Therapeutic vaccines against HIV infection

Abstract

Resistance to medication, adverse effects in the medium-to-long-term and cost all place important limitations on lifelong adherence to combined antiretroviral therapy (cART). In this context, new therapeutic alternatives to cART for life in HIV-infected patients merit investigation. Some data suggest that strong T cell-mediated immunity to HIV can indeed limit virus replication and protect against CD4 depletion and disease progression. The combination of cART with immune therapy to restore and/or boost immune-specific responses to HIV has been proposed, the ultimate aim being to achieve a functional cure. In this scenario, new, induced, HIV-specific immune responses would be able to control viral replication to undetectable levels, mimicking the situation of the minority of patients who control viral replication without treatment and do not progress to AIDS. Classical approaches such as whole inactivated virus or recombinant protein initially proved useful as therapeutic vaccines. Overall, however, the ability of these early vaccines to increase HIV-specific responses was very limited and study results were discouraging, as no consistent immunogenicity was demonstrated and there was no clear impact on viral load. Recent years have seen the development of new approaches based on more innovative vectors such as DNA, recombinant virus or dendritic cells. Most clinical trials of these new vectors have demonstrated their ability to induce HIV-specific immune responses, although they show very limited efficacy in terms of controlling viral replication. However, some preliminary results suggest that dendritic cell-based vaccines are the most promising candidates. To improve the effectiveness of these vaccines, a better understanding of the mechanisms of protection, virological control and immune deterioration is required; without this knowledge, an efficacious therapeutic vaccine will remain elusive.

Keywords: :

• therapeutic vaccine
• dendritic cell
• HIV
• functional cure
• recombinant virus
• DNA

Introduction

Combined antiretroviral therapy (cART) has become one the best tools to control HIV infection at both the individual¹ and social level.^2,3 Indeed, cART has achieved impressive reductions in morbidity and mortality among patients with advanced human immunodeficiency virus (HIV).¹ However, despite these advances in infection control, the cost of medication, the mandatory high adherence to medication for life, side effects⁴ and the risk of development of resistance remain unsolved problems. Furthermore, the HIV epidemic has severely hampered economic growth in developing countries. Although the number of HIV-infected patients under treatment in these countries has increased steadily in recent years, and although the number of new HIV infections has fallen, the demand for treatment continues to rise.⁵ In fact, the number of new infections is still higher than the number of new patients on treatment. For economic, social, cultural and political reasons, widespread treatment is difficult to implement in these countries. Given this situation, there is a need for new therapeutic alternatives to “cART for life” in HIV-infected patients.⁶

Two such alternatives which have been proposed are eradication and functional cure.⁷ There is one report of an HIV-infected patient in whom viral replication remained absent, despite discontinuation of antiretroviral therapy, after transplantation with CCR5Δ32/Δ32 stem cells.⁸ Recently published follow-ups of this patient strongly suggest that cure of HIV has been achieved.⁹ Intensification of cART,¹⁰ stem cell therapy,¹¹ gene therapy¹² and the elimination of latently infected cells are other alternatives to eradication which are currently being assessed.⁷ However, it is unlikely that therapeutic strategies leading to eradication will emerge as a valid alternative to cART, and in fact, there are no biological models to support this approach. Conversely, the concept of “functional cure,” defined as control of HIV replication without cART, is based on the observation that a small proportion of HIV-infected patients (referred to as long-term non-progressors or elite controllers) are able to control HIV replication to below detectable levels, and without abnormalities in CD4 T-cell counts, for more than 25 y. It has been proposed that the strong HIV-specific immune responses observed in these patients could account for this natural “functional cure.” Specifically, control of viral replication has been associated with HLA types B*57:01, B*27:05 and B*14, and it has been proposed that the viral peptides presented by these HLA molecules induce strong HIV-specific CD8 T-cell responses which are responsible for viral control.^13,14 These findings prompted the development of new strategies based on the use of anti-HIV vaccines to induce high frequencies and breadth of response of HIV-specific CD8 cytotoxic T lymphocytes (CTLs), together with strong CD4 T-cell help, in HIV-infected patients.^15,16 Recent reports suggest that a strong T cell-mediated immunity to HIV can indeed limit virus replication and protect against CD4 depletion and disease progression.¹⁷ Here we will review the basis of the therapeutic vaccine strategy and the different candidates tested in human clinical trials.

Pathogenic Basis for a Therapeutic Vaccine Design

HIV-specific responses against HIV infection in antiretroviral naive patients

The main question to be answered here is whether the immune system can contain viral replication without cART, even if only for limited periods. One major problem is that the types of immunological responses which are able to control viral replication are not known, although it does seem that the relevance of the different response types changes with the stage of HIV infection. Barouch et al. recently published the results of a preventive vaccine which suggest that protection against infection is mediated mainly by antibodies against env, whereas post-infection control of viral replication is associated with cellular responses against gag.¹⁸ Recent data reported by Picker's group in relation to SIV-infected rhesus macaques suggest that the induction of effector memory T-cell responses (and not central memory T-cell responses) with a CMV-vectored vaccine is essential for control of viral replication.¹⁹

