Abstract
Background Dengue human infection models (DHIMs) are important tools to down-select dengue vaccine candidates and establish tetravalent efficacy before advanced clinical field trials. We aimed to provide data for the safety and immunogenicity of DHIM and evaluate dengue vaccine efficacy. Methods We performed an open-label, phase 1 trial at the University of Maryland (Baltimore, MD, USA). Eligible participants were healthy individuals aged 18-50 years who either previously received a tetravalent dengue purified inactivated vaccine prime followed by a live-attenuated vaccine boost (ie, the vaccinee group), or were unvaccinated flavivirus-naive participants (ie, the control group). Participants in the vaccinee group with detectable pre-challenge dengue virus-1 neutralising antibody titres and flavivirus-naive participants in the control group were inoculated with dengue virus-1 strain 45AZ5 in the deltoid region, 27-65 months following booster dosing. These participants were followed-up from days 4-16 following dengue virus-1 live virus human challenge, with daily real-time quantitative PCR specific to dengue virus-1 RNA detection, and dengue virus-1 solicited local and systemic adverse events were recorded. The primary outcomes were safety (ie, solicited local and systemic adverse events) and vaccine efficacy (ie, dengue virus-1 RNAaemia) following dengue challenge. This study is registered with ClinicalTrials.gov, number NCT04786457. Findings In January 2021, ten eligible participants were enrolled; of whom, six (60%) were in the vaccinee group and four (40%) were in the control group. Daily quantitative PCR detected dengue virus-1 RNA in nine (90%) of ten participants (five [83%] of six in the vaccinee group and all four [100%] in the control group). The mean onset of RNAaemia occurred on day 5 (SD 1<middle dot>0) in the vaccinee group versus day 8 (1<middle dot>5) in the control group (95% CI 1<middle dot>1-4<middle dot>9; p=0<middle dot>007), with a trend towards reduced RNAaemia duration in the vaccinee group compared with the control group (8<middle dot>2 days vs 10<middle dot>5 days; 95% CI -0<middle dot>08 to 4<middle dot>68; p=0<middle dot>056). Mild-to-moderate symptoms (nine [90%] of ten), leukopenia (eight [89%] of nine), and elevated aminotransferases (seven [78%] of nine) were commonly observed. Severe adverse events were detected only in the vaccinee group (fever >= 38<middle dot>9 degrees C in three [50%] of six, headache in one [17%], and transient grade 4 aspartate aminotransferase elevation in one [17%]). No deaths were reported. Interpretation Participants who had tetravalent dengue purified inactivated vaccine prime and live-attenuated vaccine boost were unprotected against dengue virus-1 infection and further showed increased clinical, immunological, and transcriptomic evidence for inflammation potentially mediated by pre-existing infection-enhancing antibodies. This study highlights the impact of small cohort, human challenge models studying dengue pathogenesis and downstream vaccine development.