Watching the AIDS vaccine battle one can only ask: What is the significance of the recent bad news?

The second year of the third millennium began with the announcement that the United States has decided not to proceed with a key study of an AIDS vaccine because the prime boost strategy combining Aventis Pasteur's ALVAC-HIV (vCP1452) and VaxGen`s AIDSVAX B/B failed to meet criteria for the phase III large-scale trial (6). The chief reason for this decision is the notoriously poor immunogenicity of the ALVAC vector. This vaccine, as demonstrated in more than 40 phase I and II trials involving about 1900 volunteers, induces low-level HIV-specific CTL responses in about a third of the participants at best, and these responses are rarely directed at more than one epitope. Also, recently reported data from Harriet Robinson's group at Emory University demonstrated that adding a gp120 booster to a DNA/modified vaccinia Ankara vaccine actually reduced efficacy rather than improving it. These disappointing results, announced at the 9th Conference on Retroviruses and Opportunistic Infections (Seattle, 24-28 February 2002), were quite expectable.

More disheartening news came from vaccine trials in nonhuman primates. Two reports from studies in rhesus macaques vaccinated with vaccines that induce only cellular immunity showed only partial protection. Shiver and co-workers reported that after challenge with a pathogenic HIV-SIV hybrid virus, the animals immunized with Ad5 vector exhibited a pronounced attenuation of the virus infection (1). Despite this deficiency, the authors claim that the replication-defective adenovirus is a promising vector for development of an HIV-1 vaccine.

The recent report by Barouch et al. (2) on another primate vaccine trial that induces cytotoxic T cells (CTLs) drew less optimistic conclusions. These authors detected an escape variant in a vaccinated monkey with undetectable plasma viral RNA. This variant, with a single nucleotide mutation within an immunodominant Gag CTL epitope, subsequently caused escape from CTLs with a burst of viral replication, clinical disease progression, and death from AIDS-related complications. These data indicate that viral escape from CTL recognition may be a major limitation in CTL-based AIDS vaccines that are planned to be administrated to large human populations over the next several years.

Even if a vaccine can achieve partial protection mitigating the disease, the long-term consequences would argue against the use of imperfect vaccines. The report by Gandon and co-workers (3) studied partially effective vaccines that are available or are being tested. The authors conclude that vaccines designed to reduce pathogen growth rate and/or toxicity select for more virulent pathogens. This vaccine-driven evolution of microbes leads to higher levels of intrinsic virulence and consequently to more severe disease in unvaccinated individuals. Recently proposed massive vaccination of the most affected populations with a partially effective AIDS vaccine could select for more virulent strains of HIV at the population level. Therefore, using vaccines that offer only partial protection could be worse than not vaccinating at all.

It might be helpful to reconsider the basics of how vaccines work. Preinfection exposure to epitopes of a pathogen induces immunological memory in the B and T cell populations, with production of pathogen-specific circulating CTLs and a neutralizing antibody titer. This immunological imprinting can protect against infection as long as the infecting microbe shares the epitopes covered by the vaccine. If the infection occurs with a variant of the virus, no protection or partial protection is achieved. Mutations in most viruses and bacteria take time to spread within a given risk population, so vaccine developers can catch up with new variants. The genome of HIV-1 is highly unstable, and it only takes months to generate immune- and drug-escape variants that can be transmitted to a naïve or to a vaccinated individual. This leads to "deceptive imprinting" (4), i.e., the cross-reactivity of the variant epitope reinforces the memory response created by the vaccine but fails to stimulate a response against the infecting mutated virus. Thus, the vaccinated individual produces antibodies and CTLs specific for the original vaccine epitopes that are barely effective against the infecting variant. Furthermore, this cross-reactive recall response appears to suppress virgin immune cells from being stimulated by the infecting virus. Had that individual not been vaccinated, his disease might have been milder and delayed. We have previously reviewed literature data demonstrating that envelope-based or envelope-including vaccines are not only inefficient but also dangerous for recipients (5).

There is another danger with partial vaccine protection: It adds to the deceptive sense of false security that can lead to risky sexual practice.

Is there any hope that we will soon have an effective and safe vaccine at the present state of the art and our current knowledge of the HIV-1 infection? Our answer is no. Funds should be directed toward understanding the immune reaction in HIV-1 infection, and work has to go back to the drawing board away from premature vaccine trials. Education and avoiding unsafe sexual practices should be the response to the current AIDS epidemic. If the message that it will be years before we are able to offer an effective vaccine comes across loud and clear, then perhaps infection prevention measures will be taken seriously.

References