Enteroviruses are well-recognized as causes of severe neonatal illness.¹ Infants lacking transplacentally acquired virus-neutralizing antibodies (VNA) are at increased risk for severe enterovirus disease.² In 1982 immune serum globulin (ISG) was advocated to attenuate neonatal enterovirus infection.² In several nursery outbreaks that followed, conflicting results of ISG prophylaxis have been reported showing the attenuation of enterovirus infection^{3, 4} or the lack of protection.⁵ The use of intravenous immunoglobulin (IVIG) has been reported in a neonate who survived severe echovirus 11 disease.⁶ We report the prophylactic use of IVIG during a nursery outbreak of echovirus type 4 and 6 infections in an attempt to attenuate symptomatic echovirus infection.

Outbreak description. The outbreak occurred in a regional nursery unit near Belgrade in August, 1993. The nursery serves a population of 80 000. Total deliveries in 1993 numbered 1100 with a rate of premature births of 5.6% and perinatal mortality rate of 15 per 1000 births. The nursery had only 2 small rooms (area of 62 m²) with 30 open cribs and 1 isolette. Mothers were hospitalized apart from their babies in 6 rooms on the same floor. Hygienic conditions were poor with only 1 bathroom. The average length of stay in the nursery was 6 days (10 days for babies delivered by cesarean section). All premature babies were routinely referred to the specialized neonatal unit. Two nurses and one physician were the only staff during a day or night shift. The nursery was ultimately closed after this outbreak, and a new nursery was opened in November, 1994.

A febrile pregnant women was admitted on August 4 and 4 days later gave birth to a full term baby. Her baby became febrile on August 12 and died 1 day later. Sepsis was considered the clinical diagnosis. Virologic studies were not performed. Between August 12 and August 16 seven infants developed nonspecific febrile illness. Enterovirus was first suspected to be the cause on August 17, when a 5-day-old boy died from sepsis-like disease, and was confirmed with postmortem echovirus 6 isolation from cerebrospinal fluid (CSF) and blood specimens. The nursery was closed on August 19, 1993. Control measures were immediately instituted and the risk of hand-borne transmission of infection was explained to the staff. Fifty-eight babies were born between August 4 and August 19. Thirty infants were discharged after an average 6-day stay in the nursery. Two infants developed symptoms of infection at home and were readmitted to the general pediatric ward. Two babies with severe echovirus 6 disease were referred to the neonatal intensive care unit at the Mother and Child Health Institute, Belgrade.

Methods. Stool and throat swab specimens for viral cultures were obtained from all infants (and their mothers) born between August 4 and August 19 and repeated twice weekly thereafter. Specimens from the same sites were also taken from the personnel. CSF viral cultures were obtained in symptomatic infants. Conventional virus isolation techniques based on a primary monkey kidney cell line were used. Enteroviruses were identified by neutralization with serotype-specific antisera obtained from reference enterovirus strains (Statens Seruminstitut, Copenhagen, Denmark). Paired sera were taken from the infants and their mothers for VNA titer determination by a microneutralization test. Infection was confirmed by positive viral culture and/or a 4-fold VNA titer rise in convalescent sera. Severe enterovirus disease was defined as serious liver, heart or brain involvement with or without disseminated intravascular coagulation.¹ Mild disease was defined as fever >38.5°C with or without gastroenteritis. Bacterial cultures of blood, CSF and urine were performed by standard methods.

Before this outbreak isolation of an echovirus 6 and echovirus 4 was repeatedly reported in the community. On August 21 the regional WHO laboratory for enteroviruses reported that an enterovirus could be a possible cause of this outbreak. On the same day we decided to use a lot of IVIG (Sandoglobulin®, Sandoz Pharma Ltd., Basel, Switzerland) containing a known VNA titer of 1:32 to echovirus 6 only. Prophylactic IVIG was given in a single dose of 400 mg/kg of body weight to 21 asymptomatic infants. Informed consent was obtained from parents. The other 11 infants who had become symptomatic before IVIG administration or developed symptoms at home and in whom echovirus 6 infection was proved subsequently became the control group. The virulence of the 2 echovirus strains could have been different which would influence the results. For that reason 9 infants in whom echovirus 4 infection was proved were not included in control group. Clinical follow-up ended on September 15 and virologic studies were completed in October, 1993.

Results. Virologic studies included 57 infants (1 pair of twins) and 53 mothers. None had detectable VNA (titer >1:8) to echoviruses types 4 or 6 in the first sera samples. Twenty-nine infected infants were identified of whom 20 had echovirus 6 infection and 9 had echovirus 4 infection, respectively. Echovirus 4 infection was confirmed in 4 mothers. The mother who delivered the probable index patient had a 4-fold VNA titer increase to echovirus 4. This infant was included in a study because viral cultures were not obtained. Viral cultures taken from the staff remained negative. The source of an echovirus 6 was not identified.

