- Email: [email protected]
Long-term neurodevelopmental outcome after intrauterine transfusion for the treatment of fetal hemolytic disease
Long-term neurodevelopmental outcome after intrauterine transfusion for the treatment of fetal hemolytic disease Lynda Hudon, MD,a Kenneth J. Moise, Jr, MD,a Suzanne E. Hegemier, RN,b Reba M. Hill, MD,b† Alicia A. Moise, MD,b E. O’Brian Smith, PhD,b and Robert J. Carpenter, MDa,c Houston, Texas OBJECTIVE: The aim of the study was to assess the developmental outcome of neonatal survivors of hemolytic disease of the neonate treated with modern intrauterine transfusion techniques. STUDY DESIGN: In this prospective, observational study, auditory evoked-response tests were performed in the nursery. Neurodevelopmental evaluation with the Gesell Developmental Schedules was performed between 9 and 18 months of corrected age to assess motor skills, language development, comprehension capacity, and social skills. The McCarthy Scales of Children’s Abilities were administered between 36 and 62 months. RESULTS: Forty children who survived severe fetal hemolytic disease were followed up until 62 months old. Demographic data included gestational age at first intrauterine transfusion (26.4 ± 3.7 weeks), median number of intrauterine transfusions (4, range 1-8), lowest fetal hematocrit (20.2% ± 7.8%), peak fetal bilirubin (7.1 ± 2.1 mg/dL), incidence of hydrops fetalis (45%), and mean gestational age at delivery (35.6 ± 2.2 weeks). One case of severe bilateral deafness and 1 case of right spastic hemiplegia were diagnosed. The Gesell Developmental Schedules score was assessed between 9 and 18 months of corrected age in 22 infants. The global developmental quotient was 101.9 ± 9.5 (mean for normal population is 100). Regression analysis revealed no correlation between the global developmental quotient and gestational age at the first intrauterine transfusion, gestational age at birth, or the severity of the fetal hemolytic disease (fetal hematocrit, fetal bilirubin, presence of hydrops fetalis, total number of intrauterine transfusions, duration of neonatal phototherapy, and number of neonatal exchange transfusions). Eleven of the 40 children were followed up until they were 62 months old, and the McCarthy Scales of Children’s Abilities were administered. The mean cognitive index was 107.6 ± 9.4 (90-109 is considered average). CONCLUSION: Despite severe fetal hemolytic disease, normal developmental outcome can be expected for children treated with intrauterine transfusions. (Am J Obstet Gynecol 1998;179:858-63.)
Key words: Hemolytic disease of the neonate, hydrops fetalis, intrauterine transfusion, red blood cell alloimmunization
The intrauterine treatment of hemolytic disease of the neonate has advanced significantly since Liley reported the first successful intrauterine intraperitoneal transfusion in 1963.1 Harman et al2 later showed that intravascular transfusion markedly improved the chance for survival of the hydropic fetus compared with intraperitoneal transfusion. In their series the survival rate of fetuses with hydrops was increased from 48% to 86% through the use of intravascular transfusions. Ultrasonographically guided intrauterine transfusion, usually intravascular with or without combined intraperitoneal transfusion, re-
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology,a the Division of Neonatology, Department of Pediatrics,b and the Department of Human and Molecular Genetics,c Baylor College of Medicine. Presented at the Eighteenth Annual Meeting of the Society of Perinatal Obstetricians, Miami, Florida, February 2-7, 1998. Reprint requests: Lynda Hudon, MD, Baylor College of Medicine, Department of Obstetrics and Gynecology, 6550 Fannin No 901, Houston, TX 77030. †Deceased. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/6/92540
858
mains the only effective treatment available for fetal hemolytic disease caused by maternal red cell sensitization. This invasive treatment is not without complications, including fetal death, fetal bradycardia with the need for emergency delivery, and cerebral porencephalic cyst.3 As ultrasonographic imaging technology has improved, our ability to treat severely anemic fetuses earlier in gestation has been markedly enhanced. Fetuses in a more moribund state now routinely survive because of these advancements in treatment techniques. However, little is known regarding the neurodevelopmental outcomes of these surviving children. The aim of this study was to determine the long-term neurodevelopment and the incidence of significant handicaps among surviving children affected with hemolytic disease who underwent treatment with serial intrauterine transfusions. Material and methods Between January 1986 and December 1992, 49 singleton pregnancies affected by fetal hemolytic disease underwent treatment with ultrasonographically guided intrauterine transfusion at Baylor College of Medicine in
Volume 179, Number 4 Am J Obstet Gynecol
Houston. Nine of the 49 fetuses died during the antenatal period; 40 of the remaining fetuses were born alive and survived. These 40 fetuses comprised the study cohort. The 9 fetal deaths occurred between 1 and 4 days after an intrauterine transfusion at a mean gestational age of 23.1 weeks (18.4-31.0 weeks). Intramuscular neuromuscular blocking agents were used between 1986 and 1988; intravenous agents were used between 1988 and 1992. During the study period 5 babies received intramuscular neuromuscular blocking agents exclusively, 5 received both intramuscular and intravenous agents, and the remaining 30 infants received intravenous agents alone. In the first 2 to 3 years of the study the intrauterine transfusion technique consisted of exchange transfusions, in which equal volumes of blood were removed and transfused. After this initial period all intrauterine transfusions were performed similarly, in the following fashion. Access to the umbilical vessels, preferably the umbilical vein, was gained with a 20-gauge needle under ultrasonographic guidance. A hematocrit value was obtained for the first milliliter of fetal blood, and this determination was immediately followed by intravascular injection of a neuromuscular blocking agent. The transfusion of packed red blood cells with either maternal blood or heterologous blood (hematocrit 75%-85%) was initiated to correct the hemoglobin deficit and achieve a final fetal hematocrit of approximately 40%. In most cases an intraperitoneal transfusion was then undertaken to maintain a more stable hematocrit and prolong the interval between each procedure.4 The first 2 or 3 intrauterine transfusions were usually performed between 1 and 2 weeks apart, depending on the severity of the anemia; the interval for subsequent procedures, generally between 2 and 4 weeks, was determined according to the decline in hematocrit for each individual fetus. The last intrauterine transfusion was performed at approximately 35 weeks’ gestation. This is the preferred approach in our center because of the low probability of fetal maturity and the low procedure-related risk at this stage of pregnancy. A term delivery was planned between 2 and 3 weeks later. After the initial administration of 2 intramuscular 12-mg doses of betamethasone, corticosteroids were routinely administered on a weekly basis between 26 and 34 weeks to enhance fetal lung maturation. Maternal oral phenobarbital (30 mg 3 times daily) was prescribed after the last intrauterine transfusion for a total of 10 days to induce neonatal hepatic conjugation of bilirubin. A total of 161 procedures were performed in the study cohort. Forty-eight percent (n = 78) of these procedures were direct intravascular transfusions coupled with an intraperitoneal transfusion, 25% (n = 40) were intraperitoneal transfusions alone, 11% (n = 17) were intravascular exchange transfusions, 9% (n = 14) were direct
