PrEP and women

Introduction

The majority of people who have initiated PrEP are cisgender women.

This statement may come as a surprise to PrEP advocates in high-income countries, where PrEP awareness and provision is far higher among gay and bisexual men and trans people than it is among cisgender women.

However, an exponential growth in PrEP adoption, starting in the last quarter of 2019, has resulted in huge numbers of women starting PrEP in Africa.

The cumulative number of people who had been provided with PrEP at least once stood at around 600,000 at the end of 2019, but was 1.8 million at the end of 2021 and over 3.8 million at the end of 2022. This misses by only about 18 months the target set by the 2016 UN declaration to end AIDS by 2030, which envisaged 3 million on PrEP by the beginning of 2021.

In the UNAIDS Global AIDS Update, there were about 470,000 female and 260,000 male PrEP users in 2021. This is considerably less than the reported global total, not only because many countries don’t separate users by sex, but because high-income countries tend not to report to UNAIDS. The global total includes the US, with nearly 400,000 users, only 28,000 of whom (7%) are cisgender women. Europe adds another 150,000 users, very few of them women.

But all the African countries with over 10,000 users, apart from Nigeria, are in southern and eastern Africa, where the majority of PrEP users are women.

And a study published in The Lancet in 2021 found that of over 120,000 PrEP users in countries reporting gender, 45% were men and 62% were women. Another 1.3% were transgender. (This adds up to more than 100% because of differences in the way different countries subdivide their populations.)

Of the women PrEP users, 30% were adolescent girls and young women, i.e. aged between 15 and 25 years old. Thirty-two per cent were female sex workers – obviously, these categories overlap. Of the men, 65% were gay, bisexual or other men who have sex with men.

This picture makes it look as if under-provision of PrEP for women – who still make up 49% of new infections – might be coming to an end.

But this is to ignore an entire history of failed trials, fewer trials for women, low adherence and persistence, and lack of uptake of PrEP among women, with occasional exceptions. Awareness and availability are particularly low among women outside Africa. What works for gay men has worked, until recently, less well for women.

‘Until recently’, because in the last three years a new method of PrEP administered by injection has, excitingly, proved not only to work well for women, but to work for them even better than it does for gay men. Injectable PrEP stands on the verge of widespread provision, but matters of cost and infrastructure need to be solved first.

There also remains a number of unanswered questions around the efficacy and effectiveness of oral PrEP in women, and we have almost no evidence as to its public-level effectiveness – how much it will contribute to meeting global health targets.

Questions that need answers include the following. Some will be answered in the course of this research briefing – others need more research.

• What is the evidence for the efficacy of PrEP in women?
• Is there any evidence that oral PrEP may have lower or higher efficacy in women than in men? Or in anal vs vaginal sex?
• Does this affect recommended dosing regimens?
• What do randomised controlled trials and demonstration projects tell us about how effective PrEP is for women?
• If, as many studies have found, PrEP adherence and persistence are often lower in women, why is this? What do women themselves have to say about PrEP?
• As well as oral PrEP, are there other formulations of PrEP such as vaginal rings, injections or implants that might be more suited to women?
• What do women say about these?
• Outside sub-Saharan Africa, PrEP coverage among women (i.e. the proportion of women at risk of HIV who use it) is still low. Is this due to lack of awareness among women or their healthcare providers?
• Or is it due to difficulty in determining which women are at risk of HIV and might benefit from PrEP?

These scientific and practical questions are not the only factors that may unnecessarily restrict women’s access to PrEP. A number of countries, including large ones like Russia, report having no plans to develop national PrEP guidelines, and a number of others continue to restrict access to women, or only to certain women.

As PEPFAR says in their December 2022 five-year strategy: “Highly effective and … needs-aligned products such as oral PrEP took nearly a decade to begin scaling up after regulatory approvals, because of a lack of proactive market-shaping approaches and commitment from donors and country governments.”

How well does oral PrEP work in women? Early studies

The first completed study of oral PrEP in humans was in women; its results were announced at the International AIDS Conference in 2006 in Toronto. This study conducted in Ghana found three times more women on placebo – namely, six out of 368 – had become infected with HIV than women taking solo tenofovir disoproxil fumarate – two out of 363, yielding a 65% reduced risk. However, the number of participants in this study was too small for the results to be statistically significant.

