Canvassing CROI

The successes of ARV therapy and promising results with new HIV prevention strategies stoke excitement at recent scientific meeting

By Kristen Jill Kresge and Regina McEnery

At the opening session of the 16th Conference on Retroviruses and Opportunistic Infections (CROI)—which was held this year from February 8-11 in Montreal, Canada—the 3rd N’Galy-Mann Lecture was given in tandem by Glenda Gray and James McIntyre, executive directors of the Perinatal HIV Research Unit at the University of the Witwatersrand in Soweto, South Africa. They spoke about the challenges and opportunities associated with conducting HIV research in South Africa and provided a historical recounting of many of the successes that have been transformative in this community, including the provision of antiretroviral (ARV) therapy to HIV-infected individuals and the use of ARVs to prevent mother-to-child transmission of the virus.

Indeed it seems many hopes in combating HIV these days are pinned to ARVs, whether it is in expanding access among HIV-infected individuals worldwide, developing microbicide gels based on existing ARVs, or using them as a means of pre-exposure prophylaxis (PrEP) to block HIV infection. It is also clear that ARV therapy has its limitations. Putting aside the immediate side effects, researchers are now gaining more insights into longer-term complications associated with treatment, which were the focus of several sessions at CROI. And a pair of studies provided evidence to suggest that complete suppression of HIV replication, which is routinely measured in blood, does not necessarily correspond with lack of virus shedding in semen, suggesting HIV transmission may still be possible even if individuals are on ARV therapy. It is now also becoming more apparent that eradicating HIV from infected individuals is unlikely to be accomplished with ARVs alone.

All of this suggests that research on new HIV prevention strategies, including microbicides, PrEP, and vaccines, will continue to be top priorities. And this year’s CROI showcased some promising results from both clinical trials and animal studies evaluating microbicides and PrEP, providing a burst of enthusiasm around new HIV prevention strategies. “It’s an exciting time in the prevention field,” said Sharon Hillier, vice chairman of the Department of Obstetrics, Gynecology, and Reproductive Sciences at the University of Pittsburgh. Researchers are also still mining for clues from the STEP trial, as well as from individuals who can successfully control HIV infection, to gain insights into the types of immune responses that may be critical to vaccine-induced control of HIV.

The cause of persistent viremia

Robert Siliciano, professor of molecular biology and genetics at Johns Hopkins University, delivered the Bernard Fields Memorial Lecture this year and spoke about new ways of understanding and evaluating the efficacy of ARV therapy, which when taken properly, can often completely stop new cycles of viral replication. However, residual viremia still occurs even in individuals on completely suppressive ARV therapy. This ongoing, incredibly low-level viremia, which when quantified with super-sensitive assays is approximately 1 copy of HIV RNA/ml of blood, is coming from at least two sources, according to Siliciano. One is the long-lived, and long ago described, reservoir of latent HIV-infected CD4+ T cells. While activated T cells die rather quickly, HIV’s genome can be integrated into long-lasting memory T cells on their way back to a resting state very early in the course of HIV infection, what Siliciano calls the “perfect recipe for persistence.” He estimates that only one million of these intrinsically stable, resting CD4+ T cells harbor latent HIV, but because of their slow decay rates, “it would take over 70 years to eradicate a reservoir of a million cells,” Siliciano said.

Another still undetermined reservoir is now also believed to be contributing to the persistent, yet minute viremia. “The residual viremia is complicated,” said Siliciano. He has observed that residual viremia in individuals on suppressive ARV therapy is dominated by a small number of viral clones that are not found in resting CD4+ T cells. The goal of rooting out and eliminating these pools of virus-infected cells is still the focus of much research, but so far at least one strategy—treatment intensification—in which additional potent drugs are added to an individual’s existing ARV regimen doesn’t seem to be the answer. “We will never reduce residual viremia any further with ARV drugs,” said Siliciano.

Two other studies presented at CROI looked at whether suppression of HIV in the blood corresponds with suppression of the virus in other compartments. Several studies have established a strong relationship between HIV viral load, as measured in blood, and heterosexual transmission rates. These studies indicate that individuals on ARV therapy with very low or undetectable viral loads are less likely to transmit the virus. However, some studies now suggest that in some individuals on suppressive ARV therapy, there is still ongoing viral shedding in semen, indicating that HIV transmission may still be possible.

In the first study, researchers from the University of Toronto followed 25 HIV-infected individuals who had never been on ARVs and used the branched DNA assay, which is better at detecting HIV RNA in semen, to measure their HIV viral loads in blood and seminal plasma samples following initiation of highly active antiretroviral therapy (HAART). They found that a significant proportion—12 of the 25 individuals—had isolated HIV shedding in semen over a six-month period after initiation of HAART, which resulted in an undetectable viral load in blood (fewer than 50 copies/ml). In four of these individuals with undetectable viral loads in blood, levels of isolated semen HIV shedding were greater than 5,000 copies/ml in seminal plasma.