Anti-HIV-1 CTL responses have been detected in all studied cases of acute HIV infection, and it is believed that they reduce the peak plasma viral load (PVL) to the stabilization level (set-point) of PVL which is reached at the end of the acute phase.^20,21 It has therefore been hypothesized that CTL play an essential role in controlling viral replication during acute infection. However, it is not clear that neutralizing antibodies contribute to the control of viral replication during acute infection, since a delay in the development of these antibodies is observed which is not associated with changes in viral replication.²² Finally, and in contrast to CTL and neutralizing antibodies, a clonal depletion of HIV-specific CD4⁺ T cells occurs during acute HIV infection. It has been suggested that the lack of response of these cells is key to explaining the failure to control viral replication during this phase.^20,21 In fact, massive infection and loss of memory CD4⁺ T cells occur in multiple tissues during acute infection,²³ and HIV mainly infects HIV-specific CD4⁺ T cells.²⁴ These findings may provide a basis for explaining why a therapeutic vaccine-induced transient activation of HIV-specific CD4—but not CD8—T cells might have a detrimental effect on HIV replication,²⁵ and could offer important clues for the future design of therapeutic vaccines.

Some data suggest that strong T cell-mediated immunity to HIV can indeed limit virus replication and protect against CD4 depletion and disease progression in chronic HIV-infected patients. Although in most infected patients replication leads to the progressive destruction of the immune system and evolves inevitably toward AIDS, a small number of immunologically “privileged” individuals, or long-term non-progressors (LTNP), have a potent and sustained response of anti-HIV-1 CTL, Th cells and neutralizing HIV-1 antibodies. This is associated with control of viral replication and the presence of very low or undetectable viral concentrations in plasma in the absence of cART.²⁰ Direct data on the critical role of the CTL response in controlling viral replication have been obtained in both the infection model with macaques devoid of CD8⁺ T lymphocytes,^26,27 and in the immunodeficient murine model.²⁸ In addition, there is clear evidence from both human^17,29 and animal models³⁰ that a specific helper T response against HIV is crucial in obtaining an optimal, specific CTL response which can control viral replication. This is consistent with other reported data on chronic viral infections in murine models.³¹ Finally, studies in animal models have shown that high levels of neutralizing antibodies can block infection regardless of the route of exposure to the virus.³² It has also been reported that plasma virus continually and rapidly evolves to escape neutralization, indicating that neutralizing antibodies exert a level of selective pressure. It is probable that neutralizing antibody responses account for the extensive variation in the envelope gene observed in the early months after primary HIV infection.³³

Despite the relevance of HIV-specific immune responses in both the acute and chronic stages of infection these responses are not, in most patients, able to control viral replication to undetectable levels and prevent the infection from progressing to AIDS. Many authors hypothesize that selective escape from CD8⁺ T-cell and other immune responses is the main reason for the failure to control HIV replication.^34-42 Some studies suggest that selective escape from CD8⁺ T-cell responses represents a major driving force of HIV-1 sequence diversity. Allen et al.³⁵ assessed the relationship between viral evolution and adaptive CD8⁺ T-cell responses in four HIV-infected patients who were studied for five years after acute infection. Nearly two-thirds of the amino acid mutations identified in non-envelope antigens were attributable to CD8⁺ T-cell selective pressures. The preferential selection of individual residues for mutation and the repeated selection for identical mutations observed in this study suggest that there are significant biochemical constraints on viral evolution. In addition, some mutations induced by immune-mediated selective pressures have an impact on viral replication capacity.^39,43 These findings help to decide which epitopes should be used in a therapeutic vaccine in order to avoid or limit the viral escape to HIV CTL response.

In addition to viral escape, the lack of control of viral replication in HIV infection could be secondary to functional defects of immune responses, which may be explained in terms of alterations to the immune system. Although CD4⁺ and CD8⁺ cells capable of secreting interferon gamma (IFN-gamma) can be found in most HIV-infected patients, proliferative CD4 responses are normally absent^29,44,45 and CD8⁺ cells have defective cytolytic activity.^46,47,48 One explanation for these functional deficits of CD4 and CD8 responses could be that the antigen-presenting functions of the dendritic cells have deteriorated in these patients, which may contribute to the functional defects observed in the Th1 and CTL cellular responses.^49,50 The absence of a correct proliferation and expansion of CD4⁺ responses may, in turn, influence the lack of cytolytic activity of CD8⁺ cells.^44,51 In animal models there is a clear deficit in the secretion of cytokines by CD4⁺ cells, which starts when PVL peaks in primary infection.⁵² Lastly, the selective infection of HIV-specific CD4⁺ cells in infected patients may explain why these responses are quickly lost in HIV infection.²⁴

In summary, it seems clear that both cellular and humoral immune responses are critical for correct control of HIV infection. Viral escape and functional defects in these immune responses are likely to be the main causes of sustained viral replication in HIV infection, and constitute the two main problems that a therapeutic vaccine should solve.

