Session:

Virology and Viral Infections (Non-HIV)

Abstract No.:

40.035

Title:

A shorter time interval between first and second dengue infections is associated with protection from clinical illness in a prospective school-based cohort in Thailand

Author(s):

K. Anderson¹, R. V. Gibbons², A. L. Rothman³, R. Berkelman⁴, T. P. Endy⁵; ¹Rollins School of Public Health , Department of Epidemiology, Atlanta, GA/US, ²Armed Forces Research Institute of Medical Sciences, Bangkok/TH, ³University of Rhode Island, Institute for Immunology and Informatics, Providence, RI/US, ⁴Rollins School of Public Health, Epidemiology, Atlanta, GA/US, ⁵State University of New York at Syracuse, Department of Infectious Diseases, Syracuse, NY/US

Abstract:

Background: Despite the strong association between secondary dengue (DENV) infections and dengue hemorrhagic fever (DHF), the majority of secondary infections are asymptomatic or dengue fever (DF).  The determinants of the clinical severity of secondary infections remain unclear, though some studies have suggested a possible titer-dependent and time-dependent role of cross-protective dengue DENV antibodies. We investigate the association between the time interval separating sequential DENV infections and clinical severity and whether, among individuals with the same interval between infections, there were immunological differences that were associated with disease severity.
Methods: We used data from two phases of a prospective cohort study to detect asymptomatic and symptomatic DENV infections in school-children in Kamphaeng Phet, Thailand, conducted from 1998 to 2002 and 2004 to 2007.  Children who experienced at least one DENV infection during their enrollment were selected as the population for analysis. 
Results: 1696 children had at least one DENV infection detected during their enrollment and 268 of these children had two DENV infections detected.  A shorter time interval between the first and second DENV infections detected in the cohort was associated with an increased probability of asymptomatic infection.  The association was strongest in children who were seronegative for DENV-1 – DENV-4 by hemagglutination inhibition (HI) assay at enrollment  (average interval separating sequential infections of 2.6 years for DHF, 1.9 years for DF, and 1.6 years for asymptomatic infections, p=0.01 by exact Wilcoxon rank statistic).  In the final model combining time since first observed infection and the magnitude of the antibody response to first infection, the highest probability of being asymptomatic was observed in individuals who experienced their second infection at shorter intervals after the first infection and with a higher titer HI antibody response generated to the first infection.
Conclusion: These findings are consistent with a temporal/immunological model of disease risk where cross-reactive antibodies wane from higher-titer, protective levels to lower-titer, detrimental levels.  This is the first time that a temporary window of cross-protection following DENV infection has been demonstrated using human infection data since early observations from human challenge studies in the 1940s.