Progress in HIV vaccine development

ABSTRACT

An HIV-1 vaccine is needed to curtail the HIV epidemic. Only one (RV144) out of the 6 HIV-1 vaccine efficacy trials performed showed efficacy. A potential mechanism of protection is the induction of functional antibodies to V1V2 region of HIV envelope. The 2 main current approaches to the generation of protective immunity are through broadly neutralizing antibodies (bnAb) and induction of functional antibodies (non-neutralizing Abs with other potential anti-viral functions). Passive immunization using bnAb has advanced into phase II clinical trials. The induction of bnAb using mimics of the natural Env trimer or B-cell lineage vaccine design is still in pre-clinical phase. An attempt at optimization of protective functional antibodies will be assessed next with the efficacy trial (HVTN702) about to start. With on-going optimization of prime/boost strategies, the development of mosaic immunogens, replication competent vectors, and emergence of new strategies designed to induce bnAb, the prospects for a preventive HIV vaccine have never been more promising.

Introduction

HIV infection is a major global health issue, affecting 36.7 million people world-wide.¹ New infections continue to occur, with 2.1 million cases in 2015.¹ The number of people living with HIV on antiretroviral therapy (ART) reached 17 million in 2015.¹ ART has dramatically reduced morbidity and mortality in individuals with HIV infection² and can also prevent HIV transmission.^3-5 However, it cannot eradicate HIV infection due to the persistence of a latent viral reservoir (mean half-life of 44 months).^6,7 Thus, the need for ART is lifelong and the cost is substantial^8,9 and may be difficult to sustain economically.¹⁰ Although ART is highly efficacious in preventing transmission in the setting of mother to child transmission,¹¹ in sexual transmission through the treatment of infected partners in serodiscordant relationships,¹² through pre-exposure,^4,13,14 or post-exposure prophylaxis¹⁵ scale-up difficulties and costs may make widespread implementation challenging. Furthermore, adherence is critical to the efficacy of biomedical preventive interventions but has been varied across study populations.¹⁶ According to Fauci and Marston, “even if HIV prevention efforts were optimally implemented to achieve a new infection rate of near zero, recidivism could threaten this success.” Thus an HIV vaccine is essential as it is a more sustainable solution.¹⁷ Modeling data suggest that a 70% efficacious vaccine introduce in 2027 with strong uptake and 5 y of protection could reduce annual new infections by 44% over the first decade and by 78% in 2070.¹⁸ Therefore, an effective universal prophylactic vaccine can potentially curtail and end the worldwide HIV pandemic.

The development of a universal effective HIV vaccine is an exceptionally difficult biomedical challenge. Firstly, no case of natural eradication of HIV infection has been identified, thus causal mechanisms of protection have not been definitively established.^19-21 Therefore, immune responses induced by HIV infection may not be effective in preventing HIV infection. Secondly, the extreme diversity of HIV is a major obstacle as strains belonging to different subtypes can differ by up to 35% in their envelope (Env) proteins.^22,23 Thus, vaccine immunogens derived from a particular clade may not be effective against other clades. To generate an efficacious global vaccine, immunogens capable of generating protective responses covering most major strains are required.

This review intends to summarize data from the 6 HIV-1 vaccine efficacy trials done to date (Table 1), to delineate potential protective responses, and to explore new vaccine candidates that are currently being developed.

Table 1.

Summary of HIV-1 vaccine efficacy trials.

Study	Vaccines	Study population	Immune response to vaccine	Efficacy	Correlates of risks	Immune pressure	References
Vax 003	AIDSVAX B/E gp120 in alum	Injecting drug users in Thailand	NAb to HIV-1MN and Non-nAb to gp120	No	None		29
Vax 004	AIDSVAX B/B gp120 in alum	MSM and high risk women in north America and the Netherlands	NAb to HIV-1MN and Non-nAb to gp120	No	Higher nAb to HIV-1MN, CD4 blocking Ab and/or ADCVI levels were inversely correlated with risk		30,31,32
STEP HVTN502	MRKAd5 clade B gag/pol/nef	MSM and high risk heterosexual men and women in the Americas and Australia	Interferon-γ-secreting T-cell responses by ELISPOT	No	Uncirmcumcised men, men with pre-existing Ad5 Ab	Yes on sieve analysis	39,40,41
Phambili HVTN503	MRKAd5 clade B gag/pol/nef	Heterosexual men and women in South Africa	Interferon-γ-secreting T-cell responses by ELISPOT	No			42
RV144	ALVAC-HIV [vCP1521] and AIDSVAX B/E gp120 in alum	Community risk men and women in Thailand	Env-Specific-CD4 T cells, lymphocyte proliferation to HIV antigens, binding Ab to gp120	31%	IgG to V1V2 was correlated inversely with risk. IgA to Env correlated directly with risk but there was no vaccine enhancement of infection. In the presence of low vaccine induced Env IgA, avidity of IgG for Env, ADCC, nAb and Env-sp CD4+ T cells were inversely correlated with the risk of infection	Yes on sieve analysis	43,48,53
HVTN505	DNA (clade B Gag, Pol, Nef and clade A, B, C Env) and rAd5 (Clade B Gag-Pol and Clade A, B, C Env)	MSM with Ad5 Ab <1:18 in the USA	HIV-specific CD4+ and CD8+ T cells and non-nAb	No			66

MSM, men who have sex with men. rAd5, recombinant adenovirus 5. NAb, neutralizing antibodies. ADCVI, antibody dependent cell-mediated viral inhibition. ADCC, antibody dependent cytotoxicity.

