The Future of HIV Prevention

tray with pills

Three decades into the HIV epidemic, the number of new infections remains distressingly high. In the U.S. the number has held steady at about 50,000 for several years. On a global level, UNAIDS estimates that there were 2.7 million new HIV infections in 2010.

While most groups have seen declines in HIV incidence, rates are rising among men who have sex with men (MSM). This is largely driven by a steep upswing among young black gay men -- a 48% jump between 2006 and 2009, according to the CDC. MSM are at greater risk for HIV infection both because they are more likely to encounter potential partners who are HIV positive and because receptive anal sex is conducive to transmission. In The Lancet, Chris Beyrer calculated that if anal sex were no more risky than vaginal sex, HIV rates among gay men would be at least 80% lower.

Treatment as Prevention

It has long been known that having a low HIV viral load reduces the likelihood of transmission. As early as 2006, Julio Montaner presented findings from a mathematical model suggesting that if all people with HIV worldwide started treatment, the rate of new infections could drop by as much as 70% over 45 years, effectively ending the AIDS epidemic.

In 2008, the Swiss Federal Commission for HIV/AIDS stated that an HIV-positive person on HIV treatment with fully suppressed viral load for at least six months and no other sexually transmitted diseases (STDs) "is not sexually infectious." The Commission stressed, however, that the statement applied only to heterosexuals in stable relationships.

More definitive information came from the HPTN 052 trial, which assigned the HIV-positive partners in more than 1,700 couples either to start HIV treatment right away or to wait until their CD4 count fell to 250. Co-Principal Investigator Myron Cohen got a standing ovation at the 2011 International AIDS Society Conference when he announced that early treatment reduced the risk of HIV transmission by 96%.

There are also hints that treatment is having a prevention effect on a larger scale. Researchers from San Francisco, Washington, D.C., France, Uganda, and elsewhere have reported correlations between earlier or more widespread treatment, lower community viral load, and reduction in new HIV infections.

But often lost in the excitement around treatment-as-prevention is the understanding that transmission can still, if rarely, occur even when an HIV-positive person is on effective treatment and has a very low viral load. People on treatment sometimes experience "blips," or temporary spikes in viral load. Co-infection with STDs can raise the likelihood of both HIV transmission and acquisition. And even people with undetectable viral loads in their blood can have HIV in their semen or vaginal fluid.

A French study reported in AIDS found that about 7% of men in couples seeking assisted reproduction had detectable virus in their semen even though it was undetectable in their blood. Another French study found intermittent shedding of HIV in semen from 13% of gay men with suppressed blood viral load.

Universal Treatment

The trend toward earlier treatment was already under way by the time the HPTN 052 results hit the headlines. Part of the reason for this change was the growing body of evidence suggesting that chronic HIV infection causes persistent inflammation and problems throughout the body long before CD4 counts fall to a dangerous level.

In early 2010, the San Francisco Department of Public Health recommended that all people with HIV should be offered treatment regardless of their CD4 count. "We now know that from the start HIV is causing damage to more than just the immune system," explained Diane Havlir, chief of SFGH's Positive Health Program. "The virus is causing damage to multiple organs -- including the heart, liver, kidneys, and probably the brain -- even when a person has high CD4s and feels fine."

This past March, federal DHHS guidelines followed suit, stating that all people with HIV are eligible for treatment, removing the CD4 threshold of 500 set in 2009. The guidelines stressed that the primary reason for starting treatment early was to protect the health of people with HIV, but they also acknowledged the public health benefit of reduced transmission.

The CDC and several local health departments have recently shifted toward a "test-and-treat" approach that emphasizes routine HIV testing, linkage to care, and universal treatment. On a global level, advocates have embraced treatment-as-prevention as an added incentive for governments to increase funding to get more people in low- and middle-income countries on treatment.

But test-and-treat remains controversial in some quarters. Skeptics think it is too soon to recommend HIV meds for all people with HIV in the absence of data from trials showing that early therapy is beneficial. Such studies may find that the side effects, cost, and "treatment fatigue" associated with universal treatment outweigh its benefits. The START trial could answer many of these questions, but it is not set to report results until 2016.