Hudon et al 859
intravascular transfusions alone, and the remaining 7% of procedures (n = 12) consisted of cordocentesis alone. The parents gave their consent for their infants’ participation in the study. All infants were examined within the first 24 hours after birth. The same pediatrician (R.H.) performed serial physical and developmental examinations on the children at birth and at 1, 4, 6, 9, 18, 36, and 48 months of age. Auditory evoked response tests were performed before discharge, while the infant was in the nursery. The Knobloch-Pasamanick modification of the Gesell Developmental Schedules was administered by the study nurse (S.H.) at 9 and 18 months of age. Chronologic age was corrected for premature delivery until the infants were 24 months old. The global score consists of the average score in each of the following 5 subcategories: gross motor skills, fine motor skills, adaptative behavior, language, and social skills. Scores between 85 and 114 are considered to be within normal limits. A score >115 is considered above average, whereas a score <84 characterizes developmental delay. Scores obtained at the oldest age were used if the child was tested on >1 occasion. The McCarthy Scales of Children’s Abilities were administered when children were between 3 and 5 years old. The latter test was administered by psychologic associates at The Meyer Center in Texas Children’s Hospital. The general cognitive index was considered very superior at ≥130, superior between 120 and 129, bright normal between 110 and 119, average between 90 and 109, dull normal between 80 and 89, borderline between 70 and 79, and mentally retarded at <69. The mothers underwent interviews by the study nurse to document family history, level of education, and drug exposure. All Hispanic mothers spoke English fluently and their children were examined in their primary language, with the mother as a translator when necessary. Because hydrops fetalis has been associated with a severe hemoglobin deficit, we chose to divide our population into 2 groups, those with fetal hydrops and those with no hydrops on initial examination. To be classified as hydropic the fetus had to have either the classic ultrasonographic findings (fluid in ≥2 serous cavities) or isolated ascites in association with a hematocrit value 3 standard deviations below the norm for gestational age.5 Of the 18 fetuses with hydrops, 10 were found to have the classic findings of hydrops fetalis and 8 exhibited ascites only at initial examination. Continuous variables were compared with the Student t test when the data were normally distributed (given as mean ± SD), whereas the Mann-Whitney Rank Sum was used for categoric variables and variables with nonnormal distribution (given as median with range). Prematurity and phenobarbital exposure were tested with a 2-sample t test, and regression analysis was used to determine variables associated with a worse neurodevelopmental outcome. A P value < .05 was used to in-
860 Hudon et al
October 1998 Am J Obstet Gynecol
Table I. Maternal demographics
Type of maternal alloantibody (%) Anti-D Anti-D and anti-C Anti-Kell Other Maternal age (y)* Gravidity† Parity† Ethnicity (%) White Hispanic African American Maternal level of education (y)*
Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
67 17 6 10 30.2 ± 6.0 2 (1-14) 1 (0-9)
65 18 6 11 29.2 ± 6.0 2 (1-14) 1 (0-3)
68 16 5 11 31.4 ± 5.8 2 (1-11) 0 (0-9)
P = 1.0 P = 1.0 P = 1.0 P = 1.0 P = .24 P = .69 P = .76
70.6 23.5 5.9 13.1 ± 2.2
83.0 11.1 5.6 13.3 ± 2.4
P = .73 P = .74 P = 1.0 P = .82
76.6 17.3 5.8 13.2 ± 2.3
*Mean ± SD. †Median, with range in parentheses.
Table II. Fetal demographics
Gestational age at first intrauterine transfusion (wk)* No. of intrauterine transfusions† Nadir in utero hematocrit (%) Peak in utero bilirubin (mg/dL)*
Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
26.4 ± 3.7 4 (2-8) 20.2 ± 7.8 7.1 ± 2.1
26.0 ± 3.4 4 (2-8) 23.7 ± 7.8 7.2 ± 2.1
26.8 ± 3.9 4 (1-7) 16.5 ± 5.8 7.1 ± 2.2
P = .60 P = .17 P = .28 P = .62
*Mean ± SD. †Median, with range in parentheses.
dicate statistical significance. The Sigmastat (Jander Corporation, St Raphael, Calif) statistical software package was used. Results Maternal demographics. Table I illustrates the demographics of our population. No difference was noted between the groups with respect to maternal demographics. Anti-Rh alloantibodies represented the vast majority of maternal alloantibodies responsible for the fetal hemolytic disease. Fetal demographics. Antenatal data in Table II demonstrate the severity of the fetal hemolytic disease. Fetuses with hydrops did not appear to be more severely affected than were their counterparts without hydrops when compared for mean gestational age at first intrauterine transfusion, total mean number of intrauterine transfusions, mean lowest value for hematocrit, and mean peak bilirubin value. Neonatal demographics. Table III compares neonatal findings between infants with and without hydrops as fetuses. The group with history of hydrops had a neonatal course comparable to that of the group without history of hydrops. Overall outcome. Overall, 21 babies underwent hearing tests before discharge. The test results were abnormal in 3
cases. Two babies had mild peripheral sensitivity loss; 1 had evidence of normal hearing on repeated testing at 5 months old. The third child was noted to have severe bilateral deafness. She was born by emergency cesarean delivery at 32 weeks’ gestation with severe hydrops fetalis. Her peak in utero bilirubin value was 10.8 mg/dL, her lowest fetal hematocrit was 24%, and she underwent 4 intrauterine transfusions. Her neonatal course was complicated by sepsis, a persistent ductus arteriosus, 355 hours of phototherapy without exchange transfusion, and a hospital stay of 48 days. An electroencephalogram obtained before her discharge revealed evidence of an in utero cerebral insult; results of cerebral ultrasonography were normal. Unfortunately, this child was unavailable for follow-up. Right spastic hemiplegia was diagnosed in 1 child 2.5 years old who had developed normally except for walking difficulties from a shorter right leg. This child was born by emergency cesarean delivery at 34.3 weeks’ gestation because of fetal distress and preterm premature rupture of the membranes. At birth she weighed 2245 g and had a 5-minute Apgar score of 9 and an arterial cord pH of 7.30. The neonatal course was complicated by a hospital stay of 12 days and a total of 127 hours of phototherapy without exchange transfusion. Her Gesell Developmental Schedules global score at age 17 months was within normal limits.
