Review

The status of tuberculosis vaccine development

Lewis K Schrager, Johan Vekemens, Nick Drager, David M Lewinsohn, Ole F Olesen

Tuberculosis represents the leading global cause of death from an infectious agent. Controlling the tuberculosis Lancet Infect Dis 2020
epidemic thus represents an urgent global public health priority. Epidemiological modelling suggests that, although Published Online
drug treatments for tuberculosis continue to improve, WHO timelines to control the spread of the disease require a January 31, 2020
https://doi.org/10.1016/
new vaccine capable of preventing tuberculosis, particularly in adolescents and adults. The spread of strains resistant
S1473-3099(19)30625-5
to multiple drugs adds additional urgency to the vaccine development effort yet attempts to develop new vaccines with
Independent Consultant,
wider applicability and better, longer-lasting efficacy than BCG—the only tuberculosis vaccine licensed for use Bethesda, MD, USA
globally—have proven challenging. Results from clinical efficacy trials, particularly a completed, phase 2b trial for (L K Schlager MD); Initiative
preventing tuberculosis disease in people infected with Mycobacterium tuberculosis using the adjuvanted protein for Vaccine Research, WHO,
Geneva, Switzerland
subunit vaccine M72/AS01E give hope. We review the current status of tuberculosis vaccine candidates and outline
(J Vekemens MD); TuBerculosis
the diversified vaccine development that are underway. Vaccine Initiative, Lelystad,
Netherlands (N Drager MD);
Introduction crucial component of the global efforts to control the Center for Global Child Health
Research, Department of
Developing interventions against tuberculosis represents tuberculosis epidemic, including the spread of drug-
Pediatrics, Division of
a crucial global health priority.1 Tuberculosis is the resistant strains. Pulmonary and Critical Care
leading cause of death globally from a single infectious Medicine, Oregon Health and
agent, killing approximately 1·45 million people in 2018, Tuberculosis vaccine feasibility Science University and the
Portland VA Medical Center,
251 000 of whom were HIV-infected,2 illustrating the The existence of natural human defences against tuber­
Portland, Oregon, OR, USA
importance of developing a vaccine capable of preventing culosis suggests that it should be possible to discover and (Prof D M Lewinsohn MD);
tuberculosis in both HIV-uninfected and HIV-infected develop a preventive vaccine. Approximately 85–95% of European and Developing
individuals. Approximately one quarter of people world­ people infected with M tuberculosis can control the Countries Clinical Trials
Partnership, Hague,
wide, an estimated 1·7 billion people, are infected with infection, never developing any manifestation of tuber­ Netherlands (O F Olesen PhD);
Mycobacterium tuberculosis, the bacterium that causes culosis disease.4 Observational studies have shown that and Global Health Section,
tuberculosis.3 Of these infected people, 5–15% will harbouring latent M tuberculosis infection provides Department of Public Health,
develop tuberculosis during their lifetimes, with much some protection against tuberculosis disease developing University of Copenhagen,
Copenhagen, Denmark
higher risk of disease expected for people with com­­ from new exposure to the pathogen.9,10 Additionally, some (O F Olesen)
promised immune systems, such as those with HIV individuals never acquire M tuberculosis infection despite
Correspondence to:
infection, malnutrition or diabetes, people exposed to long-term exposure to people with active pulmonary Dr Ole F Olsen, Department of
silica dust or indoor smoke, or people who use alcohol or tuberculosis, such as household contacts and health-care Public Health, University of
tobacco.4 workers, raising the possibility of intrinsic resistance.11,12 Copenhagen, 1014 Copenhagen,
Denmark
The spread of tuberculosis caused by M tuberculosis Moreover, the possibility that some become infected with
[email protected]
strains resistant to multiple drugs represents a growing M tuberculosis but then naturally clear their infection has
threat to public health due to poor cure rates,2 high fatality,5 been raised by documented accounts of individuals who
and the extraordinary cost of treating patients with convert their tuberculin skin tests or interferon-γ release
multidrug-resistant (MDR) tuberculosis (MDR tuber­ assays (IGRAs), but later revert to negative,13–15 although
culosis is defined as M tuberculosis isolates resistant to the these reports should be interpreted with caution.16 The
two first-line treatments for tuberculosis, rifampicin and reality of the ongoing tuberculosis epidemic, however,
isoniazid) and extensively drug-resistant (XDR) tuber­ illustrates that many people are incapable of independently
culosis (XDR tuberculosis is defined as M tuberculosis generating sufficient immune control of M tuberculosis
isolates resistant to rifampicin and isoniazid, as well as infection, underlining the importance of developing a
any fluoroquinolone and at least one of the three injectable vaccine.
second-line drugs [amikacin, kana­ mycin, or capreo­ The BCG vaccine, first used in 1921, is an attenuated
mycin]).6,7 Vaccines capable of preventing tuberculosis due form of Mycobacterium bovis, the primary cause of bovine
to drug sensitive strains will probably be effective against tuberculosis, and is currently the only tuberculosis
drug-resistant tuberculosis given that the molecular vaccine that is licensed for use globally. BCG is the most
mechanisms that confer M tuberculosis drug resistance widely used vaccine in history, with more than 4 billion
should not change the immunological characteristics of doses administered since that first vaccination.17 BCG is
the organism against which an effective, vaccine-induced routinely given to infants within days of birth, or when
immune response is targeted. Accordingly, tuberculosis they first interact with health services, in most countries.18
vaccines might have important roles in future strategies Multiple studies have shown that vaccination of infants
to control MDR and XDR tuberculosis, particularly as with BCG is effective at preventing severe forms of
resistance to the new mycobacterial drugs bedaquiline tuberculosis in children.17 BCG has been shown to protect
and delamanid has been reported.8 Accelerating the infants and young children, up to approximately 10 years
develop­ment of new and effective vaccines represents a of age, from developing pulmonary and extrapulmonary