PrEP: From Trials to the Real World

woman and man holding each other

Pre-exposure prophylaxis (PrEP) refers to HIV-negative people taking HIV meds to prevent the virus from taking hold in the body if they are exposed. In the iPrEx study, Truvada reduced the risk of infection by 42% overall, rising to 92% for people who had drug levels in their blood indicating good adherence. "No one in iPrEx acquired HIV infection with a drug level that would have been expected with daily dosing," said Principal Investigator Robert Grant. But results were not as good for the male-to-female transgender people in the trial, showing no evidence of risk reduction. Researchers are still investigating the reasons for that difference.

Further analysis suggested that men may be able to reduce their infection risk by 90% or more if they take Truvada four times a week, with daily use providing nearly 100% protection.

Two other trials in Africa showed that PrEP also reduces the risk of HIV infection in heterosexuals. The Partners PrEP study of 4,700 heterosexual couples saw infection rates fall by 75% with Truvada, while the TDF2 trial showed a 62% risk reduction.

"PrEP is not a home run and not everyone was protected," said Susan Buchbinder, director of the San Francisco Department of Public Health, at a recent forum to launch the city's PrEP demonstration project. "If you were to take it perfectly every day, would it always work? We don't know the answer."

Like test-and-treat, PrEP has its skeptics. Some have raised concerns about adherence, drug resistance, side effects, decreased condom use, and cost and access issues. If even a majority of study participants who received intensive adherence support in clinical trials could not manage to take PrEP consistently, they ask, how likely is it that healthy HIV-negative people will take a daily pill?

"[PrEP] studies included mostly young, healthy people, but we see people who are older and at higher risk of kidney disease and bone loss," Harry Lampiris of the San Francisco V.A. Medical Center explained. "If they are hepatitis C coinfected they may be at higher risk of kidney problems, and smokers or drinkers are more at risk of bone loss." He added, however, that"being more predisposed to bone or kidney [problems] doesn't mean you can't take PrEP -- it just means you need to take more precautions."

The only way to answer such questions is to try PrEP in the real world. To this end, demonstration projects are starting in several cities. First out of the gate are San Francisco and Miami, which launched projects for MSM and transgender women in September. Similar efforts are under way in Chicago, Los Angeles, New York City, Oakland, and Washington, D.C.

"So far everything we know comes from clinical studies in which people didn't know whether they were taking PrEP or placebo," said Stephanie Cohen, Medical Director at San Francisco City Clinic where the demonstration project will take place. "There's still a lot to learn: Who will take it and how? Will risk factors change? How do we get PrEP out there safely? And can we provide PrEP in the busy environment of an STD clinic?"

Issues of cost and access also must be addressed. Could PrEP become a "boutique" intervention available only to people with money or good health insurance? Some private insurance companies are already covering Truvada PrEP, recognizing that for high-risk individuals it can be cost-effective compared with lifelong HIV treatment.

But with a price tag of approximately $1,000 per month, PrEP could further strain local prevention budgets and programs like Medicaid and Medicare. Acknowledging that the low-income and minority populations most heavily affected by HIV are more likely to rely on public programs, HIV Medicine Association Chair Judith Aberg stressed that PrEP "must not contribute to HIV-related health care disparities."

Most experts do not expect that large numbers of HIV-negative people will want to use PrEP for life. Rather, as Montaner explained at the 2012 International AIDS Conference, it is likely to be a "time-limited niche intervention for selected individuals at high risk for infection."

Future PrEP

Daily Truvada may turn out not to be the best PrEP regimen. The IPERGAY study and the ADAPT trial are testing intermittent Truvada taken on a regular but less frequent schedule, or "event-driven" PrEP taken just before or after sex.

Other drugs may offer equal or better protection, and a drug less widely used for HIV treatment could allay concerns about resistance. Long-lasting drugs, such as an experimental once-monthly injectable form of the HIV med Edurant, could help with adherence.

The HIV meds Isentress and Selzentry may be useful for PrEP, and the ongoing NEXT-PrEP trial is testing Selzentry in gay men. But disappointing findings presented at this year's International AIDS Conference showed that oral Selzentry did not protect monkeys from rectal infection.