Hudon et al 861
Volume 179, Number 4 Am J Obstet Gynecol
Table III. Neonatal data Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
35.6 ± 2.3 2709 ± 650 88.9 7.25 ± 0.08 9 (5-9) 78.2 ± 80.7 9.0 (3-101)
35.6 ± 2.1 2857 ± 651 63.6 7.27 ± 0.08 9 (3-9) 74.5 ± 59.4 12 (4-41)
P = .54 P = .79 P = .52 P = .33 P = .39 P = .88 P = .44
Total cohort (N = 22)
Hydrops (n = 10)
No hydrops (n = 12)
Statistical significance
14.4 ± 5.8 102.9 ± 7.8 111.6 ± 12.5 104.3 ± 11.7 105.1 ± 6.8 96.7 ± 13.6 102.1 ± 11.9
15.2 ± 5.6 101.2 ± 12.0 113.9 ± 12.4 101.2 ± 10.0 103.7 ± 6.0 100.9 ± 10.8 102.1 ± 9.6
12.8 ± 3.8 102.8 ± 8.6 110.0 ± 11.1 105.0 ± 12.0 106.3 ± 5.9 94.2 ± 12.7 103.0 ± 13.6
P = .27 P = .72 P = .45 P = .44 P = .31 P = .20 P = .86
Gestational age at birth (wk)* Birth weight (g)* Cesarean delivery (%) Arterial cord blood pH* Apgar score at 5 min‡ Phototherapy duration (h)* Hospital stay (d)‡
35.6 ± 2.2 2790 ± 646 75.0 7.25 ± 0.01† 9 (3-9) 76.2 ± 68.9 11 (3-101)
*Mean ± SD. †Arterial cord pH was determined in 31 neonates. ‡Median, with range in parentheses.
Table IV. Gesell Developmental Schedules
Corrected age at test (mo) Global developmental quotient Gross motor skills Fine motor skills Adaptive Language Social skills Data are mean ± SD.
Two infants had diagnoses of severe mental retardation, 1 with Angelman’s syndrome and 1 with Menkes’ disease. A third child had Wilson’s disease diagnosed when 6 years old on the basis of hepatomegaly and Kayser-Fleischer rings found on ophthalmologic examination. The children with Angelman’s syndrome and Menkes’ disease could not be tested with the usual neurodevelopmental tests and were therefore excluded from further analysis. Because Wilson’s disease has not been associated with below average intelligence, the neurodevelopmental scores of this child were included in the final analysis. Gesell Developmental Schedules. Twenty-two children were administered the Gesell Developmental Schedules when they were between 9 and 36 months old. Six of the infants were tested on >1 occasion. No significant differences in global Gesell Developmental Schedules or any of its subcategories were noted between the groups with and without history of hydrops (Table IV). The global developmental quotient (102.9 ± 7.8) and the performances within each of the subcategories were all within normal limits. Regression analysis showed no correlation between the global developmental quotient or any of its subcategories and gestational age at first intrauterine transfusion (correlation coefficient 0.58 ± 0.65, P = .38), gestational age at birth (–1.43 ± 1.3, P = .29), birth weight (–0.003 ± 0.004, P = .46), total number of intrauterine transfusions
(–1.94 ± 1.48, P = .21), lowest fetal hematocrit (0.3 ± 0.36, P = .41), duration of phototherapy (0.22 ± 0.05, P = .65), number of exchange transfusions in the neonatal period (1.52 ± 1.12, P = .19), and maternal level of education (0.45 ± 1.44, P = .76). There was no correlation between the global developmental quotient and the presence of hydrops fetalis (P = .49) or antenatal phenobarbital administration (P = .58), as determined by analysis of variance and 2-sample t test. An elevated fetal bilirubin level was not associated with either a lower level of language development or a lower developmental score. McCarthy Scales of Children’s Abilities. Eleven children were tested with the McCarthy Scales of Children’s Abilities between 36 and 62 months of age. The mean general cognitive index for the entire group of children tested was slightly above average (107.6 ± 9.4). The global score indicated no significant differences in the performance of children with and without history of hydrops (Table V). Although the group with history of hydrops had lower verbal skills performances (P = .05), their results were still within the normal range. The performances in the other subcategories were similar between the groups. Comment Just 6 years after the description of the first successful intrauterine transfusion, Gregg and Hutchinson6 reported a survival rate of 33% for 55 fetuses undergoing treatment with intrauterine transfusion for severe Rh dis-
862 Hudon et al
October 1998 Am J Obstet Gynecol
Table V. McCarthy Scales of Children’s Abilities
Age at test (mo) General cognitive index Verbal skills Perceptual skills Quantitative skills Memory
Total cohort (N = 11)
Hydrops (n = 4)
No hydrops (n = 7)
Statistical significance
41.5 ± 8.0 107.6 ± 9.4 53.9 ± 7.4 54.9 ± 6.5 52.3 ± 6.6 50.6 ± 7.5
53.5 ± 12 102.0 ± 7.8 49.5 ± 4.9 52.8 ± 7.1 50.0 ± 4.3 48.5 ± 8.4
36.8 ± 10.0 111.4 ± 9.1 57.6 ± 6.7 55.6 ± 6.8 53.6 ± 8.1 52.9 ± 6.2
P = .30 P = .11 P = .05 P = .54 P = .36 P = .42
Data are mean ± SD.