Influence of cART on HIV-specific immune responses

The benefit of cART on the immune system is obtained through an increase in the absolute number of circulating naive CD4 T lymphocytes, a concomitant reduction in the number of T lymphocytes with activation markers, and restoration of the response to memory antigens.⁵³ However, despite the clinical efficacy of cART^54,55 this treatment by itself is unable to eradicate the infection, even if it were administered for more than 60 y.^56,57 This is mainly due to the fact that therapy cannot eliminate latent HIV-1 in the form of integrated proviral DNA, as well as to the existence of low levels of viral replication, which makes even cell-to-cell infection possible.^58,59 Furthermore, cART is incapable of restoring the immune-specific response to HIV,^17,45 and, in fact, it leads to a fall in the specific CTL response due to the lack of antigenic exposure.⁶⁰ Some reports have shown that the helper proliferative response to HIV p24 Ag presented by some cART patients does not reflect an improvement in the immune phenotype or function of CD4 or CD8 cells, but in fact is secondary to the small increases in viremia typically observed in patients taking cART.⁶¹ This would explain the rapid ‘rebound’ of viral load that occurs within days or weeks of suspending cART, even after several years of effective therapy.^62,63 This rebound occurs even if cART is initiated in very early-stage HIV-infected patients, in whom the immune system is theoretically still well preserved (circulating CD4⁺ T lymphocytes above 500 cells/mL; PVL: 5,000–10,000 copies/ml). Similarly, these viral dynamics occur even when immune restoration is practically complete in terms of the homeostasis of T lymphocytes and their subpopulations, as well as in terms of the capacity for response to polyclonal stimuli and memory antigens with cART.^62,64 Taken together these data suggest that the first challenge for a therapeutic vaccine is to demonstrate that it is possible to induce de novo HIV-specific responses that are able to control viral replication, albeit only partially. If this is the case, it would be a proof of concept that would justify further investigation of different immunotherapies.

Therapeutic Vaccines

Objectives of a therapeutic vaccine

The main objective of a therapeutic vaccine is to induce or increase immune responses against HIV infection using a planned exposure to HIV viral antigens. In the ideal scenario, these responses would control viral replication to an undetectable level, mimicking the situation of the minority of patients who control viral replication without treatment and do not progress to AIDS. This has been termed “functional cure” and it is observed in most other infections (e.g., herpes or tuberculosis). Previous research has suggested that partial control of viral replication could be enough to delay the initiation of cART in antiretroviral naive patients or to offer “drug holidays” in already-treated patients. However, the current knowledge of HIV pathogenesis precludes this “partial remission” of infection from being a conceivable clinical objective. Even antiretroviral naive patients with a normal CD4 T-cell count could be at higher risk of suffering non-AIDS events (cancer, cardiovascular or neurological events) than the general population.^65-67 In addition, it seems that the early initiation of antiretroviral therapy significantly improves survival, as compared with deferred therapy.⁶⁸ Furthermore, the discontinuation of antiretroviral therapy in order to reduce drug-related adverse effects and to improve the patient’s quality of life has been shown to be associated with increased morbidity and mortality.⁶⁹ It has been suggested that immune activation due to viral load rebound after ART interruption would explain this increase in the rates of AIDS-defining and non-AIDS-defining malignancies or cardiovascular events among patients who discontinued cART. Untreated HIV infection is characterized by a profound and continuous state of immune activation manifested by an increased turnover of B and T lymphocytes, natural killer cells and a high level of pro-inflammatory cytokines such as IL-7 and TNF-α.⁷⁰ Another relevant feature is the elevation of activated CD8⁺ T-cells expressing the DR⁺/CD38⁺ phenotype, which is considered as a surrogate marker of disease progression.^71,72 Systemic immune activation could lead, first, to exhaustion of the immune system, including lymph node fibrosis,^73,74 retention of effector T cells in lymph nodes,^75,76 clonal exhaustion and thymic dysfunction. In addition, it could contribute to the spreading of HIV infection by generating more targets for HIV, thereby allowing ongoing HIV replication.²⁴ In summary, partial control of viral replication remains deleterious for HIV-infected patients, and the objective of a therapeutic vaccine should be complete remission of infection. Nonetheless, if a therapeutic vaccine were shown to be capable of partially controlling viral load this would be a proof of concept justifying further investigation of new candidates.

A second objective has recently been proposed for therapeutic vaccines.⁷ Based on data related to the eradication of HIV in the “Berlin patient”^8,9 it has been suggested that the combination of cART intensification with a vaccine able to induce effective HIV-specific responses could purge the HIV reservoir. This strategy, together with others using different drugs or gene therapy, is already being assessed in clinical trials.⁷

Types of therapeutic vaccines

Most of the therapeutic vaccines tested have previously been used as preventive candidates. Classical approaches such as whole inactivated virus (REMUNE) or recombinant protein (gp120) were initially shown to be useful. In general, however, the capacity of these early vaccines to increase HIV-specific responses was very limited and study results were discouraging, as no consistent immunogenicity was demonstrated and there was no clear impact on viral load.^{77,78-81,82-89} Recent years have seen the development of new approaches based on more innovative vectors such as DNA, recombinant virus or dendritic cells. Most of the clinical trials involving these new vectors have been exploratory and with a low number of patients. In fact, in randomized studies no therapeutic vaccine has yet to demonstrate lasting and potent effectiveness in terms of controlling viral load, improving CD4 T-cell count or delaying clinical progression. These weak results have precluded further development. Although a great number of candidates have been used in clinical trials, this review will focus on those summarized in Table 1.

