Recent Developments in Tuberculosis Vaccines

Dessislava Marinova; Jesus Gonzalo-Asensio; Nacho Aguilo; Carlos Martin


Expert Rev Vaccines. 2013;12(12):1431-1448. 

In This Article

Recent Developments in Clinical Trials in the Current Global Tuberculosis Vaccine Portfolio

Rational Approach to Selection & Development of TB Vaccine Candidates

Given the pipeline of TB vaccines in the three different stages of development (discovery, preclinical and clinical studies), a rational process for selection of TB vaccine candidates has been established in a joint effort between the two main initiatives in TB vaccine research in the world: the European TuBerculosis Vaccine Initiative (TBVI) and Aeras TB Vaccine Foundation, originated in the USA.[59,60] The development process of vaccine candidate is based on a stage-gating approach, which follows a linear pathway to determine if there is sufficient and robust evidence to allow further development of a specific product.[59] There are six stages that each vaccine candidate needs to go through to reach marketing authorization and these include: discovery, preclinical development, Phase I and Phase II/IIb for proof-of concept, Phase III clinical trials, marketing application and market (licensure). Decision to proceed to the next stage occurs at a gate between each stage, using multiple criteria in the selection process. The evaluation criteria for advancing past each gate are defined according to the candidate's characteristics, which divide into product quality, safety, immunogenicity, efficacy, clinical characteristics, production process, regulatory strategies, business and health impact/novelty/other. The general criteria for product quality are purity, stability and potency of the vaccine. The safety, immunogenicity and efficacy criteria require data in relevant animal models as well as in clinical trials depending on the stage-gate. Clinical characteristics criteria require definition of the target product profile (TPP), which should be integrated into the preclinical to clinical development planning strategy at a very early stage. To this aim, the production process needs to be established in compliance with current good manufacturing practices (cGMP). Regulatory strategy criteria require early contacts with relevant regulators in the country of cGMP production and clinical trials conduct. In the face of multiple similar products a matrix pathway is used where head-to-head comparison within agreed animal model systems can help decision-making as can robust critical assessment of manufacturing, regulatory issues and intended TPP. As explained in the blue print, the criteria for the first three gates are well established (up to entry into early Phase II trials), but product-specific specifications for gates 4 through 6 are still under development at this time. TBVI and Aeras are currently working with vaccine developers and expert advisors on defining the rest of the stage-gating criteria crucial for down-selection of candidates as they progress to advanced clinical development (large-scale trials and manufacture), where resource availability and clinical trial sites become more limited. Clinical experience with the current vaccines would help in the definition of such down-selection criteria.

The current global TB vaccine portfolio consists of three main types of vaccine strategies, which are either preventive or therapeutic.[23,61] The preventive strategies embrace the priming BCG-replacement vaccines and subunit BCG boosts (or enhancers). Therapeutic candidates that have reached clinical development to date comprise inactivated forms of mycobacteria. In continuation, the most recent developments in TB vaccines are discussed, focusing on the candidates in clinical development.

BCG-replacement Strategies

BCG-replacement strategies divide into two classes of live vaccines, namely recombinant BCG (rBCG) and live-attenuated MTB.[23] The rBCG candidates are designed to improve efficacy of BCG by insertion of other genes. Rationally attenuated MTB from human origin are considered a classical Pasteurian approach to human vaccinology, expected to mimic natural infection without causing disease. As presented in the TBVI Annual Report 2012,[201] today there are only two priming vaccines in clinical development, rBCG VPM1002 and the live-attenuated MTBVAC (Table 1).

