Figure 1

Dissecting direct from indirect effects of vaccination by comparing with irrelevant immunization. Each square or trapezoid (placebo controls in red, vaccinees in blue) represents the risk of infection on the y-axis. Ab response to the HIV-1 (vaccinees) or irrelevant immunogen (both groups) is represented on the x-axis. The vaccine gives zero net protection: the areas of the squares and trapezoids are identical. A. Strong Ab responses to HIV-1 immunogens reduce and weak ones increase risk. The risk of infection does not vary with the response to an irrelevant immunogen for the placebo cases. B. The risk varies similarly with Ab responses to the irrelevant immunogen in placebo and vaccine groups. The associated causal factor could be an aspect of innate immunity or a genetically determined influence on the susceptibility to infection. C. The stronger the antibody response to the vaccine, the lower the infection risk (light blue trapezoid). This protective effect is counteracted by an infection-enhancing effect (dark blue), e.g. inflammation or greater proliferation of target cells upon viral exposure (cf. [22]). Since both effects are caused by the HIV-1 immunogens, the infection risk does not vary with the irrelevant Ab response in the placebo group. D. The Ab responses are indirect markers of a vaccine-induced protective effect that is not mediated by antibodies. As in A, a high HIV-1-specific Ab response is associated with a reduced risk of infection, a low response with an increased risk. As in B these Abs are not causally responsible. Unlike in B, however, they are markers of an immune factor that is induced by the vaccine. The infection risk, therefore, does not vary with the irrelevant Ab response among the placebo controls, but it does vary with both the irrelevant and the HIV-1-specific Ab responses among the vaccinees.