Management and outcomes of pregnancies complicated by human B 19 parvovirus infection: A prospective study John F. Rodis, MD, Dorothy L. Quinn, MS, G. William Gary, Jr., MPH, DrPH, Larry J. Anderson, MD, Sally Rosengren, MD, Matthew L. Cartter, MD, Winston A. Campbell, MD, and Anthony M. Vintzileos, MD Farmington and Hartford, Connecticut, and Atlanta, Georgia During a large statewide outbreak of fifth disease in Connecticut in 1988, 39 pregnant women were identified who had serologic evidence of recent human 819 parvovirus infection. The patients were followed up prospectively with targeted fetal ultrasonographic examinations to detect signs of fetal hydrops. Of these 39 pregnant women, 37 had healthy infants and two patients had miscarriages. None of the fetuses developed hydrops. We propose that pregnant women exposed to 819 parvovirus be tested for evidence of IgG and IgM 819-specific antibodies and that targeted fetal ultrasonography be considered when IgM antibodies are found. Percutaneous umbilical blood sampling and intrauterine transfusion can be considered in cases of 819 parvovirus-associated hydrops and anemia. The overall fetal loss rate in this prospective follow-up group was 5%. (AM J OBSTET GVNECOL 1990;163:1168-71.)
Key words: Parvovirus infection in pregnancy, targeted ultrasonography, hydrops
Intrauterine infection with human B 19 parvovirus has been associated with fetal hydrops and death, '·7 although the outcomes have been normal in the majority of cases. During a large and widely publicized statewide outbreak of erythema infectiosum (fifth disease), the childhood exanthem caused by B19 parvovirus, many pregnant women were exposed to the virus. More than 1000 pregnant women had blood tests for the presence of IgG and IgM parvovirus-specific antibodies. This article pertains to the pregnant women who were identified as having evidence of recent B 19 parvovirus infection (i.e., B 19 IgM-antibody-positive). The purpose of this report is to define the fetal risks associated with maternal B 19 parvovirus infection during pregnancy in a prospective follow-up population and to propose a management protocol for pregnancies complicated by B 19 parvovirus infection. Material and methods During a large statewide outbreak of erythema infectiosum from December 1987 through December From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, and the Division of Medical Genetics, Department of Pediatrics, University of Connecticut Health Center, Farmington; the Epidemiology Section, Connecticut Department of Health Services, Hartford; and the Division of Viral Diseases, Centers for Disease Control. Received faT publication March 6, 1990; revised May 3, 1990; accepted May 24, 1990. Reprint requests: John F. Rodis, MD, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06032.
6/1/22723
1168
1988, many pregnant women contacted their physicians or the Connecticut Pregnancy Information Service at the University of Connecticut Health Center. Most of the women had either direct or indirect contact with a child with erythema infectiosum; many worked in schools with erythema infectiosum outbreaks. Patients were advised to have their blood tested for the presence of IgG and IgM parvovirus-specific antibodies. Testing was performed at the Centers for Disease Control by a previously described" enzyme-linked immunosorbent assay method. Women with serologic evidence of recent infection by B19 parvovirus (i.e., B19 parvovirus IgM-antibodypositive) were advised to undergo targeted fetal ultrasonographic examinations, with special attention given to signs of fetal hydrops (polyhydramnios, placentomegaly, pericardial or pleural effusions, ascites, skin or scalp edema). Ultrasonographic examinations were performed at 1- to 2-week intervals until at least 6 to 8 weeks had elapsed from the date of maternal exposure, when it was known. In cases in which the timing of B 19 parvovirus exposure was ill defined, ultrasonography was performed for 6 weeks from the time the B 19 parvovirus IgM antibody-positive serum was drawn. This report includes only those patients whose pregnancies were uneventful at the time of serologic testing. In other words, to avoid selection bias, patients who had miscarriages who were then sent for B 19 parvovirus testing are not included. Ultrasonography was performed by the Maternal-Fetal Medicine faculty of the University of Connecticut Health Center. Pregnancy outcomes were obtained in all cases; in-
Parvovirus infection in pregnancy
Volume 163 Number 4, Part 1
formation was obtained by the referring physicians or the pediatricians who had examined the newborns. Information obtained included birth weights, Apgar scores, gestational age at delivery, and documentation of major or minor congenital anomalies. In cases of spontaneous abortion fetal and placental tissues were examined histologically. Results
During the study period 39 pregnant women were identified who had recent evidence of B 19 parvovirus infection, all of whom were followed prospectively with fetal ultrasonographic examination. Pregnancy outcomes were available for all women. Thirty-seven (95%) were delivered of healthy newborns, 36 at term and one patient delivered of twins at 35 weeks. None of these fetuses had signs of fetal hydrops. One of the 38 infants had mild hypospadias. Two patients had spontaneous abortions for an overall pregnancy loss rate of 5%. With regard to the two losses, one occurred in a patient with infertility who had suboptimal rises in quantitative /3-human chorionic gonadotropin before her exposure to a child with erythema infectiosum early in the first trimester. She subsequently had falling /3human chorionic gonadotropin values and never required a dilation and curettage; thus no tissue was available for histologic examination. The other spontaneous abortion occurred in a woman exposed to B 19 parvovirus at 10 weeks' gestation. Fetal heart tones were auscultated and serum was tested for B 19 parvovirus antibodies. Two weeks later the patient had a spontaneous abortion. Histologic examination of the placenta revealed severe erythroblastosis. Because of formalin fixation of the products of conception, B19 parvovirus deoxyribunucleic acid probes could not be used to identify viral deoxyribonucleic acid. Table I shows the gestational age at the time of serologic testing in the 39 women. Comment
