a Department of Pediatrics and Neonatology, University of Colorado Health Sciences Center, Denver, Colorado
b Department of Pediatrics and Neonatology, Poudre Valley Hospital, Fort Collins, Colorado
c Arboviral Diseases Branch, Division of Vector-Borne Infectious Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Fort Collins, Colorado
| ABSTRACT |
|---|
|
|
|---|
METHODS. Five hundred sixty-six pregnant women who presented to Poudre Valley Hospital (Fort Collins, CO) for delivery between September 2003 and May 2004 provided demographic and health history data through self-administered questionnaires and hospital admission records. Umbilical cord blood was collected from 549 infants and screened for WNV-specific IgM and IgG antibodies with enzyme-linked immunosorbent assays, with confirmation by plaque-reduction neutralization tests. Newborn growth parameters, Apgar scores, and hearing test results were recorded.
RESULTS. Four percent (95% confidence interval: 2.45.7%) of cord blood samples tested positive for WNV-specific IgG antibodies. No cord blood samples were positive for WNV-specific IgM antibodies. There were no significant differences between infants of seropositive and seronegative mothers with respect to any of the growth parameters or outcomes measured.
CONCLUSIONS. Intrauterine WNV infections seemed to be infrequent. In our study, WNV infection during pregnancy did not seem to affect adversely infant health at birth. Larger prospective studies are necessary to measure more completely the effects of maternal WNV infection on pregnancy and infant health outcomes.
Key Words: West Nile virus pregnancy seroprevalence congenital infection perinatal transmission breastfeeding epidemiology
Abbreviations: WNVWest Nile virus PVHPoudre Valley Hospital ELISAenzyme-linked immunosorbent assay PCRpolymerase chain reaction PRNTplaque-reduction neutralization test JEVJapanese encephalitis virus SLEVSt Louis encephalitis virus LBWlow birth weight SGAsmall for gestational age YFVyellow fever virus CIconfidence interval ORodds ratio CDCCenters for Disease Control and Prevention
Since its recognition in New York in 1999, West Nile virus (WNV) has spread across North America and become a major public health concern in the United States.1 WNV is a flavivirus transmitted to humans primarily through the bite of infected mosquitoes. The first documented case of intrauterine WNV infection was reported in 2002, involving an infant who appeared grossly normal at birth but had bilateral chorioretinitis and severe cystic destruction of temporal and occipital cerebral tissue.2,3 In the same year, 4 additional WNV infections during pregnancy were reported. Three infections apparently resulted in no fetal infection or adverse outcomes for the infants; in 1 instance, the otherwise normal premature infant was not tested for WNV infection.1 Also in 2002, probable transmission via breast milk was reported, with no apparent health consequences for the infant.4
In humans, maternal flavivirus infection has been associated with adverse birth outcomes; to date, however, there has been no convincing evidence that flaviviruses cause birth defects. Dengue virus infection during pregnancy can increase the risk of premature delivery and fetal death, and congenital infection has resulted in low birth weight (LBW), prematurity, and neonatal dengue and dengue hemorrhagic fever.57 Maternal Japanese encephalitis virus (JEV) infections have resulted in spontaneous abortions.8,9 Yellow fever virus (YFV) 17-D vaccine administered during pregnancy might slightly increase the risk of spontaneous abortion but has not been associated with birth defects or neonatal illness.10,11 In animals, St Louis encephalitis virus (SLEV), JEV, and WNV infections during pregnancy have been associated with various adverse birth outcomes, including growth retardation, congenital and neurologic defects, and stillbirth.1214
WNV was first detected in Colorado in 2002, when WNV-positive birds and 14 human disease cases were reported. In 2003, Colorado experienced a large WNV epidemic, with 2947 reported human disease cases. Of these, 948 occurred in or near Fort Collins, a community of
125 000 people located 65 miles north of Denver on the Front Range of the Rocky Mountains. A single hospital, Poudre Valley Hospital (PVH), serves Fort Collins and surrounding communities and is the only birthing hospital for the area. The epidemic curve for reported WNV disease in the Fort Collins area (Larimer and Weld counties) is shown in Fig 1 (Centers for Disease Control and Prevention [CDC], unpublished data). In this study, we evaluated the frequency of WNV infection during pregnancy and assessed the risk of intrauterine WNV infection during the 2003 Fort Collins epidemic.
