fetal hydrops at initial transfusion was not associated with left ventricular mass either in a model including group (children in the case group compared with those in the control group) and adjusted for BSA (P=.990) or in a model with group excluded and adjusted for BSA (P=.321) (Fig. 2).
Although there are human data to assess the effects of anemia on fetal cardiac status, there are no long-term studies to evaluate the impact of these profound changes on survivors. This is one of the first studies to investigate the cardiac function of children from pregnancies complicated by severe red cell isoimmunization necessitating intrauterine transfusion based on the conduct of an electronic Medline search of the English language literature published between 1966 and 2010 using the key words cardiac, fetus, long-term outcome, red cell isoimmunization, Rhesus, and echocardiography. Using standardized and validated echocardiographic techniques of cardiac dimensions and function, we observed a 10% reduction in left ventricular mass in the children who had severe red cell isoimmunization compared with age- and sex-matched children in the control group, although the resting function of the left ventricle was well preserved. Although there may be limitations in using formulas to calculate left ventricular mass from M-mode measurements, with small differences in measurements either increased or decreased threefold, the same method was used for children in both the control and case groups.
The fetal response to chronic anemia is manifest by a marked increase in cardiac output, an increase in heart rate, and an increase in biventricular stroke volume.15 Oberhoffer16 assessed the cardiac changes in 30 anemic human fetuses before intrauterine transfusion (mean hemoglobin 69 g/L). Before transfusion, fetal echocardiography demonstrated myocardial hypertrophy of all ventricular walls. Left ventricular mean velocities were increased, presumably secondary to the augmented cardiac workload. Baschat17 observed coronary artery blood flow velocities to be significantly elevated in anemic fetuses. Increases in myocardial oxygen demand occur with anemia and this appears to be met by augmentation of coronary blood flow. Thus, the fetal response to chronic anemia is manifest as cardiac enlargement and expansion of the coronary vascular tree.
Cardiac myocyte hyperplasia and hypertrophy are responsible for growth of the fetal heart.18 Soon after birth, the myocytes lose their ability to divide and postnatal cardiac growth is secondary to myocyte hypertrophy and nonmuscle cell hyperplasia, resulting in a 30- to 40-fold increase in volume of the individual cardiomyocytes.19 This transition from proliferative to hypertrophic growth is secondary to terminal differentiation (karyokinesis) and it is hypothesized that with increasing maturity, the fetus has a declining proportion of myocytes capable of proliferation.18 In a fetal sheep model, Jonker et al20 observed that chronic anemia was associated with larger cardiomyocytes, increased terminal differentiation, and accelerated cardiomyocyte proliferation. This group speculated that this anemia-associated accelerated cardiomyocyte growth and maturation may impact on cardiomyocyte numbers in adulthood.
The observation in this study of a reduction in left ventricular mass may reflect the long-term impact severe anemia on the fetal heart. We speculate that the acceleration of myocyte proliferation and premature terminal differentiation secondary to anemia may be associated with a reduction in myocyte number. This may result in reduction in myocardial mass in later life. The significance of this observation however is uncertain and requires further investigation. The lack of association between left ventricular mass and severity of disease and hydrops illustrated in Figures 1 and 2, respectively, may be the result of limited statistical power and also requires further investigation with larger sample sizes.
This study also demonstrated a significant difference between left atrial area when indexed for BSA with children in the case group having smaller dimensions compared with those in the control group (P=.02). The size of the left atrium is determined by body size, filling from pulmonary venous return, the status of the mitral valve, and the diastolic function of the left ventricle. Because children in both the case and control groups had normal pulmonary venous return, normal mitral valve, and normal left ventricular diastolic function by our assessment, the difference in the indexed left atrial size may therefore be related to the overall size of the left heart and correlate with the smaller left ventricular mass, which has been documented in the children in our case group. This may be further evidence for diminished myocardial size in our participants. However, this is currently speculative and requires validation by further studies.
Finally, this study demonstrated significantly shorter mitral valve atrial duration in children in the case group compared with those in the control group. The significance of this is unknown because mitral valve atrial duration is not interpreted as an isolated measurement but is used for the assessment of left ventricular diastolic function as a ratio or difference (A delta) to pulmonary venous atrial reversal duration.21 There may be a correlation with mitral valve atrial duration and left atrial size, but this is unproven and awaits further study.
The strength of this study is the long-term follow up; we are not aware of other data on childhood cardiac function after severe fetal anemia. In Western Australia, there is only one tertiary obstetric unit, and it is the sole center for fetal transfusion with only three operators performing these procedures throughout the past two decades. Intraperitoneal transfusions were not used routinely with all procedures performed primarily through the vascular route. The cardiac assessments were performed by a single experienced operator blinded to the child's history.
A limitation of our study is the small sample size, because we were restricted by the relative rarity of severe red cell isoimmunization in our population and the ability to contact surviving children many years after they were born. In addition, there have been alterations in the perinatal management of red cell isoimmunization over the past 20 years, which may have indirectly influenced outcomes. All studies were performed at rest and to evaluate early signs of impaired cardiac function evaluation with exercise would be of interest. This is of particular significance given the expansion of the coronary vascular tree that occurs after fetal anemia.22
In conclusion, our study has demonstrated that surviving children who have received intrauterine blood transfusions for severe fetal anemia resulting from red cell isoimmunization have significantly smaller left ventricular mass and, therefore, myocardial mass. The presence of smaller left atrial size and shorter mitral valve atrial duration may be further evidence of smaller left heart size and myocardium, but this remains unproven. However, there is the implication of reduced myocardial mass from the alteration of growth and maturation of cardiomyocytes in the fetus and this may predispose to cardiac disease later in life.23 Reduced cardiomyocyte numbers may mean fewer coronary capillaries, which could predispose the individual to earlier ischemic heart disease or more vulnerability to left ventricular stress as adults. Therefore, ongoing follow-up and study of this cohort is needed.