VPM1002 (BCGΔureC::hly HmR) is an rBCG Prague with deleted urease C gene to allow optimal pH for the expression of biologically active lysteriolysin O (LLO) from Listeria monocytogenes and containing hygromycin resistance (HmR) gene.[62] LLO is expected to perforate the phagosome to allow escape of BCG into the cytosol potentially increasing the amount of bacterial-derived antigen available for presentation to CD8+ T cells through the cytosolic scavenger pathways. VPM1002 has successfully been evaluated in two Phase I clinical trials for safety and immunogenicity in healthy BCG-naïve and BCG-immune adults in Germany[63] and in South Africa.[202] The trial results showed that VMP1002 is safe and well tolerated and supported entry in a Phase IIa trial in healthy newborns in South Africa in 2012.[203] Absence of severe adverse events after VPM1002 vaccination is of particular note since in the Phase I clinical trial of the rBCG Aeras-422, some study participants had reactivation of shingles.[64] In addition to some MTB antigens, including Ag85A, Ag85B and Rv4307, Aeras-422 expresses the perfringolysin O, which possesses similar membrane-perforating activity but is devoid of the inbuilt safety mechanisms characteristic of LLO.[65] It is worth commenting that the preclinical development plan (and the regulatory opinion) has been established with VPM1002 containing the Hyg antibiotic marker. For advanced clinical development, the current state of play in cGMP manufacture of markerless strain is important and bridging studies with the final vaccine construct without the antibiotic marker will have to be considered.

MTBVAC is live rationally attenuated MTB derived from the clinical isolate MT103 belonging to the widespread Euro-American Lineage. MTBVAC contains two independent stable deletion mutations in the virulence genes phoP and fadD26 without antibiotic resistance markers[66]. PhoP controls 2% of the genome of MTB including production of immunomodulatory cell-wall lipids and secretion of ESAT-6[67] so that phoP-mutants can produce ESAT-6 but are unable to export it.[68] Deletion of fadD26 leads to complete abolishment of synthesis of the virulence lipids phthiocerol dimycocerosates (DIM) part of the lipid capsule. Rigorous preclinical characterization studies provide strong evidence that MTBVAC is at least as safe as, immunogenic and effective against MTB challenge as compared with BCG. In October 2012, MTBVAC received approval to enter first-in-human dose-escalation Phase I clinical trial for safety and immunogenicity in healthy adult volunteers in Lausanne, Switzerland.[204] Three doses of MTBVAC are currently tested compared with standard human dose of BCG.

BCG Boosting Strategies in Clinical Trials

The global pipeline currently contains 9 subunit TB vaccine candidates in clinical trials (Table 1), which are mainly aimed for use as heterologous boosts in BCG-primed individuals, where BCG is given at birth and then boosting is applied with antigens specific for MTB (but which can also be expressed in BCG). Heterologous prime-boost immunization is considered to be highly effective for enhancing humoral and cellular immune responses[69] and reviewed by Dalmia and Ramsay.[70] There are also boosting strategies in the vaccine portfolio (as explained below), which are heterologous in terms of platform or delivery technology, but homologous in terms of antigen set. Subunit vaccines are based on one or a few MTB-specific protein antigens using viral vectors or adjuvants as delivery system.[23,64] At present, only a small number of well-known MTB-specific protein antigens (some of which expressed in BCG) are commonly employed in the construction of different TB vaccine candidates (subunits or rBCG). These antigens include the mycolyltransferase antigens 85A (Ag85A) and Ag85B, ESAT-6 or the low molecular weight protein antigen 7 (10 kDa) EsxH (TB10.4), which is used instead of ESAT-6 in some candidates to avoid cross-reactivity with the ESAT-6-based interferon gamma release assays (IGRA) (QFT-IT) used for clinical diagnosis of TB infection. Other proteins used in specific candidates include immunodominant Mtb32A (Rv1196) and Mtb39A (Rv0125) in M72,[71] the latency-associated antigen Rv2660c in H56 fusion protein[72,73] or Rv1813, Rv2608, Rv3619 and Rv3620 in ID93.[74]

