Parsimonious Determination of the Optimal Infectious Dose of a Pathogen for Nonhuman Primate Models

The nonhuman primate (NHP) model is often the best experimental model for testing interventions designed to block infection by human pathogens, such as HIV, tuberculosis, and malaria. A physiological model may require the use of a limiting dose of the infectious agent, where only a fraction of animals become infected upon any given challenge. Determining the challenge dose of the pathogen in such experiments is critical to the success of the experiment: using too-high or too-low a challenge dose may lead to false negative results and an excessive use of animals. Here I define an optimized protocol for defining the dose of pathogen that infects 50% of the time (AID50); other challenge doses, e.g. AID80, can be easily calculated from the same data. This protocol minimizes the number of animals, as well as resources and procedures, while providing an estimate of the AID50 within 1.5-fold of the true value.Author Summary: Exposing nonhuman primates to infectious pathogens (such as tuberculosis, malaria, or the simian equivalent of HIV) is an important model for testing vaccines or other interventions designed to prevent infection or disease. In fact, demonstrating efficacy in animals is often a requirement before clinical testing in humans can be started. A critical variable in such testing is the dose of the pathogen used: this dose should be similar to what humans would encounter. Using too-high a dose may overcome the intervention and mask a successful approach; using too-low a dose may not be relevant. Often, an optimal dose will lead to “successful” infections only a fraction of the times the animal is exposed. A successful intervention experiment therefore needs to use a precisely calibrated dose of the infectious agent; this calibration can only be done by exposing animals to a range of doses and measuring how often they become infected. Here I define the most parsimonious method for performing this calibration: one that uses the least number of animals and procedures. Given the large number of new pathogens being tested in animal models, adoption of such a parsimonious protocol is both economically and ethically warranted, and will thereby enable favorable review of proposed animal use numbers by Institutional Animal Care and Use Committees.