The study was originally intended as a larger one, including women in Cameroon and Nigeria. Although the Nigerian arm was stopped because of technical shortfalls, research in Cameroon was halted due to the same controversy about the early PrEP trials that stopped a similar study in Cambodia in 2004.

PrEP was a very new concept at this point. It was not known whether its efficacy against HIV would outweigh side effects or whether behaviour change caused by over-confidence in the effects of PrEP might make people more vulnerable to HIV. The answer soon came: but the issue of what kind of PrEP best suits particular people, and why some types of PrEP seem to work much better than others, especially in women, has continued to be a subject of debate.

Three further studies of oral PrEP involving female participants in African countries reported their results in 2011. Firstly, the TDF2 study evaluated tenofovir disoproxil fumarate plus emtricitabine (TDF/FTC) and enrolled 1219 people, 46% of them women, in Botswana. The study was finished early because researchers were concerned that, due to lower-than-expected retention rates, a statistically significant difference in infection rates between the PrEP and placebo arms might not be found.

In the end, though, when the study results were announced in 2011, the overall effectiveness of 63% was statistically significant. Effectiveness was 80% in men but only 50% in women. This difference was not statistically significant, in other words it could have been due to random variation. When only the women who collected their pills on time and so had sufficient supplies of TDF/FTC were counted, effectiveness for them was 76%. This was significant.

However, the study that really showed that PrEP worked for women, with early results presented alongside TDF2, was the Partners PrEP study. This, as the name indicated, was conducted among 4758 sero-different regular partners – one with HIV, one without – in Kenya and Uganda. In 38% of couples, the HIV-negative partner was the woman.

This study measured the effectiveness of both solo TDF and of dual TDF/FTC against placebo. The overall effectiveness averaged over both PrEP arms was 70.5%. Effectiveness in women was 68% on solo TDF and 62% on TDF/FTC, but this difference was not statistically significant.

Participants claimed extremely high levels of adherence – 97%. Comparing drug levels in those who acquired HIV with levels in some of those who remained uninfected showed that the true levels of adherence were about 80%, but this is still the highest observed in any randomised controlled study of oral PrEP involving women.

Two studies where oral PrEP didn’t work for women

The Partners PrEP finding of 65% effectiveness in women was received with relief by PrEP advocates, because another large study, FEM-PrEP, had been closed a few months earlier due to lack of effect. The failure of FEM-PrEP in 2011 and another large study, VOICE, in 2013, posed difficult questions for PrEP researchers.

It also meant that up until 2020, a single randomised controlled trial ending in 2011 (Partners PrEP) was the only statistically reliable result backing the effectiveness of PrEP in women.

Taking the two studies side by side: FEM-PrEP enrolled 1951 women in South Africa and Kenya and compared the effectiveness of TDF/FTC against placebo. VOICE, a larger and more complex trial, enrolled 5029 women in South Africa, Zimbabwe and Uganda and compared the effectiveness of three types of PrEP against placebos: solo TDF, dual TDF/FTC and a tenofovir-containing microbicide gel for vaginal application. (Topical and other formulations will be covered later in this briefing.)

FEM-PrEP was terminated for futility in 2011 (in statistics, ‘futility’ means that reaching a definitive outcome, either positive or negative, is highly unlikely). The signs were not looking good for VOICE either, as both the solo-TDF and microbicide gel arms were also stopped for futility in that year. But the dual TDF/FTC arm continued until 2013.

In FEM-PrEP the rate of infection was practically the same (4.7% vs 5% a year) in both arms. In VOICE, there were 15% fewer infections on the tenofovir gel than there were on placebo gel, but this was also statistically non-significant. In the other two arms there were actually more infections on PrEP than on placebo: 4% more on TDF/FTC and a nearly significant 49% more on solo TDF.

The explanation for these disappointing results became apparent when adherence was measured by drug level monitoring. In FEM-PrEP, adherence measured by counting women’s leftover pills was 85%. But when drug levels were measured, less than 50% of participants had actually taken a pill in the last 12 days.

In VOICE, adherence by self-report and pill count were 90%. But only 29% of women in either of the oral PrEP arms in fact had drug levels that showed they had taken a pill recently, and only 42 to 50% of women had detectable drug levels at any trial visit.

These results needed qualitative investigation to explore women’s apparent distrust of the prevention method they were supposed to be testing.

Two FEM-PrEP analyses revealed that participants often joined the study for reasons other than HIV prevention. Some distrusted the efficacy of PrEP, while others saw PrEP and condoms as an additional benefit, but not the main point of joining the study.

