HIV vaccine development: the myth of Sisyphus modernized?

Recently, a friend and colleague admitted to me that he had always considered the making of an HIV vaccine to be a Sisyphean challenge. Sisyphus, who managed to return to the living from the Underworld, was sentenced to an eternity of frustration for his trickery by Zeus, the king of the Greek gods. He was to push a huge boulder of stone up to the top of a mountain but everytime he was getting close, the rock slipped out of his grip and rolled down to the base of the mountain again. It is true that trying to produce an HIV vaccine has, indeed, turned out to be a most formidable task. Similar to a modern Sisyphus, every time the goal seems to be within reach, there is a new pitfall and we must start all over again. Resembling Sisyphus’ case, the generation of a so utterly needed HIV vaccine may, indeed, turn out to be an eternity of frustration for modern science.

Certainly, there has been considerable progress in the HIV/AIDS field. Since the establishment of the AIDS epidemic in the early 1980s, scientists have gathered perhaps the largest body of data ever produced in such a short time frame, with more than 100,000 scientific publications in PubMed, leading to the discovery of the causative agent, HIV, and to antiretroviral drugs used for combination therapy. At present, an HIV-infection diagnosis is not a death sentence in the Western world, but rather a condition that can be treated. However, we are no closer to a licensed vaccine today than we were 10 years ago. Today, more than 40 million people are estimated to live with HIV and more than 4 million are newly infected per annum [1]. In addition, the epidemic is responsible for over 3 million deaths each year, and causes millions of children in sub-Saharan Africa to become orphans. Information campaigns and behavioral changes have had an impact on the epidemic in developed and in a few developing countries, such as Uganda and Thailand [1]. Unfortunately, widespread superstition, political and personal unwillingness to change social structures and patriarchal patterns have limited the success of such information campaigns in the developing nations where they are needed the most. To date, a prophylactic vaccine would be the most effective means of preventing epidemic spread of this disease. The main problem in designing an HIV vaccine is that the correlates of protection remain largely unknown. In fact, there is not a single well-documented case of full eradication of an established infection either by the immune system or by combination therapy [2]. The lack of success in eradicating the virus or stimulating the immune system to clear the infection is due to the intrinsic properties of the life cycle of the HIV virus. It mutates rapidly, replicates at a high rate and infects the conductors of the immune system (i.e., the CD4⁺ T cells). Upon infection, its genetic material is integrated into the host cell’s genome, where it may stay dormant for a very long period. Thus, once a viral life cycle is completed, there is very little hope of the infection being cleared. Moreover, the very fast evolution of HIV makes it extremely difficult to predict the antigenic composition of the infecting virus. To further complicate the story, most researchers believe that it is not only one virus, or even a viral strain, that infects an individual but rather a swarm of related viruses referred to as a quasispecies [3]. In addition, HIV animal models have limited value for HIV vaccine researchers for several reasons. Notably, HIV does not cause symptomatic disease in animals, possibly with the exception of chimpanzees [4]. However, for ethical reasons, research on chimpanzees is discouraged. Instead, SIV infection of rhesus and cynomolgus macaques is perhaps the most relevant model for HIV infection and AIDS. One reason that progress is held back is the low infectivity of retroviruses. Thus, to obtain significant rates of infection in experimental settings, researchers are forced to use unrepresentatively high infectious doses or unnatural conditions for experimental challenge that do not mirror the conditions during heterosexual HIV transmission between humans.

Despite these very troubling circumstances, there is some hope that it may indeed be feasible to develop an HIV vaccine. Small cohorts of multiple exposed but uninfected commercial sex-workers and discordant couples in which only one partner is infected exist (reviewed in [5]). It is believed that a combination of innate and mucosal adaptive immune responses is responsible for their immunity. However, it is currently unclear how this immunity is elicited or which components of the immune response are required for protection from infection.

Over the years, several HIV vaccine strategies have gained momentum and given hope to scientists and afflicted populations alike. However, for one reason or another, most of these strategies have been abandoned or modified. Early on, it was believed that recombinant envelope glycoprotein (gp)160, and its cleavage products gp120 and gp41 appearing on the mature virion, could be used to induce neutralizing antibody responses, just as a recombinant envelope protein had for the hepatitis B vaccine (reviewed in [6]). The only Phase III clinical trials concluded and unveiled thus far relied on this approach [7,8]. By the time this vaccine was entering efficacy testing, it had become clear that these trials were likely to fail. The changing nature of the viral envelope proteins and the importance of their trimeric tertiary structure for the induction of neutralizing antibodies, rather than the monomeric gp120 used in the vaccine formulation, were, by then, well known to the scientific community. The heated debate among HIV vaccine researchers on this vaccine concept and its clinical trials highlights the long-lasting commitments required to push a vaccine candidate from bench science to a Phase III trial. I would argue that the researchers pushing this clinical trial made a very significant contribution to the HIV vaccine field. The question is not whether this first attempt to make an HIV vaccine was successful, but rather what we can learn from its achievements and results. Above all, we gained insight and scientific evidence of what is not effective. In addition, there is plenty to be learned from the flaws in the study design and, importantly, regarding the infrastructure of a very large HIV vaccine study. To judge this study, because there may have been sounder scientific approaches to HIV immunization, is useless. There will always be a theoretically better approach by the time a prospective vaccine reaches a Phase III trial.

