Adenovirus-based HIV vaccines producing strong immune responses

HIV vaccines using adenoviruses as vectors are producing strong immune responses in early studies and showing some evidence of stimulating immune responses across different HIV-1 clades, according to studies presented on Wednesday at the AIDS Vaccine 2006 conference in Amsterdam, the Netherlands.

Adenovirus-based vaccines are being developed by Merck and by non-commercial research groups because adenovirus vectors are the most consistently immunogenic viral vectors yet tested, and because they can be manufactured and administered at vastly greater doses than pox virus-based vaccines, according to Professor Larry Corey of the HIV Vaccine Trials Network at the University of Washington in Seattle.

He said: “2006 has been a vintage year for vaccine development.”

He pointed to preclinical data from primates suggesting that a T-cell based vaccine given by injection can moderate the massive loss of memory CD4 cells from the gut seen during the first weeks of SIV infection, and the evidence of immunogenicity now being reported from phase II studies of T-cell based vaccines.

T-cell based vaccines seek to stimulate cell-mediated immune responses that can control HIV replication through CD4 and CD8 cell responses. These vaccines may protect against HIV infection or they may protect against disease progression by limiting early viral replication – no one knows which approach will work, and large-scale trials will be needed before their chief advantage becomes clear.

HIV Vaccine Trials Network study 054 tested what dose of adenovirus 5 might be needed in order to produce an immunogenic response, and also tested the safety of different adenovirus doses. Since up to 90% of the adult population may have some level of neutralising antibody to adenoviruses in sub-Saharan Africa, the HIV Vaccine Trials Network wanted to test whether it was necessary, or safe, to use a very high dose of adenovirus (a concentration of 10¹¹ PU).

The vaccine consisted of four different adenovectors expressing HIV subtype B Gag-Pol-Nef and subtypes A, B, and C Env. All participants had an adenovirus 5 neutralising antibody titre below 1.12 and were randomized to a dose of 10¹⁰ (n=20) or placebo (n=4). After the first dose proved safe, a further 24 participants were randomised in the same manner to a dose of 10¹¹ or placebo.

The higher dose proved to be much more reactogenic, causing flu-like or injection site reactions in a high proportion of those who received it. Four participants in the high dose arm had severe flu-like reactions within one day of injection, although all conjunctival and upper respiratory tract cultures for adenovirus 5 proved negative. Systemic and injection site reactions were significantly less frequent (p

Eighty-seven per cent of the lower dose group showed positive ELISPOT reactions for any HIV antigen included in the vaccine by day 28, although the proportion who registered reactions to Gag, Pol and at least one Env gene was not reported. A similar proportion (85%) showed immunogenicity in the higher dose arm, leading the investigators to conclude that a dose of 10¹⁰ PU should be taken forward.

The Merck replication-defective adenovirus vaccine, now being tested in a large three-year phase IIb study in North and South America and Australia called STEP, combines clade B gag, nef and pol.

Analysis of phase I participants who received the Merck adenovirus-based vaccine presented at the Vaccines 06 meeting showed that around 70% of participants who produced CD8 T-cell responses to HIV-1 subtype B also had T-cell responses to consensus sequences of subtype A and C as measured by ELISPOT interferon-gamma assay, although the proportion of responders to subtype B nef that were responsive to subtypes A or C nef was much lower, at less than 30%.

A number of presenters at the meeting emphasised the need to look for the most highly conserved sequences of the HIV-1 group M, to which all HIV-1 clades belong, in order to identify the antigens that could be targeted by a vaccine. Bette Korber of Los Alamos National Laboratory in the United States said that as recombinant viruses spread, it made sense to target the similarities in group M rather than developing vaccines for each different subtype. “There’s no way we’ll be able to cope otherwise,” she said.

She explained that the search for conserved sites and critical differences between viruses needn’t be as difficult was once thought.

“In conserved HIV proteins, the positions where substitutions are tolerated usually oscillate between a few common substitutions,” she said. “The observed frequencies are the result of a balance between fitness, immune escape and the evolutionary relationships of the strains we sample.”

In HIV-1 subtype B for example, her research group has found that within pol, the most highly conserved HIV antigen, only four of 1,004 nucleotide positions within the genome require as many as four amino acid changes, a finding which reduces the number of potential variations that need to be covered by a potential vaccine.

Prof. Andrew McMichael’s research group at the University of Oxford is also looking at constructing a vaccine which is immunogenic across subtypes, and reported at the meeting that a construct called HIVCON spanning the most conserved regions of subtypes A, B, C and D has proven highly immunogenic when tested against plasma from healthy volunteers and HIV-positive people.