ease. The neurodevelopmental outcome was assessed for 15 of these 18 children and reported to be normal. Although fetal survival rate improved with additional experience with the technique of intraperitoneal transfusion, Hardyment et al7 reported a survival rate of only 48% between 1966 and 1975. Follow-up of 21 of their 27 surviving infants noted 2 cases with major neurologic signs consisting of hemiparesis and coarse tremors of the upper extremities. One infant was noted to have evidence of unilateral high-frequency hearing deficit. Advances in ultrasonography led to the introduction of ultrasonographically guided intravascular transfusion, a technique associated with enhanced survival of more severely affected fetuses with the potential for poor neurodevelopmental outcome.2 The first study of fetuses surviving intravascular transfusions was not published until 1993. In that report Doyle et al8 described an overall survival rate of 73% after intrauterine transfusion, including 11 cases of hydrops fetalis in 52 fetuses. Among the 38 surviving infants the authors found 1 case of cerebral palsy, 1 case of severe developmental delay, 1 case of bilateral hemiplegia, and 1 case of severe developmental delay accompanied by seizures. There were no cases of deafness detected, although the methods for assessment of hearing were not described by these authors. Mean Bayley scores at 2 years were found to be within the normal range. In our investigation the survival rate after intrauterine transfusion was 80%, a rate comparable to those seen in other recent series of intrauterine transfusions.9 Of the 40 surviving infants, 22 were followed up until a mean age of 14.4 months. Only 1 infant was noted to have development of spastic hemiplegia. This incidence of 4.5% compares favorably with both the 10.5% incidence of major neurologic handicaps in the study of Doyle et al8 and the 4.8% incidence in the study by Hardyment et al.7 The survival rates in the 2 most recent series, ours and Doyle and coworkers,8 were much better than that in the series of Hardyment et al.7 This would lead one to conclude that the more moribund fetuses in the earlier study died of their anemia and were not available for follow-up. In addition the incidence of hydrops in our series was 45%, versus 29% in the series of Doyle et al,8 yet the incidence of major neurologic handicaps was lower in our
study. It is therefore reassuring that sicker fetuses who survive with the aid of current intrauterine transfusion techniques have an incidence of neurologic handicap similar to those seen in earlier studies of surviving infants and that hydrops fetalis at first examination does not worsen the prognosis. Hearing deficits in the neonate have been reported in association with high serum bilirubin levels.10,11 High bilirubin values associated with prematurity, hemolytic disease, and a complicated neonatal course place infants at higher risk for sensorineural hearing deficits. For this reason several of the previous studies evaluated the hearing of infants treated for hemolytic disease of the neonate. In our series a permanent hearing deficit occurred in 2 of 21 infants, which was comparable to the 1 of 21 seen in the series by Hardyment et al.7 It would therefore appear that sensorineural hearing loss is probably 5 to 10 times higher among infants affected with hemolytic disease than among healthy infants.10-12 In this investigation we chose to use the Gesell Developmental Schedules. The test is divided into 5 subcategories: gross motor, fine motor, adapatative, language, and social skills. In contrast, the Bayley Scales measure similar neurodevelopmental progress but are divided into only 2 categories, motor and cognitive skills. In this study the Gesell Developmental Schedules global score (102.9 ± 7.8) and the scores of the 5 subcategories were within the normal range. This result is similar to the description of a normal Bayley score (114 ± 16.8) in the study by Doyle et al.8 Although the group with history of hydrops in our study was considered to be severely affected by hemolytic disease, its Gesell Developmental Schedules scores were similar to those in the group without history of hydrops. Again, these data are reassuring, demonstrating that fetuses with hydrops who survive intrauterine transfusion have outcomes similar to those of other fetuses with hemolytic disease. The relatively good neonatal outcome of the hydropic group can be explained by the persistence of hydrops at birth in only 1 case, the absence of significant prematurity (mean gestational age at birth 35.6 ± 2.3 weeks), and delivery in a tertiary center. A longer hospital stay was found to be associated with a lower global developmental score. Because prematurity could be the explanation for this
Hudon et al 863
Volume 179, Number 4 Am J Obstet Gynecol
finding, we reanalyzed the data by dividing infants into premature (<36 weeks’ gestation) and term groups. No difference was noted between these groups’ global developmental scores. Despite the significant disadvantages of prematurity, lower birth weight, increased duration of phototherapy, and longer hospital stay, these children still had performances within the normal range. Unlike in previous investigations, we were able to follow up a fourth of our surviving infants until they were between 3 and 5 years old. The McCarthy Scales of Children’s Abilities was used to evaluate these children. The average cognitive index was found to be within the normal range. There was no significant difference between the children who had hydrops during fetal life and those who had no hydrops. Wilson’s disease and Menke’s disease are extremely rare in the normal population (1/60,000 and 1/200,000, respectively).13 Our finding of 2 cases of copper metabolism disorder in a cohort of 40 patients affected with hemolytic disease seems unusual and remains unexplained. One of the major limitations of this study is the lack of follow-up for all children. It is possible that the children at increased risk for severe neurodevelopmental compromise either did not return at all for evaluation or were unavailable for follow-up after a normal result of initial evaluation. The likely major explanation for the lack of sequential evaluation is the considerable geographic distance between the area of residence and the tertiary care center. Despite severe fetal hemolytic disease, it seems reasonable to say that normal neurodevelopment can be expected in 36-month-old children treated with intrauterine transfusion. The aggressive treatment of fetuses with severe hemolytic disease and hydrops fetalis is justified;
normal neurodevelopmental outcome without serious handicaps can be expected in 95% of these cases. REFERENCES
1. Green GH. Historic perspective on Liley’s fetal transfusion. Vox Sang 1985;48:184-7. 2. Harman CR, Bowman JM, Manning FA, Menticoglou SM. Intrauterine transfusion Intraperitoneal versus intravascular approach: a case-control comparison. Am J Obstet Gynecol 1990;162:1053-9. 3. Dildy G, Smith LG, Moise KJ, Cano LE, Heskesth DE. Porencephalic cyst: a complication of fetal intravascular transfusion. Am J Obstet Gynecol 1991;165:76-8. 4. Moise KJ, Carpenter RJ, Kirshon B, Deter RL, Lala JD, Cano LE. Comparison of four types of intrauterine transfusion: effect on fetal hematocrit. Fetal Ther 1989;4:126-37. 5. Leduc L, Moise KJ, Carpenter RJ, Cano LE. Fetoplacental blood volume estimation in pregnancies with Rh alloimmunization. Fetal Diagn Ther 1990;5:138-46. 6. Gregg SG, Hutchinson DL. Developmental characteristics of infants surviving fetal transfusion. JAMA 1969;209:1059-62. 7. Hardyment AF, Salvador HS, Towell ME, Carpenter CW, Jan JE, Tingle AJ. Follow-up of intrauterine transfused surviving children. Am J Obstet Gynecol 1979;133:235-41. 8. 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. 9. Schumacher B, Moise KJ. Fetal transfusions for red blood cell alloimmunization in pregnancy. Obstet Gynecol 1996;88:137-50. 10. Newman TB, Maisels MJ. Evaluation and treatment of jaundice in the term newborn: a kinder, gentler approach. Pediatrics 1992;89:809-18. 11. Valaes T, Kipouros K, Petmezaki S, Solman M, Doxiadis SA. Effectiveness and safety of prenatal phenobarbital for the prevention of neonatal jaundice. Pediatr Res 1980;14:947-52. 12. Newman TB, Klebanoff MA. Neonatal hyperbilirubinemia and long-term outcome: another look at the Collaborative Perinatal Project. Pediatrics 1993;92:651-7. 13. Rimoin DL, Connor MJ, Pyeritz RE. Volume 3: Emery and Rimoin’s principles and practices of medical genetics. 3rd ed. New York: Churchill Livingstone; 1996. p. 2037-44.