Lessons from efficacy trials and correlates of risk

Most of the licensed vaccines against both bacterial and viral infections, including diphtheria, tetanus, pertussis, Haemophilus Influenzae Type b, pneumococcus, hepatitis A, hepatitis B, varicella, measles, rubella, polio, and influenza, prevention of infection correlates with the induction of antibodies.^24,25 Furthermore, pilot studies of recombinant HIV-1 Env glycoprotein subunit (rgp120) vaccines conferred protection of chimpanzees from intravenous and mucosal challenge with homologous and heterologous HIV-1 strains.^26-28 Therefore, initial HIV-1 vaccine approaches (VAX003 and VAX004) focused primarily on the generation of neutralizing antibodies (nAb).

VAX003 and 004

VAX003 was a double-blind, randomized trial of AIDSVAX® B/E (a bivalent vaccine composed of rgp120 from subtype B, strain MN and subtype CRF01_AE, strain A244) in injection drug users (IDU) in Thailand.²⁹ VAX004 was a double-blind, randomized trial of AIDSVAX® B/B (a bivalent vaccine composed of subtype B rgp120 from strains MN and GNE8) conducted among men who have sex with men (MSM) and women at high risk for heterosexual transmission of HIV-1 in North America and The Netherlands.³⁰ Despite the development of anti-gp120 antibody responses, both vaccines did not demonstrate protection. Correlates of risk analysis found that higher nAb to HIV-1MN, CD4 blocking Ab and antibody-dependent, cell-mediated viral inhibition (ADCVI) were associated with reduced infection rates among vaccine recipients in VAX004.^31,32

Given the disappointing results from the VAX003 and VAX004 trials and data supporting the importance of cell mediated immunity in controlling viral replication in rhesus macaques (RM)^33-35 and human elite controllers,^36-38 attention turned to the use of T-cell vaccines to induce HIV-specific cellular immune responses.

STEP and phambili studies

The STEP study was a double-blind, randomized trial of the MRKAd5 HIV-1 gag/pol/nef sub-type B vaccine in individuals at high risk of HIV-1 acquisition in the Americas, Caribbean and Australia.³⁹ The vaccine consisted of a 1:1:1 mixture of 3 separate replication-defective adenovirus (Ad) 5 vectors, each expressing the gag gene from HIV-1 strain CAM-1, the pol gene from HIV-1 strain IIIB, and the nef gene from HIV-1 strain JR-FL. Despite eliciting IFN-γ ELISPOT responses in 75% of vaccinees, the vaccine did not prevent HIV-1 infection and had no effect on plasma viral load. Instead, it was associated with an increased incidence of HIV-1 acquisition in male vaccinees who were Ad5 seropositive pre-vaccination or were uncircumcised. Thus, the trial and was stopped after the first interim analysis.³⁹

Subsequent comparative analyses between cases with HIV-1 infection and non-cases did not reveal differences in HIV-specific immunologic responses.⁴⁰ However, vaccine-induced T-cell responses exerting selective pressure on breakthrough viruses was evident in sieve analyses.⁴¹

The Phambili study was a double-blind, randomized trial designed to evaluate the MRKAd5 HIV-1 gag/pol/nef sub-type B vaccine in individuals in South Africa where HIV clade C is predominant. This study was halted following the Step study's interim analysis and subsequent analysis also found no efficacy.⁴²

RV144

The next efficacy trial to occur was the RV144 trial, a randomized, double-blind trial that evaluated 4 priming injections of ALVAC-HIV [vCP1521], recombinant canarypox vector expressing HIV-1 Gag and Pro (subtype B LAI strain) and CRF01_AE (subtype E) HIV-1 gp120 (92TH023) linked to the transmembrane anchoring portion of gp41 (LAI) plus 2 booster injections, AIDSVAX® B/E (bivalent HIV-1 gp120 subunit vaccine containing a subtype E Env from strain A244 (CM244) and a subtype B Env from strain MN), co-formulated with alum.⁴³ The rationale for the prime boost strategy was to induce both cellular and humoral responses.^44,45 The RV144 trial was the only efficacy trial to date that demonstrated efficacy, 60% at 12 months (post hoc analysis)⁴⁶ that declined to 31% at 3.5 y (modified intention-to-treat analysis).⁴³

