In September 2007, the HIV vaccine field received an unexpected setback when it was announced that the phase IIb efficacy trial of a candidate developed by Merck was being stopped early due to lack of efficacy. The trial was conducted by the HIV Vaccine Trials Network (HVTN) and was referred to as the STEP study. The Merck vaccine aimed to stimulate T cell immunity against HIV, and used a novel attenuated adenovirus serotype 5 (Ad5) vector to deliver the HIV antigens Gag, Pol and Nef. The Ad5 approach was selected because it induced unprecedented levels of HIV-specific CD8 T cell responses in phase I and II trials, with >70-80% of recipients responding (the previous best was a dismal 20-30% of recipients showing low-level CD8 T cell responses after immunization with an ALVAC vector). Although HIV-specific CD8 T cell responses were not anticipated to protect against acquisition of HIV infection, evidence indicated that they might be able to suppress HIV replication and thereby increase the chances of vaccine recipients becoming elite controllers if they became HIV-infected.
The STEP study was stopped after an interim analysis by the Data Safety Monitoring Board revealed that this hoped-for salutary outcome was not being observed. Furthermore, it transpired that certain subgroups of the trial population experienced a significantly increased risk of HIV acquisition associated with receipt of the vaccine. In the overall results, this finding represented a strong trend that did not quite reach statistical significance. However, in a pre-specified analysis that evaluated results based on baseline levels of antibodies to Ad5, there was a stepwise increase in the risk associated with vaccination as anti-Ad5 antibody titers increased, strongly suggesting a real biological effect. Subsequent post-hoc studies revealed that the effect appeared entirely concentrated among uncircumcised men who have sex with men. To their great credit, the HVTN engaged in a massive and extremely transparent effort to investigate this outcome, involving both investigators affiliated with the network and the solicitation of input from external scientists with relevant expertise.
In the time since, a variety of investigations have been conducted but so far no causative mechanism has been identified to explain the STEP results. Unfortunately, along the way, some figures in leadership roles in the HIV vaccine field mistakenly attempted to suggest that the adenovirus vector had been absolved from having any role in enhancing HIV acquisition risk. For example, in 2009, Alan Bernstein (then head of the Global HIV/AIDS Vaccine Enterprise) had this to say to The Scientist:
"[This] result really rules out the possibility that it was the vaccine itself, and the fact that we used Adeno5, that was somehow increasing susceptibility to acquiring [HIV] in those volunteers."
The erroneousness of Bernstein’s public claim—and similar but less public assertions made by some scientists and policymakers in conference hallways—is highlighted by a new paper published in JAIDS. After many careful analyses, the researchers (from the HVTN and Merck) found that no potential confounder could account for the increased risk of HIV acquisition associated with receipt of the Ad5 vector. The HVTN has continued to be forthright about the issue, and it was these as-yet-unpublished analyses that led the network’s principal investigator Lawrence Corey to state in 2010:
“Even though we don’t have a mechanism [for Adeno5-mediated enhancement of HIV acquisition risk], we’re pretty sure this is a real biologic effect”
Continued efforts to understand the mechanism are important because while Merck’s Ad5-based HIV vaccine has been discontinued, adenovirus vectors from an array of different serotypes continue to be studied as potential vaccines against HIV, TB, malaria and hepatitis C, and it is currently uncertain if they might also have the potential to increase HIV acquisition risk.
The most commonly cited hypothesis to explain the STEP outcome was that the vaccine had boosted numbers of Ad5-specific CD4 T cells, thereby increasing the pool of cells potentially susceptible to HIV. But several papers have reported data that is inconsistent with this idea: blood levels of Ad5-specific CD4 T cells did not associate with acquisition risk, and these responses were also rapidly induced in study participants who lacked anti-Ad5 antibodies at baseline (yet these participants experienced no increase in risk). Another notion was that perhaps anti-Ad5 antibodies levels correlate with HIV risk for unknown reasons, but an analysis of the Multi-AIDS Center Cohort (MACS) study did not find an association, and this finding is echoed in a new Journal of Infectious Diseases paper looking at participants in several vaccine trials (abstract and link below). A study just published in PLoS One even finds the opposite: the inability of some STEP participants to make neutralizing antibodies against certain Ad5 components appeared to reflect generally weaker immunity and was associated with a greater risk of becoming HIV-infected independent of vaccination (abstract and link also below).
