Tim Lahey, M.D., M.M.Sc.
Recently we got great news from a real-world study of HIV pre-exposure prophylaxis (PrEP).
Researchers at San Francisco's Kaiser Permanente Medical Center reported on the real-world experience of 657 people who started PrEP between 2012 and 2015. Over 99% were men who have sex with men, 84% reported multiple sexual partners, and 30% had HIV-positive partners. Together they were observed for over 388 person-years of PrEP use.
The big news: Exactly zero patients contracted HIV infection! That's huge.
We already knew that PrEP works. In the seminal double-blinded placebo-controlled randomized iPrEx study, PrEP cut HIV transmission by 44%. Better adherence to tenofovir/emtricitabine (Truvada), the drug used in PrEP, led to even better protection: The efficacy of PrEP was 73% among subjects who took PrEP over 90% of the time. Subsequent studies have confirmed that PrEP is a potent weapon against HIV transmission, most notably the PROUD study, which showed 86% lower rates of HIV infection among men who have sex with men randomized to immediate instead of deferred PrEP.
The Kaiser Permanente study adds to these data by showing that PrEP works in patients in the real world, not just research subjects under the microscope of a huge clinical trial. This is important because adherence to medications often is lower in clinic compared to research studies.
Perhaps because the Kaiser Permanente study confirms what we already believe -- PrEP works! -- press coverage of the study has been particularly rabid.
In a rare concordance, the Huffington Post claimed, "Truvada Protected 100 Percent Of Study Participants From HIV" at the same time Fox News reported that PrEP was "100 percent effective." The pertinent Washington Post headline was, "100 percent of participants taking HIV prevention pill Truvada remained infection-free." Clearly inflamed by all those 100s, popular sex-positive podcaster Dan Savage said, "Clearly Truvada is insanely effective at preventing HIV infection."
It is right to be excited about the Kaiser Permanente data, but we should step back, put down those shiny 100s for a second, and take a deep breath. It is dangerous and frankly a little alarming to froth at the mouth like that.
First, the Kaiser Permanente study did not show PrEP was 100% effective. Without a comparator group the study cannot formally evaluate the efficacy of PrEP. Furthermore, we don't know how many sex partners the study patients had, or how many condomless episodes of anal intercourse they engaged in, and thus how at risk they were for contracting HIV. Were subjects who had the initiative to start PrEP more likely to serosort, or PrEP-sort, or to engage in other risk reduction activities at the same time? I bet they were a savvy group of men who protected themselves in more than one way while having fun, and we should keep that in mind when ascribing the totality of their HIV-free well-being to PrEP.
Not only is there reason to doubt the claims of 100% efficacy in the Kaiser Permanente study -- which by the way the authors themselves did not make -- but more definitive studies suggest the efficacy of PrEP is lower than 100%. The most precise estimate comes from randomized placebo controlled clinical trials like iPrEx, which control for the confounding factors that can skew the estimates of other studies. In iPrEx, overall efficacy was 44%. Even among iPrEx subjects with over 90% adherence to PrEP, efficacy was only 73%. In the open label follow-up study to iPrEx, creatively called iPrEx OLE, there were zero HIV infections among subjects taking Truvada at least four times a week. Importantly, each of these estimates are the high end of PrEP efficacy and pertain only to the small subset of study subjects who adhered to PrEP the best. It is conceivable that the men in the Kaiser Permanente study were freakishly adherent to PrEP to a degree not seen in previously published studies, but I'd bet a 55-gallon tub of lube that reality is more complicated.
It is tempting to link the purported 100% efficacy claimed by the press in the Kaiser Permanente study to the reported 86% efficacy estimate in the PROUD study and say that's about how well PrEP works in the real world. Perhaps the more confidence we have in the efficacy of PrEP the more consistently people take it, and so results over time will be better than the original iPrEx study. It's possible. But not proven, and far from certain. For instance, we should remember that PROUD too can only estimate PrEP efficacy imperfectly. Instead of comparing the efficacy of PrEP now to placebo now, the PROUD study compared PrEP now to PrEP later, which means other factors that vary with time -- like concurrent risk behaviors, condom use and serosorting -- may have changed as well in a fashion that affected the accuracy of the efficacy estimate.
I am not suggesting PrEP is bad, or ineffective, or that the Kaiser Permanente study is bunk. Far from it! I think PrEP is a huge step forward, I prescribe it, and I think the Kaiser Permanente study was a very nice piece of work. I might even frame it.
