Dissent on doxyPEP: recent guidelines becoming more cautious

Two recent statements about taking the antibiotic doxycycline up to 72 hours after sex to prevent bacterial sexually transmitted infections, a prevention intervention known as doxyPEP, are notably more cautious about recommending its use than the first guidelines from California.

The Australian Society for HIV Medicine (ASHM) released a Consensus Statement in March which, in fact, had to record that it failed to reach complete consensus on its proposed guidelines for doxyPEP use. The statement summarised a meeting on 17 March in which there was general agreement that doxyPEP should be considered “primarily” for the prevention of syphilis in gay and bisexual men who are at risk of it – but added that “for some individuals the reduction in chlamydia and lesser reduction in gonorrhoea might be important.”

Forty-five of the 50 people in the survey conducted prior to the meeting agreed with this formulation but four advocated strongly for the position that only the prevention of syphilis should be included in the final guidelines, while one person felt so strongly about this that they asked for their name and their participation to be withdrawn.

Two participants gave reasons for their dissent. One said they were concerned that using doxyPEP against gonorrhoea – where studies have reported efficacy ranging from 55% to zero – risked raising the levels of resistance to the tetracycline class of antibiotics, which includes doxycycline, not only in gonorrhoea, but in related bacteria that could acquire tetracycline resistance by gene-swapping.

The other participant thought that using doxyPEP to prevent chlamydia could only be justified in bisexual men with urethral chlamydia, where it could be passed to female partners – chlamydia is largely asymptomatic and self-limiting in men, but can cause pelvic inflammatory disease and infertility in women.

The recommendations arrived at by the meeting included both the majority formulation and the dissenting view. ASHM say that doxyPEP should be suggested for gay and bisexual men with a syphilis diagnosis in the previous six to 12 months (leaving the time period open to the physician), or with two or more STIs other than syphilis in the same period, or who plan a period of heightened STI risk, e.g. a sex party, or who are having vaginal sex with cisgender female or transgender male partners.

The recommendations also recommend avoiding multiple doses, for example someone having multiple sex partners over a weekend can just take a dose on Monday morning. Since STI risk can change, they stress that doxyPEP use be reviewed every three to six months.

They add that community organisations should be funded to provide regular updated clinical and educational information, and that funding should be provided to STI clinics and public health bodies to closely monitor antimicrobial resistance (AMR) in gonorrhoea infections, both generally and particularly in people acquiring gonorrhoea while on doxyPEP.

These Australian guidelines are by no means the most cautious. On 30 November last year, the Belgian Research HIV Consortium (BREACH) released a short statement saying they had reached a consensus “not to recommend the widespread use of doxyPEP for the prevention of STIs.”

They had two concerns. Firstly, they said that in the studies, although doxyPEP proved effective for syphilis and chlamydia, and in some cases gonorrhoea, “the number of symptomatic infections was low or not reported”. They pointed out that screening initiatives that aimed to identify and treat asymptomatic STIs, such as the Belgian Gonoscreen programme, did not produce falls in gonorrhoea incidence – so why should doxyPEP?

The second objection was to do with antimicrobial resistance which is, BREACH says, “a huge concern in STI and other bacteria”, with “some studies showing an increase in antimicrobial resistance.”

The conclude: “Considering that the number needed to treat to avoid a symptomatic infection with gonorrhoea or chlamydia is probably very high, the considerable potential to select resistance in STIs and other bacterial species and the possibility of disrupting the microbiome, it is considered that the potential individual benefits of doxyPEP are outweighed by the risks in men who have sex with men.”

They recommend that, if doxyPEP is to be prescribed, is should only be done as part of a research study.

DoxyPEP: latest findings

Before analysing some of the claims in the BREACH statement, it is worth looking at what the doxyPEP studies have found and why the enthusiasm for this prevention idea has become more qualified over time.

Although prophylaxis against STIs has a long history, dating back to the First World War, the two most important studies were the US DoxyPEP Study, conducted in San Francisco and Seattle, and the DOXYVAC study, conducted in Paris.

The DoxyPEP study was closed in 2022, a year earlier than planned, when the efficacy of doxycycline against the three STIs was found in an interim analysis to be 66% (62% in people with HIV). This compound efficacy was an average of the 88% and 87% efficacies seen against chlamydia and syphilis (in HIV-negative people), and the lower but still useful efficacy of 55% seen against gonorrhoea.

The DOXYVAC investigators had lower expectations, at least regarding gonorrhoea, as an earlier French study had found zero efficacy against gonorrhoea. For this reason, they combined doxycycline PEP with an experimental gonorrhoea vaccine.

DOXYVAC was also stopped early, prompted by the results from DoxyPEP. The efficacies against chlamydia and syphilis were exactly the same as in the DoxyPEP study. But the study also found ‘unexpectedly’ that the efficacy of doxycycline against gonorrhoea was 51%. Furthermore, the vaccine appeared to be 51% effective – and there were indications that where the PEP was especially effective against anal and urethral infections, the vaccine worked better in the throat.

These promising results produced early guidelines that were very positive about doxyPEP. These were issued not only in response to the news from the trials, but because sexual health clinics were being flooded by requests for doxycycline from users, including many who’d started to use the drug without medical supervision, and were asking about safe use.

Worryingly, some had been using other antibiotics, including azithromycin. This is still used as an alternative treatment for gonorrhoea in people allergic to ceftriaxone, the primary treatment drug, and clinicians are very keen to prevent any more resistance to it.