Taking pills might not be the answer. Another kind of PrEP is applied where it's needed -- in the vagina or rectum. Researchers have studied a large number of gels, creams, and films, looking for a microbicide that can both provide a physical barrier against HIV entry and disable the virus with antiretroviral drugs.

The CAPRISA 004 trial showed that a tenofovir vaginal gel applied before and after sex (using the applicator below) reduced the risk of infection by 39% overall, and by 54% for those with the best adherence. Though this level of effectiveness pales in comparison with the oral Truvada PrEP studies, it was considered the first real biomedical prevention breakthrough when findings were announced at the 2010 International AIDS Conference.

The same gel was also tested for rectal use. While it was found to be safe, it wasn't very well tolerated, causing gastrointestinal side effects and bloating. Fortunately, a reformulated gel containing less glycerin has been found to be much more tolerable. The new gel is being tested in the MTN-017 trial, which compares daily oral Truvada, daily tenofovir gel, and tenofovir gel applied before and after receptive anal sex.

Different PrEP formulations may work better for specific groups. For example, a small study published in Science Translational Medicine showed that tenofovir gel produced drug levels in vaginal tissue 100 times higher than those seen with oral dosing.

Researchers are also looking at vaginal rings that release drugs slowly over time. One recent study showed that a ring containing MIV-160 protected monkeys against vaginal exposure. Further along in the pipeline, the Phase 3 ASPIRE trial is testing a vaginal ring containing dapivirine in more than 3,000 women in five countries.

"We're trying to give people as many options as possible," Buchbinder said at the recent San Francisco PrEP forum. "For some, a monthly injection is the best option, for some using [an antiretroviral] lube during sex is a great option, and some would rather take a daily pill."

One of the major concerns surrounding treatment-as-prevention and PrEP is that taking a protective pill might lead people to abandon other risk-reduction strategies. But most studies have seen the opposite effect. In trials of early treatment, PrEP, microbicides, and vaccines, risky behavior decreased and HIV infection rates fell in both the active and placebo arms. In iPrEx, for example, participants across the board reported more condom use and fewer sex partners.

Behavior Still Matters

While public health officials insist that condoms, monogamy, and abstinence are the only 100% sure means of protection, gay men are voting with their dicks and taking a harm-reduction approach. An informal poll taken at a recent serosorting forum in San Francisco revealed that 97% of respondents sometimes had condom-free sex and most thought condom use was not the norm in the community.

Given this reality, serosorting and other strategies based on HIV status have become a mainstay of risk reduction for MSM -- even though they are not as effective as many people seem to think (see the previous issue of Achieve).

Overall, studies show that serosorting is more effective at preventing HIV than no risk-reduction strategy at all, but far less effective than using condoms. An analysis by Matthew Golden and colleagues found that serosorting did not lower the risk for black gay men -- the group with the highest rate of new infections.

The Achilles heel of serosorting, of course, is incomplete or incorrect information. The CDC estimates that 20% of people with HIV are not aware they are infected. Some have never been tested, but even people who get tested regularly may have been infected recently, before they produce enough antibodies to show up on a standard HIV test (several weeks to a few months after infection). Viral load can be very high during this period, and experts estimate that at least half of all new infections are transmitted by people recently infected.

Because HIV meds are so effective at preventing transmission, having unprotected sex with an HIV-positive person on stable treatment with an undetectable viral load may actually be less risky than doing so with a random person you think is HIV negative. But not everyone with HIV is fully suppressed. According to the CDC, only 28% of people with HIV in the U.S. (about 75% of those on treatment) have an undetectable viral load.


What about condom-free sex between people who are already HIV positive? Regardless of HIV status, unprotected sex can also spread other STDs. Some public health officials blame serosorting for rising rates of chlamydia, gonorrhea, and syphilis among gay men in several cities. Equally worrisome are outbreaks of sexually transmitted hepatitis C among HIV-positive gay men. The emergence of multidrug-resistant gonorrhea illustrates that treatment is not always a simple matter of a shot or a few pills, and some infections like herpes and HPV cannot be cured.

And what about superinfection? Public health campaigns that discourage condom-free sex between HIV-positive men often warn about the risk of becoming infected with additional HIV strains that might be more aggressive or drug-resistant. Study data, however, are wildly inconsistent.

