A Shot in the Arm for Therapeutic Vaccines

By Mary Rushton

Therapeutic vaccination has long been considered the “Rodney Dangerfield” of HIV research, for somewhat good reason. None of the vaccine candidates tested so far has demonstrated efficacy, and the tremendous success of treating and controlling HIV with antiretroviral therapy (ART) stymied enthusiasm for this approach. Until four years ago when regulators approved a vaccine for metastatic prostate cancer, it was questionable whether therapeutic vaccination was useful in the treatment and control of any disease.

Therapeutic vaccine research faces several of the same hurdles as preventive vaccine research. Both strategies require induction of immune responses that are qualitatively different from those induced during natural infection because in all but rare cases those are insufficient in controlling the virus. Both also suffer from lack of exact nonhuman primate or mouse models. But could therapeutic vaccines at last be earning some respect?

Italian immunologist Barbara Ensoli, who has been working on therapeutic vaccine strategies since the late 1990s, certainly thinks so. “It’s very dynamic and promising, as a number of different therapeutic approaches are in clinical testing,” says Ensoli, director of the National AIDS Center’s Istituto Superiore di Sanità in Rome. “It is quite probable that major advancements will first come from the therapeutic setting rather than the preventative side.”

The therapeutic candidates tested in clinical trials in recent years have used a variety of platforms and approaches, including DNA, viral vectors, dendritic cells, and peptides. A few induced transitory reductions in viral load in the context of treatment interruption and modest delays in the time to viral load rebound. Although the clinical benefit of this type of response is unknown at this point, recent findings suggest therapeutic vaccines might have a role in an HIV cure strategy, according to findings presented at a two-day meeting on the topic held last year in Bethesda, Maryland, and recently published (Vaccine 32, 5540, 2014). This is in large part what is fueling the resurgence in therapeutic vaccine research.

“There is definitely more interest in therapeutic vaccines than in the past,” says Yegor Voronin, senior science officer at the Global HIV Vaccine Enterprise, the New York City-based group that organized the meeting in partnership with the HIV prevention advocacy group AVAC and Treatment Action Group.

However, financial support for the present-day pipeline of therapeutic vaccine candidates is still pretty dismal. Data compiled by the HIV Vaccines and Microbicides Resource Tracking Working Group shows only US$11.5 million was spent on therapeutic vaccine research in 2013, a 45% decrease from 2011 and a fraction of the $818 million spent on preventive AIDS vaccine research in 2013. But this could all change quickly. “A lot will depend on the results of some of the trials that are now underway. People are hedging bets on what they think will likely work,” says Voronin.

Most cure strategies are combining therapeutic vaccine candidates with other immune-modulatory drugs primarily used in cancer therapy. Anthony Fauci, director of the US National Institute of Allergy and Infectious Diseases (NIAID) and a driving force behind the HIV cure renaissance, suggested at the AIDS 2014 meeting in Melbourne that therapeutic vaccination would likely need to be a component of any successful cure strategy.

Aside from their potential role in combination cure strategies, therapeutic vaccines could also be a stand-alone approach to suppress HIV in the absence of ongoing ART. Broadly neutralizing antibodies (bNAbs), which are a main focus of preventive vaccine research these days, may also be useful in therapeutic vaccination based on encouraging animal studies that suggest infusion of a single bNAb or cocktails of different bNAbs can suppress HIV replication for short periods of time in untreated animals. Two studies are also underway at NIAID and Rockefeller University exploring passive transfer, or direct injection of bNAbs in both HIV-infected and uninfected volunteers. The use of adeno-associated virus (AAV) vectors as a vehicle to deliver antibody genes rather than directly injecting the antibodies—a strategy being studied in animals and humans as a way of preventing HIV acquisition—may be another way to harness the power of the antibodies in therapeutic vaccines.

