Decisive Moments in Wake of Paris
The themes and research brought up at the International AIDS Society’s annual gathering may set the tone for years to come.
By Michael Dumiak
Glenda Gray, chief executive of the South African Medical Research Council and a well-known HIV researcher is, like most scientists, not given to hyperbole. Even so, Gray thinks that clinical trials either already underway or about to start this fall will set the stage for the design and development of HIV vaccine studies for the next decade. “The HIV vaccine field is at a pivotal moment,” Gray told a large audience at the Palais des congrès in Paris, where the 9th International AIDS Society’s (IAS) Conference on HIV Science took place over five humid and rainy midsummer days from July 23-27.
Whether or not these trials become heralds for the next decade, the Paris meeting underscored several themes that will play out in coming months and years and will likely set the direction toward developing an HIV vaccine.
New efficacy trials, but more vaccine concepts a must
The HIV epidemic is in the middle of its third decade. Life-saving antiretroviral therapy is accessed by more and more people around the globe, yet with concerns about the virus developing resistance to these therapies, and a less certain funding environment (see Spending Increases for HIV Vaccine Research, But Concern Rife for Future), prevention remains paramount. “The ultimate control of the worldwide HIV epidemic will likely require the development of a safe and effective vaccine,” says Dan Barouch, a founding member of the Ragon Institute and director of the Center for Virology and Vaccine Research at the Beth Israel Deaconness Medical Center. “Only four different vaccine concepts, though, have been tested for clinical efficacy in the 35-year history of the epidemic.”
Barouch’s team aims to add to that number. He came to Paris to report on some key data from ongoing experiments. Barouch’s group and his partners at Janssen are approaching crucial points in a 14-year effort pursuing HIV vaccine candidates based on adenovirus vectors. The team is well into parallel preclinical studies as well as early-phase clinical studies in humans with their Phase IIa APPROACH, TRAVERSE, and ASCENT trials, gauging the safety, tolerability, and immunogenicity of a variety of vaccine formulations. All use a strain of the common cold virus, adenovirus serotype 26 (Ad26), as the priming vector. The vaccines also use either HIV clade C or clade M and C gp140 as the boost to increase antibody titers, as well as, in APPROACH, a modified vaccinia Ankara (MVA) vector-based boost in certain groups (see AIDS 2016, IAVI Report, Vol. 20, No. 3, 2016; Science 349 (6245) 320, 2015; The Confidence Booster, IAVI Report, Vol. 20, No. 2, 2016). The vaccine payloads are so-called ‘mosaic’ antigens, which are synthetic proteins derived by algorithm to be effective against the many different strains of HIV in circulation.
“The goal is to develop a global vaccine,” Barouch says, meaning a single vaccine that is effective against multiple strains of HIV. In Paris, Barouch showed data from preclinical studies in rhesus macaques employing mosaic vaccine candidates like the ones which will be tested in humans. The results showed 66 percent efficacy in preventing infection following a series of viral challenges with an HIV/simian immunodeficiency virus (SIV) hybrid, or SHIV, while all the control animals became infected.
The study results are not yet published but Barouch reported in Paris that this study involved a group of 72 rhesus monkeys, which is significant in terms of size and cost. The idea was to conduct an experiment that would parallel the A004/Phase IIa APPROACH study as closely as possible. Monkeys were inoculated with six different variations of the vaccine candidate, with 12 monkeys receiving each variant. All received the Ad26 Env/Gag/Pol prime, with differential boosts. The boosts consisted of Ad26 alone, Env protein alone, Ad26 plus Env, Ad26 plus MVA, and Ad26 plus both MVA and Env. There was also a 12-monkey placebo group. The vaccine components came from the clinical seed stocks at Janssen. The Env tested was clade C gp140 with alum as an adjuvant. The prime shots came at week zero and week 12, with the boosts at weeks 24 and 52, modeling APPROACH as closely as possible. Because the study employed HIV antigens instead of antigens against SIV, researchers challenged the animals with low doses of SHIV-SF162P3 six months after the final boost.
