Erythema infectiosum (Fifth disease) and pregnancy

David Mankuta, MD; Benjamin Bar-Oz, MD; Gideon Koren, MD, FRCPC

March, 1999

Parvovirus B19 is a small DNA virus and the only strain that is pathogenic in humans. The virus binds to the antigen of the P-system blood group, known as the P antigen.¹ The P antigen is present in various cells and precursor cells, including erythrocytes, erythroblasts, megakaryocytes, and fetal liver and heart cells. Rare groups of people are P-antigen negative and, therefore, are not susceptible to parvovirus B19 infection.²

Epidemiology

Human parvovirus B19 infection occurs worldwide, most commonly among children 5 to 14 years old. The most common manifestation in children is erythema infectiosum (Fifth disease), which typically attacks in the winter and spring. It usually involves the respiratory system, but vertical transmission (mother to fetus) occurs during pregnancy. When vertical transmission occurs during the first 20 weeks of pregnancy, the risk of developing fetal hydrops from hemolytic anemia is about 10%; the rate decreases during the second half of pregnancy.³

The viremia develops approximately 7 days after inoculation and persists for 4 days. The rash typically appears 16 days after inoculation or 5 days after the virus disappears from the blood. Presence of immune globulin G (IgG) and absence of immune globulin M (IgM) in the blood confirms immunity. Up to 50% of adults acquire immunity during childhood, so 50% of women remain susceptible to parvovirus B19 infection during pregnancy. The organism is highly infectious, and up to 50% of susceptible household contacts will become infected when exposed to this childhood disease.³ Schoolteachers who are seronegative and are exposed during an epidemic have a 20% to 30% chance of developing the disease.⁴

Clinical presentation

Diagnosis of children is generally made based on the highly characteristic erythematous "slapped-cheeks" facial rash. The rash usually begins on the cheeks and then spreads to the trunk and limbs. Adults usually do not have a rash; the disease manifests with fever, arthralgia, adenopathy, and mild arthritis, particularly in the wrists, interphalangeal joints, and knees.⁵

Up to 20% of the population have evidence of seroconversion without any clinical manifestations.³ Some have only a mild respiratory illness and no rash. Children with erythema infectiosum are unlikely to be infectious after the rash appears.

Parvovirus B19 can cause transient red blood cell aplasia, which is related to lytic infection in erythrocyte precursors. The aplastic stage lasts about 7 to 10 days and is followed by increased reticulocytosis.⁵ Bone marrow usually recovers within 2 to 3 weeks. The infection sometimes affects other cell lines, including the white cell line and the thrombocytes. Mild anaemia is found in healthy older children and adults, but anemia can be severe in neonates. Other organs, such as the heart, liver, and spleen, can be affected also.⁶

Effects on the pregnancy and the fetus

Despite earlier reports of high rates of vertical transmission and morbidity and mortality, more recent reports demonstrate that, in most cases, no adverse effects on the fetus are evident.^7,8 Rodis et al⁷ reviewed 37 cases of women exposed and infected during pregnancy; 14 (38%) of the pregnancies had adverse outcomes including miscarriage, fetal death, and congenital anomalies.

The vertical transmission rate (confirmed by IgG positivity in children at 1 year of age) is reported as 16% when mothers are infected during the first 20 weeksí gestation and 35% when mothers are infected after 20 weeksí gestation.⁹ In this prospective study, 1610 women were enrolled, and 60 (3.7%) seroconverted during pregnancy. Only 30% of these 60 women reported signs or symptoms of the disease; five had spontaneous miscarriages, but evidence of the virus was found in the fetal tissue of only one of the aborted fetuses. The remaining 55 infected women delivered 56 healthy infants.⁹

Levy et al⁶ reviewed the literature to find rates of fetal loss and nonimmune hydrops fetalis (NIHF) among mothers with serologically proven parvovirus infection. Out of 334 cases, fetal death occurred in 22 (6.6%), and hydrops fetalis in two (0.6%). In a prospective cohort study, Miller et al¹⁰ investigated the risk of fetal loss and congenital abnormalities after maternal parvovirus B19 infection among 427 pregnant women with B19 infection and 367 surviving infants, of whom 129 were followed up at 7 to 10 years of age. Fetal loss occurred only in the first 20 weeks of pregnancy and was around 9%. Seven cases of fetal hydrops followed maternal infection between 9 and 20 weeksí gestation. No abnormalities attributed to B19 infection were found at birth in the surviving infants, and no late effects were found at 7 to 10 years. Most cases of fetal death occurred within 3 to 6 weeks of maternal infection, but one was reported to have occurred as late as 12 weeks after infection.¹¹