Researchers also studied 12 individuals who had been on a suppressive HAART regimen for at least four years and found that four of them also had isolated HIV shedding in semen. Prameet Sheth, a PhD student at the University of Toronto who presented this study, said the virus that was shed was potentially infectious—HIV isolated from the individual with the highest level of seminal shedding (16,000 copies/ml of seminal plasma) was capable of infecting activated CD4+ T cells in in vitro studies. “Our study shows that even though HAART will be able to reduce sexual transmission of HIV on a population level, there is still an individual risk that exists despite long-term HAART,” said Sheth.

Researchers did not find any association between the ARVs used and the level or frequency of HIV shedding in seminal plasma in these individuals, and they observed a wide differential in penetration of ARVs in both seminal and blood plasma.

In a second study, researchers at the Hospital Pitié-Salpêtrière in Paris collected blood and semen samples from 145 HIV-infected, ARV-treated males who were participating in a reproduction program that allows men to have their sperm washed of HIV for transplantation into their uninfected female partners. Over a six-year period they collected 264 paired blood and semen samples and found undetectable HIV viral loads (fewer than 40 copies/ml of plasma) in both 85% of the time. Nine of the paired samples, only about 3%, showed detectable levels of HIV in both blood and semen. However, Anne-Geneviève Marcelin, who presented the study, reported that seven men or 5% of the study population had detectable HIV levels in seminal plasma even though there was no detectable virus in blood. Marcelin said these results suggest a “small, residual risk of transmission is still possible during unprotected sexual intercourse.”

Another perennial question regarding HIV transmission is whether the virus that is transmitted and establishes infection is typically cell-free virus in plasma, detectable by viral RNA, or cell-associated virus in lymphocytes, detected by proviral DNA. To determine which of these viral reservoirs in male genital secretions is the primary source of HIV infection, David Butler, a post-doctoral student at the University of California in San Diego, studied four HIV transmission pairs, all of whom were men who have sex with men (MSM). Blood samples were collected an average of 59 days following the estimated date of HIV infection, and semen samples were collected from the infecting partner an average of 72 days after transmission. Genetic sequencing was used to analyze both the transmitted and infecting virus, and in all four cases Butler concluded that the virus that was transmitted and established infection was cell-free virus originating in the infecting partners’ seminal plasma. He is now planning to study a greater number of transmission pairs.

First hint of microbicide efficacy

Some of the more encouraging data at CROI came from a triumvirate of clinical and nonhuman primate studies with new HIV prevention strategies. The first study, known as HPTN 035, evaluated the safety and efficacy of the microbicide candidate PRO 2000, a topical gel composed of 0.5% of a synthetic polyanionic polymer that non-specifically acts to block attachment of HIV to host cells.

This Phase IIb study enrolled 3,099 women at seven clinical trial centers in Africa and the US and evaluated the efficacy of PRO 2000, as well as a second topical microbicide called BufferGel, which contains an agent designed to boost the natural acidity of the vagina in the presence of seminal fluid.

The study also had two control arms—one received a placebo gel and the other, which was unblinded, received only condoms and no gel. A no-gel arm was included in the trial over concerns that the placebo might have antimicrobial properties that could protect against HIV.

The results of this study showed that women who were randomly selected to receive both PRO 2000 gel along with condoms had 30% fewer HIV infections than those who received the placebo gel and condoms. At the conclusion of this three-year trial, there were 36 HIV infections among women in the PRO 2000 group, compared to 54 in the BufferGel group, 51 in the placebo gel group, and 53 in the no-gel group.

However, Salim Abdool Karim, a clinical infectious disease specialist who led the PRO 2000 study, cautioned that the results were not statistically significant compared to either the placebo gel or no-gel groups. “This could be a chance finding,” he said, adding that additional evidence would be necessary to “conclusively determine whether PRO 2000 is an effective microbicide.”

When researchers analyzed the data based on adherence, they found that women who reported using the gel at the last coital act at least 85% of the time, had an overall 44% reduction in HIV infection compared to women who received the placebo gel. And in women who reported using the gel that often without regularly using condoms, there was a 78% reduction in HIV infection compared to the placebo group.

There was a palpable level of excitement following Karim’s presentation, with many audience members rushing to the microphones to congratulate the researchers on the conduct and results of the trial. Karim said this excitement was understandable given the recent results from two trials of other microbicide candidates. Carraguard, made from a seaweed derivative, was found last year not to reduce the risk of HIV acquisition in a three-year, Phase III study of 3,200 women in South Africa. And a Phase III trial of cellulose sulfate that had enrolled 1,333 women was discontinued in December 2007 after early data suggested that the microbicide candidate might be contributing to an increased risk of HIV infection.

