Neurodevelopmental outcome after intrauterine red cell transfusion for Parvovirus B19-induced fetal hydrops


* Dr J. Dembinski, Department of Neonatology, Center of Pediatrics, University of Bonn, Adenauerallee 119, D-53113 Bonn, Germany.


Objective To assess long term neurodevelopmental outcome of children after intrauterine intravascular red cell transfusion (IUT) for Parvovirus B19-induced fetal hydrops.

Design Data of study children were investigated retrospectively. Neurodevelopmental evaluation was performed by appropriate standard tests (Griffiths, Snijders–Oomen, Kaufmann Assessment Battery for Children tests).

Setting Tertiary care university teaching hospital.

Sample Twenty children who had Parvovirus-induced fetal hydrops and intrauterine transfusion of packed red blood cells (IUT).

Methods Retrospective chart analysis and standard neurodevelopmental testing.

Main outcome measures Developmental quotient (DQ) and intelligence quotient (IQ) according to the age at testing.

Results Twenty survivors of Parvovirus B19-induced fetal hydrops successfully treated by IUT were followed until 13 months to nine years of age. On clinical follow up, no neurologic sequelae were evident. Neurodevelopmental scores of all children ranged within two standard deviations of a normal population (median 101, range 86–116) and exceeded one standard deviation in three children. There was no significant neurodevelopmental delay.

Conclusion Children having survived successful IUT for Parvovirus B19-induced fetal anaemia and hydrops have a good neurodevelopmental prognosis. Our results support the use of IUT for correction of Parvovirus B19-induced fetal anaemia and subsequent hydrops.


Parvovirus B19 is the only known human pathogenic member of the Parvoviridae family. The so-called Erythrovirus shows a high affinity to and selective replication in erythroid progenitor cells. A direct cytotoxic effect leads to decreased erythrocyte production and subsequent anaemia1. The virus crosses the placental barrier to infect the fetus. Fetal Parvovirus infection is associated with high fetal morbidity and mortality. The spontaneous course of fetal Parvovirus B19 disease includes anaemia with and without hydrops, persistent and transient, as well as myocarditis and cerebral complications. Approximately 40% of women of childbearing age are susceptible for Parvovirus B19 infection with annual seroconversion rates of 1.5–15% depending on the epidemiologic situation. Among susceptible women, 20–50% develop infection, which is associated with fetal infection in approximately 33%. Fetal loss occurs in 10% and hydrops in 3% following infection in the first half of pregnancy2. The diagnosis can be established by PCR in amniotic fluid, chorionic villi or fetal serum3,4.

Management with intrauterine transfusion (IUT) can correct fetal anaemia and may reduce the mortality of fetal Parvovirus B19 infection significantly. Fetal mortality rate due to Parvovirus B19-induced hydrops has been reported as high as 30% without IUT5,6. Management with IUT guided by appropriate anaemia Z-scores in the absence of fetal myocarditis may reduce mortality by more than 50%5. At our centre, IUT has been successfully used in the management of fetal hydrops due to Parvovirus B19 infection since 19857–9. There has been concern about the long term outcome of these infants and enough time has passed with a sufficient number of cases for a follow up study. We present the results of neurodevelopmental follow up of 20 patients with hydrops fetalis, the most severe form of prenatal Parvovirus infection, until nine years of age.


Sixty-seven cases of maternofetal Parvovirus B19 infection were referred to a single tertiary care centre from 1989 to 1998. Fetal infection was confirmed by PCR or specific IgM. In 30 of the 67 cases, the infection resolved spontaneously with a favourable outcome. Thirty-seven fetuses developed hydrops and had been treated with IUT. In this subgroup, there were five intrauterine deaths. Thirty-two of the IUT-treated fetuses were liveborns, one infant died during the neonatal period. The parents of the surviving infants were contacted. Informed consent for neurodevelopmental follow up was obtained from 20 infants (12 females/8 males) of the 31 survivors who had been treated with IUT. Pre- and perinatal data were recorded retrospectively by chart analysis (Table 1). Eleven of the 31 surviving infants could not be followed due to declined consent or lack of contact address, however, there was no neonatal morbidity associated with intrauterine disease in these infants. At the time of neurodevelopmental assessment, the children were 13–110 months of age (median, 52 months). The neurodevelopment was assessed from 1/1998–1/1999 by the following tests mentioned below.