The finding of vaccine efficacy in RV144, despite the induction of only weakly nAb was paradigm changing.⁴⁷ The investigation of correlates of risk using specimens from week 26 (2 weeks post last vaccination) from 41 cases of vaccine recipients who acquired HIV-1 and 205 control (vaccine recipients who did not acquire HIV-1) identified that the development of non-neutralizing, binding IgG to scaffolded gp70-variable regions 1 and 2 (V1V2) of HIV-1 Env proteins was inversely correlated with infection and Env-specific IgA was directly correlated with infection. It is important to bear in mind that neither low levels of IgG to V1V2 nor high levels of Env-specific IgA in vaccinees were associated with higher rates of infection than placebo recipients. Thus, there was no vaccine-associated enhancement of infection.^48,49 Therefore, these results suggest that IgG to V1V2 may have contributed to protection against HIV-1 infection. In the setting of low vaccine induced Env IgA, avidity of IgG for Env, antibody-dependent cellular cytotoxicity (ADCC), nAb and Env-specific CD4⁺ T cells were inversely correlated with the risk of infection.⁴⁸

Subsequently a number of studies further exploring the potential mechanisms of protection of RV144 have been performed.^50-52 Sieve analysis comparing sequences of breakthrough virus isolates found that isolates from vaccine recipients were less likely to have a lysine at K169 of the Env V2 region than placebo recipients, suggesting selective effects of vaccine induced immune responses on V1V2 on breakthrough HIV-1 viruses.⁵³ Further characterization of RV144 vaccine-induced V2 Ab found that the Ab recognized both conformational (on gp70-V1V2) and linear epitopes (residues 165–178, immediately N-terminal to the putative α4β7 binding motif in the mid-loop region of V2) and were cross-reactive with the V1V2 regions from many HIV-1 subtypes (A, B, C and CRF01_AE).⁵⁴ In addition, epitope mapping of IgG to V1V2 indicated binding also involved the lysine residue at amino acid 169 in the Env V2 region. The V2 Ab isolated from RV144 vaccinees were able to mediate ADCC against RV144 breakthrough Env-target cells, and this ADCC activity was dependent on position 169 in breakthrough Envs.⁵⁵

Though AIDSVAX® B/E was part of the vaccine regimen in both RV144 and VAX003, only RV144 showed efficacy. Overall, both vaccines predominantly elicited gp120 specific IgG1Ab. However, compared with VAX003, RV144 induced more gp120-specific IgG3, which was associated with enhance Ab mediated effector functions, including ADCC and Ab-dependent cellular phagocytosis (ADCP).^56,57 In comparison to VAX003, RV144 elicited significantly higher IgG3 to V1V2, which was also correlated with lower HIV-1 infection risk.⁵⁸ Furthermore, when compared with VAX003 and VAX004, complement activation by V1V2 Ab, measured as C3d deposition on a panel of gp70-V1V2-coated beads was stronger and detected more frequently in RV144. Positive V1V2-specific complement activating Ab was also correlated with lower infection risk in RV144 vaccine recipients.⁵⁹

Therefore, though only weakly neutralizing, Ab induced by RV144 may potentially mediate protection through aggregating or immobilizing virions in mucin layers, impeding transverse through mucosal barriers,^60-62 via IgG Fcγ receptor (FcγR) dependent ADCVI including ADCC,^55-57 ADCP,^56,57 Ab-mediated release of cytokines or chemokines and complement-mediated killing.^{50-52,59,63,64} (Fig. 1). These non or weakly-neutralizing Ab with other potential anti-viral functions will subsequently be referred to as “functional Ab.”

Figure 1.

Schematic representation of potential mechanisms of protection by vaccine induced antibodies. Broadly neutralizing antibodies (BnAb) can bind to HIV virions from diverse strains, inactivating the virions. However, bnAb may not completely block mucosal portal of entry. Beyond neutralization, Abs may mediate protection through binding to cognate epitopes on HIV, immobilizing virions in mucin layers. Ab can also elicit immune responses through complement mediated lysis as well as recognition by FC receptors on effector cells (Natural Killer, NK cells and phagocytes) and mediate antibody dependent cell-mediated viral inhibition (ADCVI), including antibody dependent cytotoxicity (ADCC), Ab-dependent cellular phagocytosis (ADCP) and Ab-mediated release of cytokines or chemokines.

The association between Env-specific IgA antibodies and lack of protection was perplexing at first⁵¹ Subsequent analysis found that there was significant enrichment of higher Env IgA/IgG ratio among infected vs uninfected vaccinees. Furthermore, some Env-specific IgA from RV144 vaccinees were able to block the binding of an IgG monoclonal Ab that mediated ADCC. In addition, an Env-specific IgA monoclonal Ab isolated from an RV144 vaccinee also inhibited ADCC, measured as the ability of natural killer cells to kill HIV-1-infected CD4⁺ T cells coated with RV144-induced IgG. Thus, these data support the hypothesis that higher IgA may modulate vaccine-induced immunity by diminishing ADCC effector function.⁶⁵

HVTN 505

The last efficacy trial conducted to date is the HVTN 505 trial, a randomized, placebo-controlled trial of a prime boost, DNA/rAd5 vaccine consisting of a 6-plasmid DNA vaccine (expressing clade B Gag, Pol, and Nef, and Env proteins from clades A, B, and C) with rAd5 vector boost (expressing clade B Gag-Pol fusion protein and Env glycoproteins from clades A, B, and C).⁶⁶ The trial was halted prematurely due to lack of efficacy. The vaccine induced both cellular and humoral responses. However, these were not associated with protection.⁶⁶

In summary, none of the vaccine candidates that have completed efficacy trials to date induced strong bnAb responses. CD8⁺ T cell responses were induced in STEP, Phambili and HVTN505 studies but were not associated with protection. Only one trial, RV144 demonstrated efficacy and protection was associated with functional binding antibodies. However, efficacy was of suboptimal magnitude and was not durable.

