Safety of a tetravalent live dengue virus vaccine in children responding to one serotype only

Dengue virus (DENV) vaccines should be designed to induce balanced protective immunity against all four dengue serotype to mitigate the risk of vaccine-mediated enhanced dengue disease. The first tetravalent vaccine (Dengvaxia) tested in humans was efficacious in children who were partially immune to DENV at baseline. In DENV-naive children, the vaccine was not efficacious and placed some naïve children at risk of experiencing more severe wild-type DENV breakthrough infections. To define dengue vaccine responses at the individual subject level and their relationship to mild and severe dengue infections, we prospectively studied a cohort of DENV-naive children who received one dose of Dengvaxia. The vaccine stimulated variable responses that neutralized 0, 1 (monotypic), or 2+ (multitypic) serotypes in individual children. We used a logistic regression model to evaluate whether vaccine status and serotype-specific NAb status at the end of study period 1 influenced the probability of experiencing a virologically confirmed dengue disease (VCD) case thereafter (months 20 - 60). Vaccinated children with NAb response to only one serotype were at greater risk of being a case compared to the DENV-naïve control group (Odds Ratio 5.07). This risk was not observed in vaccinated children with no NAb or NAb to 2 or more serotypes. We propose that individuals with durable NAb to one serotype have an abundance of serotype cross-reactive, non-neutralizing Abs implicated in the enhanced replication of heterologous serotypes. We discuss the implications of our findings for flagging vaccine candidates that are likely to pose a special risk to seronegative subjects.