Giving an effective AIDS vaccine to 70% of the population by 2030 could virtually extinguish the HIV epidemic, and even an AIDS vaccine with a modest level of efficacy provided to high-risk groups could avert one third of the HIV infections expected to occur between 2015 and 2030, analysts from the Futures Group and the International AIDS Vaccine Initiative reported last week at the AIDS Vaccine ’06 conference in Amsterdam.
The study also found that a focused vaccination programme targeting only high-risk individuals in concentrated epidemics (such as Brazil) and the whole population in generalised epidemics (such as South Africa) would have almost as much impact as a vaccination programme that targeted the general population in all countries, and would require less than 50 million vaccinations a year between 2015 and 2030, compared with up to 250 million vaccinations a year during the same period if the broader approach to immunisation were to be employed.
The introduction of an effective AIDS vaccine is likely to pose huge logistical and financial challenges, especially if the first vaccine to be used requires a number of vaccinations spread over a period of one year. During the early years vaccine doses may also be in short supply.
Delivering a vaccine to 70% of the population by 2030 in countries where HIV is an emerging epidemic may not be a realistic prospect, but lower levels of coverage are likely to be achievable if strong networks of services for prevention and treatment have developed by the time that a vaccine is ready for introduction.
The model being developed by the Futures Group and IAVI takes data on the likely course of the epidemic from South Africa, India, Nigeria, Brazil, Mexico, Russia and China, and assumes that countries will reach targets to provide universal access to HIV prevention by 2010 and to antiretroviral treatment and care by 2015. The modelling also assumes that a vaccine will become available in 2015.
Even so, the addition of an AIDS vaccine to AIDS control efforts could avert a further five to 27 million new HIV infections between 2015 and 2030.
Findings
The model shows that a vaccine with only 30% efficacy provided to 20% of the population would reduce the number of new HIV infections by 17% in the year 2030 and by a total of 11% throughout the period 2015 – 2030.
However, a vaccine with higher efficacy and higher coverage would have much greater impact:
- A vaccine with 50% efficacy and 30% coverage would reduce new infections by 53% in 2030 and avert 34% of projected infections between 2015 and 2030.
- A vaccine with 70% efficacy and 40% coverage would reduce new infections by 81% in 2030 and avert 56% of projected infections between 2015 and 2030.
A generalised vaccination programme could avert one infection for every 15-30 people vaccinated, but a targeted programme could avert one infection for every three to ten people vaccinated.
A successful vaccine which also delayed disease progression by doubling the period during which people remain free of AIDS symptoms could achieve savings of between $28 billion and $132 billion between 2015 and 2030, the model estimates.
The developers of the model say that they intend to work with national AIDS programmes to refine the estimates, and to use the findings to build support for large-scale vaccine trials and the planning of national vaccine delivery programmes.
The model
Full details of the model are due to be published shortly in an IAVI working paper, but were described briefly to the conference by John Stover of the Futures Institute.
The model used local data from on sexual activity and sexual mixing patterns, including number of sexual acts per partner, condom use, sexually transmitted infection rates, concentration of the epidemic within sub-populations such as men who have sex with men and injecting drug use, movement of the population out of sexual risk groups due to ageing or death, circumcision prevalence and availability of antiretroviral therapy. These data were used to build a model of HIV prevalence over time.
The model contains assumptions about the uptake of various interventions. For example, the model assumes that 50% of the male at-risk population in South Africa will be circumcised by 2010, and that circumcised men will continue to enjoy a similarly lowered risk of HIV infection throughout the period 2015 to 2030 to that seen in the 20 month Orange Farm study, in which circumcised men were approximately 60% less likely to become infected during the follow-up period.
The effects of a vaccine were then applied to the model, taking into account varying degrees of vaccine efficacy, variations in population coverage by age, sex and risk group, varying reductions in infectiousness and disease progression (on the assumption that a vaccine may attenuate HIV disease in people who become infected despite vaccination), and behavioural disinhibition caused by the availability of a vaccine.
Stover J. The impact of an AIDS vaccine in developing countries: the use of modeling to build support for vaccine development. AIDS Vaccine ’06, abstract S09-05, 2006.