A novel vaccine using the common virus Cytomegalovirus (CMV) as the vector or container of proteins from the simian immunodeficiency virus (SIV) protected none of a group of 24 rhesus macaques from infection. But in 13 of the monkeys vaccinated, it did produce infections characterised by an undetectable viral load.
This profound viral suppression led to an apparent decline in the number of SIV-infected cells over a period of one year after infection to the point that SIV-infected cells were undetectable in 72% of monkeys with controlled viremia. Despite this, there was no apparent waning of immune responses to SIV in the all-important effector-memory CD8 and CD4 lymphocytes over this time in 12 of the 13 monkeys.
The researchers comment that their vaccine seems to have produced “an unprecedented level of SIV control and even the possibility of progressive clearance of SIV infection over time”.
The question now is how to make a safe analogue of this vaccine for use in humans.
The vaccine tested enclosed SIV components within the shell of another virus which establishes an ongoing but non-pathogenic infection. In this case they used CMV, which is a ubiquitous infection in rhesus monkeys and is present in about 50% of humans. The vaccine, therefore, acted not as a new viral infection but as a ‘superinfection’ of a new variety of CMV.
In this experiment, 24 rhesus macaques were given the novel CMV-vector vaccine, which contained the viral proteins gag, nef, rev, tat, env and parts of the polymerase (pol) protein. Twelve of these animals were given two doses of the CMV vaccine. The other twelve were given one dose of the CMV vaccine and then one dose of a more conventional adenovirus-vector vaccine containing the gag, pol, env and nef SIV proteins.
The 24 CMV-vaccinated animals were compared with nine animals given the adenovirus-vector vaccine alone. All vaccinated animals were then challenged, 13.5 months after the first vaccination shot, by rectal introduction of a highly pathogenic SIV variant.
The course of infection was compared with 28 controls that were all unvaccinated but were challenged with SIV: eleven were challenged with SIV at the same time as the vaccinated animals, while the other 16 had been challenged previously.
The CMV vaccine did not work by protecting the animals against infection and indeed every single one of the challenged animals was infected, with no statistical difference in the number of challenges need to establish infection.
Interestingly, one of the control animals displayed a delayed infection, maintaining an undetectable viral load despite testing antibody-positive for SIV for the first 105 days. After this, however, SIV suddenly appeared in the blood and soon established itself at normal levels, and the subsequent course of infection was similar to that in the other controls.
However, in just over half (13) of the CMV-vaccinated animals, the subsequent infection resembled those seen in ‘elite controllers’: after an initial spike of virus in the blood, they quickly achieved an almost complete control over their virus, maintaining viral loads under 30 copies/ml with occasional ‘blips’, usually to no more than 1000 copies/ml. The frequency of blips declined after week 30 post-infection from 1.5 blips per ten-week period to 0.1 per ten weeks, and then stayed at that frequency for the remainder of the 700-day follow-up period.
Immune responses in CD8 T-cells to the SIV gag and pol viruses remained strong throughout the follow-up period on CMV-vaccinated animals. In contrast, responses to an SIV protein that was not contained in the vaccine, the vif protein, while starting out at the same levels, declined to 10% of its initial level over the 700 days.
The researchers hypothesised that this response to vif – which must be caused by the monkeys’ natural response to SIV – might be declining over time because the number of cells infected with SIV was declining.
This proved to be the case: after sacrificing four vaccinated animals, they tried to find cell-associated viral DNA in cells taken from the gut, lymph nodes and other tissues and found none in 72% of the animals. Excitingly, the levels of cell-associated DNA seen were almost indistinguishable from the proportion of ‘false positive’ DNA results seen in an uninfected animal and far lower – in the order of one DNA copy per 100 million cells – than the levels seen in two animals that had achieved long-term viral control in two previous vaccination experiments (about one DNA copy per half a million cells).
In the nine animals given the more conventional adenovirus vaccine, none achieved an undetectable viral load, but they did initially display a lower viral load that the control animals. However, their viral load eventually returned to normal levels.
What is different about the CMV-based vaccine? Vaccines using viral vectors and ‘fake viruses’ are now almost commonplace. However, previous vaccines using viruses such as adenovirus have produced inconsistent results in animals and in humans, the only large efficacy trial using a viral vector alone, the STEP trial, may actually have increased some people’s vulnerability to HIV. The more successful RV144 trial used another viral vector but the vaccine’s (rather weak) efficacy appeared to be generated by an antibody response to the other vaccine component, not to the viral vector.
The CMV vaccine stimulated immunity in a different group of T-cells. Previous viral vectors have stimulated immunity in the central memory T-cells that mainly dwell in sites like the lymph nodes. This vaccine however mainly stimulated immunity amongst the effector-memory T-cells that patrol the mucous membranes. The researchers hypothesise that the vaccine is able to interrupt the process of infection at an earlier stage, before the SIV has travelled to the lymph nodes and established a fully-productive infection.
However, the researchers emphasised that there is a lot they still do not understand about the immune response seen to the CMV vaccine, exactly why it produced such a powerful response and, crucially, what distinguished the 13 monkeys who responded from the eleven who did not.
Vaccine researchers welcomed the study, but said that it might be difficult to manufacture a version of he vaccine that was safe to use in human studies, given that CMV can cause a number if significant illnesses, especially in people with AIDS and compromised immunity.
Professor Sir Andrew MacMichael of Oxford University told the BBC: "CMV is not totally benign, it does cause a number of diseases. If you're giving people something you're not going to be able to get rid of should it cause problems, then that's quite a difficult risk to manage."
Lead investigator Louis Picker, of the Vaccine and Gene Therapy Institute in Oregon, said in reply that 99% of people in sub-Saharan Africa are already infected with CMV. He said: "We know a lot about it and it's mostly non-pathogenic, except in vulnerable populations like pregnant women." He added that his team were now looking to create a vaccine that had the same immune activity with a weakened version of CMV that could not cause harmful infection.
Hansen SG et al. Profound early control of highly pathogenic SIV by an effector memory T-cell vaccine. Nature, early online publication, doi:10.1038/nature10003. May 2011.