One remaining possibility that has yet to be thoroughly explored is that the STEP results related to something happening at the mucosal site of exposure, specifically the foreskin of the uncircumcised individuals that experienced the increased HIV acquisition risk. This idea is a subtle variant of the hypothesis related to Ad5-specific CD4 T cells. It is possible that Ad5 antibody titers in STEP were linked to the presence of persistent Ad5 infection, and that vaccination did not boost Ad5-specific CD4 T cell responses in blood but rather at mucosal sites of Ad5 antigen expression. One such site is known to be the gut, but because there are already huge numbers of CCR5-expressing CD4 T cells in that location it would be unlikely that a small additional influx would alter susceptibility to HIV (and this would be consistent with the risk associated with vaccination in STEP being seen in individuals reporting insertive anal intercourse but not receptive, as reported in the new JAIDS paper).
To the best of my knowledge, it is not known whether persistent Ad5 antigen expression can be detected in the foreskin, which would likely be a prerequisite for causing vaccine-boosted Ad5-specific CD4 T cell responses to traffic to—and be retained at—that site, potentially increasing susceptibilty to HIV. This type of scenario was suggested as a possible explanation of the STEP results by the research group of Steven Patterson in the discussion section of their PNAS paper in 2009, but still requires further investigation (there is one study that attempted to investigate the possibility in macaques, but because they are not susceptible to natural Ad5 the relevance of the results to humans is unclear). Of particular concern, CD4 T cells targeting adenoviruses cross-react with multiple serotypes (and simian adenoviruses), so the safety of vectors derived from variants other than Ad5 will remain uncertain until the issue is resolved.
JAIDS Journal of Acquired Immune Deficiency Syndromes:
POST ACCEPTANCE, 14 March 2012
Koblin, Beryl A. PhD; Mayer, Kenneth H. MD; Noonan, Elizabeth MS; Wang, Ching-Yun PhD; Marmor, Michael PhD; Sanchez, Jorge MD, MPH; Brown, Stephen J. MD; Robertson, Michael N. MD; Buchbinder, Susan P. MD
Background: The Step Study found that men who had sex with men (MSM) who received an adenovirus type 5 (Ad5) vector-based vaccine and were uncircumcised or had prior Ad5 immunity had a higher HIV incidence than MSM who received placebo. We investigated whether differences in HIV exposure, measured by reported sexual risk behaviors, may explain the increased risk.
Methods: Among 1,764 MSM in the trial, 724 were uncircumcised, 994 had prior Ad5 immunity and 560 were both uncircumcised and had prior Ad5 immunity. Analyses compared sexual risk behaviors and perceived treatment assignment among vaccine and placebo recipients, determined risk factors for HIV acquisition and examined the role of insertive anal intercourse in HIV risk among uncircumcised men.
Findings: Few sexual risk behaviors were significantly higher in vaccine vs. placebo recipients at baseline or during follow-up. Among uncircumcised men, vaccine recipients at baseline were more likely to report unprotected insertive anal intercourse with HIV negative partners (25.0% vs. 18.1%; p=0.03). Among uncircumcised men who had prior Ad5 immunity, vaccine recipients were more likely to report unprotected insertive anal intercourse with partners of unknown HIV status (46.0% vs. 37.5%; p=0.05). Vaccine recipients remained at higher risk of HIV infection compared to placebo recipients (HR =2.8; 95% CI:1.7, 6.8) controlling for potential confounders.
Interpretation: These analyses do not support a behavioral explanation for the increased HIV infection rates observed among uncircumcised men in the Step Study. Identifying biologic mechanisms to explain the increased risk is a priority.
