In this randomized, double-blind, placebo-controlled clinical trial, the multiclade, multiprotein rAd5 HIV-1 vaccine VRC-HIVADV014-00-VP appeared safe and was highly immunogenic among participants seronegative for prior nAb to Ad5. This study extends the prospective safety testing of recombinant type 5 adenoviral vectors specifically to include Ad5-seronegative volunteers, and provides evidence that in this group the 1010 PU dose is of comparable immunogenicity to, and better tolerated than, the 1011 PU dose.
A single dose of the vaccine induced T-cell responses against all three HIV-1 proteins encoded by the vector. Responses to one or more antigen pools were detected in 85% to 95% of vaccine recipients. Among responders, median ELISpot response magnitudes were on the order of several hundred SFC/106 PBMC, and ICS detected CD4+ and CD8+ T-cell responses on the order of 0.1% to 1% of T cells. Sixty-five percent of vaccinees showed responses in both CD4+ and CD8+ T cell subsets at 28 days post vaccination. Responses in individual participants reached magnitudes of up to 3000 SFC/106 PBMC in the IFN-γ ELISpot assay and up to 4% of total CD8+ T cells in the ICS assay. The majority of vaccinees (~80%) had IFN-γ-secreting, HIV-1-specific T cells persisting at one year post vaccination. Responses to Pol were observed at higher frequencies than reported in previous in trials that used a DNA-prime/rAd5 boost regimen, as discussed below.
All vaccinees (and no placebo recipients) developed binding antibodies against a consensus HIV-1 gp140 oligomer at day 28. The vaccine also induced binding antibodies against vaccine antigens (but not against gp120 antigens from laboratory-adapted HIV-1 strains) in the majority of recipients, and induced neutralizing antibodies at low titers in approximately half of participants in the lower dose group.
Given the diversity of circulating HIV strains and the rapidity with which variants arise within infected individuals, a vaccine's ability to induce a broad response across viral proteins may be important to the success of a T cell-based vaccine strategy. In this study, three quarters of vaccinees developed T-cell responses to two or more HIV-1 antigens detectable by stimulation with pools of global potential T-cell epitopes. The induction of both IL-2 and IFN-γ secreting cells among the CD4+ and CD8+ T-cell populations responding to the vaccine indicates that the vaccine induces T cells representing a variety of functional characteristics, including both MHC class I-restricted cytotoxic and class II-restricted helper cells.
A Phase IIb proof-of-concept trial (the Step study) evaluating an HIV-1 gag/pol/nef
vaccine using a different adenoviral vector developed by Merck Research Laboratories was stopped early after an interim efficacy analysis found the vaccine to be ineffective, despite inducing frequent and diverse immune responses 
. Moreover, although in that study no statistically significant differences in incidence of HIV infection were observed in the vaccine vs. placebo group, based on exploratory subgroup analyses uncircumcised men with prior Ad5 nAbs have been excluded in further trials with Ad5. Of note, unlike the VRC Ad5 vaccine investigated in the current study and the canarypox vector that was used in the Thai trial, the Merck vaccine did not include HIV Env antigens. Although immune correlates of protection in the Thai trial are not yet known, that study found CD4+
T cell and antibody responses against HIV Env antigen in vaccine recipients 
. In addition, the specific deletions (E1, E3, E4) used to render the VRC adenovector replication defective may have different effects on vaccine immunogenicity compared to the E1-deleted Merck vector. The addition of the E4 deletion has been found in vitro
to result in lower levels of rAd5 expression than seen in E1,E3-deleted vectors, and an E1,E3,E4-deleted vector did not prevent vaccine “take”, as measured by increases in Ad5 neutralizing antibodies four weeks following vaccination in volunteers with prior Ad5 nAb 
The immune responses following a single injection in the current study (which was performed before the results of the Step study became available) appear to compare favorably with responses in studies of multiple dose regimens of adenovectors alone or in combination with DNA plasmids, but there are some differences. For example the Step study 
of three doses of an HIV-1 gag/pol/nef Ad5 vaccine found HIV-specific T cells recognizing one or more gene products in 86% of participants with Ad5 titers ≤200, with geometric mean ELISpot response magnitudes in the range of several hundred SFC/106
PBMC at week 8, the primary immunogenicity time point 
. Similarly, analyses of T-cell responses from clinical trial subjects who received Merck vaccine candidates that included an HIV-1 gag insert – either by DNA priming followed by rAd5 boosting, or by a homologous rAd5/rAd5 prime-boost regimen – found IFN-γ ELISpot response rates of 42% to 65% in subjects not stratified by prior Ad5 immunity, with evidence of higher response rates in those with prior Ad5 Nab titers <1
. In an international Phase IIa study of a three-injection DNA plasmid priming series followed by a single boost with VRC-HIVADV014-00-VP (HVTN study 204), T-cell responses to Env and Gag peptides were seen in approximately half of participants, twice as many as responded to Pol 
. Similarly, an East African Phase I/II study that compared a prime-boost regimen using the same DNA and rAd5 products to rAd5 alone found T-cell responses to Env to be several times more frequent than those to Pol following the prime-boost regimen, while rAd5 alone elicited more frequent responses to Pol (and less frequent responses to Gag) than did the DNA prime/rAd5 boost regimen 
. These results raise the possibility that the prominent responses to Pol seen in the current study, as well as in the much larger Step study 
, may be associated with the use of an adenoviral vector alone, and that choice of vaccine vector or prime-boost regimen, or both, may affect the relative responses to different vaccine antigens.
