ABSTRACT
Dengue virus infection elicits immune responses to multiple viral antigens including antibodies to dengue non-structural protein 1 (NS1) which are rapidly induced and detected within days of infection. The recombinant, live, attenuated, tetravalent dengue vaccine (CYD-TDV; Sanofi Pasteur) uses the yellow fever vaccine virus as a back-bone but expresses dengue virus pre-membrane and envelop proteins. Since CYD-TDV does not express dengue NS1, we evaluated the utility of dengue NS1-specific IgG antibodies as biomarkers of dengue exposure in CYD-TDV recipients and controls. We optimized and evaluated a quantitative anti-dengue NS1 IgG enzyme-linked immunosorbent assay (ELISA). Parameters assessed included: accuracy, dilutability/linearity, precision, limit of quantitation and specificity. The assay specificity was further evaluated using Japanese Encephalitis virus, West Nile virus, Yellow Fever virus or Zika virus positive sera samples collected following confirmed infection or vaccination. Receiver-operating-characteristics (ROC) curves as well as sensitivity and specificity for discriminating previous dengue exposure were assessed using 1250 reference samples. Overall, the anti-dengue NS1 IgG ELISA was able to discriminate previous dengue exposure from non-exposure before vaccination with CYD-TDV (ROC area under the curveâ¯>â¯0.9). Assessment of paired samples from 2511 vaccinated participants showed high overall agreement (93%) between pre-vaccination and post-vaccination dengue serostatus classification based on the anti-dengue NS1 IgG ELISA. However, misclassification of dengue serostatus was observed after vaccination likely due to a combination of asymptomatic dengue infections, assay variability and a modest effect of CYD-TDV on the anti-dengue NS1 IgG ELISA readout.