Congenital Anemia Caused by Parvovirus B19 Infection

Parvovirus B19 exhibits a marked tropism to human bone marrow and replicates only in erythroid progenitor cells. 1 Infection can cause erythema infectiosum, arthropathy, hydrops fetalis and various hematologic disorders (aplastic crisis, chronic anemia, idiopathic thrombocytopenic purpura). 1 We present a case of parvovirus B19 infection causing severe hydrops fetalis and subsequent congenital anemia, which was corrected coincident with a combination of intrauterine transfusions and postnatal intravenous immunoglobulin (IVIG). Patient and methods. A 36-year-old woman, gravida 3, para 2, was clinically evaluated at 20 weeks of gestation because her two children exhibited a facial rash resembling erythema infectiosum. Two individual maternal blood samples were both positive for parvovirus B19 IgM and IgG with a commercial enzyme-linked immunosorbent assay (ELISA; Biotrin, Ireland). Her past obstetric and medical histories were noncontributory, and she remained healthy with no arthralgia or rash. Ultrasound examination every second week was commenced and showed no abnormalities at 20 and 22 weeks of gestation. At 24 weeks of gestation profound hydrops fetalis developed with pericardial effusion, swollen placenta and a copious amount of amniotic fluid. The weight of the fetus was estimated to be 1020 g, 65% above the mean for gestational age. A fetal blood sample obtained by ultrasound guided venipuncture revealed hemoglobin of 1.5 mmol/l (2.4 g/dl) and platelets at 37 × 10 9 /. Seven intrauterine transfusions with red blood cells and platelets were needed to correct these indices during the ensuing 3 weeks (Fig. 1). Ultrasound scans showed a gradual and partial resolution of hydrops. At 31 weeks of gestation the woman gave birth to a 2190-g male infant by spontaneous vaginal delivery. The infant’s abdomen was enlarged, and ascitic fluid was aspirated. The neonate exhibited continued anemia (5.2 mmol/l ≈ 8.4 g/dl), which was briefly treated with erythropoietin with no effect. A large number of serum samples from the mother and the fetus/child were available for examination by ELISA, dot blot and nested PCR of both the NS and VP genes (Fig. 1). 2 Results showed evidence of a persistent parvovirus B19 infection. Consequently at 1 month of age, treatment with IVIG (1 g/kg) was given for 2 days and repeated 1 month later. After 6 months of follow-up, coinciding with the clearance of virus, the hemoglobin and reticulocytes were normalized and have remained stable for an additional 7 months. The patient’s white blood cell count and serum IgG value were normal, but concentrations of IgM and IgA were marginally lowered: IgM, 0.21 μmol/l (range, 0.31 to 2.2 μmol/l); and IgA, 3.0 μmol/l (range, 5.0 to 23.0 μmol/l).Fig. 1: Hemoglobin, platelets, treatment and microbiologic findings during parvovirus B19 infection. ―――, hemoglobin values; – – – –, platelets; +, positive sample; −, negative sample.Sequencing of a 435-bp fragment of the NS1 gene (nucleotides 1093 to 1527 3) in blood samples drawn at the same time from the mother and the fetus was performed and repeated 4 months later, using the BigDye Kit (Perkin Elmer Applied Biosystems, Warrington, UK) and the ABI Prism 310 Genetic Analyzer. Results showed a perfect match of DNA sequences, confirming vertical transmission. When comparing with the published sequence by Shade et al., 3 four nucleotide substitutions were found (0.9%) leading to altered translation at sites 1291 to 1293 (cysteine → serine) and 1468–70 (valine → asparagine). A test for specific IgA was developed using VP2 capsids from a commercial IgG assay (Biotrin) in which horseradish peroxidase-labeled rabbit anti-human IgA (Dako, Glostrup, Denmark) was used as enzyme conjugate. All maternal serum samples were IgA-positive with a declining titer. Breast milk was, for the first time reported, also clearly positive for IgA, which may have contributed to limiting the spread of disease by the respiratory route because the child was breast-fed for >5 months. Discussion. Parvovirus B19 infection in utero can be lethal and is characterized by severe anemia and high output cardiac failure. 