Thus, while parvovirus infection during pregnancy can cause miscarriage and hydrops fetalis that can deteriorate to fetal death, in most cases no adverse fetal effects occur. Due to the low incidence of fetal effects, Harger et al¹² questioned the need for serologic and ultrasound surveillance. In their assessment of 618 pregnant women exposed to parvovirus, 52 (8.4%) contracted B19 infection. None of the 52 fetuses of the infected women developed NIHF. Relative risk of maternal B19 infection was 2.8 if the source was a related child living in the household. Risk of B19 infection could not be predicted by pregnant womenís occupations. The authors concluded that excluding pregnant women from the workplace during endemic periods is unjustified and that weekly fetal ultrasound evaluation yields little.

Diagnostic tests

Presence of IgG on enzyme-linked immunosorbent assay (ELISA) indicates previous infection and immunity and, if present in maternal blood, protects mother and fetus from becoming infected. The easiest way to detect infection in healthy people is to evaluate B19 IgM-specific antibody status; its presence confirms infection within the past several months. Pregnant women who are IgG and IgM negative are susceptible to infection and should be counseled to reduce their exposure to sick children, especially if they are schoolteachers or day-care staff, and more so during a Fifth disease epidemic. If they have already been exposed and the childís diagnosis has been serologically confirmed, they should be evaluated for immune status. Specific IgM antibodies begin to appear within 3 days of onset of illness and are relatively short-lived, persisting only 30 to 60 days.¹³

Elevated maternal serum a-fetoprotein (MSAFP) has been suggested to indicate development of hydrops.¹⁴ This could serve as an indirect indicator of fetal infection, with elevated levels probably arising from damage to fetal liver cells. The authors speculated that the increase in a-fetoprotein preceded ultrasonographic detection of fetal hydrops by 4 weeks. The sensitivity of this test, however, is unknown because, in several cases, MSAFP level was normal despite severe fetal infection.¹⁵ Elevated MSAFP levels on routine maternal serum screening for Down syndrome and spina bifida should suggest a parvovirus infection rather than an open neural tube defect. An ultrasound study should be able to differentiate between the two conditions.

Parvovirus B19 cannot be cultured on traditional tissue culture media. Torok et al¹⁶ reported use of a polymerase chain reaction (PCR) to diagnose in utero fetal infection with human parvovirus B19. Specimens from fetal fluid samples (amniotic fluid, ascites, pleural effusion, fetal blood) might show viral DNA. Direct identification of viral particles or genome is possible only in the viremic stage.

Fetal tests for IgM are not reliable because IgM appears in fetal circulation only after 22 weeksí gestation. Proven DNA isolation has been associated with negative results of antibody studies.¹⁷ Electron microscopy might be able to identify viral DNA particles, and viral B19 antigens can be detected by radioimmunoassay or enzyme immunoassay, but these methods are generally insensitive.

Surveillance and treatment

Most patients require supportive treatment only. Pregnant health care workers or teachers should be informed of potential risk from B19 infections and about preventive measures to reduce risk of infection (although this is still controversial, based on the low incidence of adverse fetal effects). Neither antiviral medications nor a vaccine are available yet.

Fetuses infected by parvovirus B19 should be referred to tertiary care centres experienced in fetal cord blood sampling (cordocentesis) and fetal blood transfusion. A more aggressive approach favours cordocentesis for every fetus with evidence of NIHF. Conservative physicians rely on the fact that the disease is self-limiting and resolves spontaneously in most cases and decry the adverse effects of invasive procedures.

In cases of mild hydrops or with evidence of resolution of hydrops, a conservative approach is reasonable. If hydrops worsens, a diagnostic cordocentesis and fetal blood transfusion should be considered. A reticulocyte count in a fetal blood sample could provide evidence of bone marrow recovery. Currently we have no reliable way to predict prognosis for individual fetuses. Among the 23 cases reviewed by Levy et al,⁶ six (26%) fetuses of mothers managed conservatively died, while four of 18 (22%) treated with intrauterine transfusions died. Outcomes of fetuses with NIHF caused by parvovirus are substantially better than outcomes related to other causes of hydrops. Termination of pregnancy should not be recommended.¹⁸

Unlike other infections during pregnancy (toxoplasmosis, syphilis, rubella, cytomegalovirus), parvovirus is not associated with congenital malformations. Follow up should continue for 12 weeks after diagnosis of maternal infection¹⁸ with biweekly ultrasound and fetal heart rate monitoring. For hydrops fetalis and with evidence of fetal anemia, fetal blood transfusion should be performed. If an affected fetus is older than 34 weeksí gestation, delivery should be considered, although recently, late third-trimester intrauterine transfusions have been advocated to provide fetuses with optimal hematologic conditions before delivery. If there is evidence of intrauterine recovery, it is reasonable to wait for full recovery and not induce labour prematurely.