Hitching on to a cure

One of the biggest obstacles to an HIV cure is the pool of virus in latently HIV-infected cells that constitute, at least in part, what remains the largely unchartered territory of the HIV reservoir. Antiretroviral therapy suppresses viral load but does not deplete the viral reservoir. If at any point antiretroviral therapy is interrupted, this latent virus often resurfaces, resulting in ongoing viral replication and progressive disease (see CROI: Progress on Prevention and Cure).

Attempts to force this latent virus out of hiding using a variety of substances remains an area of intense investigation. Now, researchers are evaluating whether combining this approach with a therapeutic vaccine designed to boost the immune responses against the awakened virus, a so-called “Shock and Kill” strategy, could reduce or even eliminate the viral reservoir.

Six months ago, Danish researchers began testing a peptide-based vaccine candidate called Vacc-4x with Celdene’s cancer drug romidepsin in a small study involving 20 HIV-infected volunteers. Romidepsin is among a handful of histone deacetylase (HDAC) inhibitors undergoing testing to assess their ability to roust HIV from the latently infected T cells that make up at least part of the viral reservoir. Vacc-4x, composed of four synthetic peptide sequences from within the highly conserved HIV p24 core protein, was tested alone in a Phase IIb trial but did not decrease viral loads among vaccinated HIV-infected volunteers as compared to placebo recipients after they discontinued highly active antiretroviral therapy (HAART). There was also no difference in CD4+ T-cell counts at the end of the HAART-free period between the two groups.

However, further analysis indicated Vacc-4x reached one of its secondary endpoints—the virus levels in participants who received Vacc-4x never returned to pre-treatment levels, which is what typically happens when treatment is interrupted. There was a statistically significant difference in viral load in volunteers who received the therapeutic vaccine candidate as compared to placebo recipients. Now, the hope is that the vaccine candidate in combination with an HDAC inhibitor could effectively deplete the viral reservoir, and possibly even be a step toward an HIV cure.

Cure research is growing dramatically. While funding for most categories of HIV prevention research declined in recent years, HIV cure research grew a whopping 421% over the last three years, according to the HIV Vaccines and Microbicides Resource Tracking Working Group. US investment in this area is expected to increase even more with President Barack Obama’s announcement that $100 million in US National Institute of Health (NIH) funds will be re-prioritized to launch a new HIV Cure Initiative.

In the six years since Timothy Brown, the so-called “Berlin patient,” was the first to be considered cured of HIV, an increase in cure-related studies is generating a wealth of data that are helping to characterize the viral reservoir and develop strategies to combat it. Along with therapeutic vaccination and transcriptional activators (like HDAC inhibitors), scientists are also exploring how bNAbs, epigenetic agents that can induce changes in the genes controlling behavior of HIV provirus, immune-modulators like the checkpoint protein PD-1, and immune targeting might also contribute to a combination HIV cure strategy (see Much Accomplished, Much More to Achieve).

Yet curing HIV is proving to be a huge challenge. Brown was treated for acute myelogenous leukemia, receiving two allogeneic bone marrow transplants from a donor who was homozygous for the CCR5∆32 mutation, which renders cells resistant to CCR5-tropic HIV—not what you’d consider a widely replicable approach. And his case so far is unique.

A flurry of reports over the past year underscore how difficult it will be to achieve even a functional cure—defined as the lack of detectable viral replication in the absence of ongoing ART. The groundbreaking case of the “Mississippi baby,” an infant considered potentially cured of HIV two years ago following early initiation of ARVs (see A Toddler Stole the Show), was reported to once again have detectable levels of virus following interruption of ART (see Much Accomplished, Much More to Achieve). Similar relapses occurred in a three-year-old Italian boy whose virus rebounded weeks after treatment was suspended (The Lancet 384, 1320, 2014), and two HIV-infected males from Boston who, like Brown, received stem cell transplants for cancers (see CROI: Progress on Prevention and Cure), though unlike Brown their donors were not homozygous for the CCR5∆32 mutation.

