Is the Salk principle still viable for the design of an effective HIV vaccine?

Enormous resources are currently being expended on the development of an effective HIV vaccine – for good and obvious reasons. One of the earlier vaccine approaches was to apply the principles used by Jonas Salk when his group developed the polio vaccine in the 1950s, namely to use whole but inactivated virus. HIV was stripped of its outer surface proteins, gp 120, and combined with incomplete Freund's adjuvant. The resultant vaccine, HIV-1 Immunogen (REMUNE®; Immune Response Corp., Carlsbad, CA, USA), was tested as a therapeutic vaccine in one phase III trial that failed to demonstrate an increase in HIV progression-free survival [1], and subsequent subgroup analysis failed to demonstrate consistent effects of the vaccine on HIV-RNA and CD4 cell count, although benefits were reported in somesubgroups [2]. A phase II trial also failed to find a significant difference in a composite laboratory primary endpoint [3]. Although HIV-1 Immunogen is able to boost the HIV-1-specific CD4 T-helper response [4], this body of evidence suggests that the response is not of sufficient potency to improve clinical outcome. However, HIV-1 Immunogen remains under active investigation, and data continue to emerge from previous trials arguing for a potential role after all. On balance, these reports do not, however, provide consistent evidence of therapeutic benefit. In this issue of HIV Medicine, Chantratita et al. [5] report on a long-term retrospectively performed observational assessment of outcome among patients originally included in a randomized phase II trial of HIV-1 Immunogen. The trial included asymptomatic HIV-infected persons with a CD4 count above 300 cells/μL recruited at several Thai clinics. After the completion of the 40-week study period, HIV-1 Immunogen was offered to all study participants, and, of the 297 originally enrolled patients, 75% were included. This cohort was stratified based on CD4 level at week 132, and, not surprisingly, those with higher CD4 cell count at this time did better. The clinical progression rate was low and few individuals developed resistance. Taken together, studies on HIV-1 Immunogen have taught us several important lessons regarding the immune response to HIV infection and the ability to manipulate that immune response in a clinically effective manner. First, they have taught us that it is possible to increase in vitro lymphocyte proliferative responses to the p24 antigen [4]. It has been postulated by many groups that this in vitro measurement is an important reflection of the ability of the host to suppress HIV replication. Unfortunately, secondly, the HIV-1 Immunogen studies have shown that increasing this parameter may not lead to a substantial change in host defence against HIV, and raise the question of whether this measurement is truly a manifestation of HIV protective immunity or merely a reflection of exposure to HIV antigens. A third fact highlighted by the original randomized trial [1] is that clinical endpoint studies in patients with relatively early HIV infection require large numbers of patients followed for an extended period. Any group seeking to establish the clinical impact of such an intervention needs to accept this reality in their study design and to carry through their commitment to completion of the study. To interpret the results of the present study, several factors need to be taken into account. First, as noted above, this study was neither prospective nor randomized. Thus, considerations of investigator bias must be factored into any interpretation. Secondly, the analysis of changes in viral load involves a comparison of HIV-1 Immunogen responders to nonresponders. A similar type of responder analysis indicated a clinical benefit of gp160 immunotherapy in an earlier study. This benefit was not seen in the context of a randomized controlled trial, highlighting the potential bias such an analysis may introduce. Thirdly, the comparison of event rates in a year 2000 cohort with those in pre-1995 cohorts is extremely difficult without correcting for factors such as baseline CD4 count. Several other HIV vaccine candidates are currently under development. Such candidates utilize a variety of other principles in order to boost cellular immunity. The challenge will be to rationally explore which of these candidates stands a reasonable chance of being clinically effective. A key determinant in this work will be to reach a consensus on which efficacy outcome parameter(s) to use to claim that a vaccinated population has gained clinically meaningful protection from a given vaccine. The fact that no HIV vaccine has reached the stage of approval for usage outside of research protocols, however, demonstrates that achieving such a consensus is not straightforward. Nevertheless, a consensus is necessary and is currently being explored.