Introduction
Sarah J. Glover/Philadelphia Inquirer/MCT
A patient receives an experimental cancer vaccine. Following the failure of Merck's V520 HIV vaccine, researchers are questioning the safety of certain vectors used in vaccines.
The recent failure in phase 2 of Merck's HIV vaccine was a devastating blow to the AIDS community. But the finding that the vaccine, called V520, seemed to make some study participants more susceptible to the disease was a shocking coup de grâce that has had a chilling effect on vaccine research using adenovectors.
"No one expected this," says Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID).
In the Whitehouse Station, New Jersey–based Merck trial, participants that had higher levels of preexisting immunity to the disabled cold virus used as a vector in the vaccine were more likely than those with lower immunity to become HIV infected. Although it's impossible for researchers to say whether the results are meaningful, specu-lative hypotheses abound and the results sent ripples through a vaccine community already dissatisfied with the vector, called adenovirus 5 (Ad5).
From a developmental standpoint, Ad5 was the logical choice: researchers have been safely working with it for decades, and there are few vector alternatives ready for the clinic. Considering that the trial was a proof-of-concept study evaluating the use of T-cell vaccines, researchers deemed it best to get the vaccine itself into the field as quickly as possible.
But the vector has its limitations. Because humans encounter it frequently in their everyday environment, there are substantial levels of preexisting immunity to the vector, which could potentially disable it before unloading its cargo. "I don't think anybody wants to use Ad5 in humans," says Jonathan Bramson, who studies genetic vaccines at McMaster University in Hamilton, Canada. "They're using it because they have to."
The Merck trial originally was designed to vaccinate only those with low Ad5 immunity, but researchers lifted that restriction after animal studies showed that proposed doses were likely high enough to overcome the preexisting immunity. Besides, study designers had good reason to take the chance on expanding enrollment: >80% of sub-Saharan Africans (a population with a large need for a vaccine) have high levels of Ad5 immunity.
The results from the Merck trial are "going to push people to come up with strategies in which preexisting immunity is not going to be a problem," says Gary Kobinger of the National Microbiology Laboratory, Public Health Agency of Canada, Manitoba.
It's not clear what those strategies are, but researchers have been discussing a return to excluding those with high preexisting immunity in future trials, says Gary Nabel, director of the National Institutes of Health Vaccine Research Center. For diseases like HIV and Ebola, that will mean leaving out a large part of the population that stands to benefit most from HIV vaccines. Although that sacrifice is unfortunate, "these are not licensure trials. We're not going to get 95% efficacy with just one vaccine trial," says Lawrence Corey, lead researcher on the Merck HIV vaccine trial. "Ad5 is not likely to emerge as the vaccine, but that doesn't mean testing the adenovirus prototype is wrong."
There are suggestions that using the Ad5 vector in a "prime-boost" system could reduce the influence of preexisting immunity, and some upcoming HIV vaccine trials by the NIH will explore that option. In the NIH trial, volunteers are given three shots of naked DNA plasmid vaccine, followed by one shot of a vaccine that uses Ad5 in the hope that the first shots of naked DNA—though not likely to be a strong vaccine on their own—will ready the immune system for the final Ad5 shot. The response to that final shot is then boosted, so as to be strong enough to overcome any vaccine that is eliminated by preexisting Ad5 immunity.
This approach holds a second advantage: the Ad5 vector will be given only once, rather than three times as it was in the Merck study. Even so, the NIH trials have been put on hold while researchers decide how best to respond to the Merck data. "Needless to say, we have put things on pause," says Nabel. "Almost without question there will be changes in the design of that trial."
There also have been a few clinical trials of an HIV vaccine that does not rely on Ad5 at all, instead using an adeno-associated virus (AAV). But trouble has arisen there, as well: shortly after the Merck trial data were released, AAV data were published showing that the vector can produce abnormal T-cell responses (J. Clin. Invest. 117, 3958–3970, 2007). The paper's authors say the data suggest that the result, in mice, could be a weake-ning of the immune response against the very pathogen the vaccine was designed to fight. "Just envision that you vaccinate people with this," says Hildegund Ertl, an immunologist at the University of Pennsylvania and an author on the paper. "This may be dangerous if it also happens in humans."
Targeted Genetics, a Seattle company with an AAV-based HIV vaccine in development, says that Ertl's findings, which relied on data in mice and high doses of vaccine, are unlikely to be relevant to their product. It's not clear if regulators will agree, but Ertl says the results show that safety issues with T-cell vaccines should consider not only the magnitude of the T-cell response, but also the quality of the T-cells that are activated. "I think there will be a shift of analysis," says Ertl. "There must be a test for T-cell vaccines that will also look at T-cell functionality. It [the AAV-based HIV vaccine] induced a T-cell response, but that didn't do the job. "
All of this might become moot if alternative adenovectors in development can prove their worth (Table 1). Some have suggested adenoviruses from serotypes rarely found in the environment are the answer, as preexisting immunity to those viruses is much less frequent than Ad5 immunity (though often higher in the developing world). Two serotypes are about to enter clinical trials: adenovirus-26 and adenovirus-35. "We're still figuring out how the Merck data will affect these studies," says Dan Barouch, a virologist at Harvard Medical School in Boston. An adenovirus-26 phase 1 study was expected to start in October, he says, but was postponed. "People need a little more time to digest the Merck information," says Barouch. "We're hoping it'll start in the next few months."
Also being considered are chimeric vectors that combine regions of Ad5 with Ad48. These vectors normally delete the regions of Ad5 that are common antigens. Researchers hope that this will allow the chimeras to pass into the body without detection. Clinical trials of an Ad5 chimeric vector are expected to begin in early 2008.
Others have been looking for rarer adenovirus serotypes by searching for viruses found in animal populations. Preliminary work in animals suggests that viruses constructed from chimpanzee adenoviruses may dodge elimination by anti-Ad5 immunity.
But the best way to sidestep the safety problems associated with using viral vectors is to lose the virus altogether. However, naked DNA vaccines, which consist of plasmid DNA containing little more than genes encoding the antigen and a promoter, have been plagued by low expression in human cells. "Moving away from viral vectors to plasmids would eliminate many problems," says Ertl. "But I'm just not sure that what's out there right now is quite potent enough." Companies are trying to overcome that barrier by improving the pene-tration of DNA into the cell through electroporation or particle bombardment. One of the Ebola vaccines relies on plasmid DNA with no viral vector; it is being investigated by NIAID and has finished phase 1.
Still, all alternatives need extensive safety testing before they will be of use in vaccines. For now, researchers are likely to continue to use Ad5 for the sake of establishing a proof-of-concept trial. "It's a scientific tightrope," says Nabel. "We err on the side of safety, but what you don't want to do is decide not to move forward for the wrong reasons."