www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5 1

 Review

tuberculosis, although the degree and duration of pro­ Using a whole cell-derived vaccine represents an
tection has been variable, with increased protection in attractive vaccine strategy for several reasons. Chief among
areas farther from the equator.19 Some observations also them is the proven efficacy of BCG, a live, attenuated
indicate that BCG seems to induce better protection in whole cell vaccine that provides a solid proof-of-concept.
those who are not tuberculosis skin test positive.20 Some Additionally, whole cell-derived vaccines comprise many
degree of protection against active tuberculosis might different antigenic components and offer the possibility
last for up to 20 years following school-aged BCG of stimulating a more diverse immune response than
vaccination21 and as long as 50–60 years following infant subunit vaccines due to the presence of non-protein
vaccination.22 Studies of BCG revaccination of adolescents antigens such as lipids32–34 and glycolipids35 prevalent in the
have not consistently shown a protective effect against outer layers of M tuberculosis, microbial metabolites, and
tuberculosis disease.23,24 Although a 2018, completed, phosphoantigens.36
placebo-controlled study of BCG revaccination in
South African adolescents showed a significant degree of Subunit vaccines
protection against sus­ tained, de novo M tuberculosis Subunit vaccines can be further subcategorised into
infection as defined by an IGRA conversion of 6 or more adjuvanted protein subunit vaccines and recombinant
months duration relative to the placebo arm,25 the clinical viral vectored vaccines. By contrast with whole cell
significance of these results will require further study. vaccines, subunit vaccines target a small number of
selected antigens, usually six or less. Accordingly, a major
The clinical pipeline of tuberculosis vaccine challenge to developing protein subunit vaccines is the
The tuberculosis vaccine candidates currently in clinical need to identify the optimal antigens to include in such
trials (figure) can be grouped into two general categories: vaccines. Intense efforts have been underway to identify
mycobacterial whole cell-derived vaccines (table 1) and the key antigens, expressed from approximately 4000 genes
subunit vaccines, directed against a limited number of associated with M tuberculosis, that could elicit a protective
selected antigens (table 2). immune response against the organism.37,38 Some re­
searchers have specifically selected antigens for inclusion
Mycobacterial whole cell-derived vaccines in vaccine candidates according to their expression profile
Whole cell vaccines are derived from M tuberculosis, at various stages of mycobacterial infection, on the basis of
BCG, or closely related strains of non-tuberculous metabolic activity and protein expression thought to affect
mycobacteria. This class of vaccines can be further vulnerability to the immune system.39
broken down into live vaccines that have been attenuated The antigens included in current vaccines in clinical
through genetic modification (VPM 1002,26 MTBVAC27), trials have all been tested for their ability to induce some
or vaccines derived from killed or fractionated whole form of protective immunity against tuberculosis in an
mycobacteria (Mycobacterium vaccae [Vaccae],28 Myco­ animal model of tuberculosis vaccination. However, most
bacterium indicus pranii [MIP],29 DAR-901,30 RUTI31; of these antigens have been selected for their capacity to
table 1). induce the release of the Th1 cytokines interferon-γ and
tumour necrosis factor. This choice is informed by the
clinical observation that individuals with innate or acquired
Phase 1 Phase 2a Phase 2a Phase 3 Licensed immunodeficiency in these immune pathways are highly
susceptible to tuberculosis.40 Whether these characteristics
BCG will result in the generation of protective immunity,
Live VPM 1002 however, is uncertain.41,42 A component of the advancement
Mycobacteria

MTBVAC of these purified protein antigens to clinical trials involves

VaccaeTM evidence of their ability to elicit classical T-cell responses.
Killed
MIP Non-classic T cells are intriguing in that they use host
DAR-901 molecules that do not vary between individuals, such as
RUTI CD1, MR1, and HLA-E. The ligands for these molecules
M72 + AS01E are lipid, glycolipid, vitamin meta­bolite, and peptide and
Adjuvanted
H56:IC31 glycopeptide antigens.43 The T-cell receptors from the non-
ID93 + GLA-SE classic T cells can recognise these ligands presented in the
Sub-unit

GamTBvac context of the non-classic molecules. How these non-

TB/Flu-04L classic T cells might best be used as vaccines remains to be
Viral Ad5Ag85A
vectored
determined. Vaccine strategies targeting the non-classic
ChAdOx185A responses could possibly enhance the development of
more traditional T-cell or B-cell responses. Additionally,
Figure: Tuberculosis vaccine candidates in clinical development cells using the γδ T-cell receptor can recognise glycolipid-
The indicated clinical development stages of vaccine candidates are based on an extrapolation from data in antigens and phosphoantigens.43 Tuberculosis vacci­nation
ClinicalTrials.gov. strategies predicated on the generation of anti­ bodies

