A case note review from 56 Dean Street, the largest HIV and STI clinic in Europe, has found that former PrEP users who get HIV are more likely to have a certain type of drug resistance than people who have never taken PrEP. However, this had little if any impact on their response to HIV treatment once they were diagnosed and started antiretroviral therapy.
If people take PrEP while unknowingly infected with HIV, the currently used PrEP drugs, while strong enough to prevent HIV infection, are not capable of bringing an already-established infection under control. A third or more powerful drug is generally needed to suppress viral reproduction to the point of undetectability.
Taking some, but not enough, of a drug or drug combination is what causes drug resistance. The drug still kills the non-resistant organisms, but if there are ones that happen to have mutations that confer drug resistance (as a few will at random), then they will survive. The pressure from the drug “selects” them and they soon become the predominant population.
In the case of PrEP, this could be caused either by:
- starting PrEP without having an existing (perhaps very recent) HIV infection diagnosed,
- by taking PrEP erratically, or
- by stopping it but having an HIV infection soon after stopping so that drug is still present in the body.
How often does this actually happen? To find out, the Dean Street Collaborative Group of researchers reviewed notes for all patients testing HIV positive between July 2015 and January 2019. After excluding 47 people with no HIV resistance test recorded at diagnosis, they found 22 clinic attendees who were diagnosed HIV positive and mentioned having taken PrEP within the previous year, and compared them with 917 not known to have taken PrEP.
The patients were overwhelmingly gay men (there were 13 women and 26 exclusively heterosexual men, none of whom had taken PrEP) with an average age of around 32.
Interestingly, the PrEP users had a considerably lower average viral load at diagnosis – 7700 versus 58,000 copies/ml. There is evidence from both animal studies and cases in humans that PrEP taken at the time of an acute HIV infection may result in an HIV infection with a considerably lower viral load – so low in some cases that it becomes difficult to test for.
Probably due to the lower viral load in former PrEP users, in 18% of the patients who’d taken PrEP (four out of 22), the clinic tried but failed to perform a baseline drug resistance test, compared with 2% of the 917 non-PrEP users. The researchers were looking for evidence of the two most important mutations that confer resistance to the two drugs used in PrEP, emtricitabine and tenofovir.
The M184V (or M184I) mutation confers resistance to emtricitabine (and to lamivudine). It arises very easily – within 2 to 3 weeks in someone taking either of these drugs as part of a regimen that fails to suppress HIV. As a result, it’s very common: in one recent study it arose in 88% of people who failed to suppress HIV with their first therapy regimen.
However M184V/I is not such a serious matter clinically. This is partly because the M184V mutation damages the ability of HIV to replicate: viral loads in people with the mutation who continue to take emtricitabine or lamivudine, despite virological failure, may be 50% lower than people without the mutation. It even confers hypersensitivity to some other HIV drugs. In addition, because of their susceptibility to resistance, HIV clinicians do not depend on these two drugs to be the ‘backbone’ in HIV regimens: others can be substituted.
The other mutation they looked for was K65R. This arises much more rarely but has more serious implications. It causes resistance to the drug tenofovir, used in PrEP, and also to abacavir, and to a lesser extent causes resistance to most drugs in the NRTI (nucleoside reverse transcriptase inhibitor) class. As NRTIs, the oldest class of HIV drugs to be developed, are still used as the 'backbone' of first-line therapies, K65R might cause difficulties in putting together a first-line regimen in resource-limited settings with comparatively few regimens to choose from.
But K65R is rare: in fact, it was non-existent in this study, which found no patient with it in their baseline resistance test, whether or not they had taken PrEP.
"M184V/I is not such a serious matter clinically."
In contrast, five out of the 22 previous PrEP users had the M184V/I mutation (23%), compared with only five out of 917 non-users (1%).
Of the 22 people who’d previously used PrEP, 20 started antiretroviral therapy, a median of ten days after diagnosis. Seventeen were still in care three months later and 16 had an undetectable viral load at this point.
Four of those with an undetectable viral load were people with the M184V/I mutation. The other person with M184V dropped out of care, at least at 56 Dean Street. All four people with M184V/I had emtricitabine as part of their regimen, for reasons described above.
This case note review has some limitations. It features a small group of PrEP takers who may not be typical of ones who started it more recently; it doesn’t compare treatment success rates with people who had not taken PrEP; and it doesn’t estimate the time of HIV acquisition.
It also can’t say whether people developed M184V due to taking PrEP while already having HIV – or whether they caught HIV that already had that drug resistance mutation in the first place, which was why their PrEP failed. Emtricitabine reaches high levels in rectal tissues much faster than tenofovir does, so could be offering the sole protection against HIV early on after a first dose of PrEP. This might make intermittent PrEP, especially, less effective in people exposed to HIV with the M184V mutation.
Tittle V et al. Antiretroviral resistance and management after pre-exposure to prophylaxis. The Lancet HIV 7:e34, online ahead of print, February 2020 (open access).