The Path to Protection

Clinical trials both past and present are contributing to the design of new and improved HIV vaccine immunogens

By Seema H. Bajaria

Larry Corey, who stepped down as president and director of the Fred Hutchinson Cancer Research Center (FHCRC) in June (see Back to the Bench, this issue), set the tone for the opening day of this year’s Full Group Meeting of the HIV Vaccine Trials Network (HVTN), which was held in Washington, DC, from June 3-5, when he deemed the general mood of vaccine researchers as “optimistic.” Five years have passed since the surprising results of the RV144 trial—the first to provide any evidence of vaccine-induced protection against HIV. And while researchers are still interpreting its aftermath, other areas of investigation are also advancing.

The first day of the HVTN meeting centered on four plenary sessions that covered research highlights in the areas of immune correlates, non-human primate (NHP) models, broadly neutralizing antibodies (bNAbs), and ongoing Phase I clinical trials.

Immune correlates

The ongoing immune correlates analysis of the RV144 trial, the subject of the first plenary, has so far centered largely on antibodies. Researchers have shown that plasma immunoglobulin (Ig)A antibodies to HIV Env in vaccine recipients are directly correlated with an increased risk of HIV infection, making them what researchers refer to as a correlate of risk; whereas IgG antibodies directed to the V1/V2 regions of HIV Env have been inversely correlated with infection risk (NEJM 2012, doi: 10.1056/NEJMoa1113425).

While it did not appear that the total frequency of antigen-specific T cells had a significant effect on HIV infection risk, researchers hypothesized that there were functional characteristics of the T cells induced by the RV144 vaccine that were protective for certain individuals. And it turns out this is the case. Greg Finak, staff scientist at FHCRC, presented data from a combinatorial polyfunctionality analysis (COMPASS) method that indicates polyfunctional T-cell responses are indeed independent correlates of infection risk in RV144. Further work is required to better understand this data, but one possible hypothesis Finak suggests is that multifaceted T cells could be the “help” that B cells require to be effective.

Researchers are also conducting sieve analyses of RV144, comparing the breakthrough virus sequences in vaccine and placebo recipients. This helps determine what effect immune pressure exerted by the vaccine had on selectively blocking, or sieving, the infecting virus, as well as on evolution of the virus after acquisition. Paul Edlefsen, a biostatistician at the Vaccine and Infectious Disease Division of the FHCRC, followed up on prior reported results of the significance of antibody-mediated immune pressure at amino-acid positions 169 and 181 (signature sites) in the V2 region of HIV Env (Nature 2012, doi: 10.1038/nature11519).

Researchers have also used a more expansive analysis to identify evidence of T-cell mediated pressure in regions of Gag and Pro that were included in the vaccine, and provide data that genetic differences occur outside of those immunogen regions. Volunteers who became HIV infected despite vaccination tended to have a higher level of phylogenetic divergence in Gag, a finding consistent with what was seen in volunteers in the Phambili (HVTN 503) trial, and reminiscent of the T-cell epitope divergences seen in volunteers in the STEP (HVTN 502) trial, both of which tested Merck’s MRKAd5 HIV gag/pol/nef trivalent vaccine candidate that failed to provide any protection.

Susan Zolla-Pazner, professor of pathology at New York University, compared two signature sites that Edlefsen and others identified in the V3 region of Env, with the two previously identified V2 signature sites. She found that position 307 in V3 and position 169 in V2 are similar in that both indicate the vaccine was able to protect against viruses with amino acids matching those in the immunogen. Zolla-Pazner presented evidence that these signature sites are both contact residues for vaccine-induced antibodies.

By contrast, position 317 in V3, which is similar to position 181 in V2, are both sites at which immunogen-matching residues are more frequently conserved in vaccine-recipient breakthrough viruses. Zolla-Pazner hypothesized that rather than directly impacting antibody binding, mutations in these residues reduced viral fitness and infectivity, and were thus more conserved in HIV sequences that were able to infect vaccine recipients despite the presence of antibodies targeting sites such as position 307 in V3 and 169 in V2.

NHP models

The second session of the day focused on identifying what immune responses might be required for protection using NHP studies. Dan Barouch, director of the Center for Virology and Vaccine Research at Beth Israel Deaconess Medical Center, is examining the earliest events during the so-called eclipse phase of infection, which is the time between virus inoculation and the first appearance of virus in plasma, including the pathways of virus spread and the host response, which are still poorly understood following intra-vaginal (IVAG) infection in rhesus macaques.

During IVAG infection with simian immunodeficiency virus (SIV)mac251, data show that the virus may breach the mucosal barrier as early as the first day following infection, followed quickly by systemic virus dissemination. While analyses are still underway, virologic data show that on the first day after infection, occasional lymphoid and gastrointestinal tissues tested positive for virus. Both CD4+ and CD8+ T-cell responses first appeared locally in the mucosa, but not until seven days post-infection, with systemic T-cell responses lagging even further behind, echoing earlier findings. Therefore, Barouch concluded that the eclipse phase is dynamic and complex, with virus broadcasting to distal tissues within the first few days of infection. His group now plans to use this NHP model to interrogate how neutralizing antibodies and vaccines protect and precisely when and where they might intercept the trajectory of virus spread following viral challenge.

Richard Koup, chief of the immunology laboratory at the Vaccine Research Center (VRC) at the National Institute of Allergy and Infectious Diseases (NIAID), also spoke about using NHP models to understand the development of bNAbs. Koup showed that during chronic SIV/HIV chimeric virus (SHIV)-AD8 infection of eight monkeys, bNAbs develop slowly in the germinal centers and are associated with Env sequence diversity, both the quantity and quality of Env-specific TFH cells, and the quantity of Env-specific B cells in the lymph nodes.

