AIDS vaccines: where are we?

Most effective vaccines were produced before we had an extensive knowledge of the mechanisms involved in protection. Although the majority of successful vaccines produce neutralizing antibodies, the correlates of protective immunity are often unclear. Mechanisms of immune protection following vaccination may differ significantly from those believed to be important after natural infection as recently shown by the successful development of vaccines against human papillomavirus [1].

The need for an effective HIV vaccine is manifest, and prodigious efforts in basic research, animal models and in clinical trials have so far produced little in the way of success. Thus, two schools of thought have emerged as the optimum way forward. Either continue with the serendipitous approach, using a variety of candidate vaccines in humans to see if protective immunity emerges, or return to a more basic research agenda to establish what vaccines would be required to induce effective immune responses [2].

Long-term nonprogression

As with any major infection, a small number of individuals acquire HIV but never develop disease. These individuals (elite controllers) have evidence of immune responses to the virus but the viral load remains below detectable limits (<50 copies per/Ml). In a slightly larger group of individuals the viral load is detectable but low, and while it is believed that such individuals will eventually develop disease, if they live long enough, many are asymptomatic in excess of 20 years [3].

The immunological responses of elite controllers to a wide variety of HIV epitopes have been studied intensively as a possible guide to what would be required of a vaccine. Studies of elite controllers show that there is marked genetic (HLA-antigen subtypes, individuals with deletions in the CCR5 receptor or genetic variations in the promoter region of its natural ligands) and immunologic (CD8 T cell cytokines profiles and presence of HIV neutralizing antibodies) heterogeneity among individuals who successfully suppress HIV-1 replication in the absence of antiviral therapy [4]. It is hoped that ongoing studies in this patient population may provide useful leads to the definition of an effective HIV-specific immune response and facilitate the development of a successful vaccine.

Neutralizing antibodies

Elite controllers also have strains specific to neutralizing antibodies and early data in progressive infection suggest that such antibodies occur early and evanescently in the disease course, but the viral mutations render them ineffectual rapidly. More broad cross-strain neutralizing antibodies also occurred at a slightly later stage of disease but again were not associated with viral eradication [5]. Neutralizing antibodies are particularly difficult to produce against HIV, partly because there is enormous sequence variation of target envelope proteins, which are heavily glycosylated and ‘hidden’ from immune recognition. There has been renewed interest in the potential of neutralizing antibody responses to guide vaccine design following the recent failure of the STEP trial (see later) and following data demonstrating that 25% of HIV-1 infected patients have broadly reactive neutralizing antibodies and that 1% of patients have highly potent neutralizing antibodies against a majority of HIV-1 clades [6]. Further studies are required to correlate the presence of potent broadly reactive neutralizing antibody responses with disease progression. The targets of broadly reactive neutralizing antibodies include the CD4 and CCR5 co-receptor on gp120 and the proximal membrane–proximal external region of the gp41 fusion protein [7]. One CD4 binding antibody (B12), which has been isolated from patients with HIV-1 infection, has been shown to prevent retroviral infection in nonhuman primate models of HIV-1 infection. Infusion of broadly reactive neutralizing antibodies delayed viral rebound in HIV-1 patients undergoing structured antiviral treatment interruptions [8]. Data from these studies suggest that efforts to boost humoral immunity should continue to be a focus of HIV vaccine research.

Cell-mediated immunity

Most of the HIV vaccine candidates produced over the last 20 years have concentrated on the importance of cell-mediated immune responses. The CD8 cells are the effector arm of the immune system and respond specifically to both core and surface epitopes of HIV following natural infection. The first occurrence of cytotoxic T-lymphocyte-specific responses following sero-conversion coincide with a fall in HIV viremia. This may not be a causal association, but it suggests that the cellular immune system is important in the control of HIV viremia. Perhaps more important are experiments in simian immunodeficiency virus (SIV)-infected macques where CD8 cell depletion is associated with uncontrolled viremia and rapid disease progression [9].

The strength of most of these CD8 responses during the asymptomatic period of infection parallel the height of a viral load in the plasma, while only a minority are inversely correlated with this viral load. Thus, it is possible that the latter CD8 responses directed to particular epitopes are correlates of protective immunity. The strength of the CD8 response in most instances is, however, most likely a function of the degree of antigenic stimulation. Most individuals who have apparently protective CD8 responses do eventually develop progressive disease, presumably related to further viral mutations. In recent years it has been appreciated that the likelihood that cell-mediated immune responses (CMIs) will produce sterilizing immunity is low and therefore the goal for a vaccine that might be effective has been switched from protection against infection to the ability in infected individuals to reduce the viral load set point, which might have a significant effect of both upon forward transmissibility and the natural history of HIV infection [5].