Key words: Hemolytic disease of the neonate, hydrops fetalis, intrauterine transfusion, red blood cell alloimmunization
The intrauterine treatment of hemolytic disease of the neonate has advanced significantly since Liley reported the first successful intrauterine intraperitoneal transfusion in 1963.1 Harman et al2 later showed that intravascular transfusion markedly improved the chance for survival of the hydropic fetus compared with intraperitoneal transfusion. In their series the survival rate of fetuses with hydrops was increased from 48% to 86% through the use of intravascular transfusions. Ultrasonographically guided intrauterine transfusion, usually intravascular with or without combined intraperitoneal transfusion, re-
From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology,a the Division of Neonatology, Department of Pediatrics,b and the Department of Human and Molecular Genetics,c Baylor College of Medicine. Presented at the Eighteenth Annual Meeting of the Society of Perinatal Obstetricians, Miami, Florida, February 2-7, 1998. Reprint requests: Lynda Hudon, MD, Baylor College of Medicine, Department of Obstetrics and Gynecology, 6550 Fannin No 901, Houston, TX 77030. †Deceased. Copyright © 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/6/92540
858
mains the only effective treatment available for fetal hemolytic disease caused by maternal red cell sensitization. This invasive treatment is not without complications, including fetal death, fetal bradycardia with the need for emergency delivery, and cerebral porencephalic cyst.3 As ultrasonographic imaging technology has improved, our ability to treat severely anemic fetuses earlier in gestation has been markedly enhanced. Fetuses in a more moribund state now routinely survive because of these advancements in treatment techniques. However, little is known regarding the neurodevelopmental outcomes of these surviving children. The aim of this study was to determine the long-term neurodevelopment and the incidence of significant handicaps among surviving children affected with hemolytic disease who underwent treatment with serial intrauterine transfusions. Material and methods Between January 1986 and December 1992, 49 singleton pregnancies affected by fetal hemolytic disease underwent treatment with ultrasonographically guided intrauterine transfusion at Baylor College of Medicine in
Volume 179, Number 4 Am J Obstet Gynecol
Houston. Nine of the 49 fetuses died during the antenatal period; 40 of the remaining fetuses were born alive and survived. These 40 fetuses comprised the study cohort. The 9 fetal deaths occurred between 1 and 4 days after an intrauterine transfusion at a mean gestational age of 23.1 weeks (18.4-31.0 weeks). Intramuscular neuromuscular blocking agents were used between 1986 and 1988; intravenous agents were used between 1988 and 1992. During the study period 5 babies received intramuscular neuromuscular blocking agents exclusively, 5 received both intramuscular and intravenous agents, and the remaining 30 infants received intravenous agents alone. In the first 2 to 3 years of the study the intrauterine transfusion technique consisted of exchange transfusions, in which equal volumes of blood were removed and transfused. After this initial period all intrauterine transfusions were performed similarly, in the following fashion. Access to the umbilical vessels, preferably the umbilical vein, was gained with a 20-gauge needle under ultrasonographic guidance. A hematocrit value was obtained for the first milliliter of fetal blood, and this determination was immediately followed by intravascular injection of a neuromuscular blocking agent. The transfusion of packed red blood cells with either maternal blood or heterologous blood (hematocrit 75%-85%) was initiated to correct the hemoglobin deficit and achieve a final fetal hematocrit of approximately 40%. In most cases an intraperitoneal transfusion was then undertaken to maintain a more stable hematocrit and prolong the interval between each procedure.4 The first 2 or 3 intrauterine transfusions were usually performed between 1 and 2 weeks apart, depending on the severity of the anemia; the interval for subsequent procedures, generally between 2 and 4 weeks, was determined according to the decline in hematocrit for each individual fetus. The last intrauterine transfusion was performed at approximately 35 weeks’ gestation. This is the preferred approach in our center because of the low probability of fetal maturity and the low procedure-related risk at this stage of pregnancy. A term delivery was planned between 2 and 3 weeks later. After the initial administration of 2 intramuscular 12-mg doses of betamethasone, corticosteroids were routinely administered on a weekly basis between 26 and 34 weeks to enhance fetal lung maturation. Maternal oral phenobarbital (30 mg 3 times daily) was prescribed after the last intrauterine transfusion for a total of 10 days to induce neonatal hepatic conjugation of bilirubin. A total of 161 procedures were performed in the study cohort. Forty-eight percent (n = 78) of these procedures were direct intravascular transfusions coupled with an intraperitoneal transfusion, 25% (n = 40) were intraperitoneal transfusions alone, 11% (n = 17) were intravascular exchange transfusions, 9% (n = 14) were direct
Hudon et al 859