The results show that the monkey groups receiving Env protein boosts produced higher antibody responses than the groups receiving only vector immunizations. These groups also had the highest level of non-neutralizing functional antibody activity, as well as an augmented Tier-1 neutralizing antibody response. The MVA boost, for its part, increased T-cell responses. The group showing the best protection data against the SHIV challenge, Barouch says, is the one immunized with the Ad26 prime and Ad26 plus Env protein boost, with 66 percent of animals remaining completely uninfected through six challenges, or an equivalent 94 percent per-exposure risk reduction versus placebo. Barouch says one of the most important aspects of this large monkey study is that the responses and immune profiles in the immunized macaques are similar and comparable to those they believe are protective in humans.
Hanneke Schuitemaker, Janssen’s head of viral vaccine discovery and translational medicine, was also in Paris to characterize the company’s lead vaccine candidate, developed with Barouch, that should soon be heading into a Phase IIb efficacy trial. If everything proceeds as planned, Janssen will advance a regimen consisting of a four-valent Ad26-based double-prime expressing Gag/Pol and Env mosaic inserts, followed by a double boost consisting of a mix of the four-valent Ad26 candidate co-formulated in a one-to-one-to-one-to-one ratio with a clade C gp140 protein, and a clade C mosaic protein.
Schuitemaker says the APPROACH trial, which tested eight Ad26-based vaccine regimens, is producing data showing favorable safety and immunogenicity profiles. “All vaccine regimens that we tested were very immunogenic,” she says. The upcoming Phase IIb trial will be known as HVTN 705, but first researchers are awaiting data from the TRAVERSE trial of Janssen’s lead candidate, which should wind up in the next several weeks.
“We don’t know whether this vaccine will protect humans,” Barouch says. But the data to date, he says, supports moving the vaccine candidate into an efficacy trial, which the group hopes to start before the end of the year pending that last bit of crucial data. He also expressed his wish that there were more and different kinds of advanced studies taking place. “We need more shots on goal,” he says. “We’re delighted that multiple different vaccine concepts are moving ahead.” Barouch is by nature confident but even-keeled—the HIV vaccine field is nothing if not humbling—and in Paris he allowed himself a bit of tempered hope. “These promising preclinical and early-phase clinical data, together with advances from many other investigators in the field, support a new sense of optimism that the development of an HIV vaccine might, in fact, be possible.”
HVTN 702 update and EAVI progress
Another pox virus vector is of course under study in HVTN 702, the field’s only ongoing Phase III efficacy trial. It is based on the canarypox and protein candidates tested in the RV144 study in Thailand, which is the only regimen to date to show any efficacy in preventing HIV infection—albeit a modest 31 percent. In Paris, Gray, protocol chairwoman on the HVTN 702 trial, described the aspirations for a perfect vaccine: it should be effective in a single dose, durable enough to provide lifetime protection, or at least protection for several years, should have minimal side effects, offer cross-clade protection, be administered simply and co-administered with other vaccines, and employ preparation and a supply chain that does not require special handling such as a long and intense cold chain.
However, an imperfect vaccine building on other imperfect vaccine strategies with moderate efficacy will suffice, Gray then told her audience with a smile. The 702 trial, started late last year, aims to build on RV144 and its follow-on study HVTN 100 by testing a clade C-specific candidate, a newly constructed protein boost, a new adjuvant, and an additional boost in a bid to make whatever immune responses are induced more durable. The scheme of the study calls for a prime with ALVAC-C, the canarypox-based vaccine, boosting with ALVAC and gp120 proteins, with the addition of an aluminum hydroxide gel adjuvant, and another booster at the 12-month mark, this one a GlaxoSmithKline-produced gp120 mix with a squalene-in-water emulsion adjuvant. Now 39 weeks in, Gray says HVTN 702 has enrolled 997 participants, averaging 26 a week, and is about to double capacity in the next phase. The study calls for 5,400 volunteers and its goal is to reach greater than or equal to 50 percent efficacy after three years. Results are expected in 2020.