"Several analyses over the years have found that superinfection occurs rarely, and typically happens only in people with a recent initial infection. But rarely doesn't mean never."

Several analyses over the years have found that superinfection occurs rarely, and typically happens only in people with a recent initial infection. But rarely doesn't mean never. Superinfection may go unrecognized if it does not have clinical consequences such as rising viral load or worsening symptoms, and most people do not undergo the sensitive genetic sequencing needed to reveal multiple viral strains.

The first widely reported cases of superinfection came to light in 2002. Stephanie Jost from the University of Geneva described a 38-year-old gay man who became infected with two different HIV subtypes on two separate occasions more than two years apart. That summer, Bruce Walker from Massachusetts General Hospital reported a case of a man who was infected a second time with a new HIV strain within the same subtype.

Fears about superinfection increased in 2004 when Geoffrey Gottlieb described a man who experienced rapid progression to AIDS after a more aggressive HIV strain, acquired more than a year after his first infection, overtook his original virus. And Davey Smith reported on a gay man in San Diego who was superinfected with a second HIV strain that was resistant to two classes of antiretroviral drugs, leading to a jump in viral load.

The U.S. and Europe have generally seen very low rates of superinfection, with only about 50 confirmed cases (some from studies and some from case reports). A 2004 analysis of stored blood samples from 78 recently infected people in Southern California who were not taking HIV treatment showed a superinfection rate of 5%, all occurring within a year after initial infection. But a 2011 study of 15 gay men in Amsterdam who had not yet started treatment found no evidence of superinfection, despite unprotected sex.

Superinfection is even less common in people taking HIV treatment. Matthew Gonzales of UCSF found no cases of superinfection in an analysis of HIV sequences from 718 people with HIV who were mostly on treatment. The Positive Partners study, which looked at long-term couples in which both partners were HIV positive but carrying different strains, likewise saw no evidence of superinfection.

Some studies in Africa, however, have seen higher rates of suspected superinfection. Julie Overbaugh and her team have been following a cohort of female sex workers in Kenya for nearly 20 years. By 2012 they had analyzed viral sequences from 110 women and identified 19 cases of superinfection. They calculated an annual superinfection rate of about 3%, similar to the rate of initial infection in this group. While half the cases occurred during the first two years after initial infection, two people were superinfected after five years.

In another African study, Andrew Redd from NIAID sequenced virus from 149 people who were recently infected in Uganda. They identified seven cases of superinfection for a rate of 1.4 per 100 person-years -- similar to the local initial infection rate of about 1.2 per 100 person-years.

A Comprehensive Approach

The shift from behavioral interventions to biomedical prevention in an era of scarce funding has raised concerns about putting too many eggs in one basket. If treatment-as-prevention takes priority over treatment-as-treatment, some advocates fear that people with HIV will be subtly pressured, or even coerced, into starting therapy before they are ready, for the sake of public health. "Prevention for positives" puts the onus on people with HIV to prevent transmission, while HIV-negative people risk being left out of the picture. On the HIV-negative side, shifting resources to PrEP could mean getting rid of culturally sensitive behavioral approaches tailored for specific groups.

"The problem with prevention for many is that you cannot avoid dealing with sex and drugs," former UNAIDS head Peter Piot declared at the 2006 International AIDS Conference. Biomedical prevention should not be an excuse to avoid difficult conversations, nor to neglect tried-and-true strategies like explicit sex education, condom distribution, and needle exchange that are proven effective but remain politically sensitive.

If recent studies have a common theme, it is that different strategies work for different people. There is no magic bullet, and a combination prevention approach that includes risk-reduction counseling, community support, condoms, and now biomedical interventions is likely to be most effective.

"It is easy, but somewhat naive, to say that people should just try harder to use condoms," said Dana Van Gorder of Project Inform. "We must give people at risk for HIV the ability to choose for themselves which proven prevention methods are most likely to respond to their needs."

"Choice matters," Mitchell Warren of AVAC concurred. "For the millions of men and women who remain at risk for HIV worldwide, each new HIV prevention option offers additional hope that we will achieve the end of the epidemic."

Liz Highleyman is a San Francisco-based freelance medical writer and editor-in-chief of