The way forward

The current major goal in HIV vaccine design is to elicit a protective immune response mediated primarily by antibodies that are able to recognize a range of diverse strains. One approach is focused on the induction of broadly neutralizing antibodies (bnAb) with activities against major strains that are common in human transmission. Another approach is the induction of protective functional antibodies as well as T cell responses through prime boost strategies.^67-69 A number of HIV-1 vaccines at different stages of development are currently in the pipeline (Table 2).

Table 2.

HIV Vaccine Pipeline.

				Estimated Trial duration (Year 20–)
Product description	Phase	Trial identifier	Clinicaltrials.gov identifier	11	12	13	14	15	16	17	18	19	20	21
Broadly neutralizing Ab
VRC01 (passive immunization)	I	HVTN 104	NCT02165267
VRC01 (passive immunization)	II	HVTN 703/ HPTN 081	NCT02568215
VRC01 (passive immunization)	II	HVTN 704/ HPTN 085	NCT02716675
VRC01, VRC01LS (passive immunization)	I	HVTN 116	NCT02797171
VRC01LS (passive immunization)	I	VRC 606	NCT02599896
VRC01LS (passive immunization)	I	VRC 607	NCT02840474
10-1074 (passive immunization)	I	MCA-0885	NCT02511990
3BNC117 and 10-1074 (passive immunization)	I	YCO-0899	NCT02824536
rAAV1-PG9DP (Gene delivery)	I	IAVI A003/ CHOP HVDDT 001	NCT01937455
Trials building on RV144
ALVAC-HIV (vCP1521) and/or AIDSVAX B/E late  boost	II	RV 305	NCT01435135
ALVAC-HIV (vCP1521) prime, ALVAC-HIV/ AIDSVAX B/E boost	II	RV 306	NCT01931358
AIDSVAX B/E prime and boost	II	RV 328	NCT01933685
ALVAC-HIV (vCP2438) prime, ALVAC-HIV  (vCP2438)/bivalent clade C gp120/MF59®  boost	I/II	HVTN 100	NCT02404311
ALVAC-HIV (vCP2438) prime, ALVAC-HIV  (vCP2438)/bivalent clade C gp120/MF59®  boost	IIb/III	HVTN 702	NCT02968849
New Env immunogens
CN54gp140 with GLA-AF	I	X001	NCT01966900
Trimeric gp140 with/without aluminium phosphate	I	CR104488/ HIV-V-A003/ IPCAVD008	NCT02304185
Full length single chain gp120-CD4 complex vaccine	I	FLSC-001	NCT02756208
Mosaic vaccine
MVA Mosaic HIV	I	CR100965/ HIV-V-A002/ IPCAVD006	NCT02218125
Ad26 Mosaic HIV prime, Ad26 Mosaic HIV or MVA  Mosaic and/or clade C gp140/aluminum  phosphate boost	I/II	CR106152/ HIV-V-A004/ IPCAVD009	NCT02315703
Ad26 Mosaic HIV or Ad26 Mosaic4 HIV prime,  clade C gp140/aluminum phosphate and Ad26  Mosaic HIV or Ad26 Mosaic4 HIV boost	II	CR108152/ VAC89220HPX2004	NCT02788045
Ad26 Mosaic HIV with clade C gp140/aluminum  phosphate prime and boost	I	CR108068/ VAC89220HPX1002	NCT02685020
Replicating vectors
VSV-Indiana HIV gag vaccine	I	HVTN 090	NCT01438606
Ad4-mgag and Ad4-EnvC150	I		NCT01989533
Ad4-mgag and/or Ad4-EnvC150 prime,  AIDSVAX B/E/aluminum hydroxide boost	I	HVTN 110	NCT02771730
RcAd26.Mosaic1.HIV-Env	I	rcAd001/IAVI R001	NCT02366013
DNA based
VRC-HIVDNA-016-00-VP prime,  VRC-HIVADV014-00-VP boost	I	HVTN076	NCT00955006
HIV-MAG vaccine with/without IL-12 pDNA  adjuvant electroporation prime, VSV HIV  gag boost	I	HVTN 087	NCT01578889
HIV DNA (CN54ENV/ZM6GPN) prime, MVA-C/  CN54rgp140/GLA-AF adjuvant boost	I	CRO2059	NCT01922284
DNA-HIV-PT123 prime with/without NYVAC-HIV-PT1  and NYVAC-HIV-PT4 boost	I	HVTN 092	NCT01783977
Ad35-GRIN/MVA.HIVconsv with/without pSG2. HIVconsv DNA with/without electroporation	I/II	HIV-CORE 004/ IAVI N004	NCT02099994
DNA Nat-B env or DNA CON-S env or DNA  Mosaic env prime, MVA-CMDR boost	I	HVTN 106	NCT02296541
PENNVAX®-GP HIV-1 DNA vaccine with  electroporation with/without IL-12 DNA adjuvant	I	HVTN 098	NCT02431767
HIV DNA-C CN54ENV prime with and without  electroporation, CN54gp140 boost.	I	CUTHIVAC002	NCT02589795
Lipopeptides
LIPO-5 or MVA HIV-B or GTU-Multi HIV B prime  and LIPO-5 or MVA HIV-B boost	I/II	VRI01	NCT02038842

Ad, adenovirus. GLA-AF, glucopyranosyl lipid adjuvant. MVA, Modified Vaccinia Ankara. rAAV1, recombinant adeno-associated virus 1, Rc, replication competent. Data obtained from clinicaltrials.gov.