Echoing Barouch’s call for more vaccine concepts to undergo testing, Robin Shattock, a mucosal infection and immunity researcher at Imperial College London, was in Paris to keep an eye on all the options and to propose testing them more efficiently, a theme he’s been touting for some time. “While studies are going on into these perfect vaccines,” says Shattock, referring, partly tongue-in-cheek, to the HVTN 702 regimen, “if they fail to realize the level of efficacy that the investigators would hope to reach, we need some alternatives.” Shattock is now in a place to develop potential alternatives as coordinator of EAVI 2020, the European Commission-backed European AIDS Vaccine Initiative, which launched last summer with about US$20 million in seed funding for basic research.
|HIV Envelope trimer in complex with PGT151|
Overall structure of a soluble HIV-1 envelope glycoprotein trimer in complex with the broadly neutralizing antibody PGT151. The gp120 and gp41 subunits of the trimer are depicted in white and red, respectively. The antibody (Fab) is shown in blue. The “native-like” ConM SOSIP trimer was generated at the Academic Medical Center Amsterdam for the EU H2020 EAVI2020 program and will be tested in experimental medicine trials in the near future.
Image prepared by Alba Torrents de la Peña, PhD candidate in Rogier Sanders' lab in the department of medical microbiology at the Academisch Medisch Centrum, Amsterdam, the Netherlands.
“It’s prudent to have approaches that can continue irrespective of what the findings of the [current] trials are, and if they are spectacular, we can move on to something else,” he says. Shattock endorses the use of experimental medicine trials of promising HIV candidates. “These are not different from Phase I studies—they are still about safety—but they are specifically designed to test hypotheses, and they do not have the purview of being part of a product development strategy,” he says. “That means in some ways they can move faster and be smaller scale and that allows us to accelerate some of the things we want to do.”
With funding becoming more and more fraught, Shattock expects it to become harder to put more candidates into efficacy trials. Better prevention technologies, such as oral pre-exposure prophylaxis, should also bring down incidence of HIV, which will make it necessary to do larger and larger efficacy trials to gather data properly. This will be more expensive. “We will still urgently need an HIV vaccine. That sense of urgency needs to be maintained,” he says. “Refining vaccines early in the pipeline makes sense.”
Shattock says these efforts should speed development and decrease risk of late-stage failure in other trials; address questions that can’t be definitively answered in animal models; and enable validation and sequential iteration for structural design, of which there has been an extreme blossoming of late. “The bottleneck now is getting those design concepts into humans,” Shattock says. Hypothesis testing of novel concepts prior to formal product development should be the purview of experimental medicine trials, with results set up to deliver in-depth analysis of human immune and antibody repertoire responses. And they should involve the kinds of intense sampling of blood, mucosal layers, lymph nodes, and bone marrow that are impractical in large-scale studies.
“We have no vaccine that produces any neutralizing antibodies of any breadth in humans, so a lot of work needs to be done to start changing immunization schedules that may go out for many years to types of schedules applicable to a real-world setting.” Shattock is particularly interested in utilizing experimental medicine trials to evaluate recombinant trimers that resemble HIV’s native Env structure and to understand how they can be used through sequential immunizations or as defined cocktails to drive B-cell responses towards neutralizing breadth. The idea is to be able to reproduce the chain of events that happens in those very rare infected people that produce broadly neutralizing antibodies (bNAbs), but to do it proactively in vaccinated individuals in a compressed period of time.
One approach that EAVI participants are pursuing is the use of a series of trimers isolated from an infected individual that developed bNAbs within a period of months. Shattock and the EAVI team (see Europe Invests in HIV Vaccine Research, IAVI Report, Vol. 20, No. 3, 2016) are currently producing a recombinant protein every six months, Shattock says, in service of creating a pipeline sufficiently nimble and cost effective to get things into trials faster. Some 18 months into the EAVI program, teams in Rogier Sanders’ lab at the Academisch Medisch Centrum in Amsterdam and Quentin Sattentau’s lab at Magdalen College in Oxford are preparing two candidates based on a consensus Env sequence, one that is meant to provide coverage against the majority of circulating HIV strains. These stabilized trimers will be manufactured in small batches in Austria by a company called Polymun. The trials will be Phase I safety and immunogenicity studies involving between 30 and 50 volunteers. Shattock says the first trials will examine the B-cell repertoire response, or how the protein subtypes in the adaptive immune system react toward the protein candidate after exposure. “In later developments we’ll look at lineage design approaches or approaches to broaden the response,” he adds.
