Role of the dengue vaccine TAK-003 in an outbreak response: Modeling the Sri Lanka experience

by LakKumar Fernando, Randee Kastner, Pujitha Wickramasinghe, Asvini D. Fernando, Dulanie Gunasekera, Van Hung Nguyen, Mengya Liu, Inge LeFevre, Derek Wallace, Nicolas Folschweiller, Shibadas Biswal

Background

Outbreaks of dengue can overburden hospital systems, drastically reducing capacity for other care. The 2017 dengue serotype 2 (DENV-2) outbreak in Sri Lanka coincided with vaccination in an ongoing phase 3 efficacy trial of a tetravalent dengue vaccine, TAK-003 (NCT02747927). Here, we present data on the efficacy of TAK-003 following two doses of the vaccine administered 3 months apart in participants aged 4–16 years in Sri Lanka. In addition, we have used the 2017 outbreak dynamics to model the potential impact of TAK-003 on virologically confirmed dengue (VCD) cases and hospitalizations during an outbreak situation.

Methodology/principal findings

Modeling was performed using an age-structured, host-vector, spatial and stochastic transmission model, assuming 65% vaccine coverage and 30 days until initiation of vaccination. Efficacy of TAK-003 against VCD and hospitalized VCD cases was based on data against DENV-2 from the first year of the phase 3 trial. Vaccine efficacy and safety findings in Sri Lanka were in line with those of the overall trial population. The efficacy estimates in Sri Lanka up to the first 12 months after the second dose of TAK-003 were 94.7% and 95.7% against VCD and hospitalized VCD cases, respectively. Modeling of the trial data over an extended geographic area showed a substantial reduction in cases and a flattening of outbreak curves from TAK-003 use. The baseline vaccination scenario (initiation at 30 days, 65% target coverage, vaccine effective at 14 days, 70% hospitalization rate, VE of 95% for VCD and 97% for hospitalized VCD, and 47% for asymptomatic) resulted in a 69.1% reduction in VCD cases and 72.7% reduction in VCD hospitalizations compared with no vaccination. An extreme high scenario (vaccination initiated at Day 15, 80% coverage rate, baseline VE) resulted in 80.3% and 82.3% reduction in VCD and VCD hospitalizations, respectively. Vaccine performance, speed of vaccination campaign initiation, and vaccine coverage were key drivers in reducing VCD cases and hospitalizations.

Conclusions/significance

Overall, the study and modelling results indicate that TAK-003 has the potential of meaningful utility in dengue outbreaks in endemic areas.