Ratcheting Up T-cell Respnses

A crop of vaccine candidates that are capable of inducing robust cellular immune responses garner attention at this year's AIDS Vaccine conference

By Kristen Jill Kresge

The audience of close to a thousand researchers who gathered at the start of the annual AIDS Vaccine conference in Amsterdam from August 29 to September 1 was greeted with a level of optimism that has become quite unusual in this field. Lawrence Corey of the HIV Vaccine Trials Network in Seattle called 2006 "a vintage year for vaccine development" when he gave an overview presentation on the current status of the HIV vaccine pipeline at the opening plenary session.

Over the following days support for his statement came from several presentations that showed promising immunogenicity data for some of the vaccine candidates currently in clinical trials, a welcome shot of good news for a field that has struggled for many years to develop candidates that invoke promising levels of immune responses in humans. "A few years ago it was hardly possible to stimulate T-cell responses with vaccination," said Andrew McMichael of the University of Oxford, UK. Now several Phase I and II trials with candidates based on adenovirus alone or in prime-boost combination with DNA plasmids, modified vaccinia Ankara virus, and poxvirus vectors containing HIV genes are providing encouraging results.

Researchers have been able to increasingly ratchet up the level of CD4+ and CD8+ T-cell responses; those induced by the current crop of candidates are three times higher than those possible just a couple of years ago, said Corey, and he is hopeful that this progress will continue in the coming years. While efforts to stimulate a neutralizing antibody response with immunization are still largely unsuccessful, researchers suggest advances in cellular-based vaccines may bring the field closer to a vaccine that may either prevent HIV infection or, perhaps more likely, lower viral load and therefore ameliorate disease progression and lower the infectivity in individuals who do become infected. "There's cautious optimism now" about the ability of T-cell based vaccines, said Corey.

But others who addressed the conference delegates painted a less rosy picture of the current state of research. Jaap Goudsmit of the Netherlands-based biotechnology company Crucell reminded researchers that there is yet to be a vaccine against any disease licensed on the basis of results from T-cell assays. He also recounted that the late vaccinologist Maurice Hilleman wrote in one of his last published articles that any vaccine based on cellular immunity was unlikely to be effective (Proc. Natl Acad. Sci. USA 101, 14560, 2004).

Even with the encouraging immunogenicity data in hand there are still several potential pitfalls with the current candidates. Debate continues over including the highly variable HIV env gene as a vaccine insert, which John Moore of Weill Medical College at Cornell University called "one of the most hyper-glycosylated proteins vaccinologists have ever had to deal with." And although many of the concerns about the potential for pre-existing immunity to viral vectors like adenovirus seem to have subsided somewhat, research into new vectors like cytomegalovirus (CMV) or other serotypes of adenovirus are a clear priority for the field. Several speakers also underscored other key challenges, including establishing a better pre-clinical system for evaluating vaccine candidates and streamlining the clinical trials process to expedite the development of vaccines.

Vintage data

Much of the promising immunogenicity data presented at the conference came from trials administering vaccine candidates in a prime-boost combination and several utilize DNA plasmid vaccines for the initial immunization. "Don't consider DNA as not immunogenic because they [DNA plasmid vaccines] are priming very well," said Giuseppe Pantaleo of the Centre Hospitalier Universitaire Vaudois in Lausanne, Switzerland.

One of the earliest presentations on a DNA vaccine candidate was by Eric Sandström of the Karolinska Institute, Sweden, who presented for the first time immunogenicity data from a Phase I trial that is ongoing in Stockholm with a seven plasmid DNA vaccine developed at the institute in collaboration with the Swedish Institute for Infectious Disease Control. The candidate contains env genes from clade A, B, and C; rev from clade B; gag from clade A and B; and RT from clade B. Forty volunteers in the trial were randomized to receive three injections with one of three doses (1mg, 2mg, or 3.8mg) of the DNA vaccine administered either intradermally (via the needle-less system developed by Bioject) or intramuscularly. Half of the participants also received a 0.5ml injection of the cytokine granulocyte-macrophage colony-stimulating factor (GMCSF).

Six months after the final DNA immunization, all but two volunteers also received a booster immunization with a modified vaccinia Ankara (MVA)-based vaccine candidate developed by the US National Institutes of Allergies and Infectious Diseases and manufactured by the Walter Reed Army Institute of Research (WRAIR). Side-effects in vaccine recipients were limited to fatigue and general flu-like symptoms after the DNA prime and only mild adverse events were reported after the MVA injection.

Researchers measured the level of immune responses by IFN-g ELISPOT assay and positive responders in this trial were defined as those with greater than 55 spot forming cells (SFC) per million cells. After the third DNA immunization, 11 of the 38 vaccinees had positive immune responses. But after the MVA immunization there was a significant immune response boost to both Gag and Env, with 33/36 of participants having positive ELISPOT results. The MVA vaccine appeared highly immunogenic despite several of the trial volunteers having been previously exposed to vaccinia virus through smallpox vaccination.