Broadly neutralizing antibodies

A detailed review of bnAb in HIV infection is beyond the scope of this review. Excellent reviews on this topic have been recently published.^70-72 BnAb, antibodies capable of neutralizing diverse circulating strains from multiple clade groups, can be present in 20–30% of individuals with HIV-1 infection.^73-77 They usually develop 2–4 y after HIV-1 infection, in the presence of continual antigen stimulation from viral replication.^75,77,78 HIV Env protein, composed of 3 gp120 and 3 gp41 monomers, is the main target for bnAb.^71,79 BnAb can be directed at the CD4-binding site on gp120, the glycopeptides on V1V2 of gp120, the glycans on V3 region, the membrane proximal external region (MPER) on gp41, the contiguous region of gp120-gp41 and the glycan shield.^71,79

Passive immunization using broadly neutralizing antibodies

The efficacy of bnAb as passive immunotherapy has been demonstrated in RM models. A single infusion of broadly neutralizing antibody (PGT121,^80,81 VRC01,^81-83 10E8⁸³ and 3BNC117⁸¹) can prevent infection from a single high-dose Simian/Human Immunodeficiency Virus (SHIV) challenge. A single infusion of VRC01, 3BNC117, and/or 10–1074 has also been shown to delay SHIV acquisition after repeated, low dose, weekly intrarectal challenges, for up to 23 challenges.⁸⁴

The site of antibody interception of SHIV and the mechanisms by which bnAb mediates protection have not been clearly elucidated. In a recent study by Liu et al., the protective efficacy of PGT121 against intravaginal challenge with SHIV-SF162P3 was first established in 12 female RM, with no detectable plasma SHIV RNA for over 6 months following challenge. In another group of 24 RM given the same bnAb and challenge, low levels of SHIV RNA were detectable in at least one tissue distal to the site of infection, in 7/8 monkeys necropsied at day 1, 3 and 7 post challenge. SHIV RNA and DNA were not detectable in all tissues sampled at the day 10 necropsy. Viral clearance was not associated with SHIV-specific T cell responses. Instead, it seemed to be associated with the activation of innate immune responses. These data demonstrated that PGT121 did not completely block the challenge virus at the mucosal portal of entry. Instead, some virions appeared to transit to distal tissues, but seemed to be vulnerable to immune-mediated elimination as they were progressively cleared over a period of ∼7 days.⁸⁵

VRC01, a bnAb that binds HIV-1 at the gp120 CD4 binding site, is currently the bnAb that has the most human passive immunotherapy data. A recent phase I study in healthy adults found that VRC01 administered intravenously or subcutaneously was safe and well tolerated.⁸⁶ The AMP (antibody mediated prevention) study, consisting of 2 phase II protocols, aims to evaluate the safety and efficacy of the VRC01 in reducing HIV-1 acquisition, among 2700 men and transgender persons who have sex with men in the US and South America (HVTN704/HPTN 085, clinicaltrials.gov NCT02716675) and in 1500 women in Sub-Saharan Africa (HVTN 703/HPTN 081, clinicaltrials.gov NCT02568215), is currently underway (Table 2).

The use of bnAb as passive immunotherapy in its current form will be challenging to implement widely, due to the production costs, the healthcare infrastructures necessary for infusions and the need for repeated administrations. Thus, new research is taking place to explore the introduction of bnAb using vectored immunoprophylaxis, where adeno-associated virus (AAV) vectors are used to deliver the genes encoding bnAb to muscle tissues, thereby enabling long-term production and systemic distribution. This technique has been shown to protect humanized mice as well as RM against high dose intravenous and repeated mucosal challenges.^87-89 The first human trial, IAVI A003 (clinicaltrials.gov NCT01937455) a phase I, randomized, blinded, dose-escalation study of rAAV vector coding for PG9 Ab in 24 healthy men is close to completion (Table 2).

Administration of Ab that are not broadly neutralizing may also mediate protection. α4β7 integrin monoclonal antibody (mAb) binds strongly to memory CD4+ T cells (α4β7hi), selectively inhibits interactions with MAdCAM-1 on endothelial cells, thereby interferes with the trafficking of lymphocytes to the gut. In RM models, administration of α4β7 integrin mAb to uninfected RM prevented or delayed infection after low dose vaginal simian immunodeficiency virus (SIV) challenge.⁹⁰ More recently, infusion of α4β7 integrin mAb, in RM SIV models, was associated with sustained control of plasma viremia even months after ART and α4β7 integrin mAb were discontinued.⁹¹ The mechanism for persistent virologic control remains to be defined, but was not associated with nAb or classical cell mediated immune responses. This major breakthrough show that antibody administration has the potential to modulate anti-HIV immune responses and has implications for both HIV-1 prevention and cure.