2 www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5

 Review

remain relatively unexplored but represent a valid area for proof-of-concept study assessing the potential for a vaccine
further investigation.44 to prevent the development of tuberculosis disease in
M tuberculosis-infected adults are unprecedented and
Adjuvanted protein subunit vaccines increase optimism that developing vaccines capable of
Four adjuvanted protein subunit vaccines are in clinical preventing pulmonary tuberculosis in adults is possible.
trials: M72/AS01E,45 ID93 + GLA-SE,46 H56/IC31,47 and Follow-up studies to confirm these results, to obtain a
GamTBvac48 (table 2 and appendix pp 1–4). Adjuvanted more precise estimate of the extent of protection, and to See Online for appendix
protein subunit vaccines consist of an antigenic protein, get a better sense of the duration of the protective effect will
or a linked series of antigenic proteins, administered be essential. Whether the vaccine can protect indi­viduals
along with an adjuvant—an agent that is designed to who have not been infected with M tuberculosis in the
potentiate immune system stimulation in such a way as past should also be evaluated along with other potential
to maximise the immunological impact of the antigens indications for the vaccine, such as prevention of re­
incorporated in the vaccine. current tuberculosis following completion of successful
A 2018, phase 2 clinical trial compared H4:IC31 vaccine, treatment.50
consisting of a fusion protein of two immunodominant
antigens, antigen 85B and TB 10·4, in combination with Parental mycobacterium Modification or inactivation
the IC31 adjuvant, to placebo, and BCG revaccination
Live attenuated
to placebo, in protecting South African adolescents
VPM1002 (BCG ΔureC::hly) Mycobacterium bovis BCG-Prague Integration of listeriolysin encoding gene
(uninfected with M tuberculosis but at high risk of insert, urease gene deletion
M tuberculosis infection) from developing sustained MTBVAC Mycobacterium tuberculosis Transcription factor phoP and fadD26
quantiferon conversion.25 Results of this trial showed that (phthiocerol dimycocerosate synthesis)
neither the H4:IC31 vaccine nor revaccination with the gene deletions
BCG vaccine prevented initial quantiferon conversion, Killed or extract
the primary outcome (for H4:IC31, efficacy point estimate Vaccae Mycobacterium vaccae Heat-killed
9·4%, 95% CI –36·2 to 39·7; p=0·63; for BCG, efficacy MIP Mycobacterium indicus pranii Heat-killed
point estimate 20·1%, 95% CI –21·0 to 47·2; p=0·29). (formerly: Mycobacterium w)
In light of these results, the clinical development of DAR-901 Mycobacterium obuense Heat-killed
the H4:IC31 vaccine has been discontinued. BCG re­ RUTI M tuberculosis Cell wall fragments of M tuberculosis
grown under stress conditions, in
vaccination, however, reduced the rate of sustained liposome suspension
quantiferon conversion, the secondary outcome (efficacy
point estimate 45·4%, 95% CI 6·4 to 68·1; p=0·03); the Table 1: Composition of whole mycobacterial cell-derived vaccines in clinical trials
point estimate of efficacy of sustained quantiferon
conversion for H4:IC31 was 30·5% (95% CI –15·8 to 58·3;
Antigen (antigen description) Adjuvant Adjuvant description
p=0·16).25 Future studies should be conducted with
subunit vaccines comprised of a wider and more Adjuvanted protein

diversified array of antigens, and used in combination M72 Rv1196 (PPE family), AS01 Liposome, TLR4 agonist
Rv0125 (peptidase)
with more potent adjuvants, in order to assess the full
H56 ESAT-6 (Mycobacterium IC31 Antimicrobial peptide
potential of the subunit vaccine strategy for developing a tuberculosis antigen), Ag85B KLK + oligodeoxynucleotide
protective tuberculosis vaccine. Additionally, the results (mycolyl transferase), Rv2660c ODN1a, TLR9 agonist
of the BCG revaccination arm of this study suggest that (dormancy antigen)
further investigation of the potential utility of this strategy ID93 Rv2608 (PPE family), GLA-SE Glucopyranosyl lipid
Rv3619 (virulence factor), adjuvant in emulsion of
is warranted. Rv3620 (virulence factor), squalene oil-in-water,
Final analysis of a phase 2b clinical trial of M72/AS01E, Rv1813 (dormancy antigen) TLR4 agonist
consisting of a fusion protein of two immunodominant, GamTBvac Ag85A (mycolyl transferase), CpG ODN Oligodeoxynucleotides,
highly conserved M tuberculosis proteins, M tuberculosis ESAT6-CFP10 (virulence antigens) TLR9 agonists
32A and adjuvanted with the proprietary adjuvant system Viral vector
AS01E, provided even more robust evidence of protective Ad5Ag85A Ag85A (mycolyl transferase) Ad5 Recombinant replication
efficacy of a subunit vaccine.49 In this randomised, double- deficient human
adenovirus 5 vector
blinded, placebo-controlled trial, more than 3500 IGRA-
ChAdOx185A + MVA85A Ag85A (mycolyl transferase) ChAdOx1, MVA Chimpanzee adenovirus,
positive, HIV-negative-adults age 18–50 years from South boost modified vaccinia ankara
Africa, Zambia, and Kenya, most of whom had received virus
BCG vaccination as infants, were randomly assigned TB/Flu-04L Ag85A (mycolyl transferase) Flu-04L Attenuated replication-
1:1 to receive two doses of either the vaccine or placebo deficient influenza virus
intramuscularly, 1 month apart. The primary endpoint (H1N1)

was progression to bacteriologically confirmed active PPE=proline-proline-glutamic acid. TLR=toll-like receptor.

