In Depth

Parvovirus B19 and Intrauterine Death

Richard M. Pauli, M.D., Ph.D.

Fifth Disease. Erythema infectiosum, or fifth disease, is a common illness of childhood. It comes by the curious appellation of ‘fifth disease’ since it was the fifth of the six childhood viral exanthems. In children it is a usually benign process; in this way and in the fact that it may cause serious fetal injury, it shares certain properties with rubella (German measles).

Children are most often affected between about 5 and 15 years of age. Infection in children often begins with a quite mild prodrome of sore throat, malaise and a low fever. This is followed by a characteristic ‘slapped cheek’ facial rash which then gives way to a more generalized lacy exanthem most frequently over the arms, upper legs and buttocks which persists for another 3-6 days.

Some children have simultaneous joint symptoms, but joint complaints are far more frequent in adults who are affected. Indeed as many as 75% of infected adults may have at least mild, transient, symmetric polyarthritis. Most other symptoms are less severe in adults and, so, the infection itself will frequently go unrecognized.

Parvovirus B19. The cause of fifth disease remained undiscovered until 1974 when it was demonstrated that the clinical process was caused by a human-specific parvovirus.

Parvoviruses are single stranded DNA viruses which have certain general properties that are relevant to their intrauterine effects. First, all parvoviruses require actively replicating cells in order to grow. This would suggest that the fetus is particularly at risk since the embryo and fetus have an extraordinarily large number of cells which are actively dividing. Secondly, parvoviruses often cause cytolysis of the dividing cells which they infect. This in part explains the pathologic changes which they cause in the embryo and fetus. Thirdly, they all are rather species specific. So, while other parvoviruses infect other animal species (including cats and dogs along with rats, mice, hamsters, pigs, mink, cattle, ferrets, horses etc.) humans can be infected only by human parvoviruses with which they make contact through other humans.

Epidemiology of Parvovirus B19. Infection with parvovirus B19 is common. Indeed, at least 50% of adults in most populations are immune. That high level of immunity reflects how common antecedent infections really are.

Infections with parvovirus B19 may occur year round but usually peak in the late winter and early spring. Mini-epidemics, particularly at that time of year, are associated with mutual close contact among many children, such as in preschools and schools.

Children who have become infected with parvovirus B19 are most contagious during the viremic phase. Unfortunately viremia develops before any outward signs of the infection are obvious. That means that efforts at controlling contact are generally unsuccessful. The virus has an incubation period of about 4-14 days in both children and susceptible adults.

During outbreaks around 50% of all susceptible exposed household contacts will become infected. Similarly around 20-25% of other close contacts, such as teachers, healthcare providers, daycare workers etc., if susceptible, will become infected. Curiously, cafeteria workers seem to be at greater risk than anyone other than parents that they will become infected during a local epidemic. Finally, within a community in which an outbreak is occurring, around 5% of susceptible individuals not having those close contacts with infected children will, nevertheless, become infected.

Transmission is primarily through contact with mucous secretions (and, so, like many viral respiratory infections B19 probably is spread both by aerosol droplets and by nose-to-hand-to-hand-to-nose contact). The latter suggests that good hand washing is relevant in limiting the likelihood of infection in contacts.

Parvovirus B19 and the Fetus. Parvoviruses can cross the placenta, can infect the developing fetus and so, at least theoretically, could cause harm. The first recognition of an association between parvovirus infection and intrauterine death was not in humans but in other animal species. The first apparent association of this sort in humans was reported in 1984.

Now there is no doubt that parvovirus B19 infections of a mother sometimes results in adverse fetal outcomes. There is a demonstrable increased risk of miscarriage, stillbirth, nonimmunologic fetal hydrops and intrauterine growth retardation associated with maternal infection. Parvovirus B19 likely does not cause specific congenital anomalies (although this, too, has been anecdotally suggested). On the other hand, epidemiologic studies suggest that, overall, very few stillbirths are caused by parvovirus B19 — probably less than 1% of all stillbirths — and that perhaps only 5% of all instances of nonimmune fetal hydrops are associated with such an infection (although this latter figure may be high enough to justify a search for B19 in hydropic stillborns in whom no other demonstrable cause of the hydrops has been found; in such circumstances unfixed tissues could be frozen for subsequent molecular testing.)