Table 1.  Characteristics of intrauterine therapy in 20 children with Parvovirus B19-induced fetal hydrops. Demographic data of study group. Values are median (range).
Gestational age at first IUT (weeks)23 (17–29)
Minimum fetal haemoglobin (g/dL)3.35 (2.1–8.5)
IUT (n)4 (1–7)
Gestational age at delivery37 (25–40)

Intelligence scores were assessed with the Snijders–Oomen Non-Verbal Intelligence Test in 14 children aged >2–6 years. The test consists of five subtests, the mean global intelligence is 100 with a standard deviation of 15. In three children aged >8 years at the time of testing, the Kaufmann Assessment Battery for Children was administered, which allows assessment of intellectual properties independent of socio-economic influences by age adapted subtests. A result of 50% represents an intelligence quotient (IQ) of 100, the standard deviation is 15. In three children aged 24 months or less, the developmental quotient (DQ) was assessed by the Griffiths Test for infant development. The Griffiths Test consists of five subcategories (motor skills, personal–social skills, language skills, eye and hand coordination and practical reasoning). A global score is calculated by the quotient of corrected age and developmental age, which represents the average of the subcategorical scores with a mean of 100 and a standard deviation of 11. For all tests, the standard deviation of the IQ/DQ is expressed and allows a comparison with an age-matched normal population.


Eleven infants were delivered at term, the median gestational age of nine infants delivered preterm was 33 weeks (range 25–36 weeks), median number of IUT was 3.5 (range 1–7), median age of first IUT was 23 weeks (range 17–29 weeks). Median minimum fetal haemoglobin was 3.35 (range 2.2–8.5) (Table 1). Neurologic examination revealed no clinical sequelae of fetal Parvovirus B19 disease. Mean neurodevelopmental/intelligence score of all children was 101 (84–116). In two children, the IQ/DQ fell below one standard deviation, in one child the IQ/DQ exceeded one standard deviation.

Multivariate analysis of variance revealed no independent effect on IQ/DQ of minimum fetal haemoglobin (P= 0.64), minimal fetal haematocrit (P= 0.97), total number of IUT (P= 0.1), gestational age at first IUT (P= 0.1) or gestational age at delivery (P= 0.4). Neurodevelopmental testing revealed no significant delay of children after intrauterine Parvovirus B19 infection, fetal hydrops and IUT.


IUT may reduce mortality of Parvovirus B19-induced fetal hydrops to <10%. Spontaneous resolution of the condition may occur in up to 34%, but intrauterine mortality without IUT has been reported to be 30% to >90%5,6,10–12. Parvovirus B19 may cause complications other than intrauterine fetal hydrops and even fetal death in the absence of anaemia13,14. Corneal opacification and meconium peritonitis have been associated with Parvovirus B19 and encephalitis may occur15–17. In this study, 20 surviving children of 37 IUT cases showed no complications related directly to Parvovirus B19 infection.

Neurodevelopmental long term outcome of surviving children after IUT has been described rarely. In a two-year follow up study, Doyle et al.18 reported a rate of 10.5% of severe handicaps following IUT for fetal erythroblastosis. The two children in our study whose IQ was below 85 were delivered vaginally and had an uncomplicated neonatal history. However, in these two children, the neurodevelopmental quotient ranged within two standard deviations and was considered normal. Cameron et al.12 reported normal neurodevelopment in the single survivor of 17 pregnancies complicated by NIHF due to Parvovirus B19 at two years follow up. Neurodevelopment has been assessed in 33 children with fetal haemolytic Rh disease up to 62 months of age after IUT by the Gesell Schedules, which are comparable to the Griffiths Test. In these children, one case of bilateral deafness and hemiplegia, respectively, has been observed19. In our study, 100% of the fetuses had developed hydrops compared with 45% in Hudson's study. However, subcategory tests proved no statistical differences between hydropic and non-hydropic patients. An important prognostic factor may be the total duration of anaemia, which may be less in Parvovirus B19 infection than in Rh disease. In this study, all children had developed fetal hydrops and can be considered severely affected by hypoplastic anaemia.

For neuropsychologic testing in three children <24 months of age, we have chosen the neurodevelopmental quotient (DQ) assessed by the Griffiths Test adapted for the German population20. A DQ of 100 (11) indicates normal development. In children aged >2 years, the IQ was determined by use of the Snijders–Oomen Non-Verbal Intelligence Test or the Kaufmann Assessment Battery for children. IQ levels of 100 (15) are considered normal21–23. The neurodevelopment quotient (DQ) is not equivalent to the IQ, however, there is empirical evidence that a normal DQ has a strong predictive value for normal intellectual performance in later childhood of the three children <24 months of age in this study 24,25. Median gestational age of the control group was slightly accelerated, which underlines the positive outcome in the study children. The relatively small number of study infants and the range of age at testing are a limitation of this study. Failure of follow up in 11 infants may have biased our results, as cognitive disorders or developmental delay in later life could not be excluded completely.

However, this observation supports the use of repeated IUT for correction of anaemia in Parvovirus B19-induced fetal hydrops. Reduction of mortality by aggressive management with IUT does not seem to be correlated with adverse neurodevelopmental outcome in infancy despite extremely low fetal haemoglobin levels and invasive intrauterine therapy. A normal neurodevelopmental outcome can be expected after IUT for severe Parvovirus B19 fetal disease.