Based on the results of this trial Sandström and colleagues are preparing for a Phase I/II trial in Tanzania with the same candidates, excluding GMCSF, since the DNA plasmid vaccine is based on HIV strains circulating in that country.

Results from a Phase I trial conducted by the EuroVacc consortium with another DNA plasmid vaccine tested in combination with NYVAC, a modified poxvirus vector vaccine, were also presented. A total of 40 volunteers—20 male, 20 female—were recruited for this trial at two sites in the UK and in Lausanne. Half of the participants received two injections of a DNA plasmid vaccine containing HIV clade C envgagpol, and nef genes and all volunteers received two doses of NYVAC containing the same genetic inserts.

The combination was more immunogenic than NYVAC alone, which induced positive ELISPOT responses in only 40% of volunteers and these tended to disappear by 48 weeks after immunization. Of the 20 volunteers who received the DNA/NYVAC combination, 90% had positive responses (>55 SFC). Responses to env averaged around 120 to 180 SFC after DNA priming and boosting with NYVAC, with the highest responder reaching 300 SFC. Some volunteers also generated more diverse responses to gag and pol but at much lower levels, and the immune responses induced by this vaccine regimen were maintained through 48 weeks. "These responses are in the range of what we see in long-term non-progressors," said Pantaleo. However they are primarily directed toenv, which, he pointed out, is not the case in people who are HIV infected.

Rick Koup presented an update on the DNA plasmid vaccine and adenovirus serotype 5 (Ad5) vector-based vaccine developed by the Vaccine Research Center (VRC) at the US National Institutes of Health (NIH) that is currently in Phase I/II clinical trials (see On trials, IAVI Report 10, 2, 2006). Enrollment for these trials is now two-thirds complete but is still ongoing at HIV Vaccine Trials Network (HVTN) sites in North and South America and South Africa and at WRAIR sites in Africa.

Data from several previous trials with this combination of vaccine candidates show that the DNA priming produces higher CD4+ and CD8+ T-cell responses up to the third DNA immunization and Koup reported that the CD8+ T-cell responses are boosted 5-fold after administration of the Ad5 vaccine. The T-cell responses after boosting are also more polyfunctional, as determined by expression of the cytokines IFN-g, interleukin-2 (IL-2), MIP1b, and TNFa, and expression of CD107. Some immunizations in these trials were given with the Bioject system and Koup hinted that data comparing this route of administration with traditional intramuscular injection would be available soon. For now the VRC, WRAIR, and IAVI are preparing to move these candidate vaccines into a Phase IIb test-of-concept trial, known as PAVE 100.

'Nature's adjuvant'

Adenovirus-based vaccine candidates have produced the most impressive cellular immune responses seen so far, and Gary Nabel of the VRC referred to the viral vector as "nature's adjuvant." Another advantage of this vector is that it can be administered at much higher doses than other viral vectors—1000 to 10,000—fold more particles than can be safely used with MVA, for example. But results from a Phase I safety trial, HVTN 054, indicate that at higher doses the reactogenicity of Ad5 vaccines increases without any further gain in immunogenicity. Laurence Peiperl of the University of California School of Medicine in San Francisco presented data from this trial that evaluated a single injection of the VRC's Ad5 vaccine candidate at a dose of either 1010 or 1011 particle units (PU). The recombinant Ad5 vaccine encoding HIV Env from clades A, B, and C and clade B Gag/Pol was administered to 20 volunteers at each dose. Systemic reactogenicity and flu-like symptoms or reactions at the injection site were reported in four volunteers, all of whom received the dose of 1011 PU. The side-effects peaked one or two days following injection and trailed off within a week and although none of the serious adverse events were considered related to the vaccine, Peiperl concluded that the safety profile of the lower dose seemed more favorable.

Additionally the immune responses measured by intracellular cytokine staining assay were actually higher in volunteers who received the lower dose—95% were considered responders compared to 90% at the 1011 PU dose. "It appears that less is more for immune responses to adenovirus," said Robert Seder of the VRC.

The same dose of Merck's Ad5 vaccine candidate is also being evaluated in the company's ongoing Phase IIb test-of-concept trial in collaboration with the HVTN (see Renewed promise, IAVI Report 9, 4, 2005). The highly anticipated efficacy data from this trial won't be available until 2008 but Michael Robertson of Merck provided some preliminary information about the safety of the vaccine candidate. The majority (74%) of volunteers reported mild or moderate adverse events, most of which were headache, fever, fatigue, or pain at the injection site. Serious adverse events occurred in 13 individuals and 3 of these were attributed to the vaccine, including a severe case of fever, diarrhea, and a report of possible anaphylaxis. The number of adverse events decreases after the second and third immunizations, which Robertson credited to anti-vector immunity.

Another focus of his presentation was on the experiences of conducting the trial in individuals at high risk of HIV infection either through sexual activity or injection drug use. Conducting trials in these populations will allow researchers to get preliminary efficacy results with smaller, faster, and cheaper studies, said Robertson, but some have speculated that it will be more difficult to both recruit and retain high-risk individuals in long-term vaccine trials. So far, at least, this has not been Merck's experience. Approximately 2000 of an intended 3000 volunteers were enrolled in the study at the end of July-enrollment should be complete by the end of this year—and volunteers have completed 95% of scheduled visits.