Shattock is keen to close the loop between human trials and structural vaccinology. “We can link structural design to human immunogenicity and have an iterative cycle where the human response can feed back into the structural biology, and we can design better immunogens to elicit the type of responses we want to achieve.”
Moove over mice!
Things might look quite different, though, if humans made antibodies like cows. “They make fantastic antibody responses, very quickly, with broadly neutralizing activity,” Shattock says.
Shattock and everyone else knows this thanks to recent experiments conducted by researchers at IAVI, The Scripps Research Institute (TSRI), and Texas A&M. It all started with a discussion between Devin Sok, director of antibody discovery and development at IAVI and Vaughn Smider, a bovine antibody expert and protein engineer at TSRI. Sok had been exploring and characterizing antibodies in rabbits, mice, and guinea pigs, but after discussions with Smider began thinking about cow antibodies. Smider has been studying cow antibodies for some time. Bovine antibodies have unusually long and diverse protein chains as part of their antigen recognition sites. “We thought, well, why don’t we just immunize the cows and see if we generate antibodies,” Sok says.
The two pressed ahead, and the results are impressive (Nature 548, 108, 2017). “The level of response—the titers that we saw, how potent the serum was against the virus—I’ve never seen that before,” a still-amazed Sok recalls. He participates in IAVI’s Neutralizing Antibody Consortium and has seen neutralizing antibodies developed and isolated in both humans and animal models, but there was something special about what was happening in cows (see Stumbling on Greatness). “Seeing that against one virus, and then seeing it span across multiple viruses…” He trailed off. “We’ve been trying to work on this forever. This is the first time we’ve seen that it actually did work. It’s exciting.”
Sok and Smider immunized four cows—Holsteins (two steers, two heifers) at Texas A&M—with the BG505 SOSIP HIV Env glycoprotein, an engineered immunogen that maintains the trimeric structure of native HIV Env. All four cows developed immune responses and did so quickly. After 42 days, a longitudinal serum analysis for one cow shows 20 percent neutralization breadth against 117 cross-clade isolates. At 381 days, serum analysis for the same cow showed 96 percent neutralization breadth against the same 117 isolates. A single monoclonal antibody isolate from the cow neutralized 72 percent of the cross-clade range. It also showed the unusually long chain characteristic to bovine antibodies that drew the attention of the two researchers in the first place. One of the antibody’s heavy chains reached 60 amino acids in length. A typical human antibody heavy chain reaches 17 or 18 amino acids. All four cows developed robust and reliable responses, Sok says.
“The hypothesis is that cows have these four stomachs full of bacteria,” he says. “The thinking is that cows have these long antibodies to maintain that microbiome, which could be full of pathogens. That’s a hypothesis. We haven’t tested any of that.”
Global media grazed contentedly for days on the cow study, but Sok and fellow observers in Paris are roping in expectations. The ease with which cows developed a broadly neutralizing antibody response is noteworthy, but it’s just a starting point. If, however, it becomes possible to create long-chain antibodies in humans, that could be a stepping-stone to more effectively inducing broadly neutralizing antibodies against HIV.
Common cause with cancer research?
At the beginning of September the US Food and Drug Administration approved its first-ever gene therapy, a chimeric antigen receptor-based treatment for pediatric B-cell acute lymphoblastic leukemia. Carl June, a University of Pennsylvania cancer immunotherapist and one of the lead researchers in developing this “living drug” that will be marketed by Novartis under the name Kymriah, earlier this year gave an address in Seattle at the Conference on Retroviruses and Opportunistic Infections in which he both described this antigen receptor-based treatment, and called on the HIV research field to more closely explore the techniques which produce these chimeric antigens, or CAR-T cells. The overlap and potential benefit from commonalities in both cancer and HIV research became even more explicit in Paris, featured in a special day-and-a-half-long forum on cancer and HIV cures.
A CAR-T cell is engineered to bind to the protein CD-19 expressed on B cells. Acute lymphoblastic leukemia causes overproduction of B-cell lymphoblasts in bone marrow which then multiply, causing a corresponding dropoff in the production of healthy red and white functioning blood cells and platelets. The CAR-T works by killing those B cells that are malignant, while at the same time suppressing molecules that had previously allowed the cancer cells to evade detection. The immune system’s existing T cells are also reprogrammed by the engineered CAR-T to go after the cancerous cells. As the modified T cells kill indiscriminately, normal B-cell function is replaced by an antibody therapy of gammaglobulin injections.