“These recent experiences drove home for me that we are going to need some way to survey what residual virus might persist,” says Steven Deeks, a professor of medicine at the University of California, San Francisco, who treated Brown and is studying different HIV cure strategies. “The Mississippi and Boston cases were people we thought might be cured, but many months and years later the virus rebounded and did so without people knowing it was happening. What that says to me is that we are going to need something to maintain control of the virus and the best way, I think, is through a vaccine.”

Finding therapeutic vaccine candidates that work effectively enough in HIV-infected people won’t be easy though, says Stuart Shapiro, who leads a vaccine discovery team at NIAID’s Division of AIDS. “We know that those already infected have been primed in a way for the immune responses to be inadequate. So we have to find a way of redirecting the immune response. That’s much more difficult than figuring out how to direct it in the first place,” says Shapiro. “Secondly, people who are infected have a large amount of virus in them so you need a much larger immune response than you would if you were simply trying to prevent people from becoming infected. The vaccine has to be much more potent.”

The potency issue

Potency of therapeutic vaccine candidates does seem to be a problem. Two years ago, Spanish researchers found that stimulating dendritic cell (DC) function in HIV-infected individuals briefly controlled their viral load after discontinuing HAART. Unfortunately, the reprieve observed in the Phase I trial was short-lived (Sci. Transl. Med. 5, 166ra2, 2013).

To make the vaccine candidate, the research team led by University of Barcelona scientist Felipe Garcia extracted autologous monocyte-derived dendritic cells (MD-DCs) along with HIV from the blood of 36 HIV-infected individuals on HAART, and used heat to inactivate the HIV in 22 of the 36 samples. They then vaccinated the 22 individuals three times over a six-week period with high doses of their own DCs and with their own intact HIV. The immunizations were given either before or immediately after interruption of ART. Twelve weeks after treatment interruption, researchers observed a 90% drop in setpoint viral load in 12 of the infected individuals who received DC cells pulsed with the inactivated HIV, compared to just one in the control arm. By week 45 the virus had rebounded, though. Nonetheless, the study was important in that it was the first randomized placebo-controlled study of a therapeutic vaccine candidate showing a statistically significant downward trend in viral load.

Because the logistics of developing a candidate like this for each patient is prohibitive, Garcia’s team now plans to target the DCs in vivo with a rationally designed messenger RNA-based immunogen. The HIV antigen is based on viral targets of protective HIV-specific T-cell responses that were previously identified in three large cohorts of HIV-infected individuals. The candidate also includes a TriMix of three different immunostimulatory molecules that systematically alter the activation status of DCs and enhance the induction of antigen-specific T cells. This vaccine candidate is being developed as an alternative to ART.

Like the Spanish study, results were also fleeting in a randomized, placebo-controlled study of HIV-infected individuals on HAART who received a replication defective adenovirus serotype 5 (Ad5) gag vaccine candidate that was also tested prophylactically. The therapeutic study of this vaccine candidate showed a trend toward lower viral load following treatment interruption, but the results did not quite reach statistical significance (J. Infect. Dis.(202)5, 705, 2011).

One way around the potency problem, though, might be to use therapeutic vaccines to try to eliminate only those HIV-infected cells that are reactivated from the viral reservoir—a much smaller target. This is the approach that Ole Schmeltz Søgaard of Aarhus University is taking in the Danish trial of romidepsin and Vacc-4x. Søgaard created a buzz at the International AIDS Society’s two-day cure symposium in Melbourne this past July when he reported that the HDAC inhibitor romidepsin provoked such robust bursts of virus replication in people whose virus had been suppressed for years due to HAART that their HIV became detectable on standard blood tests. Other scientists were quoted in news reports hailing the tiny six-person study as one of the key cure-related findings of the meeting.