The San Francisco Department of Public Health guidelines were the first to be released, in October 2022. They recommended doxyPEP to cisgender men and trans women who had had one or more bacterial STI in the last year, or “who report condomless anal or oral sexual contact with one or more cis male or trans female partner in the past year.”  

The California Department of Public Health followed with its guidelines in April 2023, saying that they “would like to inform all health care providers of a compelling new biomedical intervention to prevent bacterial STIs”. They recommended that clinicians proactively offer doxyPEP to “all non-pregnant individuals at risk of bacterial STIs and to those requesting it, even if [they] have not been previously diagnosed with an STI.”

The strong backing for doxyPEP in these guidelines encouraged an expansion in its use. Uptake among people offered doxyPEP was 74%, and by September 2023, 3288 people had started doxyPEP at three signature clinics in San Francisco. Positive public health effects were seen: in the first 13 months after doxyPEP implementation, syphilis cases in San Francisco dropped by 78% and chlamydia by 67%. There was no significant decline, however, in gonorrhoea cases.

The news was less good from France, where a re-analysis of the data from the DOXYVAC study found that, although the efficacy against chlamydia and syphilis was largely confirmed (at 86% and 79% respectively), doxyPEP’s efficacy against gonorrhoea fell to only 33% (still statistically significant but not very useful), while the efficacy of the vaccine was only 22% and was not statistically significant.

This lack of clear individual and public-health efficacy against gonorrhoea poses a problem in how to communicate this to existing or potential users of doxyPEP. Guidelines describing the intervention as effective against “STIs” will need to make it clear that it is unlikely to be effective against gonorrhoea.

Drug resistance – in gonorrhoea and other organisms

The lack of efficacy is due to the gonorrhoea organism already being largely resistant to doxycycline – the very reason it stopped being used as a treatment in the late 1980s. Resistance in fact varies substantially from place to place. The proportion of gonorrhoea resistant to antibiotics of the tetracycline class ranges, according to one survey of European countries, from 14% in Estonia to 94% in Portugal (and 92% in France). It is also very local, and dependent on previous prescribing practice – in Spain, next door to Portugal, it is only 18%.

However we do know from the studies that even if gonorrhoea resistance starts off low, the use of doxyPEP is likely to raise it. In the DOXYVAC trial, the proportion of gonorrhoea samples with high-level resistance was greater in the doxyPEP group (33%) than in the control group (19%). In the DoxyPEP study, although there was half as many cases of gonorrhoea in the doxyPEP arm than the control arm, they had 2.3 times the odds of being tetracycline-resistant.

While such high levels of resistance mean that doxyPEP might not prevent gonorrhoea, they are unlikely to directly imperil gonorrhoea treatment, which uses ceftriaxone, a drug from a different antibiotic family with no cross-resistance.

What worries bacteriologists more is the possibility that tetracycline resistance genes in the Neisseria gonorrhoeae bug might get transferred to other species of bacteria. This can happen when bacteria swap genes horizontally by exchanging packets of DNA called plasmids.

In the DoxyPEP study, samples taken from study participants showed fewer infections with the bacteria Staphylococcus aureus, a feared nosocomial (hospital-acquired) infection, but did find an absolute increase of 8% in the proportion of infections with tetracycline resistance. Doxycycline resistance also emerged rapidly in two other bacteria that cause nosocomial infections, Staphylococcus aureus and Klebsiella pneumoniae, when they were cultivated in the lab dish with Neisseria gonorrhoeae and doxycycline.

At present, fears about major public health consequences are theoretical. However the authors of guidelines in Australia, Belgium, Germany and British Columbia would like to see more research before authorising doxyPEP for more than a restricted group of people who are already at high risk of STIs, especially syphilis.

Are the benefits of doxyPEP worth the risks?

The other question the Belgian statement brings up is: is doxyPEP worth it? The authors’ reservations centre on two issues. The first is that about 90% of chlamydia infections and nearly 90% of rectal and throat gonorrhoea infections are asymptomatic and are often self-limiting – the immune system will eventually get rid of them. In these cases, they argue, doxyPEP is unnecessary.

However they also assert that syphilis is often asymptomatic. In fact, though the initial chancre (sore) can be missed, early syphilis symptoms including neurological and ocular  symptoms are not uncommon and can be severe, and relying on testing and treating to control syphilis would miss some of these cases. Also mentioned above is the fact that chlamydia infections can have more severe consequences in women, so special consideration may need to be given to people with both male and female partners.

The Belgian paper also asserts that “The number to treat (NNT) or avoid a symptomatic infection with gonorrhoea or chlamydia is probably very high". A a 2023 modelling study found the number of people one needed to dose with doxycycline to prevent one new chlamydia infection was 2.9 to 7.2, depending on how tightly one drew eligibility criteria (e.g. giving it to all people at risk, all PrEP/ART users, or just people with previous STIs). The NNT was 9.5 to 31.1 for syphilis because it’s less common.

However those figures are for preventing any infection, either symptomatic or asymptomatic. If, however, it it is only relevant to prevent symptomatic infections, then the NNT for chlamydia would be in the region of 30-70. This is comparable to the NNT for HIV PrEP for US gay men, which ranges from 10 to 70, depending on the population studied.

Meanwhile, public health bodies continue to give the impression of not being sure what to say about doxyPEP.  A draft set of guidelines released by the US Centers for Disease Control for comment last October have not yet been published in final form (and are no longer on their website), French authorities have not yet published any recommendations, and the British Association for Sexual Health and HIV (BASHH) has still not endorsed doxyPEP, despite noting the large numbers of people who are already self-sourcing antibiotics.