In most cases of maternal B 19 parvovirus infection the fetus is not adversely affected, whereas in othersfor reasons not well understood-miscarriages, fetal hydrops, and death may occur. The actual incidence of fetal death after maternal infection is not well established because previous studies have been retrospective and subject to selection bias. A retrospective review of all cases reported in the literature suggested a fetal loss rate of 38%,4 whereas a recent report from West Germany found a 26% rate of hydrops in 39 pregnant women. 9 However, an unpublished prospective study from the United Kingdom suggests that the B 19 parvovirus-associated fetal death rate after maternal infection is
Key words: Parvovirus infection in pregnancy, targeted ultrasonography, hydrops
Intrauterine infection with human B 19 parvovirus has been associated with fetal hydrops and death, '·7 although the outcomes have been normal in the majority of cases. During a large and widely publicized statewide outbreak of erythema infectiosum (fifth disease), the childhood exanthem caused by B19 parvovirus, many pregnant women were exposed to the virus. More than 1000 pregnant women had blood tests for the presence of IgG and IgM parvovirus-specific antibodies. This article pertains to the pregnant women who were identified as having evidence of recent B 19 parvovirus infection (i.e., B 19 IgM-antibody-positive). The purpose of this report is to define the fetal risks associated with maternal B 19 parvovirus infection during pregnancy in a prospective follow-up population and to propose a management protocol for pregnancies complicated by B 19 parvovirus infection. Material and methods During a large statewide outbreak of erythema infectiosum from December 1987 through December From the Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, and the Division of Medical Genetics, Department of Pediatrics, University of Connecticut Health Center, Farmington; the Epidemiology Section, Connecticut Department of Health Services, Hartford; and the Division of Viral Diseases, Centers for Disease Control. Received faT publication March 6, 1990; revised May 3, 1990; accepted May 24, 1990. Reprint requests: John F. Rodis, MD, Division of Maternal-Fetal Medicine, Department of Obstetrics and Gynecology, University of Connecticut Health Center, 263 Farmington Ave., Farmington, CT 06032.
6/1/22723
1168
1988, many pregnant women contacted their physicians or the Connecticut Pregnancy Information Service at the University of Connecticut Health Center. Most of the women had either direct or indirect contact with a child with erythema infectiosum; many worked in schools with erythema infectiosum outbreaks. Patients were advised to have their blood tested for the presence of IgG and IgM parvovirus-specific antibodies. Testing was performed at the Centers for Disease Control by a previously described" enzyme-linked immunosorbent assay method. Women with serologic evidence of recent infection by B19 parvovirus (i.e., B19 parvovirus IgM-antibodypositive) were advised to undergo targeted fetal ultrasonographic examinations, with special attention given to signs of fetal hydrops (polyhydramnios, placentomegaly, pericardial or pleural effusions, ascites, skin or scalp edema). Ultrasonographic examinations were performed at 1- to 2-week intervals until at least 6 to 8 weeks had elapsed from the date of maternal exposure, when it was known. In cases in which the timing of B 19 parvovirus exposure was ill defined, ultrasonography was performed for 6 weeks from the time the B 19 parvovirus IgM antibody-positive serum was drawn. This report includes only those patients whose pregnancies were uneventful at the time of serologic testing. In other words, to avoid selection bias, patients who had miscarriages who were then sent for B 19 parvovirus testing are not included. Ultrasonography was performed by the Maternal-Fetal Medicine faculty of the University of Connecticut Health Center. Pregnancy outcomes were obtained in all cases; in-
Parvovirus infection in pregnancy
Volume 163 Number 4, Part 1
formation was obtained by the referring physicians or the pediatricians who had examined the newborns. Information obtained included birth weights, Apgar scores, gestational age at delivery, and documentation of major or minor congenital anomalies. In cases of spontaneous abortion fetal and placental tissues were examined histologically. Results
During the study period 39 pregnant women were identified who had recent evidence of B 19 parvovirus infection, all of whom were followed prospectively with fetal ultrasonographic examination. Pregnancy outcomes were available for all women. Thirty-seven (95%) were delivered of healthy newborns, 36 at term and one patient delivered of twins at 35 weeks. None of these fetuses had signs of fetal hydrops. One of the 38 infants had mild hypospadias. Two patients had spontaneous abortions for an overall pregnancy loss rate of 5%. With regard to the two losses, one occurred in a patient with infertility who had suboptimal rises in quantitative /3-human chorionic gonadotropin before her exposure to a child with erythema infectiosum early in the first trimester. She subsequently had falling /3human chorionic gonadotropin values and never required a dilation and curettage; thus no tissue was available for histologic examination. The other spontaneous abortion occurred in a woman exposed to B 19 parvovirus at 10 weeks' gestation. Fetal heart tones were auscultated and serum was tested for B 19 parvovirus antibodies. Two weeks later the patient had a spontaneous abortion. Histologic examination of the placenta revealed severe erythroblastosis. Because of formalin fixation of the products of conception, B19 parvovirus deoxyribunucleic acid probes could not be used to identify viral deoxyribonucleic acid. Table I shows the gestational age at the time of serologic testing in the 39 women. Comment
In most cases of maternal B 19 parvovirus infection the fetus is not adversely affected, whereas in othersfor reasons not well understood-miscarriages, fetal hydrops, and death may occur. The actual incidence of fetal death after maternal infection is not well established because previous studies have been retrospective and subject to selection bias. A retrospective review of all cases reported in the literature suggested a fetal loss rate of 38%,4 whereas a recent report from West Germany found a 26% rate of hydrops in 39 pregnant women. 9 However, an unpublished prospective study from the United Kingdom suggests that the B 19 parvovirus-associated fetal death rate after maternal infection is
