PubMed1999
The incidence of anthrax in humans is extremely low. Human vaccine efficacy studies for inhalational anthrax cannot be conducted. The identification of a correlate of protection that predicts vaccine efficacy is crucial for determining the immune status of immunized humans. This surrogate marker of immunity can only be established by using an appropriate animal model. Numerous studies showed that protective antigen (PA) is the principle protective antigen in naturally- or vaccine-induced immunity. However, attempts to correlate the quantity of anti-PA antibodies with protective immunity in the guinea pig model for anthrax and various vaccine formulations have failed. In these studies, we used the licensed anthrax vaccine adsorbed (AVA) in rabbits. Groups of New Zealand white rabbits, 10 or 20 per group, were immunized intramuscularly (two doses, 4 weeks apart) with varying doses of AVA, ranging from a human dose to 1:256 dilution in sterile phosphate-buffered saline (PBS). Control rabbits received PBS/Alhydrogel according to the same schedule. Each rabbit was bled 2 weeks after the second dose, and antibody levels to PA measured by both the quantitative anti-PA IgG ELISA and the toxin-neutralizing antibody (TNA) assay. Rabbits were aerosol-challenged 10 weeks from day 0 with a lethal dose of Ames spores. All the rabbits that received the undiluted and 1:4 dilution of vaccine survived, whereas those receiving the higher dilutions of vaccine (1:16, 1:64 and 1:256) had deaths in their groups. All the controls died. Rabbit survival was compared with the antibody response. Statistical models were used to test for significance of the peak antibody responses to predict survival. Results showed that both the amount of anti-PA IgG and TNA titres present in the sera at the time of the peak antibody response were significant (P < 0.0001) predictors of survival. These results demonstrate that the humoral immune response to AVA can predict protection in the rabbit model of inhalational anthrax.