First study of population-level preventative impact of Medical Male Circumcision and ART on HIV incidence in a country of sub-Saharan Africa

Clinical studies have demonstrated the potential effectiveness of ART (HPTN 052) and Voluntary Medical Male Circumcision (VMMC) (Gray & Kigozi/STIs) as preventative measures against HIV.  This led WHO/UNAIDS to launch a Joint Strategic Action Framework (JSAF) setting a target in 14 priority sub-Saharan countries of 80% VMMC by 2016.

What, then, are the potential gains of ART and VMMC interventions in these countries?  Comparative ecological studies have shown the population-level impact of male circumcision as a cultural practice (MacLaren & Vallely/STIs). Various mathematical modelling studies have sought to quantify that potential effect of interventions both in the realm of VMMC (Jenness & Cassels/STIs) and ART (Shafer & White/STIs)  (though other studies have highlighted the challenges that scale-up of these interventions is likely to present (Kaufman & Ross/STIs)).

Now, for the first time, a study has sought to quantify the real-life population-level impact of these interventions.  Kong & Gray (K&G) base their study on data from the 1999-2013 Rakai Community Cohort Study (Uganda)).  Among the 45 Rakai communities (44,688 participants surveyed over 24.6 years), VMMC coverage had, by 2013, increased from 19% to 39%, and ART had risen, in males, from 0% to 21%, and, in females, from 0% to 26% – and HIV incidence had fallen, concurrently, from 1.25 per 100 person-years to 0.84 per 100 person-years in males, and from 1.25 to 0.99 in females.  As regards VMMR, each 10% increase in the rate was associated amongst males with a decline in incidence that could be quantified, on multivariate analysis, at 0.87 – though, in females, the reduction was statistically insignificant.  As for ART, the decline attributable was not statistically significant in either case, but, when ART coverage was modelled as a categorical variable, and coverage of over 20% was compared with coverage of under 20%, a decline in HIV incidence was observed in the former group of the order of 0.86 among males, and 0.77 among females.

These results are not surprising.  VMMC is, in the first instance, protective of men – though, of course, in the longer term women too will benefit from any population-level effect. (There is, in fact, a worrying possibility, investigated by Maughan-Brown & Thornton/STIs, that men could incorrectly assume that their VMMC will be directly protective of their partners, and modify their behaviour accordingly.)  As for ART, here too the (as yet) limited impact in Rakai is what we might have expected.  Tanser & Newell, in a South Africa-based study only observed significant association when ART coverage was over 30%.  However, population level decline in incidence – especially that associated with VMMC – is encouraging.  The results of this study allow us to predict that increasing VMMC coverage more than 40% could reduce male incidence by approximately 39% at population level.   A major limitation of the study, of course, is its assumption that sexual networks, and hence HIV transmissions, are internal to the community. However, a recent study by Chemaitelly & Abu-Raddad/STIs would seem to indicate that, in a context like sub-Saharan Africa the contribution of networks going beyond the wider community is likely to be limited.