Eliciting broadly neutralizing antibodies through immunization

The promising results from bnAb as passive immunotherapy in animal models have spurred great interest in the design of HIV-1 vaccines capable of inducing bnAb.^70,92 Though a number of bnAb and their target epitopes have already been identified (discussed above), generating bnAb against HIV-1 is no simple endeavor as the mechanisms for the development of bnAb have not been clearly elucidated.⁹² An immunogen that can elicit bnAb responses has still not been identified and the high levels of somatic mutations in bnAb suggest complex maturation pathways.^93-95 The SOSIP gp140 trimer is a mimic of the natural Env trimer, where the gp120-gp41 interactions are stabilized by an intermolecular disulfide bond, and the gp41-gp41 interactions are stabilized by an isoleucine-to-proline substitution at position 559 in the N-terminal heptad repeat region of gp41.⁹⁶ It can be further modified to express epitopes of bnAb and to react with bnAb.⁹⁷ Immunization with SOSIP trimers induced nAb in rabbits and to a lesser extent in RM but bnAb responses were not generated.⁹⁸ The next step that has been proposed is the use of a longitudinal series of SOSIP trimers from multiple variants to induce bnAb.^70,98

Recently, computational redesign in the HR1 (a region that undergoes drastic conformational change during viral fusion with host cells) of SOSIP trimers has led to improved yield, purity and stability of trimers that closely mimic the native, pre-fusion trimer.⁹⁹ Furthermore, these trimers, presented on nanoparticles were able to bind bnAb and trigger B cells carrying cognate VRC01 B cell receptors.¹⁰⁰ These gp140-nanoparticles can be manufactured in CHO cell systems thus scalability is possible.¹⁰⁰

An alternative pathway to elicit bnAb is B-cell lineage vaccine design.¹⁰¹ Unmutated common ancestors, the putative naïve B cell receptors of bnAb could be targeted with relevant priming Env immunogens to trigger affinity maturation, followed by boosting to develop breadth. A germline-targeting, gp120 outer domain immunogen (OD-GT8) on self-assembling nanoparticles, was able to activate VRC01-class precursors, select productive mutations, create a pool of memory phenotype B cells and induced antibodies that showed characteristics of the VRC01-class bnAb, in a germline reverted VRC01 H-chain knock-in mouse model (VRCO1 gH)¹⁰²; as well as isolate VRC01-class precursor naïve B cells from HIV-uninfected donors.¹⁰³ Furthermore, administration of boosting immunogens (BG505 core-GT3 nanoparticle (NP) and BG505 SOSIP-GT3 trimer) drove the maturation of eOD-GT8 60-mer primed B cells toward VRC01-class bnAb and induced broad neutralization of near-native (N276A) viruses and weak neutralization of a fully native virus in VRC01 gH mice.¹⁰⁴ These data showed that B-cell lineage vaccine design has the potential to induce bnAb.

Eliciting bnAb through vaccination is still at the very early stages of development and is unlikely to progress to efficacy trial in the near future. On the other hand, more evidence and experience exist for the induction of functional antibodies and T cell responses through prime boost strategies. New data and insights are emerging from studies that build on the success and refinement RV144, on the use of mosaic antigens and the development of new vectors.

Studies building on RV144

The Pox-Protein Public-Private Partnership (P5) is a diverse group of organizations committed to building on the success of RV144, with the goal to deepen the understanding of immune responses associated with HIV prevention and to produce a licensable HIV-1 vaccine, with efficacy of 50% for 3 years, that can result in significant impact in Southern Africa. Two clinical trials, RV305 (clinicaltrials.gov NCT01435135) and RV306 (clinicaltrials.gov NCT01931358) were designed to evaluate whether additional boosting with components of RV144 can increase the durability of immune responses induced by RV 144 (Table 2).

In RV305, HIV-uninfected RV144 volunteers (n = 162) who completed all vaccinations 6–8 y earlier, were randomized to receive 2 injections of ALVAC-HIV or AIDSVAX® B/E or both, or placebo, at day 0 and 6 months. The original RV305 study activities are completed, and analyses are well underway. Weak residual HIV-specific ab responses from RV144 were present at the baseline, and rose dramatically after additional boosting with AIDSVAX® B/E. Responses declined 6 months post first boost, but increased again, albeit to lower levels following the second boost. IgG responses against gp120 and scaffolded gp70-V1V2 at weeks 2 and 26 in the groups receiving ALVAC-HIV/AIDSVAX® B/E or AIDSVAX® B/E alone were significantly higher than levels at peak immunogenicity in RV144. There were no significant differences between groups receiving ALVAC-HIV/AIDSVAX® B/E vs AIDSVAX® B/E alone. ALVAC-HIV alone did not induce significant titers against any of the capture antigens. These data suggest that late boosting with AIDSVAX® B/E with or without ALVAC-HIV may overcome suboptimal efficacy induced by the RV144 regimen. Unfortunately, RV305 vaccinations did not improve the durability of Ab responses when compared with RV144.¹⁰⁵ IgG responses to HIV-1 gp120 and gp70V1V2 scaffolds were also induced in cervico-vaginal mucous and seminal plasma after boosting with AIDSVAX® B/E with/without ALVAC-HIV. IgG responses were not detectable in anogenital secretions from the group receiving ALVAC-HIV alone and were also not detectable in rectal secretion from any groups. IgA responses were absent in anogenital secretions from all groups.¹⁰⁶