intravascular transfusions alone, and the remaining 7% of procedures (n = 12) consisted of cordocentesis alone. The parents gave their consent for their infants’ participation in the study. All infants were examined within the first 24 hours after birth. The same pediatrician (R.H.) performed serial physical and developmental examinations on the children at birth and at 1, 4, 6, 9, 18, 36, and 48 months of age. Auditory evoked response tests were performed before discharge, while the infant was in the nursery. The Knobloch-Pasamanick modification of the Gesell Developmental Schedules was administered by the study nurse (S.H.) at 9 and 18 months of age. Chronologic age was corrected for premature delivery until the infants were 24 months old. The global score consists of the average score in each of the following 5 subcategories: gross motor skills, fine motor skills, adaptative behavior, language, and social skills. Scores between 85 and 114 are considered to be within normal limits. A score >115 is considered above average, whereas a score <84 characterizes developmental delay. Scores obtained at the oldest age were used if the child was tested on >1 occasion. The McCarthy Scales of Children’s Abilities were administered when children were between 3 and 5 years old. The latter test was administered by psychologic associates at The Meyer Center in Texas Children’s Hospital. The general cognitive index was considered very superior at ≥130, superior between 120 and 129, bright normal between 110 and 119, average between 90 and 109, dull normal between 80 and 89, borderline between 70 and 79, and mentally retarded at <69. The mothers underwent interviews by the study nurse to document family history, level of education, and drug exposure. All Hispanic mothers spoke English fluently and their children were examined in their primary language, with the mother as a translator when necessary. Because hydrops fetalis has been associated with a severe hemoglobin deficit, we chose to divide our population into 2 groups, those with fetal hydrops and those with no hydrops on initial examination. To be classified as hydropic the fetus had to have either the classic ultrasonographic findings (fluid in ≥2 serous cavities) or isolated ascites in association with a hematocrit value 3 standard deviations below the norm for gestational age.5 Of the 18 fetuses with hydrops, 10 were found to have the classic findings of hydrops fetalis and 8 exhibited ascites only at initial examination. Continuous variables were compared with the Student t test when the data were normally distributed (given as mean ± SD), whereas the Mann-Whitney Rank Sum was used for categoric variables and variables with nonnormal distribution (given as median with range). Prematurity and phenobarbital exposure were tested with a 2-sample t test, and regression analysis was used to determine variables associated with a worse neurodevelopmental outcome. A P value < .05 was used to in-
860 Hudon et al
October 1998 Am J Obstet Gynecol
Table I. Maternal demographics
Type of maternal alloantibody (%) Anti-D Anti-D and anti-C Anti-Kell Other Maternal age (y)* Gravidity† Parity† Ethnicity (%) White Hispanic African American Maternal level of education (y)*
Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
67 17 6 10 30.2 ± 6.0 2 (1-14) 1 (0-9)
65 18 6 11 29.2 ± 6.0 2 (1-14) 1 (0-3)
68 16 5 11 31.4 ± 5.8 2 (1-11) 0 (0-9)
P = 1.0 P = 1.0 P = 1.0 P = 1.0 P = .24 P = .69 P = .76
70.6 23.5 5.9 13.1 ± 2.2
83.0 11.1 5.6 13.3 ± 2.4
P = .73 P = .74 P = 1.0 P = .82
76.6 17.3 5.8 13.2 ± 2.3
*Mean ± SD. †Median, with range in parentheses.
Table II. Fetal demographics
Gestational age at first intrauterine transfusion (wk)* No. of intrauterine transfusions† Nadir in utero hematocrit (%) Peak in utero bilirubin (mg/dL)*
Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
26.4 ± 3.7 4 (2-8) 20.2 ± 7.8 7.1 ± 2.1
26.0 ± 3.4 4 (2-8) 23.7 ± 7.8 7.2 ± 2.1
26.8 ± 3.9 4 (1-7) 16.5 ± 5.8 7.1 ± 2.2
P = .60 P = .17 P = .28 P = .62
*Mean ± SD. †Median, with range in parentheses.
dicate statistical significance. The Sigmastat (Jander Corporation, St Raphael, Calif) statistical software package was used. Results Maternal demographics. Table I illustrates the demographics of our population. No difference was noted between the groups with respect to maternal demographics. Anti-Rh alloantibodies represented the vast majority of maternal alloantibodies responsible for the fetal hemolytic disease. Fetal demographics. Antenatal data in Table II demonstrate the severity of the fetal hemolytic disease. Fetuses with hydrops did not appear to be more severely affected than were their counterparts without hydrops when compared for mean gestational age at first intrauterine transfusion, total mean number of intrauterine transfusions, mean lowest value for hematocrit, and mean peak bilirubin value. Neonatal demographics. Table III compares neonatal findings between infants with and without hydrops as fetuses. The group with history of hydrops had a neonatal course comparable to that of the group without history of hydrops. Overall outcome. Overall, 21 babies underwent hearing tests before discharge. The test results were abnormal in 3
cases. Two babies had mild peripheral sensitivity loss; 1 had evidence of normal hearing on repeated testing at 5 months old. The third child was noted to have severe bilateral deafness. She was born by emergency cesarean delivery at 32 weeks’ gestation with severe hydrops fetalis. Her peak in utero bilirubin value was 10.8 mg/dL, her lowest fetal hematocrit was 24%, and she underwent 4 intrauterine transfusions. Her neonatal course was complicated by sepsis, a persistent ductus arteriosus, 355 hours of phototherapy without exchange transfusion, and a hospital stay of 48 days. An electroencephalogram obtained before her discharge revealed evidence of an in utero cerebral insult; results of cerebral ultrasonography were normal. Unfortunately, this child was unavailable for follow-up. Right spastic hemiplegia was diagnosed in 1 child 2.5 years old who had developed normally except for walking difficulties from a shorter right leg. This child was born by emergency cesarean delivery at 34.3 weeks’ gestation because of fetal distress and preterm premature rupture of the membranes. At birth she weighed 2245 g and had a 5-minute Apgar score of 9 and an arterial cord pH of 7.30. The neonatal course was complicated by a hospital stay of 12 days and a total of 127 hours of phototherapy without exchange transfusion. Her Gesell Developmental Schedules global score at age 17 months was within normal limits.