Parvovirus B19 appears to cause harm most often if maternal infection occurs between about 6 and 26 weeks of gestation. Probably this is true because this is a time of marked increase in red cell numbers within the fetus, which cells are the most sensitive to this virus. The virus preferentially infects red blood cell precursors. As a result of the infection, some of these cells may undergo hemolysis. More importantly, it seems the parvovirus causes a maturational arrest in the red cell precursors which, in turn, precipitates development of severe anemia over the next 1 to 6 weeks. Then changes in fluid compartmentalization and heart failure (in part secondary to the severe anemia and in part because of direct myocardiopathic effects of the virus) can lead to death.

Quantifying Risks. It has been quite challenging to arrive at reasonable estimates of risk of parvovirus B19 exposure during pregnancy. Initially, case studies suggested exceedingly high risks. More carefully controlled observations suggest that those risks are much more modest than originally estimated. Overall, the following seem to be reasonable estimates of those risks.

So, overall, the probability that a fetus whose mother is exposed to parvovirus B19 during an epidemic outbreak will die secondary to such exposure is perhaps in the vicinity of 0.4% or 1 in 250. Other estimates made in a similar way have ranged only as high as around 2%.

Given this rather low overall risk, it does not seem reasonable for women to be uniformly excluded from the workplace, even if that is a school or a daycare center, during a parvovirus outbreak.

If a Pregnant Woman is Exposed or May Be Exposed. First the low level of actual risk should be emphasized.

Secondly, immune status can be ascertained through testing offered by the State Laboratory of Hygiene. If the woman is immune, certainly nothing more need be done.

If she is non-immune, options including the following should be explored. Some may be sufficiently reassured that no action seems necessary. Good hand-washing and other hygienic efforts may significantly reduce risk. Avoidance of close contact with children during an outbreak might be prudent if feasible. One might choose to determine antibody status a couple of weeks after any documented exposures in susceptible women to search for evidence of recent infection.

Should initial or subsequent immunologic studies indicate that the mother has had a recent infection, then serious consideration should be given to using frequent repeat ultrasound (e.g. weekly) which might allow for early identification of hydrops and to assure timely intervention if that ends up being needed.

Treatment of Fetal Infections. If a fetus is infected and there is documented development of nonimmune hydrops, treatment with intrauterine exchange transfusion may be the only remaining option. In those in whom this has been successful, follow-up has shown that the liveborn children are healthy and develop normally.

In the Future. Parvovirus vaccines are already in general use for dogs, cats, pigs etc. Studies are now underway to assess whether a similar vaccine program for human parvovirus B19 is prudent policy.

Further reading*:
Enders G, Biber M (1990): Parvovirus B19 infections in pregnancy. Behring Inst Mitt 85:74-78.

Gillespie SM, Cartter ML, Asch S, Rokos JB, Gary W, Tsou CJ, Hall DB, Anderson LJ, Hurwitz ES (1990): Occupational risk of human parvovirus B19 infections for school and day-care personnel during an outbreak of erythema infectiosum. JAMA 263:2061-2065.

Gloning K-P, Schramm T, Brusis E, Schwarz T, Roggendorf M (1990): Successful intrauterine treatment of fetal hydrops caused by parvovirus B19 infection. Behring Inst Mitt 85:79-85.

Jordan EK, Sever JL (1994): Fetal damage caused by parvoviral infections. Repro Tox 8:161-189.

Public Health Laboratory Service Working Party on Fifth Disease (1990): Prospective study of human parvovirus (B19) infection in pregnancy. Br Med J 300:1166-1170.

*Copies of these and other relevant articles are available for personal use by request from WiSSP.