In the current trial, in which Søgaard and his team are evaluating the safety and tolerability of the “shock and kill” or “kick and kill” approach using romidepsin combined with Vacc-4x, the HIV peptides in the vaccine are injected with granulocyte macrophage-colony stimulating factor (GM-CSF), which sparks CD4+ and CD8+ T-cell responses to target the p24 proteins and hopefully control the virus. Three weeks after being immunized, volunteers will receive weekly infusions of romidepsin for three weeks and then discontinue HAART. The trial has many endpoints but a primary goal will be a proof-of-concept of the cure strategy’s ability to deplete the viral reservoir.

How a vaccine candidate that did not perform with flying colors following treatment interruption can be expected to hold its own in a cure setting comes down to the breadth of the required response, says Søgaard. “When therapeutic vaccines have been tested as an alternative to HAART, the vaccine-induced immune responses required when treatment was interrupted needed to be broad and potent enough to control viral replication, including immune evasion and viral evolution. That is a pretty big task,” he says. “In the kick-and-kill setting, HAART is continued during reactivation and the vaccine-induced immune response only needs to target a small number of reactivated cells. There is no or only very limited ongoing viral replication, and so it may not require as broad and as potent vaccine responses.”

A delicate balance

Beyond treatment interruption, researchers are also using therapeutic vaccine candidates to address the damaging effects of immune activation in HIV-infected individuals. Ensoli’s group, for instance, developed a subunit vaccine candidate made from recombinant HIV Clade B Tat protein that is designed to elicit antibodies against Tat epitopes.

The HIV Tat protein greatly increases the rate of viral transcription and replication, and therefore is considered a prominent player in both the establishment of infection and in the replenishment of the viral reservoir in chronic infection and under antiretroviral therapy. A recent animal study led by Ruth Ruprecht at Harvard Medical School suggests that antibodies to the HIV protein Tat might also be involved in protection against HIV (see IAVI Report blog, March 28, 2013).

The vaccine candidate used in Ensoli’s study is intended in part to restore balance to the immune systems of the HIV-infected individuals, leaving them less prone to tumors, accelerated aging, atherosclerosis, and other diseases to which HIV-infected individuals are at an increased risk.

A Phase II safety and immunogenicity study conducted in Italy several years ago found Ensoli’s Tat vaccine helped reduce immune activation and loss of regulatory T cells, and improve immune function in 168 HIV-infected individuals on HAART (PLoS One 5(11), e13540, 2010). Ensoli says the open label study helped establish the most appropriate immunological and virological parameters to monitor in future trials. The biomarkers of note included early increases in CD4+ T cells, which were associated with a reduction of effector memory cells, and an increase in central memory CD4+ and CD8+ T cells, and natural killer cells. Immune reconstitution increased progressively over time as well, says Ensoli.

Ensoli’s group recently completed a randomized, double-blinded, placebo-controlled confirmatory study involving 200 HIV-infected individuals on HAART in South Africa, from which results are expected soon. While Ensoli and her team have been using the Tat candidate in combination with ART, they are also looking into whether it can be used to simplify or delay therapy, or as a substitute for ART.

The CMV story

Another challenge in therapeutic vaccination will be finding vectors and adjuvants that produce the broadest and most potent immune responses in HIV-infected individuals. A variety of vectors are undergoing testing in the preventive realm, but one vector is garnering great enthusiasm among both vaccine and cure researchers. The vector is based on the common cytomegalovirus (CMV).

Louis Picker, professor of pathology/molecular microbiology and immunology at Oregon Health & Sciences University, and his colleagues are studying the CMV-vector based HIV vaccine in non-human primates. Their studies show the vector induces a remarkable pattern of viral control in about half of rhesus macaques vaccinated with a CMV-derived vector encoding simian immunodeficiency virus (SIV) genes. Not only did the animals suppress plasma SIV to undectable levels after repeat rectal challenge with SIV (Nature 473, 523, 2011)—suggesting the likely induction of an unusual and broad effector memory T-cell response (Science 340, 940, 2013)—they also suppressed plasma SIV to undectable levels following vaginal and intravenous challenge (Nature502, 100, 2013).