Interviews with a randomly selected 5% of participants found that they liked the regular HIV testing and used it as a reassurance they did not have HIV. The three benefits of the trial most commonly cited by participants were the monthly HIV test (32% of participants said this was an important reason for joining the trial), the behavioural support the trial gave them to stay negative (25%) and the monitoring offered for other health problems (22%).

A second study found that, despite repeated reassurances by researchers that they would not be penalised for not taking their pills, participants feared being excluded from the trial if they admitted low or no adherence, or simply feared confrontation with the trial staff and saw adherence counselling as a reprimand.

Participants also feared side effects; these fears were amplified by participants discussing the subject with each other. Some stored their pills up for use should the study report a positive effect.

Interviews with participants at the Johannesburg site of the VOICE study uncovered similar feelings. Some women had stronger priorities than PrEP, including worrying about other health priorities such as pregnancy. Others were concerned about being assumed to have HIV, especially as many did not have the privacy to take or store pills. Some had feelings of resentment at being singled out as trial subjects or had not clearly understood that they were being asked to take antiretrovirals, which were seen as being ‘for sick people’.

Lack of acceptance or understanding of PrEP in the wider community was also a factor, with participants saying they were deterred from taking the pills by partners or peers.

These early trials revealed that when women were motivated to take PrEP in a situation where they felt supported, it was effective; but that considerably more information, support and community education was required. The complex pressures in the lives of women at risk of HIV would need to be better understood in order to design PrEP interventions that might suit them.

The study that didn’t happen – TAF/FTC in women

Tenofovir has one characteristic toxicity. It is eliminated from the body by being filtered out through the kidneys rather than by being chemically broken down in the liver, as the majority of antiretrovirals are. This can sometimes irritate the kidney tubules and lead to a decrease in the efficiency with which they filter the blood (kidney dysfunction or renal insufficiency). While tenofovir does not cause problems for most people, toxicity is more common in older people and those with previous kidney disease. Injury to the kidneys also leads to too much calcium being filtered out, which can cause deficiency of the calcium in bones that gives them their stiffness and density.

Because of this, TDF‘s manufacturers devised a second prodrug of tenofovir, tenofovir alafenamide or TAF. This is incorporated much faster into cells, so less is left in the bloodstream to damage the kidneys.

TAF was licensed for HIV treatment in various different combinations; TAF/FTC, with the brand name Descovy, was licensed for people with HIV in Europe and in the US in April 2016.

An obvious question was whether TAF/FTC could be used as PrEP and might have less risk of side effects than TDF/FTC. A large international study, DISCOVER, was conducted in over 5000 gay and bisexual men and trans women. It reached inconclusive results regarding effectiveness: although there were 46% fewer HIV infections in people taking TAF than in those taking TDF, this difference was not statistically significant. There were significantly lower levels of renal insufficiency and bone mineral loss, however.

On the basis of DISCOVER, US drug regulators approved TAF/FTC for use as PrEP in gay and bisexual men and trans women in October 2019.

But not for cisgender women. The DISCOVER 2 study was proposed for cisgender women and adolescent girls in Africa but never materialised; as a result, TAF/FTC has never been approved as PrEP for them.

This situation is finally being remedied because the PURPOSE 1 study, which is investigating the efficacy of the injectable drug lenacapavir as PrEP in adolescent girls and young women, also includes a direct comparison of TDF/FTC and TAF/FTC for the first time. Primary results are expected in March 2024. See more in ‘Future directions’ below.

Does the efficacy of oral PrEP differ in women?

Barriers to the effectiveness of PrEP in women may not be entirely social or behavioural. There may be biological differences between men and women – or rather, between vaginal and anal sex – which affect oral PrEP’s efficacy.

A study in 2011 was the first to uncover these differences. It gave 15 volunteers, seven of them women, a single dose each of the two drugs then used for PrEP, TDF and FTC.

In the women, levels of FTC were detectable for about ten days. From the start they were consistently tenfold higher in cervical than in rectal tissues and somewhat higher in vaginal tissues.

However, drugs have to be phosphorylated – have phosphate groups added to them – to be able to enter cells and inhibit HIV within them. Levels of phosphorylated FTC (known as FTC-TP) were only detectable for two days in vaginal tissue before disappearing.

