Reassessment: Qdenga and the seronegative risk signal

My colleague pointed me to a modeling study by Cracknell et al. that I had not seen before my ViVa 29 publication on dengue vaccines and the discontinuation of the Butantan-DV program in Brazil. Summary Imperial College researchers built a Bayesian cohort survival model calibrated to Qdenga’s full published phase III trial dataset (TIDES), linking vaccine-induced neutralizing antibody titers to disease risk across serotype, serostatus, age, and time. They then embedded this model into a dengue transmission model to estimate public health impact. Through 57 months after the first dose, reported VE against DENV-3 in seronegative recipients was −15.5% (95% CI, −108.2 to 35.9), and against DENV-4 it was −105.6% (95% CI, −628.7 to 42.0)—both statistically nonsignificant but with negative point estimates. Hospitalization VE for DENV-3 in seronegative recipients was also negative (−87.9%; 95% CI, −573.4 to 47.6), and vaccinated seronegative recipients experienced more severe/hemorrhagic DENV-3 cases than placebo seronegative recipients. The modeled evidence...