But we should report the study results responsibly, in part because tossing around unfounded 100% efficacy estimates can have real danger. If at-risk men hear they are 100% protected by PrEP, and as a result decide they no longer need to use condoms or take any number of complementary risk reduction steps that men in the Kaiser Permanente study likely took, then our sloppy oversell could lead directly to new cases of HIV despite PrEP.
A sobering side note in the Kaiser Permanente study makes this point nicely: After 12 months of PrEP use, fully 50% of people who took Truvada were diagnosed with a sexually transmitted infection and 41% indicated their condom usage had decreased. This raises the possibility, not seen in prior studies but perhaps a realistic reflection of our evolving understanding of PrEP, that PrEP use could undercut other risk reduction activities. Surely the more effective people think PrEP is the greater the chance they could put down the condoms and throw caution to the wind.
PrEP works, it works safely, and it works particularly well if taken consistently. More at-risk people should take PrEP, and more insurers should pay for it. But let's admit it: PrEP shouldn't put the condom industry out of business just yet, and PrEP works best if taken with other risk reduction activities and, most importantly, a daily dose of common sense.
Comment by: John
Tue., Oct. 27, 2015 at 5:30 pm UTC
I think people are finding it hard to get their heads around the fact that HIV is combatable without using condoms due to slut shaming and the control the sexual health professionals have over Gay and MSM (and others).
It's a problem of homophobia and control. Gay and bisexual men are the devil incarnate - how dare they have pleasure or slip up every so often. Their desires must be CONTROLLED by the judging of society as a whole.
They are not of our approved risk takers like skiiers, rugby and football players, grand prix racers and horse racing jockeys.
Once they have produced vaccines for HIV, and a great doxycycline for all the other nasties I'm not sure what some of these people will say: don't do it because it's immoral - no sex before marriage!
Comment by: Ryan Racino, PharmD, AAHIVE
(San Diego, CA)
Tue., Oct. 6, 2015 at 1:33 am UTC
Thank you SO MUCH for writing this article. I've been saying this for years that the media sensationalizes the important work that researchers do and with attention spans getting shorter and shorter, abbreviating important study results like the Kaiser study or PROUD (or any of them) creates such a dangerous scenario as you so eloquently describe here. This is the message that needs to get disseminated to the general public with emphasis and less about the real data. Educators, physicians and others can educate on the numbers and the raw data, but what people really need to learn on a fundamental level is that they need to scrutinize the media sources they get their information from. The headlines you quote in your article are prime examples of this. If any of these news articles (as well-intentioned as they may be) misinformed just one person and led to just one infection, that is one infection too many.
Comment by: Kevin Paul
(San Diego, CA)
Thu., Sep. 24, 2015 at 7:45 pm UTC
Nothing is ever 100%, but I do believe PrEP is at least 98% over-hyped. Gilead, the makers of Truvada, on their own web page tell users they "must" continue with the use of condoms. They say users "can still become infected with HIV-1," and if it happens, you must start another medication because Truvada by itself could make "your HIV-1 harder to treat." Along with the possible "severe liver damage," the hype 'has' been one sided. Getting facts from social media, Truvada is being presented as an 'alternative' to condoms.
We now have a generation taking pharmaceuticals to lower viral loads and/or to block viral 'loads.' It is a generation now capable of creating a drug resistant pathogen creating a health crisis that will make the 80's look like a bad case of scabies.
Gilead is taking advantage of the gay community. No one should be surprised if there are federal agencies looking into the pharmaceutical's promotional campaign.
Comment by: Gus Cairns
Wed., Sep. 23, 2015 at 2:25 pm UTC
I'm sorry, but think the only conclusion to be reached if you use the evidence form this study is that yes, PrEP is 100% effective - at least here - and that the study was not over-hyped at all. Despite all the caveats about reduced condom use and high STI rates (though, without a baseline comparison, we do not know if these have increased) this is a remarkable result, demonstrating 100% effectiveness for PrEP, in a ‘real world’ community setting, in those who start it. I think this should calm concerns that as PrEP usage spreads, adherence and effectiveness rates will fall. The increased concern about other STIs is entirely justified, but is a measure of how successfully PrEP appears to be at dealing with HIV infection. I'd like to know how many other health interventions that are precribed by primary care physicians have a 100% success rate.