Based on these observations, Koup concluded that antigenic changes will be required for Env immunogens to drive the degree of somatic hypermutation required for the development of bNAbs, and he thinks NHP studies can be effectively used to inform the design and efficacy testing of different Env immunogens.

Neutralizing antibodies

The focus on bNAbs continued with a presentation by John Moore, professor of microbiology and immunology at Weill Cornell Medical College. He described the work of a team that, as reported previously, has produced a stable trimer structure, BG505 SOSIP.664, that is able to bind multiple bNAbs (see CROI: Progress on Prevention and CureIAVI Report, Vol. 18, Issue 1, 2014). Pooled data from two rabbit studies shows that this trimer structure induces antibodies that neutralize the autologous tier-2 virus, as well as more sensitive, but less relevant, heterologous tier-1A and tier-1B viruses. In these studies, Moore found that induction of tier-1 NAbs, which are predominantly directed to the V3 region of gp120, is not predictive of the induction of tier-2 NAbs. Moore suggested it is possible that tier-1 NAbs may even distract from or interfere with the induction of antibody responses against tier-2 viruses, which should be the goal of vaccine design strategies.

Future experimentation on this point seems justified, Moore said, while emphasizing that much work remains to be done to devise immunization strategies that might lead to the induction of cross-reactive tier-2 bNAbs with properties similar to those found in about 20% of HIV-infected individuals. New immunization and trimer design strategies could include cocktails of SOSIP trimers from the same or different subtypes, according to Moore, as well as the application of structure-guided improvements that are now being explored in the team’s laboratories. A consortium including the International AIDS Vaccine Initiative, the Bill & Melinda Gates Foundation (BMGF), and the Center for HIV/AIDS Vaccine Immunology-Immunogen Design (CHAVI-ID) has been established to create BG505 SOSIP.664 trimer cell lines by the end of this year, with the aim of testing this trimer in humans. New and improved SOSIP trimer candidates will also yield additional immunogens for testing in animals in the coming year, Moore said.

Leonidas Stamatatos, scientific director at the Seattle Biomedical Research Institute, continued the discussion of Env immunogens as he described his efforts to elucidate why the generation of narrowly neutralizing antibodies (nNAbs) is more predominant than bNAbs. To understand how most recombinant Envs activate only those B-cell receptors (BCRs) that induce nNAbs, Stamatatos stressed the importance of understanding dynamics within the germinal centers where high affinity B cells are selected. As HIV has evolved to avoid detection by B cells that express BCRs giving rise to bNAbs, trimers that avoid recognition by nNAbs are necessary to elicit antibodies with broadly neutralizing capability, according to Stamatatos.

Phase I program

While immunogen development in the laboratory continues, others are already evaluating different antibodies, vectors, and immunogens in clinical trials.

Mark Connors, chief of the HIV-specific immunity section at NIAID, provided an update on the need for small, flexible clinical trials to evaluate which immunogens drive B-cell somatic hypermutation and T-cell cytotoxic capacity, including the testing of Ad4-based HIV vaccine candidates. In a Phase I clinical trial, an Ad4 vector-based H5N1 influenza vaccine was well tolerated and induced neutralizing antibodies and somatic hypermutation (Lancet Infect. Dis. 2013, doi: 10.1016/S1473-3099(12)70345-6; and other unpublished observations, according to Connors), making Connors optimistic about the potential for an Ad4 vector-based HIV vaccine to also induce neutralizing antibodies.

He and colleagues are now evaluating the immunogenicity of replicating Ad4 vectors in humans, including an Ad4-mGag (designed to induce T-cell responses) and an Ad4-EnvC150 (designed to induce antibody responses). This Ad4-HIV trial, which began in January, is part of a long-term collaboration between NIAID, NIAID’s Division of AIDS (DAIDS), and the biotechnology company PaxVax, Inc.

Meanwhile, Marina Caskey, assistant professor of clinical investigation at Rockefeller University, and colleagues are evaluating the direct administration of two monoclonal antibodies, 3BNC117 (directed to the CD4 binding site) and 10-1074 (directed to the V3 region of HIV Env), discovered by Michel Nussenzweig’s laboratory at Rockefeller. Based on the initial success of certain doses of 3BNC117 and 10-1074 to prevent virus acquisition in NHPs, as well as transiently decrease viremia in chronically infected macaques (Nature 2013, doi: 10.1038/nature12746), a Phase I dose-escalation study to evaluate the safety, pharmacokinetics, and antiretroviral activity of 3BNC117 began in February. Two other anticipated Phase I studies include a dose escalation, safety, and pharmacokinetics study of 10-1074 and a safety study of 3BNC177 and 10-1074 combined. Future studies could evaluate the efficacy of 3BNC117 or 10-1074 to prevent acquisition of HIV, as well as to treat chronically HIV-infected individuals during antiretroviral therapy interruption or in addition to traditional antiretroviral therapy.

Lastly, George Lewis, co-director of basic science and vaccine research at the Institute of Human Virology, discussed progress in developing full-length single chain gp120-CD4 complex Env immunogens. Data from five completed NHP protection studies will be published soon, according to Lewis, and a Phase I trial is slated to begin by early next year in collaboration with several partners, including Profectus Biosciences, the US Military HIV Research Program, DAIDS, Sanofi Pasteur, the National Institutes of Health, and BMGF.

Seema Bajaria reports on infectious diseases, systems biology, and public health and is based in New York City.