Although low level SIV RNA and DNA, and replication-competent SIV could be found in the tissues of protected monkeys early after challenge, its presence waned over time. By 70 weeks after challenge all evidence of the SIV infection was gone, despite extensive analysis using the most sensitive assays. Thus, for the first time, an AIDS-causing virus in these monkeys was cleared by immunologic mechanisms, says Picker.

Picker, who began working with the CMV vector over a decade ago, says he wasn’t focused at first on its potential as a therapeutic tool. “That came after we found that CMV-vector vaccinated monkeys were able to clear infection.”

Questions that arose out of Picker’s original findings are now being addressed in a new round of studies. A large challenge study underway is comparing the efficacy of the original RhCMV vectors that elicited the unconventional CD8+ T-cell responses with modified RhCMV vectors that elicit otherwise similar responses targeting conventional epitopes. The original vectors were genetically modified wild-type strains that lacked two genes (UL128 and UL130). These modifications are responsible for the unconventionally targeted CD8+ T-cell responses, according to Picker. While the modified vectors—with the two genes repaired—have been shown to have conventional epitope targeting, they still demonstrated a higher breadth of responses than those elicited by other vectors, or SIV itself. Results due out in six to eight months will hopefully shed light on how the unusual responses induced by the original RhCMV vaccine candidate helped 50% of the vaccinated animals suppress SIV after vaginal, rectal, and even intravenous challenge. “We, at this point, don’t know whether or not the unconventional responses are required for protection,” says Picker. “That’s the reason why we’re testing the repaired vectors.”

Picker and his colleagues are also getting closer to understanding the CMV genes responsible for differential CD8+ T-cell targeting, but he says it is more difficult and time-consuming to determine the mechanisms by which these genes work. And it may not even be necessary, he says. “Mechanisms of protection have been defined for few licensed vaccines,” he says. “It would be helpful to have a strong immune correlate of protection to guide clinical development, but ultimately vaccine efficacy in humans will have to be shown in human efficacy trials.”

Picker is also vaccinating SIV-infected animals on ART to look at how well they control the virus following treatment interruption. Pending the manufacture of the prototype CMV vectors, completion of toxicology tests, and regulatory approval in 2015, a Phase I safety and immunogenicity trial in healthy, HIV-uninfected individuals could be launched in 2016, says Picker. But developing the CMV vaccine as a prophylactic candidate will take much longer, in part because of the populations in the trials. “The process of vaccine development will be slower for prevention than cure because giving a vaccine to millions of healthy people requires a higher safety bar than therapeutic use under medical supervision,” says Picker. “Bottom line: I strongly believe our CMV vector approach has a shot at contributing to both, but it will be seven to ten years before we know for certain.”

Other researchers are encouraged by Picker’s work, but still have reservations. “I think CMV is one of the most exciting things that has happened to the field in the last dozen years,” says Shapiro. “One of the very promising aspects of the vaccine is that it is a persistent vector, meaning that it if it works it may not need to be boosted. However, I’m not sure it’s going to work in people. First of all, the vaccines we have tested thus far have worked much better in NHPs [non-human primates] than in people. To simply say that because this has worked so well in non-human primates it will work the same in people is like ignoring all of history.”

Deeks likes the CMV vector as well but says the lack of an activator—something to awaken the latently infected cells—and a biomarker for the size of the reservoir complicates proof of concept studies to evaluate this strategy in the setting of cure research. Finding a quick method of measuring the size of the latent reservoir requires a biomarker that can distinguish cells in the reservoir from other infected cells (eLife 3, e04742, 2014), and this remains elusive.

Other challenges

In addition to a lack of a biomarker for the viral reservoir, the poor predictive value of the animal models used to study HIV is another obstacle to studying therapeutic vaccines. Scientists have struggled to identify the immune responses that correlate best with viral control. There also isn’t universal agreement on which assays are the most valid and reliable ones at gauging the immunogenicity of vaccine candidates.

