Why aren’t more people with HIV accessing doxyPEP?

Two studies presented at the Conference on Retroviruses and Opportunistic Infections (CROI 2025) in San Francisco last week found that so far relatively few people with HIV are taking doxyPEP (using the antibiotic doxycycline to prevent bacterial STIs), both in absolute terms and in the proportion of doxyPEP users who have HIV.

This seems surprising. In the US and elsewhere, people with HIV have much higher rates of STIs than similar HIV-negative people, especially gay and bisexual men and transgender women. For instance, people with HIV, mainly gay and bisexual men, account for 47% of syphilis cases in the US even though no more than 12% of gay and bisexual men have HIV.

Although there is no evidence that STI acquisition makes it more likely that people with HIV on treatment will become virally unsuppressed and able to transmit HIV, STIs do raise the viral load in people off antiretroviral therapy (ART).

And although the evidence is clearer for viral STIs like HPV and mpox, the bacterial STIs that doxycycline can prevent may take a more aggressive course in people with HIV, especially if they have low CD4 counts.

The first pilot study of doxycycline prophylaxis (as PrEP) was conducted in men with HIV, but the only study that compared the efficacy of doxyPEP in people by HIV status was the pivotal US DoxyPEP study in 2022, in 501 people, 35% of them with HIV. It found that doxyPEP was pretty much as effective in HIV-positive as HIV-negative participants, with an overall efficacy against the three STIs (gonorrhoea, chlamydia and syphilis) of 66% and 62%, respectively.

DoxyPEP prescriptions in Washington DC and San Francisco

Despite this, relatively few people with HIV in the US have so far used doxyPEP. A study presented by Dr Amanda Castel of George Washington University looked for doxyPEP prescriptions among the medical notes of the DC Cohort Study, which includes over 13,000 people with HIV, from all risk groups, attending 14 clinical sites in Washington, DC.

Between 2019 and 2024, 1564 people had at least one STI diagnosis. Among this group, the annual incidence of any STI was 56% and for syphilis was 42%.

Since June 2024, having had at least one STI diagnosis in the past year is the criterion for being offered doxyPEP in US Centers for Disease Control and Prevention (CDC) guidelines. But medical records showed that only 64 people – 4% of the total – were prescribed doxyPEP between 2019 and 2024. Eighty-seven per cent were gay or bisexual men.

Seventy per cent of those prescriptions – 46 – were issued in 2024. Yet the medical records show that 300 individuals, not 46, were eligible for doxyPEP that year.

This is only 15% of those eligible for it. Because this study was of prescription records, it cannot be shown whether some of the under-prescription was due to people declining the offer of doxyPEP, but it seems likely most was due to under-prescribing.

Now that the CDC guidelines have been published, it could be assumed that the low proportion of eligible people with HIV who are prescribed doxyPEP will increase.

This is what might have been expected from San Francisco, the first city to roll out doxyPEP in 2022 and which issued its own recommendations in the same year.

Yet even here, people with HIV appear to be falling behind when it comes to receiving doxyPEP. Dr Mike Barry of the San Francisco AIDS Foundation presented a ‘doxyPEP cascade’ of attendees at the Magnet clinic in 2022-2024. Magnet sees about 6000 people a year and is responsible for about 16% of the syphilis and gonorrhoea diagnoses in San Francisco.

In the three years, 7436 people who attended Magnet had an indication for doxyPEP according to the criteria in the city’s guidelines.

Of these, 59% were prescribed doxyPEP. Of those with a prescription, 61% (36% of all eligible) reported that they took at least one dose of doxyPEP. And of them, 61% (36% of those starting doxyPEP) reported taking doxyPEP correctly for every potential exposure to an STI. Overall, 22% of those with an indication for doxyPEP took it consistently.

Not surprisingly, people without HIV who were already on HIV PrEP were more likely to be prescribed doxyPEP – 71% of those eligible – while those not on PrEP were less likely (44%). As for people with HIV, 56% were prescribed it.

