Vaginal ring can substantially impact HIV prevention, reveals mathematical model
New mathematical modelling study suggests that the dapivirine vaginal ring could substantially reduce HIV infections among South African women and be an affordable intervention depending on roll-out strategy.
A vaginal ring containing the antiretroviral drug dapivirine (DPV) could prevent up to 11% of new HIV infections among South African women at a high risk of infection, when compared to a scenario where this form of prevention is not available.
A mathematical modelling study was used to understand the impact of the DPV ring on preventing new HIV infections among at risk women aged between 22 to 45 years in KwaZulu-Natal, South Africa.
This study follows several trials and demonstration projects of oral PrEP, which have cast doubt on effectiveness of oral PrEP (tenofovir plus emtricitabine) among women in sub-Saharan Africa. In these scenarios, adherence to PrEP is highlighted as the principle challenge.
There is hope that long-acting PrEP delivered either through injections or vaginal rings could increase the effectiveness of PrEP in these groups, reducing problems around adherence. Already two studies (ASPIRE and the Ring Study) have shown partial effectiveness of a monthly vaginal ring containing the non-nucleoside reverse transcriptase inhibitor (NNRTI) dapivirine (DPV) for PrEP among women aged 22 to 45.
In this study a mathematical modelling approach was used to better understand the projected epidemiological impact of the DPV vaginal ring (DPV-VR) PrEP, taking into account budget and cost-effectiveness of the implementation.
The researchers modelled the impact of DPV-VR roll-out in four different scenarios, where the ring was prioritised for different groups of women and achieved different levels of coverage. In the first scenario DPV-VR implementation was hypothesised to cover 10% to 40% of HIV-negative women aged 22 to 45 years (this was the deprioritised DPV-VR scenario). In the second, DPV-VR was prioritised by age, covering 20% to 75% of all HIV-negative women aged 22 to 29 years; while in the third it was prioritised by incidence, covering 20% to 75% of HIV-negative women in age groups and behavioural risk groups with HIV incidence ≥3% per year. Finally, DPV-VR effectiveness was modelled for roll out among female sex workers, covering 25% to 75% of all HIV-negative female sex workers.
The investigators also considered effectiveness across the four scenarios when DPV-VR was combined with VMMC (at 80% coverage) and universal ART (wherein 90% viral suppression was achieved by 2020 and 95% by 2030).
They also looked to consider effectiveness at different adherence levels: firstly, high adherence at 75% (this is similar to what was achieved among participants in the earlier ASPIRE study) and then in a low adherence scenario of 49% (as was seen in the Ring Study). Both adherence scenarios were compared to a reference where there was no vaginal ring implementation.
The preventative effects were estimated for 2016 to 2030 (the UNAIDS ‘Fast-Track’ years) and over the course of a lifetime.
Among their multiple findings, the authors report that DPV-VR implementation is affordable and could have substantial population-level HIV prevention effect if prioritised to at-risk women. These interventions could also be cost saving when prioritised to sex workers.
HIV prevention was most effective when the ring was prioritised on the basis of incidence, with age-based prioritisation also achieving high levels of HIV prevention.
At 15% coverage of HIV-negative women, where adherence is high, incidence‐based DPV-VR PrEP prevented 11.3% of infections; age-based DPV-VR PrEP prevented 9.6% of HIV infections and deprioritised DPV-VR PrEP prevented 7.1%. DPV-VR for sex workers prevented a smaller proportion of HIV infections at 3.5% - although this is primarily explained by the smaller sample size.
Where adherence was lower (at 49%) approximately 40% fewer infections were prevented across all implementation strategies.
The authors also examined whether implementing DPV-VR would lead to increases in HIV drug resistance. They found that any increases were modest. Where adherence was high, accumulated drug resistance was less than the reference scenario across all strategies, but in the scenario with reduced adherence, modest increases in drug resistance were revealed across all strategies except for unprioritised PrEP. Increases in drug-resistance were most notable where DPV-VR PrEP was prioritised for sex workers, with levels rising by 2% compared to the reference scenario.
Both costs and impact from DPV-VR PrEP implementation increased proportionately with coverage. When excluding female sex workers from the analysis, HIV expenditure from PrEP implementation increased by 5% over ten years when it was scaled-up to cover 15% of adult HIV-negative women. When PrEP spending was prioritised to sex workers, HIV funding decreased overall.
The analysis also found that incidence-based PrEP was more cost-effective than age-based PrEP or deprioritised PrEP. The authors noted that without increased funding of PrEP, and further decreases in the price of the ring, funding this intervention may not be affordable in the context of South Africa’s other HIV health investments.
“As countries contemplate PrEP implementation, questions of impact and cost-effectiveness, which mathematical modelling can address, must be considered alongside issues of equity and acceptability,” commented the authors in their discussion. “While DPV-VR PrEP prioritized by incidence or age may be less cost-effective than ART, VMMC or condom provision, it could be a viable option for HIV-negative women who may not benefit directly from ART or VMMC, nor ably negotiate condom use.”