The frequency of HIV-specific interferon-gamma producing T cells remains a widely used criterion because they are readily found in individuals with detectable virus, but the magnitude and breadth of the responses do not necessarily correlate with CD4+ T-cell count or viral load, says Lucy Dorrell, a senior clinical lecturer at the University of Oxford’s Jenner Institute who has been studying therapeutic vaccination for about a decade.

Dorrell says the quality of CD8+ T cells may be a better predictor of disease progression. In a recent study she used a viral inhibition assay developed by Asier Sáez-Cirión, an HIV cure scientist at the Pasteur Institute. The test measured the capacity, ex vivo, of HIV-specific CD8+ T cells to suppress HIV infection of autologous CD4+ T cells in 50 HIV-infected individuals with diverse disease progression rates. The study found antiviral inhibitory capacity of CD8+ T cells to be highly predictive of CD4+ T-cell loss in early HIV infection (J. Infec. Dis. 206, 552, 2012). The viral inhibition assay will also be used in two upcoming studies of HIV-infected individuals in Barcelona and London, says Dorrell.

Passive immuinization

While most of the current therapeutic vaccine candidates are focusing on the T-cell side, one of the most promising strategies in preventive vaccine research is inducing bNAbs. These antibodies may also have therapeutic value. In the last two years, studies in humanized mice and NHPs found that infusions of either single or combined bNAbs were capable of suppressing viral replication following infection (Nature 492, 118, 2012;Nature 503, 224, 2013). On the heels of those findings come several clinical studies that are looking at bNAbs as a therapeutic strategy.

An open label, dose-escalation study led by Rockefeller University is testing the safety, pharmacokinetics, and antiretroviral activity of 3BNC117, a potent monoclonal antibody that binds to the CD4 binding site, in both HIV-infected and uninfected volunteers. Participants will receive a single intravenous infusion in three increasing dose levels.
A separate open-label, dose-escalation trial is being conducted by NIAID’s Vaccine Research Center (VRC). In this trial, up to 25 HIV-infected individuals will receive two infusions of the VRC’s VRC01 bNAb either intravenously or subcutaneously. Samples will show if the antibody is detectable in mucosal secretions and blood of participants and how long VRC01 can be detected in the blood after dosing.

The advantages of passive immunization are that one doesn’t have to wait for development of a vaccine that is able to induce the highly affinity-matured bNAbs; a person is armed immediately to fight infection. But passive immunity is short-lived, which means that individuals would have to be boosted routinely to stay protected.

Georgia-based biotech GeoVax Labs is considering using bNAbs in combination with its T-cell based therapeutic vaccine candidate, GOVX-B11, and a latency reversing agent to try and cure HIV. In a recently completed Phase I study, the DNA/Modified Vaccinia Ankara (MVA) therapeutic vaccine candidate demonstrated mixed results in a study of nine HIV-infected individuals following treatment interruption. GOVX-B11 induced enhanced CD8+ T-cell responses in almost all participants, but ultimately failed to prevent the virus from re-emerging or to remain at levels that minimize immune escape.

But GeoVax Chief Scientific Officer Harriet Robinson says by combining GOVX-B11 with the infusion of bNAbs (or perhaps protective non-neutralizing antibodies) you might be able to boost CD8+ T-cell responses and prevent viral rebound after HAART is halted. “The vaccine [GOVX-B11] would add potentially protective T cells as well as boosting the host’s antibody responses, and the passive antibodies would bring specificities that scientists had selected for protective efficacy,” says Robinson.

Shapiro thinks the development of a monoclonal antibody cocktail that can be used therapeutically could be only five years away. But in all likelihood it will mean that HIV-infected individuals will need to be boosted once a month for the rest of their lives, he says. Still, he remains optimistic about both therapeutic and preventive vaccine development. “I am hopeful about both,” he says. “I just think it is going to take a lot longer than people think.”

Mary Rushton is a freelance writer based in Cambridge, Massachusetts.