Technical Annex 1. The potential impact of prophylactic HIV vaccination as a function of vaccine properties: results of the Imperial College Model.

The influence of a vaccine with a range of different properties has been explored in a mathematical model of the spread of HIV in a heterosexual population reflecting the generalized epidemics observed in sub-Saharan Africa. A number of key conclusions can be derived from this work: 1) A vaccine with a low (i.e. 50%) efficacy can have a significant epidemiological impact. 2) The effectiveness of the vaccine is markedly reduced if its duration of protection is short relative to the age range over which high risk behaviours are maintained. 3) For a rapid turn around in the incidence of HIV infection a ‘catch-up’ vaccination programme amongst sexual active adults would be desirable. 4) The type of protection provided by the vaccine has an influence on the problem of increasing risk behaviour in response to vaccination. If the vaccine reduces the risk of acquiring infection per exposure then its beneficial effects can be overturned. Alternatively if a fraction of the population is protected from all challenges then any increase in risk cannot lead to the infection of the protected individuals and the aggregate effect of the vaccine is unlikely to be undermined by behavioural reversals. 5) Targeting can be the most effective vaccine strategy per vaccinated individual but may not be the most effective use of a vaccine. Targeting is less likely to be appropriate for a ‘leaky’ vaccine where the number of challenges influence the chances of a breakthrough infection. (excerpt)