Tenofovir was a different story. Levels of the non-phosphorylated drug were well over 100 times higher in rectal tissues than in vaginal tissues and remained so: tenofovir was still detectable 14 days after the single dose in rectal but not in vaginal tissues.

The levels of the phosphorylated version (tenofovir-DP) were originally lower in vaginal than rectal tissue but increased after that and were equal to rectal levels seven days after the single dose. Furthermore, levels in all tissues persisted at inhibitory concentrations until day 14, at which point they were somewhat higher in vaginal tissues.

Several other studies confirmed this pattern, such as a 2015 analysis of HPTN 067, the ADAPT study. This found that on continuous dosing, while tenofovir levels in rectal tissues reached peak levels after only two to three doses, they took seven to ten days to reach peak levels in cervical tissues and even then were tenfold lower.

A modelling study in 2016, taking data from a number of different studies, found that while protective levels of tenofovir in rectal tissues could be maintained after as few as two doses per week (taken consistently), it would take at least six doses a week to maintain the same levels in vaginal tissue.

Studies like these have been the basis for the recommendation that women should not try to use event-based dosing for oral PrEP, such as taking it on the '2-1-1' pattern (two pills 24 to two hours before sex, then single doses 24 and 48 hours after the first).

A recent comparison of HIV infections seen in the TDF/FTC arms of the HPTN 083 and HPTN 084 studies (see more on these later) found that while just two or more doses of PrEP a week seemed to be enough to prevent infection in the gay men and trans women in HPTN 083, daily PrEP was necessary to protect cisgender women in HPTN 084.

However, in the cases of the women, this observation was based on just one ‘breakthrough’ infection in a woman who had drug levels concordant with taking six to seven doses a week – because very few women were in fact that adherent. What comes through most strongly, yet again, is that very few women had anything like adequate adherence to oral PrEP. So, while tissue concentrations may be related to efficacy, low adherence and all the social factors that lead to it, may be at least as strongly associated with ultimate effectiveness.

In the gay men and trans women in HPTN 083, the percentages who had adherence of zero, less than two, two to three, four to six, and seven pills a week were 9%, 9%, 10%, 39% and 34%, respectively, meaning that 83% had adequate drug levels.

In the women in HPTN 084, the percentages go in the opposite direction: 38%, 26%, 18%, 15% and 3%. In other words, even in this more recent study, among a group of women at least some of whom were reasonably knowledgeable about PrEP, very few had adequate drug levels – certainly no more than 18% and probably less.

On the other hand, another recent study using a completely different methodology – a  meta-analysis of the relationship between adherence and infection risk in 2955 women in a number of different studies and demonstration projects in Africa – found that the risk of HIV infection was only 13% higher in women taking four to six tablets a week than in women taking seven a week, and that this difference was only barely statistically significant.

The best conclusion to draw is that while the lower or zero effectiveness of oral TDF/FTC PrEP seen in some studies in women is largely due to lower adherence and lower persistence, themselves caused by many different socio-economic and personal factors, the probably greater need for women to maintain more consistent adherence than gay men adds another barrier to achieving protection.

Microbicide gels – and vaginal rings

‘Topical’ PrEP refers to germ-killing medicines or barrier substances designed for use in a specific part of the body. Development started even earlier than PrEP pills, which are ‘systemic’ (distributed throughout the body).

Unlike tablet-based PrEP, these topical preparations were always seen as products that might specifically benefit women. This is partly because of anatomy; the vagina/cervix is essentially a closed-end pouch more suited to gels and devices than the open-ended tube of the rectum, where devices may not stay in place. In addition, the epithelium (mucous membrane) lining the vagina is several cells thick, whereas the rectal epithelium is only one cell thick, so it was thought there was less risk of a topical product causing inflammation and irritation.

The first randomised controlled study of a microbicide upset this assumption. This study, of the spermicide nonoxynol-9 used as a vaginal microbicide among 700 female sex workers in Côte d’Ivoire, started as early as 1996. When it was ended in 2001 it was found that, far from it protecting the women against HIV, it actually made HIV transmission more likely. As aidsmap commented at the time, “The results of the trial have raised ethical questions that may affect the future study of all microbicides” and affected attitudes towards early studies of systemic PrEP in women.

Microbicides consisting of neutral gels containing antiretrovirals were thought more likely to succeed. One development pathway conducted studies of a tenofovir-containing gel for vaginal application that culminated in the CAPRISA 004 study, which announced a positive result in 2010.