ANSWER

Pregnant women exposed to parvovirus should be tested for immunity. Susceptible women (IgG-negative) exposed to parvovirus who have not demonstrated evidence of seroconversion can be reassured. Women with evidence of seroconversion are still only at low risk of adverse fetal effects, but fetuses should be closely monitored by ultrasound for development of NIHF. Even if fetuses are affected by the virus, treatments are available and spontaneous recovery is common.

References

1. Brown KE, Anderson SM, Young NS. Erythrocyte P antigen: cellular receptor for B19 parvovirus. Science 1993;262:114-7.
2. Brown CK, Hibbs JR, Gallinella G, Anderson SM, Lehman ED, McCarthy P, et al. Resistance to parvovirus B19 infection due to lack of virus receptor (erythrocyte P antigen). N Engl J Med 1994;330:1192-6.
3. Committee on Infectious Diseases, American Academy of Pediatrics. Parvovirus B19. In: Peter G, editor. Red book. 24th ed. Elk Grove Village, Ill: American Academy of Pediatrics; 1997. p. 383-5.
4. Gillespie SM, Cartter ML, Asch S, Rokos JB, Gary GW, Tsou CJ, et al. Occupational risk of human parvovirus B19 infection for school and day-care personnel during an outbreak of erythema infectiosum. JAMA 1990;263:2061-5.
5. Blacklock HA, Mortimer PP. Aplastic crisis and other effects of the human parvovirus infection. Clin Haematol 1984;13:679-91.
6. Levy R, Weissman A, Blomberg G, Hagay ZJ. Infection by parvovirus B19 during pregnancy: a review. Obstet Gynecol Surv 1997;52(4):254-9.
7. Rodis JF, Quinn DL, Gary GW Jr, Anderson LJ, Rosengren S, Cartter ML, et al. Management and outcomes of pregnancies complicated by human B19 parvovirus infection: a prospective study. Am J Obstet Gynecol 1990;163:1168-71.
8. Anand A, Gray ES, Brown T, Clewley JP, Cohen BJ. Human parvovirus infection in pregnancy and hydrops fetal. N Engl J Med 1987;316:183-6.
9. Gratacos E, Torres PJ, Vidal J, Antolin E, Costa J, Jimenez de Anta MT, et al. The incidence of human parvovirus B19 infection during pregnancy and its impact on perinatal outcome. J Infect Dis 1995;171:1360-3.
10. Miller E, Fairley CK, Cohen BJ, Seng C. Immediate and long term outcome of human parvovirus B19 infection in pregnancy. Br J Obstet Gynecol 1998;105:174-8.
11. Public Health Laboratory Service Working Party on Fifth Disease. Prospective study of human parvovirus (B19) infection in pregnancy. BMJ 1990;300:1166-70.
12. Harger JH, Stuart PA, Koch WC, Harger GH. Prospective evaluation of 618 pregnant women exposed to parvovirus B19: risks and symptoms. Obstet Gynecol 1998;91:413-20.
13. Leads from the MMWR. Risks associated with human parvovirus B19 infection. JAMA 1989;261:1406-8.
14. Bernstein IM, Capeless EL. Elevated maternal serum alpha-fetoprotein and hydrops fetalis in association with fetal parvovirus B19 infection. Obstet Gynecol 1989;74:456-7.
15. Saller DJ, Rogers BB, Canick JA. Maternal serum biochemical markers in pregnancies with fetal parvovirus infection. Prenat Diagn 1993;13:467-71.
16. Torok TJ, Wang QY, Gary GW Jr, Yang CF, Finch TM, Anderson LJ. Prenatal diagnosis of intrauterine infection with parvovirus B19 by polymerase chain reaction technique. Clin Infect Dis 1992;14:149-55.
17. Pryde PG, Nugent CE, Pridjian G, Barr M Jr, Faix RG. Severe nonimmune hydrops secondary to parvovirus B19 infection spontaneous reversal in utero and survival of a term infant. Obstet Gynecol 1992;79:859-61.
18. Barrett J, Ryan G, Morrow R, Farine D, Kelly E, Mahony J. Human parvovirus B19 during pregnancy. J SOGC 1994;16:1253-8.