Among 21 asymptomatic infants who received IVIG, echovirus 6 infection was confirmed in 9 infants (4 had positive stool and throat viral culture; 5 had a 4-fold VNA titer rise at 1:32), and another 5 infants had both positive stool culture and a 4-fold VNA titer rise to echovirus 4. In the remaining 7 infants serology and virology were negative. In the nonprotected (control) group of the 11 infants, an echovirus 6 was isolated in 4 infants (3 had positive CSF and stool viral cultures; 1 had positive stool viral culture) whereas 7 infants had a 4-fold VNA titer rise at 1:32 in paired sera. Among 9 infants with echovirus 4 infection, meningoencephalitis was diagnosed in 4 infants, mild disease in another 4 and 1 infant remained asymptomatic. Five infants with echovirus 4 infection who were given IVIG had a post-IVIG VNA titer to echovirus 6 of 1:8. As reported previously with the use of ISG, we have also found that the use of IVIG did not significantly alter titer determination of neutralizing antibody in convalescent sera.⁷

Three of the 9 infants with echovirus 6 infection who were given IVIG developed mild disease 12 to 18 h after IVIG administration and the other 6 infants remained asymptomatic. In the control group of 11 infants severe echovirus 6 disease occurred in 3 (1 lethal outcome from meningoencephalitis and 1 from liver failure with fulminant disseminated intravascular coagulation) whereas the remaining 8 infants developed mild disease. Of the infants who developed echovirus 6 disease after IVIG, the proportion with severe disease was different from that of infants who became ill before IVIG (0 of 5 vs. 3 of 9, respectively) (Fig. 1). The risk for symptomatic echovirus 6 infection was 5% for infants who received IVIG vs. 19% for infants in the control group. Attack rates for echovirus 6 infection were 15% for infants who received IVIG, 19% for the control group and 34% overall during this outbreak.

Discussion. Several of the most prevalent strains of enteroviruses circulate during the enterovirus season in temperate climates.¹ Nursery outbreaks are usually caused by a single strain. This nursery outbreak is unusual in that it was concurrently caused by echovirus types 6 and 4.

Our impression is that IVIG may attenuate enterovirus infection when it is given early in the course of infection. In our 3 patients symptoms of mild echovirus 6 disease occurred soon after IVIG prophylaxis, suggesting that IVIG was given at the time of advanced incubation, whereas the other 6 infants who received IVIG remained asymptomatic. Moreover low risk for symptomatic echovirus 6 infection (5%) with the use of IVIG in our study indicates that IVIG may be protective from clinical illness. Nagington et al.³ reported previously that the use of ISG during an echovirus 11 nursery outbreak had been protective against severe disease. By contrast Kinney et al.⁵ observed a similar frequency of severe disease during an echovirus 11 nursery outbreak in both ISG-protected and control groups, although it was possible that ISG was given too late to modify infection. Selected lots of IVIG may be more effective than conventional immunoglobulin, providing the larger amount of specific antibodies at the time of minor viremia and before the involvement of susceptible organs.

Immunoglobulin has been advocated for passive immunization of neonates at increased risk for enterovirus infection^2-4 and we administered IVIG only to asymptomatic infants. Risk factors for severe enterovirus disease include transplacental transmission of virus and undetectable maternal antibody.^{1, 3} Factors associated with the virus also influence the frequency and severity of infection. Echovirus 11 has been most frequently reported as the cause of severe disease,⁸ and other serotypes such as echovirus 6 have caused fatal neonatal disease.⁹ Johnston et al.⁶ reported a good outcome of disseminated echovirus 11 infection in a high risk neonate without detectable neutralizing antibodies who was treated with the high dosage IVIG of 1 g/kg. More recently Abzug et al.¹⁰ reported the 11 symptomatic infants with enterovirus disease who, after being given IVIG at the dose of 750 mg/kg containing a high neutralizing titer of ≥1:800, had a more rapid cessation of viremia and viruria compared with controls. The same authors observed faster disappearance of irritability, jaundice and diarrhea in infants who received IVIG but there was no effect on clinical outcome. We suggest, in absence of specific antiviral agent, that the effectiveness of IVIG in treatment of neonatal enterovirus disease should be further evaluated.

It has also been suggested that passive protection with ISG could prevent a viral spread during nursery outbreaks.¹¹ By contrast the set of twins who received IVIG in our study had the first positive stool cultures for echovirus 6 twenty-four hours later. Both infants remained asymptomatic as they continued to excrete an echovirus 6 during the next 4 weeks. Similar attack rates for echovirus 6 infection for both the control and IVIG groups point out that IVIG did not control viral transmission within nursery. Oral administration of IVIG to decrease viral shedding has not been attempted during an echovirus outbreaks and might be a more rational method. Cohorting of infants with identified enterovirus infection and attention to hand washing are still extremely important control measures.^{5, 8, 12} Closing of the nurseries for new admissions has not been recommended⁸ but is a reasonable measure during nursery outbreaks in developing countries.

Srdjan Pasic, M.D., M.Sc.

Borisav Jankovic, M.D., Ph.D.

Mario Abinun, M.D., Ph.D.

Borisava Kanjuh, M.D., Ph.D.

Mother and Child Health Institute; Belgrade, Yugoslavia (SP, BJ)

Department of Pediatrics; Newcastle General Hospital; RVI and Associated Hospitals; Newcastle upon Tyne, United Kingdom (MA)

Institute for Virusology and Immunobiology; Torlak; Belgrade, Yugoslavia (BK)