The gel reduced HIV infections by 39% – not a result that would be hailed as a great success today, but which at the time (two years before Partners PrEP) was seen as a major advance in prevention technology.

Unfortunately, however, two later studies produced negative results: the microbicide arm of the VOICE study, already mentioned, and FACTS 001, the open-label (no placebo) extension of the CAPRISA 004 study, which announced no effectiveness in 2014.

Dapivirine was another antiretroviral trialled as a microbicide gel, but it had somewhat more success when delivered as a vaginal ring. Similar to a contraceptive ring, this is a silicone rubber ring designed to fit inside the vagina and over the cervix. After several years of development work, two parallel phase III placebo-controlled trials, RING and ASPIRE were launched in 2012, recruiting a total of 4588 women aged 18-45 in Malawi, South Africa, Uganda and Zimbabwe.

When effectiveness was announced in June 2015, it seemed disappointingly low in both studies – just 27% in ASPIRE and 31% in RING. In the case of ASPIRE, there was zero effect in young women aged 18-21, but 51 to 56% in women over that age.

Later adherence data showed that the majority of women did not keep the rings in all month, as recommended. Data from ASPIRE showed that on average the women took their rings out for a third of the time and efficacy during that period was zero. However, during the 42% of follow-up time when adherence was medium to high, efficacy was 65%.

Of considerable importance, however, was that some women liked the ring. One study participant summed up her feelings thus:

“I like them because once you wear it, you don’t feel it, and nobody can suspect you are wearing something. I like it because nothing changes regarding how we live as women.”

Women, in short, liked a prevention method that was discreet and long-lasting, so kept them safe even at times when, as it is often for some women, sex was uninvited or coerced, and which left them with a sense of being in control of their health.

It was noticeable that retention in ASPIRE was excellent, and the impression that women liked this prevention method and became more expert at using it was confirmed by the open-label extension studies that took place between 2015 and 2018.

HOPE was the open-label extension of ASPIRE and DREAM the open-label extension of RING. Combined together, they involved 2307 women.

Both retention, with 80% remaining in the studies at nine months, and adherence, with 90% of women with efficacious usage levels, were good. Effectiveness, as measured against what, researchers calculated, would have been the HIV infection rate if the rings had not been used, was 54%.

Later estimates of the ring’s effectiveness have ranged widely, from 37 to 90% with perfect adherence. Clearly, even with the best possible adherence the ring’s effectiveness does not match optimal oral PrEP use. But there are several points in its favour: it is discreet, does not require daily adherence and women can insert it themselves; it is safe in pregnancy and breastfeeding; the majority of male partners, in one study, supported its use; researchers are currently working on rings that also provide a contraceptive and/or that will last for three months; and they are also looking to see if removing the ring for several days during menstruation – as many women already do informally – will compromise its effectiveness against HIV.

Given these features, it is perhaps surprising that the ring has not been licensed for use worldwide, but only in Africa. The European Medicines Agency (EMA) declared a ‘positive benefit-risk opinion’ on the ring in July 2020 but did so under article 58, an approval pathway unique to the EMA which allows it to license a drug intended for use in lower-income countries with the same rigour as a drug destined for the European market.

The International Partnership for Microbicides (IPM) announced it would not be seeking a licence for the ring in Europe itself, and in December 2021 the US Food and Drug Administration (FDA) let IPM know it would be unlikely to succeed if it pursued approval in the US.

These decisions have caused controversy. The FDA gave, as its reason for advising IPM not to pursue approval, that a product would have to match oral PrEP in terms of effectiveness if it were to succeed in the US market. This has led prevention experts to question why, if the product is not good enough for Europe or the US, it is good enough for Africa. They also point to the evidence that the ring is popular among some women who might not otherwise consider oral PrEP at all, and that maximising choice should be one of the factors considered.

In Africa, meanwhile, Kenya, South Africa, Uganda, Zambia and Zimbabwe have already approved the ring and have started the administration needed for rollout, with Zimbabwe probably furthest ahead.

How many women are using PrEP in high-income countries?

A 2023 study tracked cumulative PrEP prescriptions in the US between 2013 and 2022. Between those dates there had been 381,784 prescriptions. Of those, 26,725 (7%) went to women.

