At currently planned levels of treatment coverage in Africa, transmission of drug resistant virus will remain far below the levels seen in the developed world for at least the next ten years, according to a comprehensive modelling exercise by biomathematicians at the University of California published in the January 3rd edition of AIDS.
The model also predicts that the introduction of antiretroviral therapy will have a negligible effect on the rate of HIV transmission, and that the vast majority of drug resistance will emerge as a result of treatment failure, not transmission of drug resistant virus.
Dr Sally Blower and colleagues note that a much lower proportion of the HIV-infected population in Africa is likely to receive treatment in Africa (15-20%) compared to gay communities in North America and Europe, where most studies of sexual risk behaviour and transmission of drug resistant virus have taken place.
Their modelling suggests that even if 25% of the total HIV-positive population are receiving treatment within five years of its introduction, only 5% of prevalent infections will be drug resistant (including cases of drug resistance acquired as a result of treatment). If drug resistance remains below this level the World Health Organisation recommends that national surveillance for drug resistance is not necessary.
However, if treatment coverage extends to 30-40% of the HIV-positive population, the prevalence of drug resistance will increase, to 10-15% at five years and 25-40% at ten years. The authors say that projections beyond this point are misleading because different drug regimens with unknown resistance risks are likely to be in use by that point.
They also project that if only 10% of HIV-positive individuals are treated, HIV prevalence will not have decreased after ten years and the incidence rate (new infections) will have decreased by less than 5%. Nevertheless in the context of 3 million new infections each year, the authors note that this would reduce the burden of new infections by 143,000 – 473,000 infections in the first five years of roll out, with higher coverage having a correspondingly greater effect on incidence (up to 1.3 million infections might be averted in the first five years if 25% of individuals are treated).
Although transmitted resistance will be a less important source of drug resistance than resistance acquired through treatment, up to 300,000 cases of transmitted resistance could occur if treatment coverage of 10% is achieved.
The authors conclude that large scale surveillance systems for detecting and monitoring transmitted drug-resistant virus in Africa will be unnecessary, and that sentinel surveillance at treatment centres with high treatment coverage (> 10% of HIV-positive people) are likely to prove sufficient for monitoring the development of resistance.
Blower S et al. The antiretroviral rollout and drug-resistant HIV in Africa: insights from empirical data and theoretical models. AIDS 19: 1-14, 2005.