pulmonary tuberculosis. After 3 years of follow-up, vaccine
Table 2: Composition of subunit vaccines in clinical trials
efficacy was 49·7% (95% CI 2·1–74·2). The results of this

www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5 3

 Review

Recombinant viral-vectored vaccines Preventing tuberculosis disease

Three viral-vectored subunit vaccines candidates are in Due to the small percentage of people infected with
clinical trials: TB/Flu04L,51 Ad5 Ag85A,52 and ChAdOx185A/ M tuberculosis expected to progress to active tuberculosis
MVA85A (NCT03681860) (table 2 and appendix pp 1–4). during their lifetimes, studies to assess vaccine efficacy
Recombinant viral-vectored vaccines are similar to for a PoD indication face distinct challenges, including
adjuvanted protein subunit vaccines as they serve as the necessity for a large sample size, a trial duration of
delivery systems for selected M tuberculosis protein 3 years or more, and the high costs associated with
antigens thought to be important in generating a protective meeting these conditions. Additionally, an important
immune response against tuberculosis. In this vaccine uncertainty inherent in developing a vaccine for a PoD
strategy, however, the viral vector is intended to trigger a indication is whether a single vaccine could prevent
robust and long-lasting immune response to the antigens tuberculosis disease both in people who receive the
being presented, obviating the need for an exogenous vaccine before becoming infected with M tuberculosis (ie,
adjuvant. people who are IGRA-negative), and in people
Many of the same issues inherent in developing protein immunised subsequent to becoming latently infected
subunit vaccines, including the challenge in identifying a with M tuberculosis (people who are IGRA-positive).
few, critical antigens, also apply to viral-vectored delivery. Given these challenges, vaccine developers have used
Additionally, the vector must be safe enough to prevent other clinical development strategies before embarking on
the occurrence of vector-based pathogenesis. Ensuing resource-intensive PoD trials. These strategies might be
safety is particularly challenging when designing viral used to bolster confidence in the clinical potential of a
vector-based vaccines for possible administration to vaccine candidate and to explore the possibility of seeking
immune suppressed individuals, including people other indications for tuberculosis vaccines.54
with HIV, in whom a vector proven as non-pathogenic in
immune competent individuals might cause disease. Preventing recurrent tuberculosis disease
A further challenge to a viral vectored vaccine occurs if Individuals completing treatment for drug-sensitive
more than one vaccination is required to reach the tuberculosis are at elevated risk of experiencing recurrent
targeted degree of immune response. In this situation, tuberculosis disease, either due to relapse of incompletely
immunity to the viral vector, developed during the initial treated disease or recurrent M tuberculosis infection.55
administration of the vaccine, could potentially result in Risks of recurrent tuberculosis disease increase with
the premature, immune-based destruction of the vector shortened treatment, poor compliance, and in HIV-
during sub­sequent vaccine administrations, preventing infected individuals.
the ex­pression of the antigen and thereby rendering the A PoR trial is less onerous to conduct than a PoD trial
vaccine ineffective.53 A similar situation could arise if the given that the specific population at risk for recurrent
people being vaccinated were infected or might become disease—those completing drug treatment for tuber­
exposed to the virus being used as a vector during the culosis—can be identified, and considering the high rate
course of a vaccine administration regimen. These of recurrent disease occurring in a relatively short time
problems might be circumvented using a heterologous (1 or 2 years) after cessation of drug treatment. These
prime-boost vaccination strategy, in which the boosting factors make it feasible to use smaller sample sizes, and
vaccine contains the same antigens as the priming shorter trial durations, than would be required to assess
vaccine but uses a delivery strategy different from the a PoD outcome. PoR trials can thus provide a general
vectored vaccine used to deliver the priming vaccination. indication of clinical protective activity, while also
supporting a potential licensable indication. However,
Assessing clinical efficacy of tuberculosis some individuals with relapsed or recurrent infection
vaccines might have immunological abnormalities that make their
When assessing the efficacy of tuberculosis vaccine protection via vaccination a greater challenge. Individuals
candidates, investigators generally choose one of three who have had tuberculosis can possibly also not be
possible clinical endpoints: prevention of tuberculosis protected with an immune-modulating vaccine, given that
disease (PoD), prevention of recurrent tuberculosis disease tuberculosis disease is associated with a reduction in lung
in people completing a course of anti-tuberculous drug function, damage of the lung matrix, and creation of
therapy (PoR), or prevention of established M tuberculosis immune-privileged sites. Additionally, uncertainties exist
infection in individuals entering the trial without evidence over the optimal timing of PoR vaccine administration,
of pre-existing M tuberculosis infection (PoI). A fourth whether the vaccine should be given before or after
potential use of tuberculosis vaccines, as immuno­thera­ completion of drug treatment, adding degrees of risk to
peutic agents to increase the effectiveness and reduce the this strategy.
duration of drug therapy for active tuberculosis disease, is
another endpoint of increasing interest to the field but is Preventing established M tuberculosis infection
beyond the scope of this review (the RUTI vaccine is being Preventing established M tuberculosis infection with a
developed specifically for this purpose). vaccine (indicated by persistent conversion of tuberculin