Almost all of those were for oral PrEP. Use of injectable PrEP started very recently and up to September 2022 there had only been 2695 prescriptions for injectable cabotegravir (0.7%). Interestingly, however, a higher proportion of these went to women than for oral PrEP – 12.5%.

A 2020 study estimated that a similar number of women in the US were on PrEP – 28,000 out of a total of over 300,000 PrEP users. However, it is also estimated that the number of women whose HIV risk would make them eligible for PrEP is roughly the same as the number of gay men – about 1.2 million. The US HIV surveillance report of 2019 estimated PrEP coverage for cisgender women at risk of HIV was just under 10% compared with 27% for gay men and 32% for transgender women.

The one exception is adolescents under 17. According to a 2018 study, only 1.2% of people using PrEP in the US are aged 12 to 17, but 80% of those were girls and young women. This, however, more likely reflects the culture of stigma faced by adolescent boys and men under the age of consent for gay sex, rather than a positive drive to get PrEP to female adolescents.

In Europe the situation for women at risk of HIV is worse. For example, in the UK, although 2.5% of people in the UK are of Black African ethnicity and 4% of any Black ethnicity, 42% of heterosexual HIV infections are among Black Africans. In 2020, for the first time in a decade, diagnoses in heterosexuals exceeded those in gay men and 28% of all new diagnoses in the UK are now among cisgender heterosexual women.

Despite this, there are very low rates of knowledge and use of PrEP among women, including Black women, in the UK. Even among women who are aware of PrEP, few have used it. For instance, in Invisible No Longer, a national study of women’s experiences of HIV published in 2018, while 74% of 151 women in a survey of women ‘concerned’ about HIV were aware of PrEP, not one of them had used it.

Going back further, the large Flash!PrEP study of attitudes towards PrEP in Europe, which was published in 2016, attracted 15,880 participants. Most were gay and bisexual men but 907 (6%) were cisgender women and 247 (2%) were transgender people, who in this analysis were not split into men and women.

Three-quarters of the non-German gay and bisexual men (German men were analysed separately because of over-representation), 55% of the trans people and 47% of the cisgender women had heard of PrEP. Eighteen per cent of the cisgender women and 44% of the trans people were interested in using it. But even though this implies that 163 women respondents were interested, only four women and five trans people had actually used PrEP.

The UK’s public health agency now categorises people as being in ‘PrEP need’ if they fulfil certain criteria outlined in the UK’s monitoring and evaluation framework for PrEP. During 2021, 7% of people using sexual health services were in PrEP need. The majority of gay men (64.5%) fitted the criteria; in contrast only 1.4% of heterosexual men and 0.5% of heterosexual women did.

But there was substantial under-identification in women. Of people whose behaviour and clinical characteristics, as entered in sexual health clinic records, should have identified them as being in PrEP need, 81% of gay men had their needs recognised at their clinic visit, compared to 49% of heterosexual men and only 33% of heterosexual women. ‘PrEP need’ is a gateway to PrEP provision, because among all groups 70% of those whose need was identified by a clinician ended up receiving PrEP.

The PrEP need criteria have been criticised as too restrictive by the National AIDS Trust, which found that although 143 local authorities responding to a survey reported an average of 203 users per authority (to a total of over 25,000), no local authority reported more than five women using PrEP (implying a maximum total of 600).

Studies have found that while conventional definitions of HIV risk, such as number of sex partners and condom use, may work for gay men, women may be less likely to be aware that they are at risk from partners. Wider risk factors should be taken into account, including intimate partner violence, a history of child sexual abuse, socioeconomic stress, insecure housing, undocumented migrant status and so on. Current intimate partner violence is particularly strongly associated with HIV diagnosis.

Anecdotal evidence suggests that awareness and uptake of PrEP among women may be even lower in other European countries. A recent study in Belgium found that many medical professionals held stigmatising attitudes towards migrant women’s understanding of and ability to use PrEP.

One exception in Europe is Ukraine. Before the current war with Russia, PrEP was already being scaled up and this has continued despite the war; approximately 25% of the people now receiving PrEP in the country are women. Most of them are the partners of HIV-positive men but some are female sex workers or drug users.

Two issues now raise HIV risk for Ukrainian women, many of whom are now dispersed across Europe. Within the country, there has been a large increase in cases of rape, some of it used as a weapon of war. Outside the country, there are reports that migrants, most of them women, take steps to conceal either HIV-positive status or risk behaviour (such as drug use) from people accommodating them and from local services, especially as many are being settled in rural communities they perceive as more conservative.