Table 1. Therapeutic vaccine clinical trials

References	Immunogen	Virological responses	Immunogenicity responses
Whole inactivated vaccine
87–90	Remune: inactivated whole virus.	No clinical benefits	Induction of gag-specific helper responses
Subunit vaccines
91–100	Recombinant envelope protein	No clinical benefits	Improvement in specific lymphocyte proliferation
101–103	HIV p24-like peptides (Vacc-4x)	No clinical benefits	Induction of HIV-associated specific responses in 90% of patients
104–106	Tat protein	Not assessed	Induction of Tat-specific T helper cell responses Consistent decrease in cellular and biochemical activation markers Persistent increases in regulatory T cells Stable increases in the percentage and absolute numbers of both CD4 T cells and B lymphocytes Reduced percentage of CD8 T cells and NK lymphocytes
107	A combination of Nef-Tat fusion protein and envelope glycoprotein gp120 formulated with AS02A adjuvant	Not assessed	Induction of strong gp120-specific CD4^+ T-cell responses. Induction of HIV-1-specific CD4^+ T cells and CD8^+ T-cell proliferation
Vaccines using DNA as a vector
111, 112	DNA constructs encoding the nef, rev or tat regulatory genes of HIV-1	No clinical benefits	Induction of HIV-1-specific cellular responses
113	DNA encoding env/rev and gag/pol genes	Reduction of viral ‘blips’	Poor capacity to boost virus-specific CD4 and CD8 T-cell responses
114	DNA encoding HIVA, an immunogen comprising HIV-1 clade A p24/p17 fused to a string of cytotoxic T-cell epitopes	Not assessed	Poor capacity to boost virus-specific CD4 and CD8 T-cell responses
115, 116	DermaVir, a single plasmid DNA (pDNA) immunogen expressing 15 HIV antigens formulated in a nanoparticle developed for topical application	No clinical benefits	Induction of HIV-specific immune responses
Viral vector vaccines
124	ALVAC-HIV vCP1452 [expresses HIV-1 env and gag gene products and nef and pol gene products encompassing known human HLA-A2–restricted cytotoxic T cell (CTL) epitopes from these genes] with Remune	No clinical benefits	Significantly increased HIV-1-specific CD4(+) and CD8(+) T-cell responses
77, 125	ALVAC-HIV vCP1452 and recombinant gp160	No clinical benefits	Significant increases in anti-gp120 or anti-p24 antibody titers, Transient augmentation of T-cell proliferation responses to gp160 and/or p24. Increased HIV-1-specific CD8 T cells
129	ALVAC-HIV vCP1452	No clinical benefits	Modest increases in the median lymphoproliferative response
25, 130	ALVAC vCP1452	Increased virus production following treatment interruption	Weakly immunogenic, inducing activated CD4 T-cell responses
131	ALVAC vCP1452 with Remune	Tended to delay viral load rebound Associated with a longer time to meet pre-set criteria to restart ART	Induction of HIV-specific cells, not correlated with viral load rebound
126, 132	ALVAC-HIV vCP1433 (expresses several HIV genes for gp120, gp41, 55 polyprotein, for a portion of pol encoding the protease and for genes expressing cytotoxic T lymphocyte peptides from pol and nef) with Lipo-6T vaccine (a mixture of the tetanus toxoid TT-830–843 class II-restricted universal CD4 epitope combined with some peptides of Gag, Nef and Pol) and followed by three cycles of subcutaneous interleukin-2	24% of the Vac-IL-2 group lowered their viral set-point compared with 5% of controls	Vaccine-elicited immunological responses predictive of virological control
128	ALVAC-HIV vCP1433	Prolongation of treatment discontinuation	Significantly increased HIV-p24-specific lymphoproliferative responses, HIV-gag-specific CD8 T cells boosted in 55% of patients tested
133	Fowl pox vector containing gag/pol sequences or fowl pox vector containing gag/pol sequences and a gene which encodes human interferon gamma	No clinical benefits	No differences in anti-HIV-specific immune responses between placebo and vaccine study arms
140	HIV-1 nef-expressing MVA	After interruption of cART viremia remained below the pre-cART viral load in 9/14 patients	Recognition of new CTL epitopes
141–143	MVA expressing HIV-1 gag p24, p17 and a CD8+ T cell epitope string	Not assessed	Significantly increased the frequencies of CD8^+ and CD4^+ T cells secreting IFN-gamma Enhanced proliferative responses to gag peptides
144, 145	Recombinant adenovirus serotype 5 HIV-1 gag	Reduction of 0.5 log10 of PVL after a 16-week ATI o	Induction of HIV-1 gag-specific CD4 cells correlated with control of viral replication in vivo
Dendritic cell-based vaccines
162, 165–168	Heterologous peptides or ALVAC-HIV-1 vCP1452^167	No change in median VL	Induced novel epitope-specific CD8 T cells and/or CD4 T cells
163, 164, 169–171	Autologous whole inactivated HIV or RNA encoding CD40L and autologous HIV-1 antigens^169^,^171	Median VL drop 0.5–0.7 log, maintained 48 weeks	Increased HIV-specific CD8/CD4 (ELISPOT/ICS) and correlated with VL changes