The delivery systems employed in current subunit vaccines comprise replication-deficient viral vectors or adjuvants.[6,23] The viral vectors include modified vaccinia Ankara (MVA), or adenovirus subtypes of human origin (AdHu). More recently, a modified simian adenoviral vector (chAdOx1), selected for low seroprevalence[75] and the Fowlpox virus, suggested to improve boosting of antigen-specific CD8+ T cells by malaria vaccine trials,[76] have been employed in some TB vaccine subunit regimens as described in continuation. Adenovirus vectors are known for their natural tropism for the respiratory epithelium with ability to induce potent Th1-type responses and in recent works shown to have ability to differentially imprint innate cells at the mucosal site of immunization.[77,78] The main advantage of MVA vector is its satisfactory safety profile in HIV-infected individuals and its ability to induce polyfunctional, durable CD4 responses.[79,80] Adjuvant systems used for optimal delivery of subunit TB vaccines are mainly proprietary to the candidate and are not yet licensed. IC31® developed by Crucell has been used with several TB subunit candidates in clinical trials (Table 1). This adjuvant is a combination of a single-stranded oligodeoxynucleotide and an immunopotentiating peptide and serves as a depot formulation for slow antigen release.[81]

Boosts Delivered by Viral Vectors

MVA85A employs MVA to deliver Ag85A. To date, MVA85A has been evaluated by different routes of administration (e.g., intradermal, intramuscular) in an array of more than 20 different Phase I and Phase II clinical trials for safety and immunogenicity in different target age group populations ranging from healthy adults, adolescents, children and infants to HIV-positive adults with or without latent MTB infection in the UK and in high-burden African countries.[79,80,82–91] All these trials show that MVA85A is safe and immunogenic with ability to induce robust, antigen-specific polyfunctional and durable CD4+ T-cell responses thought to be important for protection. In line with these results, the South African Tuberculosis Vaccine Initiative (SATVI) recently evaluated intradermal MVA85A in a Phase IIb trial for safety and efficacy in healthy BCG-vaccinated infants (6 months old) in Worcester.[92] MVA85A was safe and well tolerated but there was no further improvement of BCG efficacy following intradermal MVA85A vaccination. After 2 years of follow-up, 39 (2.8%) of 1395 infants in the placebo group and 32 (2.3%) of 1399 infants in the MVA85A group satisfied the primary definition of active TB disease. MVA85A is currently ongoing a protective efficacy study against TB disease in HIV-infected adults[205] as well as a Phase IIa safety and immunogenicity in infants born to HIV-positive mothers where MVA85A is given at birth, and not as a booster.[93] More recently, MVA85A entered safety evaluation in combination with other viral-vectored vaccines in clinical development,[206] as well as with the new chimp adenovirus ChAdOx1 85A[207] or the Fowlpox virus FP85[76,208] both expressing Ag85A as in MVA85A with the idea to optimize immunogenicity and induce protection.[94]

Ad5Ag85A is a recombinant human-type AdHu5 used to express Ag85A as in MVA85A.[6,95] Although, AdHu5 is considered one of the most robust gene transfer vehicles for in vivo T-cell activation, underlying seroprevalence of AdHu5 humoral immunity in humans may pose risk for interfering with TB vaccines of this kind (reviewed by Rowland and McShane.[23] The results of the recently completed Phase I study for safety and immunogenicity of a single intramuscular dose of Ad5Ag85A in BCG-immunized and in BCG-naïve healthy adult volunteers in Canada were recently presented at the Tuberculosis Vaccines 3rd Global Forum in Cape Town, South Africa (25–27 March 2013).[209,210] The vaccine was safe, well tolerated and immunogenic with more potent T-cell immunity induced in BCG-vaccinated subjects. Pre-immunization anti-AdHu5 humoral immunity was found in most subjects but did not significantly dampen the potency of the candidate. Data support further clinical development of this candidate.

Ad35/AERAS-402 uses human Ad35, which was selected for low-level seroprevalence[96] to express a fusion protein of the three common antigens Ag85A, Ag85B and TB10.4.[97] To date, the candidate is being developed for use in BCG-vaccinated individuals by the intramuscular route. In healthy BCG-vaccinated adults in South Africa and in the USA, Ad35/Aeras 402 was well tolerated, safe and induced Ag85A/Ag85B and TB10.4-specific immune responses.[98] The candidate has completed a Phase IIa trial for safety and immunogenicity in HIV-infected, BCG-vaccinated adults in South Africa,[211] and is currently in a multi-center (Kenya, South Africa and Mozambique) Phase II dose-finding study for safety and immunogenicity in healthy BCG-vaccinated HIV-negative infants.[212] At the Tuberculosis Vaccines Third Global Forum 2013, data from South Africa were presented showing the candidate is safe and immunogenic and double administration of the highest dose may be the optimal vaccination strategy, as it induced persistent specific polyfunctional CD4+ T cells thought to be important in TB immunity.[213]