It was recognized from early mouse models of chronic viral infection that the maintenance of a CD8 response was also dependent upon generation of CD4-specific responses [10]. Thus, most attempts at development of a successful vaccine program have been directed at producing both specific CD4 and CD8 responses. GlaxoSmithKline have recently developed a vaccine that elicits a CD4-specific response only. However, the utility of this approach in practice is uncertain.

Animal models

The roles of animal models in predicting which vaccines might be effective in man remain controversial. The only naturally infectable animal with HIV is the chimpanzee and, as they are a protected species, relatively few experiments have been performed in this model, which often does not produce disease anyway. Infection of macques with a naturally occurring infection SIV is followed by marked immunodeficiency and death. Many researchers believe that vaccines that protect against SIV might serve as a model to guide the development of human vaccines.

Perhaps the most commonly used virus for animal research until recently has been a human/SIV chimera with a SIV envelope and a HIV core; such viruses are highly pathenogenic to monkeys and produce an accelerated disease course. How far results from such animal models are referable to man remains unclear [11]. Particularly as protection in this model was shown by the Merck vaccine (see later), which was found to be ineffective in man. There is also the difficult issue of what route to use for challenge. Until recently, most animals given the vaccine were then challenged by high-dose intravenous use of a virus, which is clearly not applicable to the human condition. More recently, mucosal challenges, either by the vagina or rectal mucosa, have been used.

Human trials

Performing Phase III clinical trials to prove that a vaccine provides protective immunity or reduces the viral set point is a massive undertaking, requiring many thousands of patients and also requiring a high incidence of HIV acquisition in the volunteers. The studies are therefore performed in high-risk groups and in the underdeveloped world. Self evidently, such individuals have to be fully informed about risk reduction methods, which often results in a reduction to the seroconversion rate. A further problem is that the predominant virus in the developed world clade B, is not the major circulating variant in resource-poor countries; while it is likely that a successful vaccine would provide cross clade protection, this is by no means certain.

GP 120

The first large-scale Phase III study vaccine used was monomeric GP 120 [12]. Although it was recognized during the course of this study that humans were naturally exposed to trimeric GP 120, the study was allowed to continue and eventually showed no protective effect and no significant neutralizing antibody or CMI immunity.

The STEP study

Attenuated viral vaccines produce life-long immunity and are therefore attractive vaccine candidates. Attenuated strains of HIV have occurred naturally (e.g., Nef deletion); however, the potential to revert to virulence and the effects in neonates of vaccinated rhesus macaque’s has made researchers wary of performing human studies [13]. An alternative is to use an attenuated vector which would produce HIV epitopes but not contain a complete HIV viral genome so the emergence of virulent infection could not occur. Candidate vectors include cow pox, canary pox, modified vaccinia or adenovirus. The latter was used by Merck to develop a vaccine that was capable of producing strong CMI responses in Phase II human studies. It was recognized at the time that this effect was attenuated in those who had high levels of adenovirus neutralizing antibodies as a result of natural infection.

The results of the large Phase III study of this vaccine were disappointing and have caused a major rethink in the HIV vaccine field. The study was stopped prematurely. It was recognized that despite good CMI responses against HIV this did not produce protective immunity and did not result in the reduction in the viral load set point of those who seroconvert [14]. This study further suggested that those individuals who have been vaccinated were at higher risk of acquiring HIV infection. In post-hoc analysis these were mainly black American males having sex with other men, particularly those who had high natural adenovirus antibody titers. This study has cast significant doubt on the ability of CMI directly against HIV epitopes to produce a useful immunity, although, of course, this should only apply to those epitopes used in the vaccine.

Although we await the results of a further Phase III study, the present status of the HIV vaccine field is that success remains a distant and elusive prospect, but this has not deterred funding agencies from continuing major research efforts.

Additional information

Additional data have recently been published. Data from a very large HIV vaccine conducted in Thailand, which tested a combination of two different HIV vaccines demonstrated a 31.2% reduction in newly acquired HIV infection compared to placebo. This result was just statistical significant (p = 0.039), however, the number of end points (new HIV-1 infections in vaccinated and placebo groups) was very low. It is possible that the findings were due to chance, and further analysis will be required to explore the possibility of confounding variables. Nevertheless, the results of this trial are important as it is the first time that there has been a hint of efficacy in HIV vaccine trials. The findings from this trial were a big surprise as neither of the two HIV vaccines, when tested alone, had shown any efficacy and will stimulate further studies into potential mechanisms of correlates of immune protection and should encourage further progress in this field.

Financial & competing interests disclosure

The authors have 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.