The Merck HIV vaccine efficacy trial was conducted in collaboration with the National Institutes of Health-sponsored HIV Vaccine Trials Network (HVTN). At Keystone, the head of the HVTN, Larry Corey, provided an update on the ongoing analyses of data from the study and offered his thoughts on the implications for future research. Most intriguingly, Corey reported that some hints of correlations between vaccine-induced CD8 T cells and viral load setpoints have emerged, suggesting that the magnitude of the response may be important. Specifically, among the participants with anti-Ad5 antibody titers of less than 1:18 units, the magnitude of CD8 T cell responses to Gag (as measured by interferon gamma ELISpot four weeks after the second immunization) showed an inverse correlation with viral load setpoint (p=0.01). Responses to Pol displayed a similar trend (p=0.07) while those targeted against Nef did not (p=0.22). It must be emphasized that Corey cautioned that the numbers are small and the significance is fragile (additional data points may alter the findings), but the data is consistent with prior SIV challenge studies demonstrating inverse correlations between the magnitude of pre-challenge SIV-specific T cell responses and post-challenge viral load levels. However, no such correlation was seen among individuals with anti-Ad5 titers over 1:18 and exactly how pre-existing immunity would impact the relationship between CD8 T cell responses and viral load is unclear. Additional analyses of relationships between pre-infection HIV-specific CD4 T cell responses and post-infection viral load setpoints might shed light on this question, as one speculative possibility is that individuals with pre-existing immunity developed weaker HIV-specific CD4 T cell responses due to interference from vector-specific memory CD4 T cells.
In addition to the issue of magnitude, Corey also pointed out two other potentially significant limitations of the T cell responses induced by the Merck vaccine that could conceivably have contributed to its failure: breadth and functionality. Biometric data indicates that HIV’s Gag protein contains ~150 potential T cell epitopes that should be present in a substantial proportion of circulating clade B HIVs (>15%), but STEP participants only developed a CD8 T cell response to an average of one epitope from the full-length Gag protein encoded by the vaccine (similarly, participants only responded to one epitope in each of the other two vaccine-encoded proteins, Pol & Nef). Corey cited several strategies that are being pursued to increase breadth, including the use of genes from multiple different HIV variants in prime-boost vaccines (heterologous inserts), protein or peptide boosting and prime-boost with differing vaccine vectors (heterologous vectors). In terms of functionality, Corey offered his own “revisionist bit of history,” pointing out that it is now understood that Ad5 vectors given alone bias the CD8 T cell response toward a terminally differentiated phenotype with poor proliferative capacity. In support of this argument, he showed data comparing the phenotype of CD8 T cells induced by priming and boosting with Ad5 alone compared to DNA/Ad5 and also cited unpublished data from Rafi Ahmed’s lab showing that an Ad5 vector offered poor protection against LCMV due to the induction of T cells with limited proliferative capacity. It is currently unclear if this problem is also seen with other adenovirus serotypes (some recent macaque data indicates that there is some serotype-specific variability in the phenotype of vaccine-induced T cell responses).
Corey also described a suite of studies the HVTN is either conducting or planning to conduct in order to better understand the outcomes of both the STEP and Phambili Merck vaccine trials, including looking at the interaction between Ad5 immunization and pre-existing Ad5-specific immune responses at mucosal sites, sequencing the viruses in individuals that became infected, and analyzing the impact of HLA types on both susceptibility to HIV infection and the immune responses to the vaccine.
Looking at the bigger picture, Corey echoed recent statements about the importance of discovery research for developing neutralizing antibody-based or other approaches that may lead to complete protection against infection. But he also stressed that the evidence still strongly supports pursuing the nearer-term goal of trying to increase the chance that an individual who becomes infected will control the virus immunologically, thus reducing the risk of onward transmission and preventing or slowing progression to AIDS.
Keystone Symposia: HIV Vaccines: Progress and Prospects
Banff, Alberta, Canada, March 27-April 1, 2008
Abstract #002
An Overview of the SteP trial
L. Corey
MRK HVTN Study Team, University of Washington, Seattle, WA
Background: A collaborative study between Merck, the HVTN, and NIH was initiated in 2004 to perform a randomized, multi-center, double-blind, placebo-controlled test-of-concept study in 3000 HIV seronegative volunteers at high risk of acquiring HIV. Volunteers were randomized (1:1) to receive 3 injections of either the MRKAd5 HIV-1 gag/pol/nef vaccine (a replication defective Ad5 vector) or placebo. Volunteers were tested ~every 6 months for acquisition of HIV, in volunteers who became HIV infected, viral load and CD4 cell counts were measured at multiple time points post diagnosis. Results: The study was stopped in September 2007 at the first planned interim analysis; 82 endpoints were noted in men and only 1 in women. Among male volunteers with low (<200) baseline Ad5 neutralizing antibody titers (the primary analysis population), there were 24 infections among the 741 vaccinees compared to 21 infections among the 762 placebo recipients (using a modified “intent-to-treat” approach), and among the volunteers who became HIV infected, the geometric mean plasma vRNA level was ~40,000 copies/mL in the vaccine groups compared to ~26,000 copies/mL in the placebo group. Vaccine and placebo recipients were well-matched in demographic and risk characteristics at baseline within Ad5 strata. In ongoing multivariate analyses, the following factors appear to be more strongly associated with HIV-1 acquisition; treatment (vaccine vs. placebo), hazard ratio (HR) 1.9, 95% CI (1.1, 3.2); region (North America vs. other), HR 3.2 (1.5, 6.7); and unprotected receptive anal sex (yes vs. no), HR 4.4 (2.4, 8.7). With interaction terms in the model, the treatment HR (vaccine vs. placebo) was 3.1 (1.5, 6.5) in Ad5 seropositives vs. HR 1.0 (0.5, 2.0) in Ad5 seronegative; the treatment HR was 4.5 (1.8, 11.4) in uncircumcised men vs. 1.0 (0.6, 1.8) in circumcised men. A 25% sample of all recipients and all persons who subsequently acquired HIV was performed at week 8 after 2 doses of vaccine. The frequency of CD8+ T cell responses to HIV genes (76%) was found in persons who subsequently acquired HIV-1; similar to that seen in persons without HIV acquisition. Similarly the GMT of responses was also similar in vaccinees who acquired and did not acquire HIV. Conclusions: There was no evidence thatvaccination prevented infection or lowered viral setpoint; in fact, there were more infections in the vaccine group. In multivariate analysis of baseline risk factors the Merck Ad5 trivalent vaccine appeared to be associated with an increased risk of HIV-1 acquisition in men with pre-existing Ad5 immunity and in uncircumcised men. The mechanism for potential increased acquisition will be discussed.
Corey cited several strategies that are being pursued to increase breadth, including the use of genes from multiple different HIV variants in prime-boost vaccines (heterologous inserts), protein or peptide boosting and prime-boost with differing vaccine vectors (heterologous vectors).
This is particularly frustrating to me since my R01 grant proposal, designed to develop ways of making a broader T cell response, was just turned down, with the main reviewer claiming that immunodominance is simply not important or worth looking at in humans.
Posted by: iayork | April 10, 2008 at 03:45 PM
Sorry to hear that Ian, but it's helpful to know - TAG is working on some recommendations and responses to the current sturm and drang in HIV vaccine research and clearly one recommendation needs to be that study sections are informed about the key current issues; increasing the payline is obviously vital, but it's not going to help much if important grant proposals get rejected due to ignorance on the part of the reviewers!!
Posted by: Richard Jefferys | April 10, 2008 at 04:17 PM