4 www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5

 Review

skin tests or IGRA from negative to positive) would non-tuberculous mycobacteria. Only one vaccine in clinical
intuitively appear to be an effective way to prevent testing, MTBVAC, is derived directly from M tuberculosis.27
tuberculosis disease. It remains unclear, however, whether The vectored and adjuvanted protein vaccines in clinical
vaccine success or failure against a PoI endpoint would trials express, at most, four antigens, whereas the pre-
reflect success or failure against a PoD endpoint. It is clinical cytomegalovirus-vectored tuber­ culosis vaccines
possible that a vaccine might only prevent infection in expresses six.65 The need to generate a interferon-γ
people who otherwise would never develop disease.25 PoI producing Th1 type response has typically been a primary
is therefore not universally accepted as an indication for consideration for selection of antigens, but antigen
licensure of a tuberculosis vaccine. This endpoint is identification has recently sought to select M tuberculosis
nevertheless being used by developers in phase 2 trials to antigens that would result in a diversification of immune
gain greater confidence in the likelihood of the candidate’s responses beyond CD4 Th1. Particular efforts have focused
success, before proceeding to expensive and resource upon identifying antigens capable of stimulating non-
intensive, later-stage PoD trials.25 CD4 T-cell responses, mostly MHC-I-restricted, CD8 T-cell
mediated responses.66–68 Identification of antigens capable
Status of the preclinical pipeline of engendering B-cell responses and subsequent antibody
Several promising tuberculosis vaccine candidates are in expression, are also drawing attention.44 Ongoing research
advanced pre-clinical development. These include into the functions of various M tuberculosis genes, and the
vectored vaccines using chimpanzee adenoviruses,56 a effect of their deletion to allow safe administration
human parainfluenza virus vectored vaccine,57 the to humans while preserving critical components of the
replication-deficient, modified vaccinia Ankara antigen vaccine candidate’s immunogenicity, has the potential to
delivery platform,58,59 a new adjuvanted subunit vaccine identify promising new antigens. Nucleic acid-based
(H64) containing six highly expressed M tuberculosis vaccines, including DNA and RNA vaccines, represent
proteins,60 and a recombinant, attenuated BCG.61 Addi­ another route to vaccine diversification.69
tionally, early work is proceeding with several live,
attenuated M tuberculosis vaccines, including concepts Diversification of immune response types
with deletions of ΔsigH,62 ΔmosR,63 or ΔechA763 genes, and M tuberculosis is an intracellular pathogen, infecting
with a variety of next-generation vaccines based on BCG.64 pulmonary-based antigen presenting cells such as
One of the most promising pre-clinical vaccine candidates macrophages and dendritic cells before progressing to
uses an attenuated cytomegalovirus vector that has been a stage of established infection within granulomas.
modified with inserted genes for selected M tuberculosis Tuberculosis vaccine development has therefore focused
antigens thought to be important for stimulating pro­ on stimulation of cell mediated immune responses.
tective immunity.65 This new cytomegalovirus-vectored Accordingly, immunogenicity assessments essential to the
tuber­culosis vaccine has demonstrated an overall vaccine advancement of tuberculosis vaccine candidates—whole
effect of approximately 70% in rhesus macaques, with cell and subunit—currently in clinical trials focused on
complete prevention of tuberculosis disease in more than their ability to stimulate interferon-γ, interleukin 2, and
40% of the vaccinated animals.65 A crucial feature of tumour necrosis factor responses in CD4 T cells obtained
the cyto­megalovirus vector is that it persists, perhaps from vaccinated animals and humans, following ex-vivo
indefinitely, whereas other viral vectors are eliminated exposure of these T cells to tuberculin protein, tuberculosis
within a few replication cycles. The high degree of peptides, or other tuberculosis vaccine antigens. Although
protection by this vaccine construct might therefore result the T-cell interferon-γ response is believed to be essential
from its ability to elicit and maintain lung-based immune in protecting against M tuberculosis disease, it has not been
responses capable of intercepting M tuberculosis at the confirmed to be either a vaccine-induced correlate or
onset of infection rather than at later stages when the surrogate of protection against tuberculosis.70
infecting M tuberculosis organisms are better capable of Attention has focused on the potential benefits of
circumventing host immune responses.65 assessing immune responses generated via non-classic
presentation of non-protein antigens with limited
Diversification of the tuberculosis vaccine polymorphism and molecules that are unique to
portfolio M tuberculosis.71 Examples include HLA-E (which can
An examination of the current tuberculosis vaccine present peptide and glycopeptide antigens),67,72 group 1
candidates reveals the need for establishing a broader (CD1a-c) restricted T cells (which recognise CD1 glycolipid
and more robust portfolio through diversification of and lipid antigens found within the outer mycobacterial
antigens, vaccine platforms, administration routes, and membrane),34,73 mucosal associated invariant T cells
types of immune responses induced. (which are enriched in the lungs),74,75 natural killer T cells
(which recognise glycolipids in the context of CD1d),34
Diversification of antigens and vaccine platforms and γδ T cells.76 As non-classic presentation molecules are
Most tuberculosis vaccine candidates are variations of highly conserved, they offer the advantage that a vaccine
vaccine subunit technology or whole cell vaccines based on targeting these responses would work across all

