UCLA

Dose-response models are a key component of quantitative microbial risk assessment and can be used to estimate the infectious and lethal doses of novel emerging pathogens to help inform control and prevention measures. Unfortunately, obtaining estimates of infectious and lethal doses in humans can be difficult due to ethical constraints and limited data from experimental challenge studies of relevant animal models such as non-human primates (NHPs). NHP challenge studies tend to have small sample sizes and there are often only one or two dose levels within a single study, which makes dose-response modeling infeasible using data from single studies. Here, by using Bayesian computational methods, we developed an approach to aggregate NHP pathogen load data across multiple challenge studies in order to simultaneously analyze the dose-response relationship and within-host kinetics. We tested our approach by aggregating NHP viral load data across six SARS-CoV-1 challenge studies, and we obtained the first-ever ID50 estimates for SARS-CoV-1 in NHPs. Our work demonstrated the value in reusing previous data from animal experiments, and the modeling framework we developed can be applied to other pathogens, especially in cases where data is limited within individual studies.