RV306 recapitulated the original RV144 regimen in 360 vaccine naïve participants with an additional late boost of AIDSVAX® B/E at week 48 or ALVAC-HIV/AIDSVAX® B/E at week 48, or 60 or 72. The aim was to provide more in depth characterization of vaccine induced innate, cell-mediated and humoral immune responses in the systemic as well as mucosal compartments. Clinical activities are completed, and analyses have begun. Preliminary results showed that IgG to gp120 and gp70V1V2 scaffolds declined significantly at week 50 if no additional boost was given. Additional boost at week 48 maintained IgG to cognate gp120 proteins at levels similar to the RV144 series at week 26, and significantly increased IgG to gp70V1V2 scaffolds. Responses were similar between groups receiving AIDSVAX® B/E, with/without ALVAC-HIV.¹⁰⁷ Tier 1 nAb responses were also improved in all groups that received additional boost and was higher with boost at week 60 and 72 when compared with boost at week 48.¹⁰⁸ Furthermore, IgG to gp70V1V2 CRF01_AE scaffolds and IgG to HIV-1 gp120 was detected in all anogenital secretions at peak immunogenicity (week 26), with the highest IgG levels to HIV-1 gp120 in cervico-vaginal mucus followed by seminal plasma and rectal secretions.¹⁰⁹

Therefore, preliminary data from RV305 and RV306 suggest that additional boosting could potentially improve efficacy of RV144. Furthermore, vaccine-induced Ab may play a role in mediating protection in anogenital mucosa, the sites of initial viral entry (Fig. 1).

HVTN 100 (clinicaltrials.gov NCT02404311, currently ongoing) is a phase I/II randomized, double blind, placebo-controlled trial using a prime-boost regimen that is a variant of RV144, based on clade C HIV-1 with ALVAC-HIV [vCP2438] prime (expressing clade C gp120 and clade B gp41, gag and protease) and bivalent subtype C gp120 boost, with MF59® (water-in-oil emulsion) instead of alum as the adjuvant, in HIV-seronegative low risk adults in South Africa. This vaccine was designed to increase subtype C coverage, improve antibody durability and to investigate whether IgG to V1V2 is a correlate of protection. Preliminary data presented at AIDS 2016 showed that the vaccine has passed all predetermined criteria for moving on to phase III efficacy study (HVTN 702), namely i) prevalence of IgG binding antibodies to at least 2 of 3 gp120 vaccine antigens ≥ 75%; ii) non-inferior IgG binding antibody magnitude to gp120 vaccine antigens when compared with RV144; iii) non-inferior response rate of Env-specific CD4+ T cells expressing IL-2, IFN-Υ or CD40L when compared with RV144; and iv) prevalence of IgG binding antibodies to at least 1 clade C V1V2 Env ≥ 56%.¹¹⁰ HVTN 702 (NCT02968849), aiming to assess efficacy and safety of ALVAC [vCP2438] prime, gp120 with MF59® boost in 5400 participants in South Africa is currently enrolling (Table 2).

Mosaic vaccine

All the HIV-1 vaccines that have progressed to efficacy trials to date have predominantly been regional and clade-specific. The goal of mosaic HIV-1 vaccine is to generate immune responses that cover the diverse spectrum of circulating HIV-1 isolates, potentially resulting in a single vaccine that can be rolled out globally. Polyvalent mosaic immunogens are generated from natural sequences via computational optimization so that they resemble natural proteins but systematically include common potential epitopes, providing diversity coverage comparable to that afforded by thousands of separate peptides.¹¹¹

Mosaic HIV-1 antigens delivered by replication-incompetent Ad26 vectors¹¹² or DNA prime-recombinant vaccinia boost regimens¹¹³ have been shown to augment both the breadth and depth of antigen-specific T cell responses when compared with consensus or natural sequence HIV-1 antigens in RM. In RM, Ad/MVA (modified vaccinia Ankara) or Ad/Ad prime boost vectors expressing bivalent HIV-1 mosaic Env/Gag/Pol elicited neutralizing, and functional non-neutralizing Ab, robust T cell immune response and afforded significant reduction in the per-exposure acquisition risk following repetitive, intrarectal SHIV challenges.¹¹⁴ Furthermore, Ad26 mosaic prime and gp140 boost was associated with 40% protection after 6 intrarectal challenges in RM.¹¹⁵ HIV-V-A002 (NCT02218125), a phase I study of MVA mosaic vaccine in 25 participants in the USA has recently been completed. HIV-V-A004 (NCT02315703), an international phase I study using Ad26 mosaic vectors prime, followed by boosting with Ad26 or MVA mosaic vectors and/or high, low or no clade C gp140 in 394 volunteers is at the final data collection stage (Table 2). Results from these studies will inform the optimization of mosaic vaccine regimens that may advance into efficacy trials in 2017.

