Back HIV Prevention Pre-exposure (PrEP) PrEP by the Numbers: Interpreting Conflicting Study Data

PrEP by the Numbers: Interpreting Conflicting Study Data


A recent ad campaign from the AIDS Healthcare Foundation has kicked off a new round of controversy about HIV pre-exposure prophylaxis (PrEP) and how to interpret apparently conflicting numbers from prevention studies.

The AHF ad-- entitled "PrEP Facts" -- features a chart showing overall efficacy in 8 major studies of PrEP using Truvada (Gilead Sciences' tenofovir plus emtricitabine coformulation) or tenofovir alone in pill or gel formulations. Figures range from a high of 75% and 62% in the Partners PrEP and TDF2 trials of heterosexual couples in Africa, to 49% in the Bangkok Tenofovir Study of people who inject drugs, to less than 10% in the Fem-PrEP study of African women.

"Overall efficacy was low in all studies due to low adherence," the ad states. "[AHF] has taken the position that the scientific data do not support the large-scale use of Truvada as a community-wide public health intervention." AHF President Michael Weinstein went so far as to suggest that "the government-sanctioned widespread scale-up of PrEP appears to be a public health disaster in the making."

In contrast, proponents say PrEP is highly effective -- in the 90% to 100% range -- when used consistently.

Once-daily Truvada was tested in clinical trials, and this is the regimen approved by the U.S. Food and Drug Administration (FDA) in July 2012 and recently recommended by the U.S. Centers for Disease Control and Prevention (CDC) and by the World Health Organization. Further analysis of study data has shown that Truvada can still provide a high level of protection if people miss some doses. Intermittent or "as needed" PrEP taken before and after sex is currently being studied, but it is not yet known whether this is effective.

The dueling data arises from the different ways researchers can report medical study results.

In randomized, controlled clinical trials, the gold standard is "intent-to-treat" results, or how well did a drug work among everyone assigned to take it. This includes people who may not have used it regularly or who stopped early due to side effects or for other reasons. This is usually the primary type of data reported in medical journals and at scientific conferences.

Another method, known as "as-treated" or "per-protocol" results, considers only the subset of study participants who actually took the drug as directed -- that is, those with good adherence.

In general, as-treated results will make a treatment look better than intent-to-treat results. However, intent-to-treat data are thought to better reflect outcomes in the real world. A drug may have strong antiviral potency, for example, but if many people find it too inconvenient or too toxic to use regularly, its actual effectiveness in practice may be low.

Numbers from iPrEx

The debate about PrEP numbers mainly revolves around the various results from the iPrEx study, which tested Truvada PrEP in gay and bisexual men and a small number of transgender women. Many women at risk for HIV are also interested in PrEP, but so far they have less data to rely on.

iPrEx enrolled about 2500 HIV negative participants at 11 sites in Brazil, Ecuador, Peru, South Africa, Thailand, and the U.S. They were randomly assigned to take either Truvada or an inactive placebo once-daily. The study was blinded, meaning neither the participants nor the researchers knew who was taking what. Everyone also received regular HIV testing and a package of prevention services including risk-reduction counseling and free condoms. After the main study ended, participants could continue taking Truvada in the iPrEx open-label extension, or OLE.

As Robert Grant from the Gladstone Institutes and colleagues reported in the December 30, 2010, New England Journal of Medicine, during an average follow-up period of about 1 year, the risk of HIV infection was 44% lower overall in the Truvada group compared with the placebo group in an intent-to-treat analysis of the main iPrEx study -- the figure shown in the AHF ad.

About half of iPrEx participants reported taking Truvada as directed most of the time, and in this group the risk of infection dropped by 73%. Among people with blood drug level measurements indicating good adherence, risk reduction was 92%. A mathematical model later estimatedthat risk reduction would reach 90% if men took Truvada at least 4 days per week, and could hit 99% with daily use -- the figure often cited by PrEP advocates.

About 1600 participants opted to join the iPrEx OLE. About three-quarters chose to keep taking Truvada, while the rest served as an untreated control group. Unlike the original randomized part of the trial, OLE participants knew that they were getting the active drug and were informed about findings showing that it reduces the risk of HIV transmission.

As Grant reported at the recent International AIDS Conference in Melbourne, the overall risk reduction in the OLE was 49% in the Truvada group compared with the untreated control group. Among people who had blood drug levels indicating that they took Truvada at least 4 times per week, there were no new infections -- 100% efficacy. However, only one-third managed to reach this level of adherence.People who took Truvada 2 or 3 times per week had a risk reduction of 84%, while those who took fewer than 2 doses saw no significant protection.

What Does It All Mean?

With so many numbers being thrown around, many people are understandably confused about which figures to rely on. Yet the issue is not whether the 44% figure or the 99% figure is "true," but rather what these numbers mean in real life.

Taken together, studies of gay men (iPrEx), heterosexual couples (Partners PrEP and TDF2), and people who inject drugs (Bangkok Tenofovir Study) show that Truvada -- or even tenofovir alone -- is highly effective with consistent daily use.

"We know that people who did take the drug had very high levels of protection," stressed Jim Pickett from the AIDS Foundation of Chicago. "This drug works extraordinarily well to prevent HIV if you use it."

But randomized studies and open-label demonstration projects have found that many people do not maintain good adherence. A U.S. PrEP demonstration project found that while 92% of San Francisco participants had blood drug levels indicating that they took Truvada at least 4 times per week, this fell to 57% in Miami. At the other extreme, adherence levels in 2 studies of women in Africa -- VOICE and Fem-PrEP -- were so low that they could not demonstrate a protective effect.

While Weinstein argues that condoms are the best way to prevent HIV infection, PrEP advocates stress that most gay men do not use condoms all the time. The level of protection in condom studies also varies widely. While effectiveness in preventing pregnancy approaches 100% with perfect use, a study by Dawn Smith from the CDC and colleagues found that gay men who reported using condoms all the time were about 70% less likely to become infected than those who said they use them sometimes or not at all.

"Does that mean we tell guys [condoms] are only 70% effective? No, partly because some guys probably didn't use them religiously or correctly," explained David Evans of Project Inform. "So what we say is that they are highly effective if used correctly and used every time. That's what most of us are saying about PrEP."

Project Inform has prepared it's own version of the illustration used in the AHF ad adding on the efficacy levels seem among participants with high levels of adherence.

Importantly, iPrEx and other studies have found that using PrEP does not make people less likely to use condoms (known as risk compensation), but in fact is associated with less sexual risk behavior. And compared with condoms, PrEP is more forgiving of less-than-optimal adherence.

"Having used a condom yesterday provides no protection if you don't use a condom today," said Grant. But the iPrEx OLE results "demonstrate that PrEP remains highly effective, even in real-world circumstances in which adherence may not be perfect."

"It's important to note that PrEP is not simply a pill, it's a program," Pickett emphasized. "It's a program that includes quarterly testing for HIV, regular STD screens (and STD treatment as necessary), medical monitoring, adherence support, and sexual health counseling."

Damon Jacobs, a marriage and family therapist in New York City, explained that difficulty maintaining good adherence is hardly unique to PrEP. For example, a CVS study of drugs for high blood pressure, high cholesterol, diabetes, and depression found that fewer than 50% of people with prescriptions took their medications consistently.

"We don't see Michael Weinstein telling people not to take their blood pressure meds because less than half of people can adhere in the real world," Jacobs said. "Let's talk about the barriers and what can we do to promote adherence."