June was stirring with his case in Seattle, saying advances in cancer immunotherapy and gene therapy will provide methods that can be adapted for HIV. This remains to be seen—there are several aspects about CAR-T technologies that may not be equally matched to the challenges posed by HIV. But cancer research as a whole, some of the leading lights in both fields said in Paris, maps over HIV more than enough that it is worth a more determined effort to find areas of collaboration and common ground.
“When you talk about draconian ways to modify a disease when you have a pretty simple way to do so—namely, if you have a sensitive virus like HIV and you can suppress it one pill a day forever and not worry about it—the question is, should you put resources into something like gene editing?” asked Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases at the US National Institutes of Health (NIH). “I think the answer is yes.”
Fauci was part of a roundtable discussion in Paris on cancer and cure research. “You want to push to the point of proving a concept. Once you prove the concept, you try to simplify it so it is scalable,” he says. About 36 million people in the world live with HIV. More than half of them have access to antiretroviral drugs. About 17 million, however, do not. “If you have an intervention that is not scalable, it is interesting, but it is not going to address the global epidemic,” Fauci says. “We should not let that take away the importance of pursuing it. You never know whether you are going to be able to get it to scale. That’s why the NIH supports this kind of research.”
Sharon Lewin, an infectious disease physician and director of the Dougherty Institute at the University of Melbourne, Australia, is monitoring a select number of HIV-infected patients who are also burdened with cancers, a set of illnesses that appear ready to increasingly beset the HIV field as the HIV-infected continues to age. There are wholesale changes in how oncology researchers are pursuing immunotherapies and gene therapy, with more than 1,000 clinical trials underway pursuing checkpoint blockers, interferons, and the rebuilding of adaptive immunity, fields which all have either tangential or direct uses for HIV research.
Given the wide variety of cancer immunotherapy and gene therapy initiatives, Lewin expressed a widespread concern for the HIV world, which is the issue of toxicity. “We really worry about it in these interventions,” she says. The tolerance for potentially toxic treatments is much different for a patient facing a terminal cancer diagnosis than one beginning a burdensome but much more manageable course of lifelong antiretroviral therapy (ART). She and Fauci, though, see a clear need for more collaboration between to the two fields. Fauci suggested HIV labs take on a young postdoc with cancer training if the opportunity arises. “When HIV started it was all virologists who were involved,” he says. “One day we thought: ‘maybe, we need some immunologists.’ It’s the same thing.”
The University of California, San Francisco’s (UCSF) Steve Deeks, a prominent researcher in the HIV cure field, was also in Paris, where he and UCSF colleague Timothy Henrich presented research on two HIV-infected individuals examining whether extremely early initiation of ART leads to temporary remission or even cure. The small-scale study showed HIV relapse despite starting treatment at one of the earliest stages possible of HIV infection. While the treatment did lead to nearly complete loss of detectable HIV in blood and tissue, it did not lead to remission without treatment—perhaps due to the persistence of the reservoir.
This was not the case for a South African girl. The University of Witwatersrand’s Avy Violari and pediatrician Mark Cotton of Stellenbosch University presented a rare case of remission in a nine-year-old girl, who has had undetectable levels of HIV for the eight-and-a-half years since she stopped ART 40 weeks after being diagnosed with HIV at the age of one month.
This immediately made global headlines, as the South African girl is one of only three to report a long-lasting remission so far. Only one, the “Berlin Patient,” Timothy Brown—who was in Paris for the IAS meeting—remains HIV free, and he underwent a grueling bone marrow transplant because of concurrent acute leukemia. The South African girl is remarkable in many ways, including her youth. Violari says they can only detect traces of provirus in the girl. “By studying this case, we hope we will one day understand how it’s possible to stop treatment.”
While many expressed delight over this report, other cases of remission reported to date ended with individuals needing to resume ART, most notably the Mississippi baby. While these lightning-in-a-bottle cases are rays of light and strike hope about the possibilities of an HIV cure, for now, in any case, they leave researchers with little more than wonder.
Michael Dumiak reports on global science, public health and technology and is based in Berlin.