New vectors

Vectors are critical elements of a successful vaccine. They deliver HIV-1 antigens into host cells and stimulate immune responses.^116,117 A number of new vectors, including Ad26, Ad35, replication competent Ad4 and CMV vectors are currently being developed.

Results from the 3 efficacy trials using of replication-defective Ad5 vectors have been disappointing.^39,42,66 Furthermore, Ad5 seropositivity was associated with transient increased incidence of HIV-1 infection in male vaccinees in the STEP study.^39,118 Though the mechanism for this has not been clearly elucidated, the hypothesis that expansion of Ad5-specific CD4⁺ T-cells in mucosal tissues with Ad5 vaccination, leading to increased target cell availability and thus increased risk of infection has been proposed.¹¹⁹ Therefore, concern regarding preexisting vector immunity prompted the development of other Ad vectors with lower seroprevalence, including Ad26 and Ad35.^120,121 Ad26 and Ad35 studies have shown potential in phase I/II human vaccine trials.^122-125 A recent randomized, double-blind, placebo-controlled trial in 218 healthy adults using Ad26.EnvA.01 and Ad35.Env vectors in both homologous and heterologous combinations elicited humoral and cellular immune responses in nearly all participants. Furthermore, preexisting Ad26 or Ad35 neutralizing antibody had no effect on vaccine safety and no significant impact on immunogenicity.¹²⁵

Replication competent, enteric-coated, Ad4 has been used to prevent acute respiratory disease from adenoviruses by the US military.¹²⁶ Two phase I trials of replication competent, rAd4 HIV-1 vaccine, NCT01989533 (Ad4 Mosaic Gag/Ad4 clade C Env150 via intranasal or oral routes) and NCT02771730 (Ad4 Mosaic Gag or Ad4 clade C Env150 or both prime, AIDSVAX® B/E/aluminum hydroxide boost) are currently underway (Table 2).

Replicating rhesus (Rh) CMV vectors have shown promise in RM models. Vaccination of RM with RhCMV/SIV vectors led to the establishment of indefinitely persistent, high-frequency, SIV-specific effector memory T-cell responses at potential sites of SIV replication and was associated with durable aviremic control of SIV infection in 50% of RM,¹²⁷ and eventual absence of detectable plasma- or tissue-associated virus in some RM.¹²⁸ Currently, no human data are available on the safety and efficacy of CMV vector vaccines.

In addition to technical challenges, another major obstacle to HIV vaccine product development is the requirement for increasingly large numbers of participants and resources needed for efficacy assessment. HIV prevention efficacy trials can be conducted efficiently only in large target populations with HIV incidence ≥ 2 per 100-person years, as low event rates would necessitate very large studies and incur huge costs to achieve statistical power.¹²⁹ Implementation of non-vaccine prevention modalities (NVPM) will lead to very welcome declining HIV incidence in many populations. Lower incidence complicates preventive vaccine trial design by either driving enrollment requirements up substantially, or forcing identification of populations or sub-populations where NVPM implementation is not lowering incidence for some reason beyond the control of the study team. Use of NVPM by vaccine trial participants will also need to be addressed in future vaccine efficacy trials including powering trials to account for background NVPM use and inclusion of NVPM as part of the study intervention where appropriate.¹³⁰

Conclusion

A great deal has been learned from HIV vaccine research over the last 30 years. Data from completed HIV-1 vaccine efficacy trials support the role of functional Env Ab in reducing infection risk.⁵² Replication in a follow-up efficacy trial is pending. The field eagerly awaits results from HVTN702, which will provide an opportunity to assess this pox-protein regimen in light of RV144 correlates of risk as well as generate novel hypotheses on mechanisms of protection. Data from phase II studies of passive immunization with bnAb will become available in the next few years and will be critical for proof of concept that bnAb can prevent HIV acquisition in humans. The elicitation of bnAb through vaccination is still in preclinical stages. The recent improvements in the design and generation native ENV trimer will likely aid advancement. Polyvalent mosaic antigens expressed on viral vectors and replicating viral vectors have shown great promise in RM models, are currently under investigation in phase I/II studies and have high potential to expand the HIV vaccine pipeline. The knowledge that has been gained, the results that show partial efficacy, and the robust pipeline give us hope. However, the immense complexity and multitude of unanswered questions remain enormous challenges that have to be overcome.

Disclaimer

The opinions expressed herein are those of the authors and do not represent the official position of the U.S. Army, or The Department of Defense. Trade names are used for identification purposes only and do not imply endorsement.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgment

We would like to thank Dr Sandhya Vasan for her thoughtful suggestions on the design of Fig. 1.

Funding

This work was supported by the U.S. Army Medical Research and Materiel Command (Military Infectious Diseases Research Program) through Cooperative Agreements (W81XWH-11–2–0174) with the Henry M. Jackson Foundation for the Advancement of Military Medicine.