Hudon et al 861
Volume 179, Number 4 Am J Obstet Gynecol
Table III. Neonatal data Total cohort (N = 40)
Hydrops (n = 18)
No hydrops (n = 22)
Statistical significance
35.6 ± 2.3 2709 ± 650 88.9 7.25 ± 0.08 9 (5-9) 78.2 ± 80.7 9.0 (3-101)
35.6 ± 2.1 2857 ± 651 63.6 7.27 ± 0.08 9 (3-9) 74.5 ± 59.4 12 (4-41)
P = .54 P = .79 P = .52 P = .33 P = .39 P = .88 P = .44
Total cohort (N = 22)
Hydrops (n = 10)
No hydrops (n = 12)
Statistical significance
14.4 ± 5.8 102.9 ± 7.8 111.6 ± 12.5 104.3 ± 11.7 105.1 ± 6.8 96.7 ± 13.6 102.1 ± 11.9
15.2 ± 5.6 101.2 ± 12.0 113.9 ± 12.4 101.2 ± 10.0 103.7 ± 6.0 100.9 ± 10.8 102.1 ± 9.6
12.8 ± 3.8 102.8 ± 8.6 110.0 ± 11.1 105.0 ± 12.0 106.3 ± 5.9 94.2 ± 12.7 103.0 ± 13.6
P = .27 P = .72 P = .45 P = .44 P = .31 P = .20 P = .86
Gestational age at birth (wk)* Birth weight (g)* Cesarean delivery (%) Arterial cord blood pH* Apgar score at 5 min‡ Phototherapy duration (h)* Hospital stay (d)‡
35.6 ± 2.2 2790 ± 646 75.0 7.25 ± 0.01† 9 (3-9) 76.2 ± 68.9 11 (3-101)
*Mean ± SD. †Arterial cord pH was determined in 31 neonates. ‡Median, with range in parentheses.
Table IV. Gesell Developmental Schedules
Corrected age at test (mo) Global developmental quotient Gross motor skills Fine motor skills Adaptive Language Social skills Data are mean ± SD.
Two infants had diagnoses of severe mental retardation, 1 with Angelman’s syndrome and 1 with Menkes’ disease. A third child had Wilson’s disease diagnosed when 6 years old on the basis of hepatomegaly and Kayser-Fleischer rings found on ophthalmologic examination. The children with Angelman’s syndrome and Menkes’ disease could not be tested with the usual neurodevelopmental tests and were therefore excluded from further analysis. Because Wilson’s disease has not been associated with below average intelligence, the neurodevelopmental scores of this child were included in the final analysis. Gesell Developmental Schedules. Twenty-two children were administered the Gesell Developmental Schedules when they were between 9 and 36 months old. Six of the infants were tested on >1 occasion. No significant differences in global Gesell Developmental Schedules or any of its subcategories were noted between the groups with and without history of hydrops (Table IV). The global developmental quotient (102.9 ± 7.8) and the performances within each of the subcategories were all within normal limits. Regression analysis showed no correlation between the global developmental quotient or any of its subcategories and gestational age at first intrauterine transfusion (correlation coefficient 0.58 ± 0.65, P = .38), gestational age at birth (–1.43 ± 1.3, P = .29), birth weight (–0.003 ± 0.004, P = .46), total number of intrauterine transfusions
(–1.94 ± 1.48, P = .21), lowest fetal hematocrit (0.3 ± 0.36, P = .41), duration of phototherapy (0.22 ± 0.05, P = .65), number of exchange transfusions in the neonatal period (1.52 ± 1.12, P = .19), and maternal level of education (0.45 ± 1.44, P = .76). There was no correlation between the global developmental quotient and the presence of hydrops fetalis (P = .49) or antenatal phenobarbital administration (P = .58), as determined by analysis of variance and 2-sample t test. An elevated fetal bilirubin level was not associated with either a lower level of language development or a lower developmental score. McCarthy Scales of Children’s Abilities. Eleven children were tested with the McCarthy Scales of Children’s Abilities between 36 and 62 months of age. The mean general cognitive index for the entire group of children tested was slightly above average (107.6 ± 9.4). The global score indicated no significant differences in the performance of children with and without history of hydrops (Table V). Although the group with history of hydrops had lower verbal skills performances (P = .05), their results were still within the normal range. The performances in the other subcategories were similar between the groups. Comment Just 6 years after the description of the first successful intrauterine transfusion, Gregg and Hutchinson6 reported a survival rate of 33% for 55 fetuses undergoing treatment with intrauterine transfusion for severe Rh dis-
862 Hudon et al
October 1998 Am J Obstet Gynecol
Table V. McCarthy Scales of Children’s Abilities
Age at test (mo) General cognitive index Verbal skills Perceptual skills Quantitative skills Memory
Total cohort (N = 11)
Hydrops (n = 4)
No hydrops (n = 7)
Statistical significance
41.5 ± 8.0 107.6 ± 9.4 53.9 ± 7.4 54.9 ± 6.5 52.3 ± 6.6 50.6 ± 7.5
53.5 ± 12 102.0 ± 7.8 49.5 ± 4.9 52.8 ± 7.1 50.0 ± 4.3 48.5 ± 8.4
36.8 ± 10.0 111.4 ± 9.1 57.6 ± 6.7 55.6 ± 6.8 53.6 ± 8.1 52.9 ± 6.2
P = .30 P = .11 P = .05 P = .54 P = .36 P = .42
Data are mean ± SD.