Perhaps, owing to the failure of recombinant envelope proteins to induce protective immunity to HIV, the next wave in HIV vaccine development focused on the induction of CD8⁺ cytotoxic T lymphocytes (CTLs) with little or no emphasis on antibody responses. Indeed, cell-mediated immunity is thought to play a major role in the host defense against HIV infection [9,10] and is able to control the viral load in symptom-free, chronically infected individuals [11]. In addition, strong CTL responses have been correlated with nonprogression in long-term nonprogressors. In addition, potent CTL responses seem to be able to control viral load and lower the viral setpoint in immunized rhesus monkeys (the viral load during the asymptomatic period of HIV infection is referred to as the viral setpoint, and is prognostic for disease progression, infectiousness and response to treatment) [12,13]. It is important to note, however, that none of the CTL-based vaccine approaches would provide sterilizing immunity, but would allow HIV to enter cells and replicate before the immune system could become activated. Owing to the nature of the HIV life cycle, which means that the virus may persist inactively and stay hidden in immune cells for years before initiating opportunistic active replication, eradication of established infection may be unfeasible. Instead, the idea behind the CTL-based vaccines would be to keep virus replication to such a low level that disease progression to AIDS would be prevented and the risk of disease propagation reduced. In experimental models, there is, however, evidence that CTL escape mutations can occur, allowing the virus to outrun the immune system and leading to disease progression [14]. CTL escape is also correlated with AIDS progression in some HIV-infected patients [15,16].

Today, the focus of HIV vaccine development is on all branches of the adaptive immune system and it is generally believed that both neutralizing antibodies and CTL responses will play a role in a successful vaccine candidate. Therefore, new ways of inducing potent HIV-specific CTL responses, as well as new strategies to elicit broadly neutralizing antibodies, are investigated. To mount a potent antibody response capable of recognizing conserved envelope structures and neutralizing most circulating HIV strains, a number of groups hope to produce an HIV ‘superantigen’ [17–19]. This pioneering idea represents the kind of thinking out of the box that is necessary for the field. However, it is a difficult path full of pitfalls and preclinical data leading to clinical trials based on this strategy have yet to be demonstrated. In addition, there is considerable activity aimed at developing new vaccine delivery strategies and adjuvants that could be applied to the most promising antigens or antigen combinations.

Nevertheless, the remaining central question is: what does a successful vaccine need to do? The HIV epidemic has no precedent in the total lack of documented cases of infected individuals who have been able to clear the infection, and the very low proportion of infected people who are able to control it without combination therapy. There is also a considerable risk that no animal model will help to answer this central question completely, and that we will, therefore, have to turn to more tedious and expensive large-scale clinical trials for tentative answers. Naturally, the ultimate goal for an HIV vaccine is to provide protection from the initial infection, thus, sterilizing immunity. Sterilizing immunity has been achieved in experimental settings when rhesus macaques have been inoculated with very high concentrations of highly neutralizing monoclonal antibodies against the challenge virus [20–22]. Since sterilizing immunity is not a probable outcome for first-generation HIV vaccines, we may have to settle for a compromise. It is conceivable that a vaccine inducing both CTL and potent antibody responses could be generated. If such a vaccine could stimulate the immune system to control the initial burst of viremia and achieve a lower viral setpoint, the immune system might also be able to control viral replication on a longer term. Most experts agree that the less circulating virus that can be detected, the lower the risk of transmission [1]. HIV is considered to be less contagious than most other sexually transmitted diseases, with an estimated transmission rate of 0.1–1% per heterosexual exposure. Therefore, a decreased transmission rate would have a tremendous impact on epidemic spread [23,24].

Some researchers argue that a deeper understanding of the HIV virus can only be achieved through more basic research and that we must clarify the correlates of protection before we stand a chance of developing a rational and functional HIV vaccine. It is possible that they are right, but the conditions used in existing challenge models may not be predictive of the outcome in a clinical trial. In fact, the extremely stringent conditions in most experimental challenges may very well underestimate the efficacy of a vaccine candidate. It is also important to remember that in existing animal models we are not dealing with the wild-type HIV, but with viruses specific for other species or chimeric viruses. Therefore, it is possible that some of the information that we urgently need to design an effective HIV vaccine can only be generated in clinical efficacy trials. After all, we do not know what is needed to achieve full or partial protection in humans, nor do we know how the virus would react to a pre-existing, vaccine-induced immune response at the time of primary infection. It seems prudent not to put all our eggs in one basket, but rather to encourage research in all aspects of the HIV life cycle, immune responses and protection in animal models, as well as in clinical trials. We need to promote creative approaches to HIV vaccine development wherever they occur. Unfortunately, it may prove impossible to make an HIV vaccine that protects from primary infection or even disease, but we are under a moral obligation to keep performing this Sisyphean task. Perhaps, then we can too, like Sisyphus, outsmart death.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.