1 Viral myocarditis may also contribute to the pathogenesis of the observed hydrops fetalis. Recent studies of pregnant women with confirmed parvovirus B19 infection have found a low rate of hydrops (0 to 1.6%) in the involved fetuses. 4–6 The risk of adverse fetal outcome is the greatest during the first 16 weeks of gestation. 7 Among series of hydropic fetuses, the case fatality rate may be close to 50%, with the survivors showing no hematologic abnormalities or congenital defects. 5, 8, 9 In view of this, Schild et al. 10 found that intrauterine transfusion with packed red blood cells may prove beneficial even in severely anemic fetuses lowering the mortality rate to 18%. Hematologic evaluation has revealed anemia in all and thrombocytopenia in most, of the examined fetuses. 8–11 Surprisingly no case of congenital anemia has been reported when studying the outcome of series of hydropic fetuses. 5, 8, 9, 11 Previous reports have highlighted the role of parvovirus B19 as a diagnostic pitfall among children with hematologic abnormalities. We have reported on the ability of parvovirus B19 infection to mimic Diamond Blackfan anemia, a variant of congenital anemia. 12 The parvovirus B19-positive patients were the only children among the patients diagnosed with Diamond Blackfan anemia who experienced a long-lasting remission. Only a few additional cases of congenital anemia associated with parvovirus B19 infection have been published. We described three infants with hydrops, hypogammaglobulinemia and congenital anemia caused by transplacental parvovirus B19 infection who were treated with IVIG. 13 In all three the sera lacked parvovirus B19 DNA, but viral DNA was found in bone marrow. One infant died and the other two remained persistently anemic despite continued IVIG treatment. Recently an infant developed congenital anemia caused by a possible parvovirus B19 infection. 14 At the age of 10 months IVIG therapy was commenced (1 g/kg every 3 weeks), and 8 months later the anemia gradually resolved. In the present case parvovirus B19 infection caused hydrops and severe intrauterine anemia, followed by prolonged congenital anemia. Our patient represents the first case of parvovirus B19-associated congenital anemia in which intrauterine blood and platelet transfusions, combined with postnatal IVIG treatment was associated with improvement. In other reported cases of congenital anemia caused by parvovirus B19, DNA could be detected only in the bone marrow. In contrast we identified virus in serum samples by PCR examination. Urine also contained parvovirus B19-specific DNA, whereas patient saliva and maternal breast milk were DNA-negative. This may be significant to consider when evaluating the risk of infection to nursing staff and social contacts. The disappearance of IgM and the detection of IgA and IgG with a declining titer is further evidence of a recent infection. 15 Given the pathophysiology of parvovirus B19 infection and the derived number of cases of hydrops, the apparently low risk of associated congenital anemia is puzzling. This may be explained by parvovirus B19 causing severe disease in only the first two trimesters. Ordinarily infection probably leads to one of two outcomes: lethal hydrops or a milder course of disease in which the virus is eradicated and the ill effects are ameliorated before term. However, the paucity of cases of congenital anemia may also reflect underdiagnosis. Either way the risk of congenital anemia as a result of intrauterine parvovirus B19 infection cannot be determined from the current literature. In conclusion, based on previous reports combined with our findings, active treatment consisting of pre- and postnatal transfusions and IVIG may be warranted in cases of hydrops and anemia caused by parvovirus B19. ELISA is helpful as for diagnosis, whereas PCR can be used when monitoring the time course of viremia. Acknowledgments. We thank Jesper Christensen, Ph.D., for expert technical advice and Carol Vogelius for critical review of the manuscript. Dr. H. Kerzel Andersen and Dr. P.C. Grauballe generously supplied us with additional stored clinical samples. T. Dilling-Hansen assisted in preparing the figure.