|
| METHODS |
|---|
|
|
|---|
Because WNV-specific IgG may not be detectable for several weeks after infection, mothers who delivered in September and October were asked to provide a serum sample before delivery, to be tested for maternal IgM antibodies to WNV. If recent infection was indicated by WNV-specific IgM antibodies in a maternal serum sample but the infant's cord blood tested negative for both WNV-specific IgM and IgG antibodies, then additional serologic testing was performed for the child at 1 month of age. Mothers who delivered after October were not tested for WNV-specific IgM antibodies because there would have been ample time for the cord blood to show maternal IgG antibodies resulting from WNV infections that occurred near the end of the transmission season (Fig 1). All laboratory tests for WNV and other flaviviruses were performed at the CDC (Fort Collins, CO).
For deliveries that occurred between November and May, we considered negative WNV-specific IgM and IgG results in cord blood as showing no evidence of maternal or fetal WNV infection. Cord blood samples that were positive for only WNV-specific IgG antibodies were considered to indicate maternal infection. Because IgM antibodies do not usually cross the placenta,17 any cord blood samples that were positive for WNV-specific IgM antibodies were presumed to indicate congenital WNV infection.
Data on weight, length, head circumference, hearing, and Apgar scores for all children were collected from hospital records. If WNV-specific antibodies were detected in cord or maternal serum samples, then the infant's individual medical record was reviewed; if the infant demonstrated abnormal physical or hearing examination results, then additional clinical evaluations were performed on a case-by-case basis, following published guidelines.18
Infants were classified as preterm if they were born at <37 full weeks of gestation. LBW was defined as birth weight of <2500 g (5.5 lb). Short stature and small head circumference were defined as measurements >2 SDs below the mean body length and head circumference derived from reference growth parameters for children in the United States.19 Infants were classified as small for gestational age (SGA) if their weight was less than the 10th percentile of the mean birth weight according to gestational age, race, and gender for United States-born children.20
To evaluate relationships between participants with positive and negative WNV serologic tests, with respect to demographic variables,
2 tests were used. Prevalence odds ratios (ORs) were used to measure the strength of association between maternal WNV infection and the neonatal outcomes of preterm birth, LBW, SGA status, short stature, small head circumference, and low Apgar scores. All statistical analyses were performed with SAS version 9.2 (SAS Institute, Cary, NC). This study was approved by the human subjects institutional review board of PVH.
| RESULTS |
|---|
|
|
|---|
|
Serum samples were collected from 184 women in September and October, at the time of delivery. Of these, 5 tested positive for WNV-specific IgM serum antibodies, which suggested that their WNV infections had occurred recently; all of these women also had IgG antibodies to WNV in their serum and in their infant's cord blood (included in the 22 infants described above). No cord blood samples were positive for IgM antibodies to WNV, providing no evidence of congenital WNV infections.
Nine of the 22 WNV-seropositive mothers provided sufficient breast milk or colostrum for antibody tests. One breast milk sample and 1 colostrum sample tested positive for WNV-specific IgM and neutralizing antibodies and were confirmed as having WNV-specific antibodies with PRNTs; both women also had WNV-specific IgM antibodies in their serum. In addition, the single breast milk sample yielded equivocal results for WNV RNA in a PCR assay. These equivocal results could be attributable to a very low level of RNA in the sample or could be a false-positive finding related to the testing of an unconventional sample.
Of 533 women with questionnaire information, 18 (3%) reported having had a blood test for antibodies to WNV during their pregnancy, including 5 (23%) of the 22 women who were seropositive in this study. Eight women (2%) reported being told by a physician that they were infected with WNV, including 3 women who did not report having had a blood test and 2 women who were seronegative in the current study. Of the 22 WNV-seropositive women, 6 (27%) reported being told by a physician that they had a WNV infection. Ninety-four (18%) of 533 women reported having had a fever sometime during pregnancy; 34% of these 94 reported that their fever occurred during summer, 13% during autumn, 40% during winter, and 8% during spring (5% did not answer this question). Of the 22 WNV-seropositive women, 7 (32%) reported having had a fever during pregnancy; 5 (23%) women experienced this symptom during Colorado's WNV transmission season (between June and November). Two hundred ninety (54%) of 533 women reported previous travel outside the United States, including 14 (64%) of the 22 WNV-seropositive women. Twenty women reported prior vaccination against YFV or JEV, including 1 of the WNV-seropositive women.