How many women are using PrEP in Africa?

The number of women initiating PrEP has grown exponentially in Africa, as mentioned at the start of this research briefing. But initiation does not mean persistence, and some studies document extremely low levels of PrEP persistence and adherence among women who start it – even though women often express positive attitudes towards it.

For example, during interviews with Ugandan sex workers who had been given HIV self-testing kits and PrEP pills, participants expressed very positive attitudes towards PrEP, saying it gave them power and control over risk. But adherence was in fact very low, varying from 10 to 30% at clinic visits. One year after starting PrEP, not a single study participant had efficacious levels of PrEP in blood samples.

This is not an isolated example. The largest PrEP provider in South Africa recently reported on 28,100 adolescent girls and young women and 12,581 female sex workers who started PrEP in South Africa. In the year 2020 alone, 14,489 young women and 4618 sex workers initiated PrEP. At clinics serving adolescent girls and young women, 55% of those who tested HIV negative started PrEP, while at clinics serving sex workers, 19% of women who tested HIV negative started PrEP. Just over half the adolescents and young women were aged 19 to 24, while 39% of sex workers were in that age group and a similar number aged 25 to 34.

The majority of women who started PrEP only ever took one month’s supply, if they took it at all. Only 38% of young women and 41% of sex workers returned after their first month’s dispensation. There was considerable variation by site: for instance, at one clinic two-thirds of the young women returned at months two and three for more PrEP but at another clinic only 10% did. Nonetheless, over half of both groups had stopped taking PrEP by month four at even the clinics with the highest persistence rates.

Some women re-started PrEP, in what the researchers called ‘cycling’. Overall, only 10% did and rates of restarting among adolescent girls and young women were generally low, varying from 5 to 27% within 12 months, depending on the clinic. There was considerably more variation in sex workers: from 16% to as many as 80%, depending on clinic, restarted PrEP during the year after first taking it.

One could presume that some of the women were adjusting their use of PrEP to align with periods of actual or assumed risk: while some rapidly discontinued PrEP, others used it in an on-off pattern. However, without some qualitative studies of a proportion of users, it is impossible to go into any more detail about their motivations.

Not all African studies report findings as disappointing at this. In another South African study PrEP adherence levels among adolescent girls and young women in one community PrEP programme in the city of East London were relatively high – 43% had efficacious levels of PrEP in blood samples. This study underlined the extreme importance of disclosure and acceptance of their PrEP use by family and partners, with higher adherence levels reported by women whose use of PrEP was supported by people around them.

A large study of both men and women in Kenya and Uganda – mainly older than in the South African studies – found that only 27% of people identified as being at high risk of HIV initiated PrEP. Of these, 36% dropped out after their first appointment but 56% maintained consistent attendance during the 72 weeks of the study. Two-thirds of this group took drug levels consistent with efficacy.

In this study by far the strongest predictor of whether people initiated, adhered to and persisted with PrEP was whether they thought they were at risk of HIV. This is of course dependent on social mediators such as disclosure of status by partners, whether it is easy to talk about it with partners, friends and family, housing situation and other factors.

A new possibility: injectable PrEP

So the story of PrEP in women, it is fair to say has, ever since the early studies, been marked by disappointing levels of effectiveness, awareness and engagement, lack of access and some scepticism in the wider community that PrEP is an option most women want or can use effectively.

However, some of these assumptions have been demolished by the results of the HPTN 084 study, which were announced in late 2020.

The results and some of the implications were covered in detail in another research briefing. But in brief, HPTN 084 was the phase III study set up to measure the effectiveness of injections every two months of a long-acting formulation of an antiretroviral, cabotegravir, which is already one of the two drugs used in injectable HIV treatment.

Results in the companion study, HPTN 083 in gay and bisexual men and trans women, were already very promising – the cabotegravir injections were 70% more effective than TDF/FTC pills.

But HPTN 084’s results were not so much promising as groundbreaking. The effectiveness of the injections was over 90% – meaning that for every one infection stopped by the oral TDF/FTC PrEP in the other study arm, the injections stopped at least nine.

(In a mathematical model of the study, which compared the number of infections seen with the number that would be expected on no PrEP at all, the TDF/FTC PrEP was estimated to be only about 15% effective. This is not a surprise, given the low adherence figures mentioned above under 'Does the efficacy of oral PrEP differ in women?')