1.Van Kamp IL, Klumper FJ, Oepkes D, Meerman RH, Scherjon SA, Vandenbussche FP, et al. Complications of intrauterine intravascular transfusion for fetal anemia due to maternal red-cell isoimmunization. Am J Obstet Gynecol 2005;192:171–7.
2.Doyle LW, Kelly EA, Rickards AL, Ford GW, Callanan C. Sensorineural outcome at 2 years for survivors of erythroblastosis treated with fetal intravascular transfusions. Obstet Gynecol 1993;81:931–5.
3.Janssens HM, de Haan MJ, Van Kamp IL, Brand R, Kanhai HH, Veen S. Outcome for children treated with fetal intravascular transfusions because of severe blood group antagonism. J Pediatr 1997;131:373–80.
4.Hudon L, Moise KJ Jr, Hegemier SE, Hill RM, Moise AA, Smith EO, et al. Long-term neurodevelopmental outcome after intrauterine transfusion for the treatment of fetal hemolytic disease. Am J Obstet Gynecol 1998;179:858–63.
5.Grab D, Paulus WE, Bommer A, Buck G, Terinde R. Treatment of fetal erythroblastosis by intravascular transfusions: outcome at 6 years. Obstet Gynecol 1999;93:165–8.
6.Harper DC, Swingle HM, Weiner CP, Bonthius DJ, Aylward GP, Widness JA. Long-term neurodevelopmental outcome and brain volume after treatment for hydrops fetalis by in utero intravascular transfusion. Am J Obstet Gynecol 2006;195:192–200.
7.Barker DJ, Osmond C, Golding J, Kuh D, Wadsworth ME. Growth in utero, blood pressure in childhood and adult life, and mortality from cardiovascular disease. BMJ 1989;298:564–7.
8.Davis L, Thornburg KL, Giraud GD. The effects of anaemia as a programming agent in the fetal heart. J Physiol 2005;565:35–41.
9.Bigras JL, Suda K, Dahdah NS, Fouron JC. Cardiovascular evaluation of fetal anemia due to alloimmunization. Fetal Diagn Ther 2008;24:197–202.
10.Oberhoffer R, Grab D, Keckstein J, Högel J, Terinde R, Lang D. Cardiac changes in fetuses secondary to immune hemolytic anemia and their relation to hemoglobin and catecholamine concentrations in fetal blood. Ultrasound Obstet Gynecol 1999;13:396–400.
11.Gottdiener JS, Bednarz J, Devereux R, Gardin J, Klein A, Manning WJ, et al; American Society of Echocardiography. American Society of Echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr 2004;17:1086–119
12.Quiñones MA, Otto CM, Stoddard M, Waggoner A, Zoghbi WA; Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. Recommendations for quantification of Doppler echocardiography: a report from the Doppler Quantification Task Force of the Nomenclature and Standards Committee of the American Society of Echocardiography. J Am Soc Echocardiogr 2002;15:167–84.
13.Anderson B. Echocardiography: the normal examination and echocardiographic measurements. 2nd ed. Brisbane (Australia): MGA Graphics; 2007.
14.Haycock GB, Schwartz GJ, Wisotsky DH. Geometric method for measuring body surface area: a height–weight formula validated in infants, children and adults. J Pediatr 1978;93:62–6.
15.Davis LE, Hohimer AR. Hemodynamics and organ blood flow in fetal sheep subjected to chronic anemia. Am J Physiol 1991;261:R1542–8.
16.Oberhoffer R, Grab D, Keckstein J, Hogel J, Terinde R, Lang D. Cardiac changes in fetuses secondary to immune hemolytic anemia and their relation to hemoglobin and catecholamine concentrations in fetal blood. Ultrasound Obstet Gynecol 1999;13:396–400.
17.Baschat AA, Muench MV, Gembruch U. Coronary artery blood flow velocities in various fetal conditions. Ultrasound Obstet Gynecol 2003;21:426–9.
18.Jonker SS, Zhang L, Louey S, Giraud GD, Thornburg KL, Faber JJ. Myocyte enlargement, differentiation, and proliferation kinetics in the fetal sheep heart. J Appl Physiol 2007;102:1130–42.
19.Oparil S, Bishop SP, Clubb FJ Jr. Myocardial cell hypertrophy or hyperplasia. Hypertension 1984;6:1138–43.
20.Jonker SS, Giraud MK, Giraud GD, Chattergoon NN, Louey S, Davis LE, et al. Cardiomyocyte enlargement, proliferation and maturation during chromic fetal anaemia in sheep. Exp Physiol 2010;95:131–39.
21.O'Leary PW, Durongpisitkul K, Cordes TM, Bailey KR, Hagler DJ, Tajik J, et al. Diastolic ventricular function in children: a Doppler echocardiographic study establishing normal values and predictors of increased ventricular end-diastolic pressure. Mayo Clin Proc 1998;73:616–28.
22.Davis L, Roullet JB, Thornburg KL, Shokry M, Hohimer AR, Giraud GD. Augmentation of coronary conductance in adult sheep made anaemic during fetal life. J Physiol 2003;547:53–9.
© 2010 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
23.Thornburg KL, Louey S. Fetal roots of cardiac disease. Heart 2005;91:867–8.