The high (approximately 85%) rate of vaccine-induced antibody responses detected by the Abbott HIV diagnostic assay even 12 months following vaccination, and lower (but non-zero) rates with BioRad and bioMerieux diagnostic assays illustrate the importance of appropriate screening algorithms for HIV infection in the setting of vaccine trials, as participants obtaining routine antibody screening outside of the study site are at risk of unblinding, or facing uninterpretable or false-positive results. In this study, participants were counseled throughout the trial not to obtain HIV testing outside of the clinical research units, which provided HIV testing through a central laboratory that was able to distinguish between vaccine-induced seropositivity and actual HIV infection. Although anti-Env binding and neutralizing antibodies were elicited with the VRC Ad5 vaccine, these responses were low in titer and not broadly neutralizing. They are lower than antibody responses induced by DNA priming prior to rAd5 boosting, vector prime and Env gp120 subunit protein boosts, and Env subunit boost alone 
Limitations of this study include the use of an Ad5 nAb-seronegative study population, which, although intended to provide optimal sensitivity for assessing the safety of the vaccine in a small study population, would be expected to result in more frequent or higher magnitude T-cell responses than would be anticipated in volunteers with substantial pre-existing nAb immunity to Ad5. It is therefore not possible to extrapolate these results to individuals with prior antibody immunity to Ad5. However, larger studies indicate that vector immunity attenuates, but does not eliminate, immune responses. In earlier dose-escalation clinical trials of HIV-1 clade B gag
Ad5 monovalent 
and HIV-1 gag/pol/nef
Ad5 trivalent 
vaccines developed by Merck Research Laboratories, the frequency of injection-site reactions as well as systemic adverse effects was dose dependent, and systemic effects (but not local reactions) occurred more frequently in subjects with baseline Ad5 nAb titers <1
200. The proportion of vaccine recipients mounting IFN-γ-secreting T cells recognizing HIV-1 antigens, and the magnitudes of these responses, were generally higher in those with lower baseline Ad5 nAb titers. However, the effect of prior vector immunity on vaccine response in these studies was diminished at higher doses 
, although not uniformly 
. While response rates and magnitudes in Step participants with Ad5 titers ≤1
18 appeared comparable to those with titers ≤1
200, responses in those with titers >1
200 appeared somewhat lower 
. Analysis of immune responses from a Phase IIa clinical trial of a prime-boost regimen consisting of DNA plasmids followed by VRC-HIVADV014-00-VP (HVTN study 204) found that preexisting Ad5 neutralizing antibodies blunted both CD4+
T-cell responses to Gag, but did not produce a significant effect on Env responses 
Additional limitations of this study include its modest sample size, which leaves open the possibility of infrequent vaccine-associated adverse events that might be observed in a larger study, and the enrollment of participants at only US sites, where self-reporting of safety outcomes may differ from that in other locations.
To the extent that prior Ad5 nAb immunity dampens vaccine-induced T-cell responses, such an effect might potentially be addressed by priming with other viral vectors or with DNA plasmids followed by boosting with an Ad5 vector 
, or by developing adenoviral vectors of alternative serotypes to which prior antibody immunity is less frequent. In any of these approaches, predicting the vector dose that optimizes safety, particularly in vector-naïve individuals, yet provides optimal immunogenicity will facilitate the design of clinical trials.
The current study supports the clinical tolerability of adenoviral vector vaccines in the absence of prior vector nAb immunity, and shows that a single, well tolerated 1010 PU dose of the VRC adenoviral vector can produce T-cell responses of high magnitude, breadth and durability, high-frequency CD8+ and CD4+ T-cell responses, as well as antibody responses, against HIV-1 antigens.