Maternal antibody and viral factors in the pathogenesis of dengue virus in infants

Abstract

The pathogenesis of dengue in infants is poorly understood. We postulated that dengue severity in infants would be positively associated with markers of viral burden and that maternally derived, neutralizing anti-dengue antibody would have decayed before the age at which infants with dengue presented to the hospital. In 75 Vietnamese infants with primary dengue, we found significant heterogeneity in viremia and NS1 antigenemia at hospital presentation, and these factors were independent of disease grade or continuous measures of disease severity. Neutralizing antibody titers, predicted in each infant at the time of their illness, suggested that the majority of infants (65%) experienced dengue hemorrhagic fever when the maternally derived neutralizing antibody titer had declined to <1 : 20. Collectively, these data have important implications for dengue vaccine research because they suggest that viral burden may not solely explain severe dengue in infants and that neutralizing antibody is a reasonable but not absolute marker of protective immunity in infants.

Figure 1

Decay of maternally derived, dengue-reactive antibody in healthy Vietnamese infants. Shown are the median and interquartile ranges of anti–dengue virus (DENV) IgG in 55 infants at birth (cord blood) and again at 6, 9, and 12 months. The mean half-life of dengue-reactive IgG in each infant was 42 days (range, 21–77 days). All infants were IgM negative at all time points. The dashed line represents the limit of detection in the indirect ELISA, and the values above the symbols represent the percentage of infants with detectable dengue-reactive IgG at each time point.

Figure 2

Maternal neutralizing antibodies and dengue hemorrhagic fever. A, Age of individual infants with primary dengue virus (DENV) at hospital presentation vs. the corresponding maternal plasma 50% plaquereduction neutralization test (PRNT₅₀) titer against the serotype of DENV that infected each infant. Only full-term infants in whom the serotype of the infecting virus was known and whose mothers who did not have evidence of recent DENV infection were included in this analysis (n = 64). B, Predicted PRNT₅₀ titer of maternally derived antibody in the infant at the time of infection vs. the age of individual infants at hospital presentation. The predicted PRNT₅₀ at the time of infection was calculated using the maternal PRNT₅₀ titer and the half-life of maternal antibody as suggested by the change in measles virus–specific IgG in each mother-infant pair. Only full-term infants <9 months of age in whom the serotype of the infecting virus was known and whose mothers who did not have evidence of recent DENV infection were included in this analysis (n = 37). Infants older than 9 months of age were excluded because of the confounding effect of routine measles vaccination at 9 months of age in Vietnam.

Figure 3

Viremia in infants with dengue virus (DENV). Shown are the kinetics of viremia by day of illness in individual infants with dengue hemorrhagic fever (DHF) grade II or dengue shock syndrome (DSS) when caused by primary DENV1 (A and B), primary DENV2 (C and D), or primary DENV3 (E and F). The limit of detection was 650 cDNA equivalents/mL. Patients diagnosed with dengue fever (n = 2) had undetectable viremia.

Figure 4

NS1 antigenemia in infants with dengue virus (DENV). Shown are the kinetics of NS1 antigenemia by day of illness in individual infants with dengue hemorrhagic fever (DHF) grade II or dengue shock syndrome (DSS) when caused by primary DENV1 (A and B), primary DENV2 (C and D), or primary DENV3 (E and F). The limit of detection was 45 ng/mL. Patients diagnosed with dengue fever (n = 2) had undetectable antigenemias.

Figure 5

Plasma viremia vs. NS1 antigenemia in infants with dengue virus (DENV). Shown is the relationship between plasma viremia and NS1 antigenemia in individual infants with primary DENV and at least 1 measurable parameter at all time points assessed. There was no significant correlation between viral load and NS1 concentration.

Figure 6

Relationship between illness history and serological and virological markers of infection. Shown is the proportion of patients with primary dengue virus (DENV) with plasma that tested positive by either polymerase chain reaction (PCR), NS1 antigen detection assay, or IgM or IgG capture ELISA, according to the day of blood sampling.