“We are at the end of a series of disappointments,” Karim said. “We need something that gives us hope. The HPTN 035 trial results represent that hope.” A Phase III study of PRO 2000 conducted by the Microbicide Development Programme in the UK is nearing completion in South Africa, Tanzania, Uganda, and Zambia. This trial has enrolled 9,000 women, and results, which are expected in late 2009, will provide additional data on whether PRO 2000 is effective at blocking HIV transmission.

New animal data on PrEP

Other excitement came from two nonhuman primate studies, conducted by the US Centers for Disease Control and Prevention (CDC), which provided additional evidence for the effectiveness of PrEP. Studies in nonhuman primates have shown that ARVs administered systemically prior to exposure to a simian immunodeficiency virus (SIV)/HIV hybrid known as SHIV can prevent infection, although the success of this intervention seems to vary based on both the challenge model and the ARVs (see PrEP Work, IAVI Report, Nov.-Dec. 2008).

One study at CROI compared intermittent oral PrEP use—a strategy referred to as iPrEP—with daily dosing. In this study, CDC scientists administered the human equivalent doses of oral Truvada—a combination pill of two ARVs, tenofovir and emtricitabine (FTC). Four groups of six macaques received two doses of Truvada, one prior to and one following rectal SHIV challenge. The first dose was administered as long as seven days before or as soon as two hours prior to challenge, and the second dose was administered either two or 22 hours after challenge. These animals were then compared to 32 untreated controls.

All animals were challenged once a week with SHIV over a 14-week period, with a dose of Truvada administered before and after each of the 14 challenges. It took a median of two challenges to infect the untreated control animals. However, three of the six animals in the groups that received Truvada either two hours before and 22 hours after, or seven days before and two hours after challenge were protected against SHIV infection throughout the 14 weeks.

The best results were seen in the group that received Truvada either 22 hours before and two hours after, or three days before and two hours after challenge. In these two groups, five of the six animals were completely protected against SHIV infection over the 14-week period. J. Gerardo García-Lerma, the CDC researcher who presented these findings, reported that comparable levels of both tenofovir and FTC were seen in the infected and uninfected animals. He also said researchers observed a blunted level of acute viremia in the macaques that were infected despite PrEP, as compared to controls.

All of the ongoing PrEP trials are testing the efficacy of daily dosing, but there is interest in iPrEP because of the concern that daily adherence could prove to be a major barrier to PrEP effectiveness. Intermittent PrEP use would also dramatically slash the cost of providing this intervention. Just how the drugs are given intermittently could be driven by exposure or based on a fixed-dosing schedule.

Results were also presented from a topical PrEP study that compared the effectiveness of gels containing either 1% tenofovir or a combination of 1% tenofovir/5% FTC against repeat, low-dose vaginal SHIV challenge in female pigtailed macaques. Two groups of six macaques received either the 1% tenofovir gel or the 1% tenofovir/5% FTC combination gel. These groups, as well as two animals who received no gel and nine who received a placebo gel, were challenged twice a week over the course of the 10-week study.

The two animals who received no gel were infected after three and five SHIV challenges respectively, while eight of the nine animals who received the placebo gel were infected after a median of four challenges. However, both groups of six animals who received either the tenofovir or tenofovir/FTC combination gel were completely protected against SHIV infection after 20 challenges.

Concentration levels of the drugs in blood were measured in the treated animals 30 minutes after each gel application. Charles Dobard, the CDC researcher who presented this study, reported that only about 0.3% of the drug was absorbed systemically. This could be an advantage of topical PrEP, in that less systemic drug absorption could lead to fewer potential side effects and make it less likely that drug-resistant strains of HIV would develop in individuals who become HIV infected despite using the gel.

There are currently six clinical trials of PrEP involving nearly 21,000 volunteers. The VOICE study, which involves 4,200 women in Africa, is comparing the safety and acceptability of oral PrEP to a topical microbicide formulation. The first data on the effectiveness of PrEP from clinical trials will be available next year.