ease. The neurodevelopmental outcome was assessed for 15 of these 18 children and reported to be normal. Although fetal survival rate improved with additional experience with the technique of intraperitoneal transfusion, Hardyment et al7 reported a survival rate of only 48% between 1966 and 1975. Follow-up of 21 of their 27 surviving infants noted 2 cases with major neurologic signs consisting of hemiparesis and coarse tremors of the upper extremities. One infant was noted to have evidence of unilateral high-frequency hearing deficit. Advances in ultrasonography led to the introduction of ultrasonographically guided intravascular transfusion, a technique associated with enhanced survival of more severely affected fetuses with the potential for poor neurodevelopmental outcome.2 The first study of fetuses surviving intravascular transfusions was not published until 1993. In that report Doyle et al8 described an overall survival rate of 73% after intrauterine transfusion, including 11 cases of hydrops fetalis in 52 fetuses. Among the 38 surviving infants the authors found 1 case of cerebral palsy, 1 case of severe developmental delay, 1 case of bilateral hemiplegia, and 1 case of severe developmental delay accompanied by seizures. There were no cases of deafness detected, although the methods for assessment of hearing were not described by these authors. Mean Bayley scores at 2 years were found to be within the normal range. In our investigation the survival rate after intrauterine transfusion was 80%, a rate comparable to those seen in other recent series of intrauterine transfusions.9 Of the 40 surviving infants, 22 were followed up until a mean age of 14.4 months. Only 1 infant was noted to have development of spastic hemiplegia. This incidence of 4.5% compares favorably with both the 10.5% incidence of major neurologic handicaps in the study of Doyle et al8 and the 4.8% incidence in the study by Hardyment et al.7 The survival rates in the 2 most recent series, ours and Doyle and coworkers,8 were much better than that in the series of Hardyment et al.7 This would lead one to conclude that the more moribund fetuses in the earlier study died of their anemia and were not available for follow-up. In addition the incidence of hydrops in our series was 45%, versus 29% in the series of Doyle et al,8 yet the incidence of major neurologic handicaps was lower in our
study. It is therefore reassuring that sicker fetuses who survive with the aid of current intrauterine transfusion techniques have an incidence of neurologic handicap similar to those seen in earlier studies of surviving infants and that hydrops fetalis at first examination does not worsen the prognosis. Hearing deficits in the neonate have been reported in association with high serum bilirubin levels.10,11 High bilirubin values associated with prematurity, hemolytic disease, and a complicated neonatal course place infants at higher risk for sensorineural hearing deficits. For this reason several of the previous studies evaluated the hearing of infants treated for hemolytic disease of the neonate. In our series a permanent hearing deficit occurred in 2 of 21 infants, which was comparable to the 1 of 21 seen in the series by Hardyment et al.7 It would therefore appear that sensorineural hearing loss is probably 5 to 10 times higher among infants affected with hemolytic disease than among healthy infants.10-12 In this investigation we chose to use the Gesell Developmental Schedules. The test is divided into 5 subcategories: gross motor, fine motor, adapatative, language, and social skills. In contrast, the Bayley Scales measure similar neurodevelopmental progress but are divided into only 2 categories, motor and cognitive skills. In this study the Gesell Developmental Schedules global score (102.9 ± 7.8) and the scores of the 5 subcategories were within the normal range. This result is similar to the description of a normal Bayley score (114 ± 16.8) in the study by Doyle et al.8 Although the group with history of hydrops in our study was considered to be severely affected by hemolytic disease, its Gesell Developmental Schedules scores were similar to those in the group without history of hydrops. Again, these data are reassuring, demonstrating that fetuses with hydrops who survive intrauterine transfusion have outcomes similar to those of other fetuses with hemolytic disease. The relatively good neonatal outcome of the hydropic group can be explained by the persistence of hydrops at birth in only 1 case, the absence of significant prematurity (mean gestational age at birth 35.6 ± 2.3 weeks), and delivery in a tertiary center. A longer hospital stay was found to be associated with a lower global developmental score. Because prematurity could be the explanation for this
Hudon et al 863
Volume 179, Number 4 Am J Obstet Gynecol
finding, we reanalyzed the data by dividing infants into premature (<36 weeks’ gestation) and term groups. No difference was noted between these groups’ global developmental scores. Despite the significant disadvantages of prematurity, lower birth weight, increased duration of phototherapy, and longer hospital stay, these children still had performances within the normal range. Unlike in previous investigations, we were able to follow up a fourth of our surviving infants until they were between 3 and 5 years old. The McCarthy Scales of Children’s Abilities was used to evaluate these children. The average cognitive index was found to be within the normal range. There was no significant difference between the children who had hydrops during fetal life and those who had no hydrops. Wilson’s disease and Menke’s disease are extremely rare in the normal population (1/60,000 and 1/200,000, respectively).13 Our finding of 2 cases of copper metabolism disorder in a cohort of 40 patients affected with hemolytic disease seems unusual and remains unexplained. One of the major limitations of this study is the lack of follow-up for all children. It is possible that the children at increased risk for severe neurodevelopmental compromise either did not return at all for evaluation or were unavailable for follow-up after a normal result of initial evaluation. The likely major explanation for the lack of sequential evaluation is the considerable geographic distance between the area of residence and the tertiary care center. Despite severe fetal hemolytic disease, it seems reasonable to say that normal neurodevelopment can be expected in 36-month-old children treated with intrauterine transfusion. The aggressive treatment of fetuses with severe hemolytic disease and hydrops fetalis is justified;
normal neurodevelopmental outcome without serious handicaps can be expected in 95% of these cases. REFERENCES
1. Green GH. Historic perspective on Liley’s fetal transfusion. Vox Sang 1985;48:184-7. 2. Harman CR, Bowman JM, Manning FA, Menticoglou SM. Intrauterine transfusion Intraperitoneal versus intravascular approach: a case-control comparison. Am J Obstet Gynecol 1990;162:1053-9. 3. Dildy G, Smith LG, Moise KJ, Cano LE, Heskesth DE. Porencephalic cyst: a complication of fetal intravascular transfusion. Am J Obstet Gynecol 1991;165:76-8. 4. Moise KJ, Carpenter RJ, Kirshon B, Deter RL, Lala JD, Cano LE. Comparison of four types of intrauterine transfusion: effect on fetal hematocrit. Fetal Ther 1989;4:126-37. 5. Leduc L, Moise KJ, Carpenter RJ, Cano LE. Fetoplacental blood volume estimation in pregnancies with Rh alloimmunization. Fetal Diagn Ther 1990;5:138-46. 6. Gregg SG, Hutchinson DL. Developmental characteristics of infants surviving fetal transfusion. JAMA 1969;209:1059-62. 7. Hardyment AF, Salvador HS, Towell ME, Carpenter CW, Jan JE, Tingle AJ. Follow-up of intrauterine transfused surviving children. Am J Obstet Gynecol 1979;133:235-41. 8. 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. 9. Schumacher B, Moise KJ. Fetal transfusions for red blood cell alloimmunization in pregnancy. Obstet Gynecol 1996;88:137-50. 10. Newman TB, Maisels MJ. Evaluation and treatment of jaundice in the term newborn: a kinder, gentler approach. Pediatrics 1992;89:809-18. 11. Valaes T, Kipouros K, Petmezaki S, Solman M, Doxiadis SA. Effectiveness and safety of prenatal phenobarbital for the prevention of neonatal jaundice. Pediatr Res 1980;14:947-52. 12. Newman TB, Klebanoff MA. Neonatal hyperbilirubinemia and long-term outcome: another look at the Collaborative Perinatal Project. Pediatrics 1993;92:651-7. 13. Rimoin DL, Connor MJ, Pyeritz RE. Volume 3: Emery and Rimoin’s principles and practices of medical genetics. 3rd ed. New York: Churchill Livingstone; 1996. p. 2037-44.