Robertson presented data collected about the reported risk behaviors of the individuals to establish the level of risk in this cohort. Men screened for the Phase IIb study across all sites reported a median of 6 different sexual partners in the last 6 months, and 12% of them reported having unprotected anal intercourse with a partner who they knew was HIV infected. Women reported an average of 28 different sexual partners over the previous 6 months and 5% had unprotected vaginal intercourse with an HIV-infected partner. Also, 15% of the women screened for the trial reported having another sexually-transmitted disease in this same time period. During the screening process the HIV prevalence rates among males was 4% and for females was around 3%, though Robertson explained that these figures vary greatly from site to site.

Pre-existing anxiety

By the time the Merck trial is fully enrolled there will be an equal number of volunteers with both low and high levels of pre-existing antibody immunity to Ad5. This will allow researchers to draw conclusions about how the immunogenicity of the vaccine is compromised by anti-vector immune responses. As many as 95% of people in sub-Saharan Africa and southeast Asia have high antibody titers to Ad5 and this has been a major concern for the AIDS vaccine field. However preliminary data suggests that the immune responses induced by current vaccine candidates are only modestly affected by prior Ad5 infection. Immune responses to the VRC's DNA/Ad5 vaccines remain high even in volunteers with very high anti-Ad5 antibody titers (>1:7000).

Data from Merck's earlier Phase I and II trials also show that the immune responses in volunteers with high levels of pre-existing immunity are only blunted by an average of 15% when compared to volunteers without Ad5 seropositivity. This is encouraging to many researchers. "The data is limited, but it doesn't look like pre-existing immunity is going to be a show stopper," said Koup.

Still, many research groups are continuing to develop vectors based on other serotypes of adenovirus. Both Nabel and Dan Barouch of Beth Israel Deaconess Medical Center in Boston presented on recombinant adenovirus vectors that retain the immunogenicity of Ad5, yet can circumvent problems with pre-existing immunity. Nabel and colleagues at the VRC are studying the mechanism by which Ad35, a serotype of adenovirus with a much lower seroprevalence worldwide, targets dendritic cells to see if this influences its immunogenicity. "If we understand it, we may be able to engineer our way around it," he said.

Nabel also argued that researchers might not gain anything using Ad35 as a vector, especially if its immunogenicity is further compromised in people previously infected with Ad5. But for now there is a paucity of data in non-human primates. "Human data is the only thing that's going to matter here," said Jerald Sadoff of the US non-profit Aeras Global TB Vaccine Foundation, "and it's immunogenic enough to move forward."

Other groups are attempting to develop novel viral vectors. Louis Picker of the Oregon Health and Science University in the US presented on the development of cytomegalovirus (CMV) vectors for AIDS vaccines. CMV is a herpesvirus that is ubiquitous throughout the world and, once established, infection persists indefinitely. Infection with CMV has very few consequences in healthy individuals and there are rarely any symptoms but it can cause serious disease in immunocompromised individuals, including those with HIV/AIDS.

Attributes of CMV that make it attractive to AIDS vaccine researchers are its immunogenicity and the induction of T-cell responses at mucosal surfaces. Reinfection with the virus is also highly efficient so it easily avoids any potential problems with pre-existing immunity. The virus also has a large genome that could be customized to both increase immunogenicity and attenuate the virus to enhance safety, which would be an important issue moving forward with this viral vector. "We're hoping to manipulate these viruses to make them unquestionably safe," said Picker.

Moving forward

Researchers are also now focusing on ways to streamline the AIDS vaccine pipeline and push promising candidates forward as quickly as possible (see Vaccine Briefs). One way to accomplish this is to improve the ability to evaluate pre-clinical candidates so that fewer non-immunogenic candidates are moved into costly Phase I trials. Advancements in this area were reported in Amsterdam by J. Victor Garcia-Martinez of the University of Texas Southwestern Medical Center, who has developed a humanized mouse model that can be used to test vaccine candidates.

His research group transplanted human CD34+ stem cells into immunodeficient NOD/SCID mice that had already received implants of human thymus and liver tissues. This resulted in long-term repopulation of human B cells, monocytes/macrophages, T cells, and dendritic cells in the mice. When infected with HIV, these mice produced HIV-specific human antibodies and experienced progressive disease, marked by a depletion of CD4+ T cells and an increase in CD8+ T cells, which mirrors human infection.

Further analysis showed that CD4+ T cells in the humanized mice were also depleted in the lung and gut, which is now known to be a primary site of HIV replication early in human infection. Models like this could vastly improve the ability of researchers to understand a candidate's immunogenicity since it more closely mimics human HIV infection.

In the meantime, the field is eagerly anticipating the results of the ongoing trials to answer some of the critical questions about cellular immunity. "The next few years are going to be very interesting in this field," said Robertson. "Hopefully we will have good news on T-cell based vaccines."