Remune

The Remune vaccine initially aroused considerable expectation. This is a vaccine of an inactivated whole virus in which the envelope protein has been removed during the process of inactivation, which is performed by synthesis and formulated with incomplete Freund adjuvant. The vaccine stems from a virus originally obtained in Zaire and contains a type-A envelope and type-G gag. It has been administered to more than 3,000 people with an antiviral-controlled virus. The results showed that it was able to induce gag-specific helper responses which were sometimes very potent, but did not suggest a capacity for immunological control of viral replication.^87-90 The best and most extensive study aimed to determine whether the addition of Remune would confer added clinical efficacy to that achievable by cART. This was a multicenter, double-blind, placebo-controlled, randomized trial conducted at 77 centers in the United States, and involved 2527 cART-treated, HIV-infected patients with a CD4 T-cell level between 300 × 10⁶/L and 549 × 10⁶/L. There were no differences in HIV progression-free survival, changes in HIV RNA or CD4 cell counts when comparing Remune and placebo recipients.⁸⁹

Subunit vaccines

Multiple strategies have been used to design therapeutic HIV vaccines based on peptides or recombinant proteins, including vaccination with a recombinant envelope protein,^91-100 HIV p24-like peptides (Vacc-4x),^101-103 a Tat protein^104-106 and a combination of Nef-Tat fusion protein and envelope glycoprotein gp120 formulated with AS02A adjuvant.¹⁰⁷ The basis for the use of these candidates is that a therapeutic HIV vaccine that induces HIV-1-specific CD4⁺ T-cell responses could improve antiviral immunity, resulting in delayed disease progression.

A number of clinical trials using recombinant gp120 or gp160 as immunogens have been performed both in combination and without antiretrovirals.^91-100 Overall, these trials have not demonstrated clinical benefit, although an improvement in specific lymphocyte proliferation is observed in some of them.

Vacc-4x consists of four synthetic peptides corresponding to conserved domains of the HIV-1 protein p24 that preferentially include HLA-A2 restricted elements. To ensure optimal exposure of the immunogen to antigen-presenting cells (APCs) the vaccine was given intradermally together with granulocyte-macrophage colony-stimulating factor (GM-CSF).¹⁰¹ Forty HIV-infected patients on cART were vaccinated with either low-dose or high-dose Vacc-4x over 26 weeks. HIV-associated specific responses were safely induced in 90% of patients in a dose-dependent manner and were influenced by the HLA haplotype.¹⁰³ Thereafter, cART was discontinued in patients for 14 weeks. Patients with the strongest delayed-type hypersensitivity (DTH) responses before treatment interruption tended to achieve lower actual HIV RNA levels.¹⁰¹ Finally, long-term HIV-specific immune responses and clinical outcomes were evaluated. Vacc-4x-induced cellular immune responses were unchanged 1.5 y after completing immunization, and 62% of patients were still off cART.¹⁰²

Based on efficacy studies in animal models¹⁰⁸ and those suggesting that the presence of an anti-Tat immune response correlates with a slower progression to disease,¹⁰⁹ some authors have proposed that a preventive and therapeutic Tat protein-based vaccine deserves to be investigated. The first randomized, double blind, placebo-controlled phase I vaccine trial based on the native Tat protein was conducted in HIV-infected asymptomatic individuals.¹⁰⁶ The vaccine was well tolerated and induced and/or maintained Tat-specific T helper cell responses in all subjects, with a wide spectrum of functional anti-Tat antibodies. In addition, 2.5 y after the last immunization Tat-specific humoral and cellular immune responses were still detectable in vaccinees.¹⁰⁵ These data provided the basis for a phase II study in 87 successfully-treated HIV-infected patients, the results of which have recently been reported.¹⁰⁴ It was confirmed that immunization with Tat was safe and immunogenic. In addition, there was a consistent decrease in cellular and biochemical activation markers and persistent increases in the number of regulatory T cells. This was associated with stable increases in the percentage and absolute numbers of both CD4 T cells and B lymphocytes, as well as with a reduction in the percentage of CD8 T cells and NK lymphocytes. The authors concluded that Tat immunization in patients on cART could help to restore immune homeostasis. The improvement in immune system function (as opposed to functional cure) is a benefit that some other investigators have claimed as an alternative primary end-point for therapeutic vaccines.^90,110 However, the clinical relevance of the recovery of these parameters is not clear. Until this issue is clarified, functional cure remains the best objective for a therapeutic vaccine.

A clinical trial of the gp120/NefTat/AS02A therapeutic vaccine in chronically HIV-1-infected volunteers on suppressive ART found the vaccine to be safe, well tolerated and able to induce strong gp120-specific CD4⁺ T-cell responses, and a higher number of vaccinees developed both HIV-1-specific CD4⁺ T-cell responses and CD8⁺ T-cell proliferation.¹⁰⁷ Therapy interruption to determine whether these immunological responses might lead to improved control of HIV-1viral load was not performed.

Vaccines using DNA as a vector

DNA vaccines are based on bacterial plasmids that express the desired antigen in situ, which leads to induction of an immune response of the Th1 type. The first trial, reported in 1998, included nine asymptomatic HIV-1-infected patients who were immunized with DNA constructs encoding the nef, rev or tat regulatory genes of HIV-1.¹¹¹ The vaccination induced HIV-1-specific cellular responses, but was not able to control viral replication or change CD4⁺ T-cell counts. These results were confirmed with the same candidate in patients on cART.¹¹² A further two candidates have been tested in HIV-infected patients on cART, the first encoding env/rev and gag/pol genes¹¹³ and the second encoding HIVA, an immunogen comprising HIV-1 clade A p24/p17 fused to a string of cytotoxic T-cell epitopes.¹¹⁴ These candidates were safe but their capacity to boost virus-specific CD4 and CD8 T-cell responses was poor. Given these results, novel means of vaccine delivery have been proposed for further improvement of DNA vaccine potency. DermaVir, a single plasmid DNA (pDNA) immunogen expressing 15 HIV antigens, has been formulated in a nanoparticle with polyethylenimine-mannose and glucose, and has been developed for topical application to deliver DNA-encoded antigens into Langerhans cells.¹¹⁵ This candidate has recently been tested in the context of structured therapy interruption in a pilot clinical trial.¹¹⁶ Although HIV DNA immunization induces broader and higher magnitudes of HIV-specific immune responses compared with structured therapy interruptions alone, this candidate did not have any effect on the dynamics of viral rebound, set-point viral load or CD4 T-cell counts.

