The potential value of PrEP as an intervention in Western MSM populations has recently been conclusively demonstrated (PROUD – 86% (PROUD study/STIs/blog)), despite results from a whole series of trials in sub-Saharan African populations that seemed to cast doubt on its likely efficacy (CAPRISA 004 – 39%; iPrEX – 44%; CDC TDF2 – 62%; Partners-PrEP – 75%; FEM-PrEP – 0%; VOICE– 0%). The difference in outcome achieved in these two contexts would seem to have been the result of very different levels of compliance on the part of the study participants (VOICE D/STIs/blog). Evidently, the biologic efficacy of PrEP holds out hope for reducing transmission among MSM populations. But the jury remains out over the potential value of PrEP in combating the 50% of infections represented by the sub-Saharan hetero-sexual epidemic. An approach that has seemed promising in view of the evident problems of compliance posed by a daily or pre-coital regime of oral tenofovir is to deliver the ARV drug in the same way that contraception or HRT is sometimes delivered, via a ring that is inserted monthly into the vagina. Last month saw the final publication of the results of one of the two major studies that have been trialing this technology: MTN-020 – ASPIRE (Baeten & Hillier (B&H)).

The results of B&H – as indeed those of the other major sub-Saharan dapivirine ring study (IPM/Ring Study) – are disappointing and hard to interpret. Disappointing, because HIV incidence in the intervention arm was only 27% lower than in the control arm (though, when data from two (out of 15) study sites were excluded on the grounds of reduced retention and adherence, that figure rose to 37%). Hard to interpret, because – in response to problems of compliance that emerged in the earlier studies mentioned above – the study built in measures for the objective assessment of adherence including quarterly plasma sampling for dapivirine, and, after the first year of the trial, testing for residual dapivirine in used rings; and results from the latter suggested levels of compliance between 70% and 85%, depending on age group (with age positively associated with compliance).

One might be tempted to conclude that the dapivirine ring showed relatively poor efficacy. On the other hand, a post hoc age-stratified analysis of the data indicated levels of protection of 56% among those aged over 21 years, and zero protection for those aged 21 years or under. These figures are most easily interpreted as suggesting that non-compliance was at least an important factor – and, indeed, B&H themselves admit that their thresholds for non-compliance were set too low. If we disregard the objective assessments of compliance, then the post hoc analyses of levels of protection by age group would seem to suggest that the younger participants (aged 21 and under), almost unanimously removed their rings after each clinic visit, only to re-insert them, shortly before the next. The fact that an unknown proportion of the older participants might have been doing the same thing makes it impossible to reach any adequate conclusion as to the biologic efficacy of the dapivirine ring in this population.

Why the non-compliance – especially of younger participants? When VOICE D sought to investigate the non-compliance in earlier trials of oral tenofovir by confronting study participants with the objective evidence, the reason most frequently given was fear of the side effects of the drug, fuelled by peer participants and relatives and by the negative attitudes of community members. This might explain the positive association between compliance and age, if we assume a greater responsiveness on the part of younger participants to peer pressure. If this is the case, then the solution to these problems may be the development of more accurate measures of objective adherence.