J Infect Dis. (2012)
First published online: April 5, 2012
Kathryn E. Stephensona, John Huralb, Susan P. Buchbinderc, Faruk Sinangild and Dan H. Barouch*,a,e
aDivision of Vaccine Research, Beth Israel Deaconess Medical Center, Boston MA
bFred Hutchinson Cancer Research Center, Seattle, WA
cSan Francisco Department of Public Health, San Francisco, CA
dGlobal Solutions for Infectious Diseases, South San Francisco, CA
eRagon Institute of MGH, MIT, and Harvard, Boston MA
The Step study of a recombinant adenovirus serotype 5 (Ad5)-based HIV-1 vaccine revealed an increased risk of HIV-1 acquisition in vaccinees who were Ad5 seropositive at baseline. We therefore investigated whether preexisting Ad seropositivity to seven different Ad serotypes was associated with increased risk of HIV-1 infection in three HIV-1 vaccine efficacy trials. In a case-control study involving 1,570 adults enrolled in the VAX003 and VAX004 trials of a recombinant protein subunit HIV-1 vaccine and in the Step study, we observed that preexisting seropositivity to multiple Ad serotypes was not intrinsically associated with increased risk of HIV-1 acquisition.
PLoS ONE 7(4): e33969. doi:10.1371/journal.pone.0033969
Cheng Cheng1, LingShu Wang1, Jason G. D. Gall3, Martha Nason2, Richard M. Schwartz1, M. Juliana McElrath4,5, Steven C. DeRosa4,5, John Hural4,5, Lawrence Corey4,5, Susan P. Buchbinder5,6, Gary J. Nabel1
1 Vaccine Research Center, National Institute of Allergy and Infectious Diseases, National Insititutes of Health, Bethesda, Maryland, United States of America, 2 Biostatistics Research Branch, National Institute of Allergy and Infectious Disease, National Institutes of Health, Bethesda, Maryland, United States of America, 3 GenVec, Inc., Gaithersburg, Maryland, United States of America, 4 Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, Washington, United States of America, 5 HIV Vaccine Trials Network, Seattle, Washington, United States of America, 6 HIV Research Section, San Francisco Department of Public Health, San Francisco, California, United States of America
The Step trial raised the possibility that uncircumcised men with pre-existing Ad5 neutralizing antibodies carried an increased risk of HIV infection after vaccination. Thus, understanding Ad seropositivity in humans is important to the development of an AIDS vaccine. Here, we analyze the impact of different Ad5-specific neutralizing antibodies on immune function and clinical outcome.
Methods and Findings
Ad seropositivity in the Step trial volunteers was analyzed using chimeric rAd5/35 vectors to characterize their specificity for Ad5 fiber and non-fiber external (capsid) proteins. Immune responses and HIV seropositivity were correlated with the specificity of Ad5-neutralizing antibodies. Neutralizing antibodies induced by the vaccine in Ad5 seronegative subjects were directed preferentially to Ad5 capsid proteins, although some fiber-neutralizing antibodies could be detected. Pre-vaccination Ad5 serostatus did not affect the capsid-directed response after three vaccinations. In contrast, anti-fiber antibody titers were significantly higher in volunteers who were Ad5 seropositive prior to vaccination. Those Ad5 seropositive subjects who generated anti-capsid responses showed a marked reduction in vaccine-induced CD8 responses. Unexpectedly, anti-vector immunity differed qualitatively in Ad5 seropositive participants who became HIV-1 infected compared to uninfected case controls; Ad5 seropositive participants who later acquired HIV had lower neutralizing antibodies to capsid. Moreover, Ad35 seropositivity was decreased in HIV-infected subjects compared with uninfected case controls, while seroprevalence for other serotypes including Ad14, Ad28 and Ad41 was similar in both groups.
Together, these findings suggest that the case subjects were less immunologically responsive prior to infection. Subjects infected during the Step trial had qualitative differences in immunity that increased their risk of HIV-1 infection independent of vaccination.