www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5 5

 Review

populations. Developing vaccines capable of stimulating Biomarkers and biosignatures

these and other so-called unconventional T-cell responses, Although certain elements of effective immunity to
along with conventional CD4 and CD8 T-cell responses, tuberculosis are well understood, the full picture of
represents a new frontier in tuberculosis vaccine research. an immune response capable of protecting against
The role of antibody-mediated protection against M tuberculosis infection or tuberculosis disease remains
tuberculosis has gained renewed interest.77 Indirect obscure.70 A better understanding of correlates of
effects of humoral immunity can have a key role in protection—a measurable immune response, or group of
increasing the efficiency of cell mediated immune responses, demonstrated to protect against M tuberculosis
control of pathogens hiding within cells, as M tuberculosis infection or tuberculosis disease—would be a major step
does within macrophages and dendritic cells. forward.84 This knowledge could then guide antigen
selection and accelerate vaccine development by early
Diversification of administration routes selection of promising candidates in smaller clinical
All tuberculosis vaccines currently beyond phase 1 in the trials, including bridging studies across geographical
clinical pipeline are administered via injection, either areas, age groups, and risk groups. Many of the most
intramuscularly or intradermally. The initial site of advanced biometric techniques, including transcrip­
M tuberculosis infection, however, occurs in the lung, and tomics, proteomics, and metabolomics, are being used,
delivering a tuberculosis vaccine via aerosol, potentially individually and in combination, to identify potential
stimulating both mucosal and peribronchial immune biosignatures, combinations of a variety of biomarkers,
responses against M tuberculosis, is being investigated. of immune protection.84,85 However, applying these
Several phase 1 studies are exploring the safety and techniques requires that samples from clinical trials are
immunogenicity of administering various tuberculosis collected, bio-banked, and available and this aspect is not
vaccines, including BCG and MVA85A, via aerosol,78,79 always adequately addressed or budgeted for in clinical
while preclinical studies explore the efficacy of aerosol trials.
application, either administered alone or in combination
with an injected tuberculosis vaccine.80 Animal models
Older and more recent experiments in rhesus macaques The tuberculosis vaccines in clinical development have
that used intravenous route of BCG administration all undergone prior testing in animal models, mostly
showed superior protection than BCG administered intra­ mice, guinea pigs, and non-human primates. The testing
dermally or through the aerosol route.81,82 The heightened has typically involved an aerosol exposure to M tuberculosis
protective effect of intravenous administration might from hundreds to more than 1000 colony-forming units,
therefore be due to the generation of immunological but such supraphysiological conditions have created an
control through increased levels of CD4 effector T-cell unnaturally difficult challenge for vaccine candidates to
activity within the lung or increased activation of innate overcome. Recent efforts to improve animal challenge
components of the cellular immune response.82 models for tuberculosis vaccine testing have attempted
to better approximate the low level of exposure occurring
Diversification of administration strategies in natural human infection. Most animal challenge
In addition to diversifying routes of vaccine adminis­ models are now using low-dose, aerosolised M tuberculosis
tration, the potential utility of homologous and hetero­ inocula, ranging from 25 to ten colony-forming units,
logous prime-boost approaches is also being explored. In or so-called ultra-low-dose approaches, where fewer
a homologous prime-boost strategy, the same vaccine than ten colony-forming units of M tuberculosis are
platforms, delivering the same tuberculosis antigens, are administered.86,87
used in the priming and boosting phases of vaccine Mice have typically been used as a first screening model
administration; in a heterologous prime-boost strategy, in vivo due to the relatively inexpensive and rapid results,
different vaccine platforms are used to deliver the same but the extent to which the results of challenge
antigens.83 The heterologous prime-boost strategy has experiments in mice predict human outcomes of
become increasingly attractive given the potential for vaccination remains uncertain.88 Guinea pigs have not
stimulating a broader immune response when delivering been used as extensively as mice for M tuberculosis
antigens in different ways, and as means to overcome the challenge experiments, but they are noted for their
problem of anti-vector immunity, a known challenge exquisite sensitivity to low-dose M tuberculosis infection.89
to vectored vaccine platforms requiring a number of These experiments often represent the final step in
administrations over time.83 preclinical assessment, given the similarities in genetics,
tuberculosis pathophysiology and overall immunology,
Critical tools are missing between non-human primates and humans.90 A sub­stantial
Developing better tools for testing and selecting vaccine variability of the outcome of M tuberculosis infection
candidates represent, together with diversification of among non-human primate species has been noted,
the portfolio, another crucial goal towards developing a emphasising the importance of selecting the most
tuberculosis vaccine. appropriate species for challenge experiments based on

