Evaluating simian immunodeficiency virus (SIV) in nonhuman primates is an important step in developing effective HIV vaccines. In the past, vaccine regimens were tested in primates by a single high dose of the virus; however, recent studies have begun to use repeated low-dose viral challenges of the primates to more closely mimic viral exposure in natural settings and model how those who are vaccinated may become exposed to the virus. To better determine the design parameters of the low-dose challenge primate studies, VIDI members Dr. Steven Self and Dr. Peter Gilbert, together with colleagues at Harvard, University of North Carolina, and National Institutes of Health, conducted simulation studies based on computational models to estimate the statistical requirements for these studies. They found that a total of 50 animals in repeated low-dose challenge studies would be adequate to detect a 50% reduction in the probability of infection per exposure. Furthermore, the power to detect vaccine-induced effects increases with 1) the per-exposure risk of infection in control animals, 2) the fraction of animals able to be infected, and 3) the maximum number of times the animal is challenged. Importantly, this work identifies the best statistical methods for the analysis of this type of study design. These parameters will aid in the design and analysis of future studies using repeated, low-dose challenges.
Power to Detect the Effects of HIV Vaccination in Repeated Low-Dose Challenge Experiments. Hudgens MG, Gilbert PB, Mascola JR, Wu CD, Barouch DH, Self SG. J Infect Dis. 2009 Aug 15; 200(4):609-613.