Data from phase I safety & immunogenicity studies of Merck’s adenovirus serotype 5 (Ad5) HIV vaccine has just been published online in Clinical Infectious Diseases. The data has been presented at various conferences but this is the first relatively comprehensive report to appear in the literature. The paper reinforces several of the observations that were made during the development of the vaccine: first and foremost, it represented a great leap forward in terms of the proportion of recipients that actually developed a detectable CD8 T cell response to vaccine antigens. Previous bests were around 20-30% of recipients (with an ALVAC canarypox vector). What seems to have been entirely lost in recent coverage of the Merck vaccine failure is that, before this Ad5 construct came along, no one had the luxury of worrying about the impact of issues like the magnitude, breadth, function & specificity of the CD8 T cell response because, for the most part, there was no CD8 T cell response there to analyze!
Secondly, although the issue is only mentioned in passing in the paper and would benefit from a more detailed presentation of the data, it is clear that the T cell responses induced by the vaccine were heavily biased toward CD8 T cell responses and the importance of a balanced CD4/CD8 T cell response is now better recognized. Merck’s decision to drop the DNA priming immunization – criticized at the time by Jerry Sadoff and Bette Korber during an NIAID AIDS Vaccine Research Working Group meeting – may have exacerbated this problem, as DNA vaccines typically bias toward CD4 T cell responses.
Thirdly, the tendency of for Ad5 vectors to persist and thus maintain an activated effector memory CD8 T cell population – which has been reported in small animal models over the past few years – may have been mirrored in humans: the figures in the paper showing the magnitude of HIV-specific CD8 T cell responses over time do not provide much evidence of a contraction phase (click on the image to see the figure), normally a prerequisite for the development of a “central memory” CD8 T cell population endowed with robust proliferative capacity.
Finally, the paper unfortunately offers little data on the breadth of the vaccine-induced responses and yet the authors suggest that it was “broad” – this seems inconsistent with the data from the STEP trial showing that recipients only responded to an average of one epitope from each vaccine-encoded protein.
Clinical Infectious Diseases 2008;46:000–000
Frances H. Priddy,1 Deborah Brown,2 James Kublin,3 Kathleen Monahan,2 David P. Wright,4 Jacob Lalezari,5 Steven Santiago,7 Michael Marmor,8 Michelle Lally,9 Richard M. Novak,10 Stephen J. Brown,6 Priya Kulkarni,2 Sheri A. Dubey,2 Lisa S. Kierstead,2 Danilo R. Casimiro,2 Robin Mogg,2 Mark J. DiNubile,2 John W. Shiver,2 Randi Y. Leavitt,2 Michael N. Robertson,2 Devan V. Mehrotra,2 and Erin Quirk,2 for the Merck V520-016 Study Groupa
1Department of Medicine, Emory University School of Medicine, Atlanta, Georgia; 2Merck Research Laboratories, West Point, Pennsylvania; 3Fred Hutchinson Cancer Research Center, Seattle, Washington; 4Central Texas Clinical Research Center, Austin; 5Quest Clinical Research Laboratories, San Francisco, and 6AIDS Research Alliance, West Hollywood, California; 7Care Resource, Miami, Florida; 8New York University School of Medicine, New York; 9Warren Alpert Medical School of Brown University, Providence, Rhode Island; and 10University of Illinois at Chicago, Chicago
Background. The safety and immunogenicity of the MRK adenovirus type 5 human immunodeficiency virus type 1 clade B gag/pol/nef vaccine, a replication-incompetent adenovirus type 5–vectored vaccine designed to elicit cell-mediated immunity against conserved human immunodeficiency virus proteins, was assessed in a phase 1 trial.
Methods. Healthy adults not infected with human immunodeficiency virus were enrolled in a multicenter, dose-escalating, blind, placebo-controlled study to evaluate a 3-dose homologous prime-boost regimen of the trivalent MRK adenovirus type 5 human immunodeficiency virus type 1 vaccine containing from to viral particles per 1-mL dose administered on day 1, during week 4 and during week 26. Adverse events were recorded for 29 days after each intradeltoid injection. The primary immunogenicity end point was the proportion of study participants with a positive unfractionated Gag-, Pol-, or Nef-specific interferon-γ enzyme-linked immunosorbent spot response measured 4 weeks after administration of the last dose.
Results. Of 259 randomized individuals, 257 (99%) received 1 dose of vaccine or placebo and were included in the safety analyses. Enzyme-linked immunosorbent spot results were available for 217 study participants (84%) at week 30. No serious vaccine-related adverse events occurred. No study participant discontinued participation because of vaccine-related adverse events. The frequency of injection-site reactions was dose dependent. Vaccine doses of viral particles elicited positive enzyme-linked immunosorbent spot responses to 1 vaccine component in >60% of recipients. High baseline antibody titers against adenovirus type 5 diminished enzyme-linked immunosorbent spot responses at all doses except the viral particle dose.
Conclusions. The vaccine was generally well tolerated and induced cell-mediated immune responses against human immunodeficiency virus type 1 peptides in most healthy adults. Despite these findings, vaccination in a proof-of-concept trial with use of this vaccine was discontinued because of lack of efficacy.