Viral vector vaccines

Given the poor results obtained with Remune, subunit vaccines and vaccines using DNA as the vector it has been proposed that outcomes could be improved by using viral vectors as immunogens. Live recombinant vaccines are made of a live viral or bacterial vector that is engineered to carry the genes that encode HIV antigens. The antigens are presented on the surface of the cell as externalized peptides in the context of MHC class I molecules, thus priming for CD8⁺ T-cell responses. The best studied vaccine vectors in humans are the pox viruses.¹¹⁷ Vaccinia virus engineered with HIV-1 genes has been shown to induce virus-specific cellular and humoral immune responses in immunized macaques, as well as protection against simian immunodeficiency virus (SIV) infection when immunization with such constructs has been followed with boosting by recombinant proteins. However, due to the development of life-threatening disseminated vaccinia infections in immunosuppressed individuals there is a reluctance to use replication-competent vaccinia virus as a vector system in large human trials. Particular attention has, however, been focused on pox viruses with limited in vivo replicative capacity and which, therefore, are non-pathogenic in animal models and humans^118,119; examples include MVA and NYVAC, which carry deletions of the genes associated with pathogenicity, or avian pox viruses (canary pox and fowl pox). These vectors are restricted to the early stages of virus replication in human cells but initiate protein synthesis and thus elicit specific immune responses to the recombinant antigen.^118,120-123

The best studied vector has been canary pox (ALVAC).^{25,77,124-128,129-131} The international QUEST study¹²⁴ was performed on patients who started treatment during the acute phase. After a mean of two years of virological control, 79 patients were randomized to receive either immunization with ALVAC-HIV vCP1452, with ALVAC-HIV vCP1452 plus Remune, or placebo. ALVAC-HIV vCP1452 expresses HIV-1 env and gag gene products and nef and pol gene products encompassing known human HLA-A2–restricted cytotoxic T-cell (CTL) epitopes from these genes. The choice of vaccine candidates for this study was dictated by their safety records in HIV-1-infected subjects and by immunogenicity. The combination of vaccines in a single treatment arm was selected for its potential to enhance CD4 T helper cells (Remune) and CTLs (ALVAC-HIV vCP1452). After 24 weeks’ immunization, cART was interrupted. There was no difference between the groups in terms of viral rebound dynamics or viral load figures. These results confirm the findings of a previous small randomized pilot study of HIV-infected patients treated with cART during acute infection and who received ALVAC-HIV vCP1452 and recombinant gp160.^77,125

Similar results with ALVAC-HIV vCP1452 have been obtained in chronic HIV-infected patients. The ACTG 5068 included 99 subjects receiving successful cART who were randomized to undergo continued cART, structured therapy interruptions (STIs), ALVAC-HIV vCP1452 immunization, or STIs and ALVAC-HIV vCP1452 immunization. Subjects in the two STI arms had significantly enhanced immunological control of HIV replication compared with subjects in the two arms without STIs. ALVAC-HIV vCP1452 did not affect viral load measures.¹²⁹ A separate placebo-controlled study conducted in chronically-infected patients suggested that the immunogenicity conferred by the ALVAC vCP1452 candidate vaccine alone could be deleterious, since it was associated with higher virus production following treatment interruption.¹³⁰ Further research suggested that ALVAC vCP1452 was only weakly immunogenic, it failing to elicit significant HIV-specific CD8 T-cell responses and inducing activated CD4 T-cell responses that could favor greater viral rebound after treatment interruption.²⁵ These results appear to conflict with those of a clinical trial reported recently by Angel et al.¹³¹ This was a randomized, placebo-controlled, double-blind study in which effectively-treated HIV-infected individuals with high CD4 T-cell counts were randomized to receive ALVAC vCP1452 with Remune, ALVAC vCP1452 alone or placebo over 20 weeks. At week 24, participants interrupted cART. ALVAC with or without Remune did not lower the viral load set-point, although it tended to delay viral load rebound and was associated with a greater time to meet pre-set criteria to restart ART. In contrast to the findings reported by Autran et al.¹³⁰ and Papagno et al.²⁵ the induction of helper CD4 T-cell responses was not associated with a worse virological outcome. At all events, further research is required to determine whether the best strategy to induce viral control in chronic HIV-infected patients should be based on inducing robust and broad CD8 T-cell responses but only modestly activating CD4 T-cell responses.