Boosts Delivered by Adjuvants

M72/AS01E is a recombinant fusion protein of the immunostimulatory MTB-specific antigens Mtb32A (RV1196) and Mtb39A (Rv0125) in combination with adjuvant system AS01E.[71] AS01E is composed of two immunostimulants, 3-deacylated monophosphoryl lipid A (MPL) and QS-21, a detergent purified from the bark of Quillaja saaponaria, and a liposomal preparation. As presented at the Tuberculosis Vaccines Third Global Forum, in several Phase I and Phase II trials for safety and immunogenicity in PPD-negative, PPD-positive and HIV-positive adults in different TB-endemic countries, intramuscular M72/AS01 has demonstrated safety and tolerability.[214] In both TB-infected and -uninfected adults, the vaccine induces strong and persistent M72-specific CD4+ T-cell and antibody responses.[71,99] M72/AS01 is currently enrolled in the largest multi-country, Phase IIb clinical trial for safety and efficacy, designed to administer two doses of M72/AS01E or placebo to HIV-negative adults over a 3-year follow-up period.[214] The study plans to enroll 7000 participants powered at 80% for a true vaccine efficacy of 70%, thus requiring 28 pulmonary TB events as confirmed by GeneXpert MTB/RIF or TB culture. India and sub-Saharan countries were selected to provide geographical diversity, trial capacity and TB incidence rate of 3/1000 persons per year in their HIV-negative populations.

Hybrid 1(H1) is a recombinant fusion protein of Ag85B and ESAT-6 initially designed to prevent acute TB disease[72] and developed to prevent reactivation of existing latent infection in adolescents or adults as well.[215] This candidate is currently being evaluated in early clinical trials formulated in two different adjuvants, IC31 (Intercell) or CAF01.[214]

Intramuscular H1-IC31 has successfully completed several Phase I trials in BCG-naïve and BCG-immune adults with latent TB infection (LTBI).[215] Despite initial concerns with the transient false-positive responses developed to ESAT-6/CFP-10 QuantiFERON®-TB Gold (Cellestis) diagnostic test, development of the candidate has not halted and the vaccine is currently in Phase II clinical trials including HIV-infected individuals.[215]

H1-CAF01 was recently evaluated in a Phase I trial to determine the safety of the CAF01 adjuvant in healthy adult volunteers, comparing Hybrid 1 alone with Hybrid 1 with three escalating CAF01 dose levels.[216] CAF01 consists of the immune-stimulating synthetic cord factor from MTB glycolipid trehalose-dibehenate (TDB), as TLR-independent immunomodulator, incorporated into cationic surfactant dimethyldioctadecylammonium bromide (DDA) liposomes offering a depot formulation for slow release.[100]

H56-IC31: H56 is another recombinant protein, which in addition to early-secreted Ag85B and ESAT-6 (comprising H1) also contains the latency-associated antigen Rv2660.[72,73,101] H56 is delivered in IC31 by the intramuscular route and like H1 it is aimed to prevent acute TB disease as well as reactivation of existing LTBI in adolescents or adults.[215] The candidate is in a Phase I safety and immunogenicity trial in adults (healthy and with LTBI) in South Africa and plans to start recruitment for a Phase I/IIa safety and immunogenicity of multiple dosage levels and dosing regimens of H56-IC31 in HIV-negative adults with and without latent infection in late 2013 are underway.[215,217]

Hyvac 4-IC31 is a recombinant fusion protein of Ag85B and TB10.4[102] formulated in IC31 and intended as a subunit boost to an existing BCG-induced immunity in infants and children.[215] Intramuscular H4-IC31 has undergone four Phase I safety trials in healthy BCG-immune adults, three of which were for adjuvant and antigen dose definition and one for safety and immunogenicity following recent BCG immunization in adults.[215,218] Aeras is preparing for participant recruitment for a multi-center Phase I/IIa trial in South Africa in HIV-negative, BCG-primed infants for dose selection and assessment of interaction with EPI vaccines.[219]

It is important to note that H1, H56 and Hyvac 4 are three similar strategies and delivery systems in the clinic, and the clinical development plan of these three products signals important considerations. Of particular challenge may be interpretation of clinical data from early clinical trials with these vaccines, where small group size presents limitation in statistical extrapolation of data, important in the decision-making process to progress the best products to advanced clinical development where availability of resources and clinical trial sites become more limited/stringent.