HPTN 084 demonstrated not only by far the best effectiveness seen in any study of PrEP in women, it is the best ever reported for any randomised, placebo-controlled study of PrEP.

Since then, the promise of injectable PrEP for women has only been magnified by recent findings. During a further year of open-label follow-up, no more women acquired HIV while on cabotegravir. Indeed, one woman originally classed as a ‘breakthrough’ infection was found to have acquired HIV before joining the study, raising the effectiveness from 88% to 92%.

In HPTN 083 (the study with gay and bisexual men and trans women), there were a handful of puzzling HIV infections in participants who became infected despite having efficacious blood levels of PrEP. There have been none so far among the women taking part in HPTN 084, and only one infection in a woman with cabotegravir levels even slightly above averagely protective levels.

This suggests that injectable cabotegravir might not just be more effective in women, it might be more intrinsically efficacious. A 2023 study found that once the two-monthly injections began, drug levels in women were 32% higher than in men. In only 2% of women who had injections up to a month late did drug concentrations in blood fall to levels that might even possibly be ineffective. This opens the door to quarterly PrEP injections for women – which would synchronise with existing regimens for injectable contraceptives.

It is important to emphasise that this does not just attest to the efficaciousness of injectable PrEP. Retention in the study was excellent, with only 6% of women (in either arm) discontinuing permanently in the first year. In women on the cabotegravir arm, 11% missed one or more of their injections or had it more than four weeks late, but most only missed one. This means that the injections did not just work medically; they fitted in with the lifestyles and priorities of the women who took part in this study, who were happy to come to the clinic to receive them.

In terms of licensing, the US FDA approved cabotegravir injections for both men and women in December 2021. The World Health Organization issued guidelines urging other countries to license them in July 2022 and has launched a new coalition to accelerate global access to injectable cabotegravir. In November 2022, Zimbabwe became the first African country to approve injectable PrEP.

The drug’s manufacturers, ViiV, applied for authorisation by the EMA in October 2022 and approval is expected soon, though UK approval may take longer as it is no longer part of the EMA system.

As so often in the history of HIV therapy and treatment, cost may be a barrier to early adoption. A 2021 cost-effectiveness study found that to be cost-effective in the US market, injectable cabotegravir must not cost over $2000 a year more than the price currently paid for generic TDF/FTC.

A 2023 study based on the South African market found that injectable cabotegravir would need to cost less than $9 per injection ($54 a year) to be cost-effective in a situation where uptake was high (for example 10% of adolescent girls and young women and 20% of female sex workers). With a somewhat lower level of coverage, it would need to cost $15 per injection ($90 a year). These costings did not include overheads, which will be higher than for oral PrEP, due to extra clinic time and space, staff training and enhanced follow-up for people who miss injections.

Future directions

Injectable cabotegravir will not be the end of the road in the search for longer-lasting and more convenient biomedical prevention methods for women.

Currently, lenacapavir, a drug that can be given by subcutaneous injection (and therefore possibly self-administered) is in a large phase III study, PURPOSE 1, among 5000 women in Africa. PURPOSE 2 is a parallel study assessing lenacapavir in 3000 gay and bisexual men in South Africa and the Americas.

As previously mentioned, PURPOSE 1 also includes, for the first time, an evaluation of the efficacy of oral TAF/FTC in women. The design of the study will allow for comparison of the efficacy of injectable lenacapavir, oral TDF/FTC and oral TAF/FTC. Both PURPOSE studies are expected to report primary results in spring 2024.

Several studies are assessing vaginal rings which can be used for three months at a time and/or also provide contraception. Further down the line, TAF is under study as a subdermal (under the skin) implant, which might only need to be replaced every six months to a year (and the advantage of implants is that they can also be removed easily). Work is ongoing on a dissolvable insert or suppository containing the drugs islatravir and elvitegravir that could be inserted into the vagina or rectum.

Broadly neutralising antibodies as injectable or infusible drugs have already been studied as a form of PrEP. Although some strains of HIV either have pre-existing resistance to these natural products or seem to find it easy to develop resistance to them, there is a large research pipeline looking at combinations of more potent and broad-spectrum antibodies, some in combination with vaccines.

The future of PrEP for women, then, looks brighter than it did during the previous decade. Hopefully it will not be long before women start to see the benefits, not only to their sexual health but to their peace of mind, that gay men using PrEP attest to.