The best results for a viral vector have been reported with ALVAC vCP1433 in combination with Lipo-6T vaccines, followed by three cycles of subcutaneous interleukin-2.^126,132 ALVAC-HIV vCP1433 expresses several HIV genes for gp120, gp41, 55 polyprotein, for a portion of pol encoding the protease and for genes expressing cytotoxic T-lymphocyte peptides from pol and nef. The Lipo-6T vaccine is a mixture of the tetanus toxoid TT-830–843 class II-restricted universal CD4 epitope combined with certain peptides of Gag, Nef and Pol. Chronic HIV-infected patients were randomized to receive either cART alone (n = 37) or four doses of ALVAC-HIV vCP1433 and Lipo-6T followed by three cycles of IL-2 (n = 34). Vaccination improved the proliferative response against p24 Ag, as well as CD8 T-cell responses. After cART interruption, 24% of the Vac-IL-2 group lowered their viral set-point, compared with 5% of controls. Vaccine-elicited immunological responses were predictive of virological control. Vaccination with the same product (ALVAC-HIV vCP1433) in a French clinical trial was safe, immunogenic and associated with prolongation of treatment discontinuation.¹²⁸

A pilot clinical trial of a therapeutic vaccination based on a fowl pox vector and involving patients with primary infection has been reported by Emery et al.¹³³ After a mean of four years of cART, 35 patients with controlled viral replication were randomized to be vaccinated with a fowl pox vector free of HIV sequences, a vector containing gag/pol sequences, or a vector containing gag/pol sequences and a gene which encodes human interferon gamma. Surprisingly, there were few differences between the groups in terms of the presence of CTL cells, as measured by ELISPOT, or in cytolytic responses after vaccination and before interruption of treatment. Treatment was not interrupted in ten patients. There were no differences in the control of viral replication between the placebo group and the group vaccinated with gag/pol. However, patients immunized with gag/pol and interferon gamma had better control of viral replication, with a mean viral load of 0.8 log₁₀ less than the other two groups, although the differences between groups were not significant. The absence of immune responses in the two vaccinated groups is disappointing, while the response in the interferon group is surprising and merits further investigation.

The modified vaccinia virus Ankara (MVA) was developed toward the end of the smallpox eradication campaign in the 1970s, the aim being to obtain a vaccine with a better safety profile than the ones in use at that time. MVA was dereived from vaccinia by over 570 passages in chicken embryo fibroblast cells (CEF).¹³⁴ The complete genomic sequence has been sequenced and has a length of 178 kb. It consists of 193 open reading frames (ORFs), corresponding to 177 genes, of which 25 are split and/or have suffered mutations resulting in truncated proteins. These numerous mutations, affecting host interactive proteins and some structural proteins, explain the attenuated phenotype of MVA.¹¹⁹ There is clinical experience with MVA as a vaccine against smallpox in over 120,000 people.^118,120 During extensive field studies, including with high-risk patients, no side effects were associated with the use of the MVA vaccine.^118,120 The combination of a very good safety profile and the ability to deliver antigens in a highly immunogenic way makes MVA suitable as a vaccine vector. It can stimulate antibody and T-cell responses even in the presence of pre-existing antibodies,¹³⁵ and it has been shown to be effective in primates and humans for several viruses, including SIV and SHIV.^121,136-139 Harrer et al.¹⁴⁰ investigated the safety and immunogenicity of an HIV-1 nef-expressing MVA in 14 cART-treated HIV-infected patients. Vaccination with MVA-nef was safe, was associated with recognition of new CTL epitopes, and after interruption of cART viremia remained below the pre-cART viral load in 9/14 patients. In another pilot study, vaccination of 16 HIV-1-infected cART-treated individuals with MVA expressing HIV-1 gag p24, p17 and a CD8⁺ T-cell epitope string was not only safe but also significantly increased the frequencies of CD8⁺ and CD4⁺ T cells secreting IFN-gamma and enhanced proliferative responses to gag peptides.^141-143

Finally, in ACTG A5197, HIV-1-infected individuals who received a recombinant adenovirus serotype 5 HIV-1 gag therapeutic vaccine had a 0.5 log₁₀ lower PVL after a 16-week ATI, compared with those who received placebo.^144,145

In summary, although most viral vector vaccines have been able to induce HIV-specific immune responses in clinical trials they have shown very limited efficacy in terms of controlling viral replication. New therapeutic strategies are therefore warranted.

Dendritic cell-based vaccines

Dendritic cells (DC) are the most potent professional antigen-processing and presenting cells (APC), with the unique ability to induce both primary and secondary immune responses of CD4⁺ and CD8⁺ T lymphocytes.^146,147 A wide variety of data in experimental in vivo and in vitro systems have shown that DC are able to engulf exogenous soluble proteins, tumor cell lysates, and inactivated virus and virus-infected apoptotic cells, as well as presenting antigens not only in the MHC-class II pathway to helper CD4⁺ T cells (Th)¹⁴⁶ but also in the MHC-class I pathway to cytotoxic CD8⁺ T lymphocytes (CTL), a phenomenon also known as “cross-presentation” or “cross-priming”^148-150.

In vitro culture of human blood monocytes (MO) with GM-CSF and IL-4 gives rise to myeloid DC.^151,152 Some data suggest that these human myeloid DC not only activate naive CD4+ T cells but also promote their differentiation into Th1 cells, preventing the differentiation into Th2.¹⁵³ In addition, antigen-pulsed DC secrete antigen-presenting particles (termed exosomes) which express functional MHC-class I and II and T-cell co-stimulatory molecules, and they are able to prime CTL in vivo and eradicate or suppress the growth of established murine tumors.¹⁵⁴ Autologous DC pulsed in vitro with a variety of inactivated pathogens have also been shown to induce a potent protective immunity in murine models of human infections.^154-158 All these data indicate that in vitro-generated DC may be the most potent cellular adjuvant for a vaccine preparation designed to induce effective Th1 and CTL responses to tumor antigens and intracellular pathogens.