ID93/GLA-SE: ID93 is a recombinant protein composed of four MTB antigens Rv3619, Rv1813, Rv3620 and Rv2608 linked in tandem. This fusion protein is adjuvanted with GLA-SE, which is a synthetic monophosphoryl lipid A (TLR-4 agonist) formulated in a stable oil-in-water emulsion.[74] In preclinical animal models, ID93/GLA-SE enhances Th1 responses and confers optimal protection in mice, guinea pigs and non-human primates when applied as a boost to BCG. A Phase I safety, tolerability and immunogenicity of intramuscular ID93/GLA-SE in healthy adults in the USA is underway.[220]

Therapeutic Vaccine Candidates

Candidates that have reached clinical development as potential therapeutic strategies aim to reduce duration of the TB treatment[103] and to date they comprise inactivated forms of mycobacteria (MTB or NTM).

RUTI consists of liposomes containing detoxified fragments of heat-inactivated virulent MTB, grown under stress conditions and has been successfully evaluated in two clinical trials: a Phase I safety, tolerability and immunogenicity for dose definition in healthy adults in Spain,[104] and in Phase II trial for safety, tolerability and immunogenicity following 1 month of isoniazid treatment in subjects with latent TB infection in South Africa.[221] Archivel Farma is currently looking for a financial partner to test a single dose of RUTI in a Phase III trials as an adjunct to chemotherapy against active TB disease in people with LTBI; and second, to prevent relapse episodes in active TB patients by administering the vaccine in the continuation phase of treatment.[209] No clinical trials with RUTI are currently ongoing.

Mycobacterium vaccae is an inactivated whole cell M. vaccae (NTM). Initially, agar-grown inactivated M. vaccae was evaluated as a therapeutic vaccine in HIV-positive and in HIV-negative, MTB sputum-positive populations in different countries.[105–109] After the first Phase III efficacy trial of M. vaccae (the DarDar Trial) in which a 5-dose regimen was safe, immunogenic and effective in preventing TB in HIV-infected adults with prior BCG immunization,[105] development of inactivated M. vaccae has shifted its TPP to a prophylactic strategy.[222]M. vaccae (DAR-901) is a broth-grown heat-inactivated polyantigenic M. vaccae manufactured by Aeras and is targeted as a TB prophylactic 3-dose series suitable for both HIV-infected and non-infected adults who have previously received BCG.[222] Non-clinical toxicology and immunogenicity studies have been initiated by Aeras to support entry into clinical evaluation for safety and immunogenicity in the USA and Tanzania in 2013.

Mw is a whole-cell heat-inactivated saprophytic non-pathogenic Mycobacterium indicus pranii, initially coded as Mycobacterium w,[110] with an established immunotherapeutic role used beneficially in the treatment of leprosy.[111,112]Mw has also shown promising immunotherapeutic and immunoprophylactic potential against pulmonary TB disease in clinic.[113,114] Sputum conversion has been reported to appear more rapidly in patients who received DOTS + Mw as compared with DOTS alone.[114,115] The Drug Controller of India has already licensed Mw for use in humans.[209] A large-scale multi-centric Phase III evaluation sponsored by the Department of Biotech (DBT) (Government of India) and Cadila Pharmaceuticals to assess the immunotherapeutic effect of Mw as an adjunct to first-line antimicrobial therapy in pulmonary TB retreatment patients (category I and category II TB patients, and individuals with TB pericarditis) has recently been completed. Results have yet to be published (Table 1).[201,209,223]