Characteristics of WNV-seropositive and WNV-seronegative women are compared in Table 2. Self-reported fever during pregnancy was the only variable associated strongly with WNV seropositivity (OR: 2.7; 95% CI: 1.07.1). Other demographic variables, including zip code of residence, were not correlated with WNV seropositivity (Fig 2 and Table 2).
|
|
|
6 weeks of age showed no WNV-specific IgM antibodies and decreasing titer of neutralizing antibodies to WNV. Two other children born with WNV-specific IgG antibodies in cord blood were observed to have abnormalities (unilateral cryptorchidism and spontaneous pneumothorax) at birth that were considered to be unrelated to maternal infection with WNV. | DISCUSSION |
|---|
|
|
|---|
None of the newborn infants in our study had detectable WNV-specific IgM antibodies. However, just as the detection of specific IgM antibodies has limited sensitivity in diagnosing other congenital infections, the sensitivity, specificity, and predictive value of standard serologic tests for WNV in cord blood and newborn serum are uncertain.2326 It is also possible that some congenitally infected infants (such as perhaps those infected during the first trimester) do not produce detectable concentrations of WNV-specific IgM antibodies in response to infection.27 Among normal infants, the total concentration of serum IgM antibodies increases markedly at
6 days of age and does not reach adult levels until
12 months of age.17
Among participants in this study, maternal infection with WNV did not seem to influence the risk of preterm birth, LBW, or infant growth, but the number of infected women was low, limiting the statistical power to detect small differences in these and other outcomes of pregnancy. Questionnaire data indicated that seropositive women (23%) were more likely to report fever during the WNV transmission season (summer or autumn) than were seronegative women (8%). Furthermore, the proportion of asymptomatic infections (77%) reported among this cohort was similar to that published previously.21 Nonparticipating women had more adverse birth outcomes and longer hospital stays than did participants, which suggests that women with uncomplicated conditions on admission to the hospital might have been more likely to participate. It is possible that undetected WNV infections among nonparticipants caused adverse events that were not recorded in this study.
The finding of infants with physical abnormalities born to WNV-infected mothers raises the possibility that maternal WNV infection could cause adverse effects among infants in the absence of laboratory evidence of intrauterine infection. Although not performed in this study, monitoring such infants clinically and measuring WNV-specific IgG antibodies in serum at 6 to 12 months of age, by which time maternal IgG antibodies transferred transplacentally to the infant should be disappearing, might help in further evaluating the possibility of congenital WNV infection. In addition, studies of placental pathologic conditions after maternal WNV infection and dysmorphologic and developmental follow-up monitoring of infants born to WNV-infected mothers might help to define more completely the effects of WNV infection during pregnancy on infant health. Given the design of the current study, enrolling mothers at the time of delivery, we were not able to measure the effects of WNV infection on spontaneous abortion or other predelivery complications of pregnancy.
The results of this study provide some reassurance that WNV infection during pregnancy does not result commonly in congenital WNV infection or serious adverse birth outcomes. Preliminary results from the CDC WNV national pregnancy registry indicated a similarly low incidence of apparent congenital infection.28 However, larger prospective studies will be necessary to evaluate more precisely the effects (some of which may be rare) of maternal WNV infection on pregnancy outcomes and on the health of the fetus and newborn.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
Address correspondence to Jan E. Paisley, MD, Department of Pediatrics and Neonatology, Poudre Valley Hospital, 1024 S. Lemay, Fort Collins, CO 80524. E-mail: pais{at}pvhs.org
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
T. F. Tsai Congenital Arboviral Infections: Something New, Something Old Pediatrics, March 1, 2006; 117(3): 936 - 939. [Full Text] [PDF] |
||||
![]() |
D. R. O'Leary, S. Kuhn, K. L. Kniss, A. F. Hinckley, S. A. Rasmussen, W. J. Pape, L. K. Kightlinger, B. D. Beecham, T. K. Miller, D. F. Neitzel, et al. Birth Outcomes Following West Nile Virus Infection of Pregnant Women in the United States: 2003-2004 Pediatrics, March 1, 2006; 117(3): e537 - e545. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||