Mining for vaccine clues

Meanwhile, researchers are continuing to mine data from the STEP trial, which showed that Merck’s adenovirus serotype 5 (Ad5)-based vaccine candidate known as MRKAd5 had no effect on HIV infection and may have enhanced susceptibility to HIV infection in certain sub-groups of volunteers, including those with pre-existing antibody immunity to Ad5. One possible explanation for the increased susceptibility to HIV infection is that individuals with higher Ad5 antibody levels would also have higher levels of Ad5-specific T cells, which following vaccination, would expand and become activated, creating more target cells for HIV. But a study presented at CROI cast doubt on this hypothesis. Using flow cytometry, researchers measured levels of Ad5-specific T cells from volunteers in an earlier Phase I trial with MRKAd5 administered at the same dose and schedule as in the STEP trial. They analyzed samples from 25 volunteers prior to vaccination—some of whom subsequently acquired HIV—and found a similar magnitude of Ad5-specific T-cells among all trial volunteers, as measured using IFN-γ ELISPOT assay, regardless of their pre-existing Ad5 neutralizing antibody levels. And after vaccination, there was no significant difference in the level of Ad5-specific CD4+ T cells between individuals who remained uninfected or seroconverted during the course of the study. Although this suggests Ad5-specific T cells are unlikely the cause of an increased susceptibility to HIV, researchers were unable to rule out preferential trafficking of activated Ad5-specific T cells to the mucosal sites as a mechanism for increased risk of HIV infection.

All volunteers in this Phase I trial who seroconverted became infected after a single vaccination. Yet participants with no pre-existing Ad5 immunity developed both antibody and Ad5-specific cellular responses following the first vaccination and researchers still did not observe an enhanced susceptibility to HIV infection among these volunteers after subsequent vaccinations. This would suggest that it is unlikely that Ad5 immunity increases susceptibility to HIV infection following Ad5-based vaccination, said Natalie Hutnick, a molecular biologist from the University of Pennsylvania who presented the findings.

Another study, presented by David Heckerman, senior director of eScience at Microsoft Research, described what he called “a hidden success in the STEP trial.” Heckerman, along with colleagues at the Ragon Institute including its director Bruce Walker and Florencia Pereyra, set out to identify what best predicts viral control in HIV-infected individuals. Using a predictive model they developed, researchers analyzed 148 HIV controllers—individuals who maintain viral loads of less than 2,000 RNA copies/ml blood without treatment—and 102 chronic progressors. Certain previously identified human leukocyte antigen (HLA) alleles are associated with control of HIV, yet some individuals with these alleles still have high viral loads, Heckerman said. This led researchers to hypothesize that it is not the HLA allele but rather the specific cytotoxic T lymphocyte (CTL) epitopes they target that are responsible for control of virus. They found that, among these individuals, recognition of optimally defined CTL epitopes were the best predictor of viral control—even better than HLA class I alleles.

Heckerman and colleagues then identified six of what they referred to as “good or protective” HIV epitopes in these viremic controllers that were significantly associated with viral control. Some of these epitopes correspond to HLA alleles that were not previously thought to be protective. Researchers then analyzed post-immunization, pre-infection immune responses in a group of 19 participants from the STEP trial, who subsequently became HIV infected, to see if targeting of these six specific epitopes was associated with a lower setpoint viral load among vaccinated volunteers. Of the 19 participants, not a single person had responses to more than one of these six epitopes, but those who responded to at least one had lower viral loads than those who didn’t.

Ten of the 19 individuals studied had the A*02 allele, but only four responded to LV10 on Nef, one of the protective epitopes they identified. Heckerman suggested that this could be because their immune systems were distracted by targeting other epitopes. The four participants with A*02 that did respond to LV10 had lower viral loads. Heckerman concluded that the design of a successful immunogen may therefore hinge on inclusion of good epitopes as well as exclusion of others that could just distract the immune system.

Clinical data on protein vaccine

Data was also presented from a Phase I/II dose-escalation study of the vaccine candidate F4/AS01, developed by GlaxoSmithKline Biologicals (GSK), which consists of a recombinant fusion protein (F4) comprised of four clade B HIV antigens—Nef, reverse transcriptase from Pol, and p24 and p17 from Gag. The vaccine candidate was administered along with the company’s proprietary AS01 adjuvant to 180 volunteers at three different doses (10 µg, 30 µg, and 90 µg). Marguerite Koutsoukos, project leader of the HIV vaccine program at GSK, reported that all volunteers who received two intramuscular injections of the vaccine candidate at the lowest 10 µg dose developed CD4+ T-cell responses to at least three antigens in the vaccine candidate and 80% had responses to all four, as measured by intracellular cytokine staining for expression of interleukin (IL)-2 and at least one other marker of activation, including either expression of TNFα, IFN-γ, or CD40L. The majority of F4-specific CD4+ T cells secreted IL-2 alone, or in combination with TNFα, IFN-γ, or both, and were persistent. “A substantial CD4+ T-cell response was maintained throughout the entire study period,” said Koutsoukos.

Very low CD4+ T-cell responses were observed in volunteers who received the F4 vaccine without the adjuvant, and no CD8+ T-cell responses were observed in any volunteers. GSK is now evaluating the candidate in a Phase I trial in HIV-infected individuals to explore the potential of the candidate in a therapeutic setting. According to Koutsoukos, since the candidate only induces a CD4+ T-cell response, and not a CD8+ T-cell response, the company would consider testing the candidate prophylactically in “a more complex regimen including other strategies.”