However, people with HIV who were prescribed doxyPEP were considerably less likely to report using it for every exposure. Compared with the average of 61% consistent usage, younger people (below 30) were slightly less likely to use it (58% adherence), but people with HIV were considerably less likely to (50% adherence).

Regarding the entire cascade: compared with an average of 22% of eligible service users who consistently used it, 17% of young people reported consistent use, 19% of people of Black ethnicity, and only 14% of people with HIV. Only people who were homeless or with unstable housing had lower rates of consistent use (13%) than people with HIV.

San Francisco data show efficacy against gonorrhoea

Another presentation from San Francisco modelled the declines in STI incidence that were due to doxyPEP. Initial figures dating from November 2022 were reported at last year’s CROI. One notable difference in this year’s figures was that while the drop in gonorrhoea infections was not statistically significant last year it was significant this time round. Diagnoses of syphilis and chlamydia fell dramatically.

Dr Hyman Scott of San Francisco’s Department of Public Health looked at STI diagnoses in Magnet clinic attendees in people who started doxyPEP, comparing diagnoses in the five quarters before they started, compared with the five quarters after they started. Their data were compared to diagnoses in people who did not start doxyPEP in equivalent time periods.

It is important to emphasise that this study was solely conducted among the 4592 Magnet service users who were taking HIV PrEP, so conclusions about usage and effectiveness among people with HIV, and HIV-negative people not on HIV PrEP, can’t be drawn.

Scott told aidsmap that the researchers limited their analysis to PrEP users because they have regular visits every three months, allowing them to evaluate the impact of doxyPEP in a relatively stable cohort.

“Non-PrEP clients don’t have such regular visits, so it would introduce more bias: people would be more likely to come in for a sexual health screen if they have symptoms or an exposure, and would likely have bigger gaps between their visits,” he said.

Among the 4592 HIV PrEP users, 77% were taking daily oral PrEP, 14.6% on-demand oral PrEP, and 2.7% were on injectable PrEP.

Demographic characteristics of people starting and not starting doxyPEP in this population were remarkably similar, with a little variation in terms of ethnicity. Eighty-nine per cent were cisgender men, 6% identified as non-binary, 3% transgender women, 1% transgender men, and there were a handful of cisgender women. In the five quarters before starting doxyPEP, annual STI incidence in doxyPEP users rose sharply from 8% to 16%, while in non-users it fell, from about 11% to 5%. The odds ratio between users and non-users was 3.78. This much higher likelihood of STI diagnosis is hardly unexpected, given that higher STI incidence is a criterion for starting doxyPEP.

However, in the five months after starting doxyPEP, STI incidence in doxyPEP users fell back to 7%, only slightly, and non-significantly, above incidence in non-users (odds ratio 1.01).

Comparing before doxyPEP versus after it in users, the incidence rate of any STI went down 66%, those of syphilis and chlamydia by 89% and 81%, respectively. Incidence of gonorrhoea declined by 44%, and this figure was statistically significant, unlike the figure from last year, where gonorrhoea only declined by a non-significant 11%. Incidence rates of individual STIs in non-PrEP users did not change significantly.

The efficacy of doxyPEP against gonorrhoea has varied tremendously in studies, from zero in the French IPERGAY substudy to 68% in the Canadian DISCO study (which was of an ongoing daily regimen, i.e. doxyPrEP). This has been ascribed to different rates of antimicrobial resistance in the gonorrhoea bacterium Neisseria gonorrhoeae in different locations.

Scott suggests a different, or complementary reason, which the DISCO study may also help to explain.

“We think in the case of gonorrhoea it might be crucial to treat early. High concentrations of doxycycline may be able to forestall infection by Neisseria gonorrhoeae in the first 24 hours after exposure. The fact that most of the infections that did occur were rectal or pharyngeal and few urethral may also back this idea, because concentrations of drug are higher in the urethra.

“For whatever reason, we’ve recommended that people take doxyPEP as soon as possible after exposure – which implies they have it on the bathroom shelf.

“Our motto has been – ‘make doxyPEP your last load of the night.’”