HIV treatment as prevention: optimising the impact of expanded HIV treatment programmes

Abstract

Until now, decisions about how to allocate ART have largely been based on maximising the therapeutic benefit of ART for patients. Since the results of the HPTN 052 study showed efficacy of antiretroviral therapy (ART) in preventing HIV transmission, there has been increased interest in the benefits of ART not only as treatment, but also in prevention. Resources for expanding ART in the short term may be limited, so the question is how to generate the most prevention benefit from realistic potential increases in the availability of ART. Although not a formal systematic review, here we review different ways in which access to ART could be expanded by prioritising access to particular groups based on clinical or behavioural factors. For each group we consider (i) the clinical and epidemiological benefits, (ii) the potential feasibility, acceptability, and equity, and (iii) the affordability and cost-effectiveness of prioritising ART access for that group. In re-evaluating the allocation of ART in light of the new data about ART preventing transmission, the goal should be to create policies that maximise epidemiological and clinical benefit while still being feasible, affordable, acceptable, and equitable.

Figure 1. HIV transmission and mortality by CD4 count.

(A) HIV transmission rate per 100 person-years (PYs) by CD4 count for the infected partner in discordant couples enrolled in a randomized controlled trial of acyclovir . (B) Mortality rate by CD4 category in ART-naïve HIV-positive individuals enrolled in research cohorts in West Africa . In both panels, the width of the bars represents the proportion of ART-naïve HIV-positive 15- to 64-year-olds by CD4 count in a nationally representative household survey in Kenya .

Figure 2. The transmission potential of individuals as a function of set-point viral load.

(A) Infectiousness (per unit calendar time) and (B) duration of asymptomatic infection are estimated by fitting to various sources of data as described in . (C) The product of these is the transmission potential, the average number of people an infected individual is expected to infect over the whole of asymptomatic infection. The transmission potential measures the relative prevention effect of treatment as prevention targeted to an individual with a given SPVL. Adapted from Fraser et al. .