6 www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5

 Review

the ultimate study goals.91 Developments in scanning

technologies, particularly a combination of PET and Search strategy and selection criteria
CT scanning, offer the potential of making quantifiable We searched PubMed for articles published in English from
comparisons between vaccinated and unvaccinated Jan 1, 2000, to June 30, 2019 to identify papers on
non-human primates of the development of pulmonary tuberculosis vaccine development, by use of the terms
tuberculosis infection and disease within days to weeks “tuberculosis”, “vaccine”, “Mycobacterium”, “immunization”,
after challenge.92 This comparison might provide an or “vaccination”. We further searched the ClinicalTrials.gov
opportunity to obtain earlier assessments of vaccines at database, provided by the US National Library of Medicine,
far lower cost than were possible before.93 for relevant entries using the search terms “tuberculosis” and
“vaccine”. Additionally, we reviewed relevant articles cited in
Controlled human infection model references from the above search extracts and included them
The development of a controlled human infection as primary sources when appropriate. We reviewed articles
model of tuberculosis would represent a breakthrough resulting from these searches and their references. The
for tuberculosis vaccine research. These models have number of relevant studies was too large for all to be cited,
played important roles in advancing vaccine development but we considered all prioritising clinical trials, antigen
efforts, particularly for malaria and influenza.94 Controlled identification, and pre-clinical validation records.
human infection models permit a carefully controlled
administration of pathogens to humans in a safe manner
to assess the clinical effect of vaccines or drugs at an intermittent mass vaccination campaigns would be
early stage of vaccine development and at relatively low logistically and economically feasible.
cost. Optimally, a controlled human infection model for New information about the molecular structure
M tuberculosis would permit an assessment of vaccine and mechanisms driving the disease-producing activity
protection against a low dose M tuberculosis exposure via of M tuberculosis, the immunological responses to
the aerosol route. Developing a challenge strain of live M tuberculosis infection, the basic pathogenesis of
M tuberculosis sufficiently safe for experimental aerosol tuberculosis disease in humans and animal models, and
administration to humans, and detectable by sensitive the efficacy or lack thereof, shown for a variety of
assays in serum, urine, or in the breath, remains a preclinical and clinical vaccine candidates, point the way
primary goal for the controlled human infection model of to several notable, interrelated directions for future
tuberculosis. Additionally, a model using intradermal tuberculosis vaccine research and development. At the
BCG assessed via punch biopsy and PCR and colony- same time, considerable efforts are being made to
forming units quantification has been used in humans establish a global framework for optimal development of
and is undergoing further evaluation.95 a new vaccine. Organisations central to global tuberculosis
vaccine development such as the TuBerculosis Vaccine
Tuberculosis vaccine development: a global effort Initiative, the International AIDS Vaccine Initiative, the
Ambitious goals for ending the global epidemic of European Commission, the European Investment Bank,
tuberculosis and vastly reducing tuberculosis mortality the European and Developing Countries Clinical Trials
on or before the year 2035 have been set by the UN Partnership, and the Bill and Melinda Gates Foundation
within its Sustainable Development Goals, the WHO have formed the Global TB Vaccine Partnership and
End TB Strategy, and the End TB Goals of the Stop TB created a set of mutually agreed stage gate criteria for
Partnership.2 The first United Nations High Level advancing early-stage tuberculosis vaccine concepts into
Meeting on tuberculosis was held on Sept 26, 2018, and animal testing and subsequently into progressive clinical
called for accelerated and concerted efforts against the development phases.99 Adherence to these criteria is
disease in a declaration that was subsequently adopted by intended to optimise the use of resources by proposing
the UN General Assembly.96 None of these ambitious a predefined prioritisation framework, and thereby
goals, however, can be met without the development of decrease the risk of candidate failure in expensive, later-
new and better drug treatments for tuberculosis, better stage clinical efficacy trials. In parallel with this, WHO
point-of-care diagnostics, and the implementation of an has developed preferred product characteristics for
effective vaccine.2 tuberculosis vaccines that will provide additional
Economic models have suggested an extraordinary cost- guidance regarding charac­teristics and efficacy targets to
effectiveness from tuberculosis vaccines on both health tuberculosis vaccine development efforts.100
system and societal perspectives.97 In one such modelling
exercise, a tuberculosis vaccine targeting adolescents and Conclusion
adults in low-income countries, with 60% efficacy over Developing a tuberculosis vaccine remains a critical
a 10-year duration could prevent 17 million cases of global health priority given the deadly nature of
tuberculosis (range 11–24 million) by 2050.98 Even if the the ongoing tuberculosis epidemic and the spread of
resultant vaccines demonstrated efficacy rates as low drug-resistant M tuberculosis strains. The significant
as 20%, assuming a duration of effect of 10 years, degree of protection against tuberculosis disease shown