It has been reported that an inactivated SIV-pulsed therapeutic dendritic-cell vaccine was able to elicit effective and lasting SIV-specific cellular and humoral immunity. This study found that after three immunizations administered at two-week intervals the SIV cellular DNA and the SIV plasma RNA levels decreased by 50-fold and 1,000-fold, respectively. This decrease in viral load was negatively correlated with SIV-specific T-cell responses. Neutralizing antibody responses also increased significantly and remained increased throughout the follow-up period.¹⁵⁹ Finally, two studies in a hu-PBL-SCID mouse model suggest that mice immunized with inactivated virus-pulsed dendritic cells elicit a potent immune response against HIV, which protects the animals from virus challenge.^160,161

There are at least ten published clinical trials of DC immunotherapy for HIV infection in humans.^{162-169,170,171} Most of them found that DC immunotherapy elicited immunological responses, even though the design of the trials and the HIV antigens used to pulse DC were very different. It should be noted, however, that virological responses were only observed in four of these trials.^{163,164,170,171}

Two of these four trials were performed in untreated patients. The first was a preliminary, non-controlled, non-randomized study that showed striking results for a MD-DC-based vaccine loaded with high doses of chemically-inactivated (with aldrithiol-2) autologous virus in untreated HIV patients.¹⁶³ Plasma viral load (PVL) decreased by 90% for at least one year in eight of 18 patients. This decrease in PVL was associated with strong and sustained HIV-1-specific cellular responses. The second trial was a double-blind, placebo-controlled study using the same schedule and immunizing the same type of patients (HIV-1-infected patients off cART).¹⁷⁰ Only a modest virological response, which was maintained for at least 48 weeks in the vaccinated group, was observed.¹⁷⁰ This change in VL in the immunized patients was small (ranging from 0.30 to 0.60 log₁₀) but clinically significant when compared with the placebo group, and it was maintained up to week 48. However, this small decrease in VL in immunized patients was inversely correlated with a weak increase in HIV-1-specific CD8⁺ T-cell responses.

The other two trials involved antiretroviral-treated patients. The first was an open, randomized (2:1) clinical trial using heat-inactivated autologous virus in cART-treated HIV-1-infected patients, and reported partial and transient control of viral replication after interruption of antiretroviral therapy.¹⁶⁴ In this study, autologous DC vaccine pulsed ex vivo with heat-inactivated autologous HIV-1 elicited only weak Th1- and HIV-1-specific CD8⁺ T-cell responses. Recently, Routy et al.,¹⁷¹ in an uncontrolled study, reported the final results of an immunotherapy regimen consisting of MD-DC electroporated with mRNA encoding autologous HIV-1 antigens (Gag, Nef, Rev, Vpr); this was administered monthly in 33 subjects in four intradermal doses in combination with cART, followed by two more doses during a 12-week cART interruption (STI). They found that eight out of 24 subjects (33%) met the criteria for the primary endpoint of three instances of viral load < 1,000 copies/mL, while 11 out of 24 subjects (46%) met the criteria for the secondary endpoint of three instances of viral load < 10,000 copies/mL. At week 12 of the STI, 16/24 subjects responded with a mean reduction in viral load of –1.2 log compared with their pre-cART viral load value. These data are impressive and are similar to those reported by Lu et al.¹⁶³ with HIV-1-infected patients who were vaccinated when treated with cART. These data need to be confirmed in a randomized, placebo-controlled study. A review of DC-based vaccine clinical trials has recently been published.¹⁷²

Conclusions

Our knowledge of the immunological control of HIV viral replication and the causes of cellular and humoral immune deterioration is limited, and we lack immunological methods able to demonstrate an efficacious immune control of HIV in vivo. The efficacy of immune therapy and therapeutic vaccines has been modest in the best of cases, although some preliminary results with dendritic cell-based vaccines seem promising. Efforts must now be redoubled in order to improve our understanding of the mechanisms of protection, virological control and immune deterioration, as without this knowledge an efficacious therapeutic vaccine remains a long way off. Given the toxicity and long-term efficacy problems associated with current drugs the search for therapeutic vaccines must be seen as a priority line of investigation.

Acknowledgments

This study was partially supported by the following grants: FIS PS09/01297, TRA-094, EC10–153, FIS PI10/02984, SAF2006–26667-E, RIS*, HIVACAT**, ORVACS***. F.G. was recipient of a research grant from IDIBAPS****, Barcelona, Spain. M.P. was supported by contract FIS 03/0072 from the Fundació Privada Clínic per a la Recerca Biomèdica, in collaboration with the Spanish Department of Health. *Red Temática Cooperativa de Grupos de Investigación en Sida del Fondo de Investigación Sanitaria (FIS). **HIVACAT: HIV development program in Catalonia. ***ORVACS: Objectif recherche vaccin sida. ****IDIBAPS: Institut d’Investigacions Biomèdiques August Pi I Sunyer.