www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5 7

 Review

by the M72/AS01E candidate in a phase 2b clinical trial 12 Verrall AJ. G. Netea M, Alisjahbana B, Hill PC, van Crevel R.
strongly suggests that tuberculosis vaccines are feasible Early clearance of Mycobacterium tuberculosis: a new frontier in
prevention. Immunology 2014; 141: 506–13.
and the encouraging preclinical results from advanced 13 Pai M, O’Brien R. Serial testing for tuberculosis: can we make
vaccine candidates, such as a cytomegalovirus-vectored sense of T cell assay conversions and reversions? PLoS Med;
tuberculosis vaccine construct, offer the prospect of 4: e208.
14 Hill PC, Brookes RH, Fox A, et al. Longitudinal assessment of an
further progress. Diversifying tuberculosis vaccine- ELISPOT test for Mycobacterium tuberculosis infection. PLoS Med
generated immune responses, routes of administration, 2007; 4: e192.
and vaccine platforms represent novel strategies likely 15 Hawn TR, Day TA, Scriba TJ, et al. Tuberculosis vaccines and
prevention of infection. Microbiol Mol Biol Rev 2014; 78: 650–71.
to increase the chance of future vaccine success. Parallel
16 Andrews JR, Hatherill M, Mahomed H, et al. The dynamics of
improvements in preclinical and clinical testing stra­ QuantiFERON-TB Gold In-Tube conversion and reversion in a
tegies will permit greater information about potential cohort of South African adolescents. Am J Respir Crit Care Med
vaccine efficacy to be collected in a shorter time and at 2015; 191: 584–91.
17 Smith KC, Orme IM, Starke JR. Tuberculosis vaccines.
less cost. Success in this endeavour is not optional In: Vaccines. Philadelphia, PA: Elsevier, 2013: 789–811.
given the health burden and cost of the tuberculosis 18 Dockrell HM, Smith SG. What have we learnt about BCG
epidemic; a persistent, robust, global effort must be vaccination in the last 20 years? Front Immunol 2017; 8: 1134.
maintained until a safe, efficacious tuberculosis vaccine 19 Abubakar I, Pimpin L, Ariti C, et al. Systematic review and
meta-analysis of the current evidence on the duration of protection
is achieved. by bacillus Calmette-Guérin vaccination against tuberculosis.
Contributors Health Technol Assess 2013; 17: 1–372.
All authors contributed to the conception of the Review. LKS and 20 Mangtani P, Abubakar I, Ariti C, et al. Protection by BCG vaccine
OO conducted the literature search, data extraction and data synthesis, against tuberculosis: a systematic review of randomized controlled
trials. Clin Infect Dis 2014 ; 58: 470–80.
created the tables and figure, and wrote the manuscript. All authors
contributed to the interpretation of the data and revision of the 21 Mangtani P, Nguipdop-Djomo P, Keogh RH, et al. The duration
of protection of school-aged BCG vaccination in England:
manuscript.
a population-based case-control study. Int J Epidemiol 2018;
Conflicts of interests 47: 193–201.
We declare no conflict of interests. 22 Aronson NE, Santosham M, Comstock GW, et al. Long-term
efficacy of BCG vaccine in American Indians and Alaska
Acknowledgments Natives: a 60-year follow-up study. JAMA 2004; 291: 2086.
This work was supported by the European and Developing Countries 23 Rodrigues LC, Pereira SM, Cunha SS, et al. Effect of BCG
Clinical Trials Partnership (EDCTP) 2 programme under contract revaccination on incidence of tuberculosis in school-aged children
WP2018/3.7A. The EDCTP2 programme is supported by Horizon 2020, in Brazil: the BCG-REVAC cluster-randomised trial. Lancet 2005;
the EUs Framework Programme for Research and Innovation. 366: 1290–95.
References 24 Leung CC, Tam CM, Chan SL, Chan-Yeung M, Chan CK,
1 European Commission. G20 leaders’ declaration: shaping an Chang KC. Efficacy of the BCG revaccination programme in
interconnected world. Hamburg, Germany; July 7–8, 2017. a cohort given BCG vaccination at birth in Hong Kong.
http://europa.eu/rapid/press-release_STATEMENT-17-1960_en.htm Int J Tuberc Lung Dis 2001; 5: 717–23.
2 WHO. Global tuberculosis report 2019. Geneva: World Health 25 Nemes E, Geldenhuys H, Rozot V, et al. Prevention of
Organization, 2019. M tuberculosis infection with H4:IC31 vaccine or BCG
revaccination. N Engl J Med 2018; 379: 138–49.
3 Houben RM, Dodd PJ. The global burden of latent tuberculosis
infection: a re-estimation using mathematical modelling. PLoS Med 26 Nieuwenhuizen NE, Kulkarni PS, Shaligram U, et al.
2016; 13: e1002152. The recombinant Bacille Calmette–Guérin vaccine VPM1002:
ready for clinical efficacy testing. Front Immunol 2017; 8: 1147.
4 WHO. The global plan to end TB 2016-2020: the paradigm shift.
Geneva: Stop TB Partnership, United Nations Office for Project 27 Gonzalo-Asensio J, Marinova D, Martin C, Aguilo N. MTBVAC:
Services, 2015. attenuating the human pathogen of tuberculosis (TB) toward a
promising vaccine against the TB epidemic. Front Immunol 2017;
5 Chung-Delgado K, Guillen-Bravo S, Revilla-Montag A,
8: 1803.
Bernabe-Ortiz A. Mortality among MDR-TB cases: comparison with
drug-susceptible tuberculosis and associated factors. PLoS One; 28 Yang X-Y, Chen Q-F, Li Y-P, Wu S-M. Mycobacterium vaccae as
2015; 10: e0119332. adjuvant therapy to anti-tuberculosis chemotherapy in never-
treated tuberculosis patients: a meta-analysis. PLoS One 2011;
6 Marks SM, Flood J, Seaworth B, et al. Treatment practices,
6: e23826.
outcomes, and costs of multidrug-resistant and extensively
drug-resistant tuberculosis, United States, 2005–2007. 29 Saqib M, Khatri R, Singh B, Gupta A, Kumar A, Bhaskar S.
Emerg Infect Dis 2014; 20: 812–21. Mycobacterium indicus pranii as a booster vaccine enhances BCG
induced immunity and confers higher protection in animal models
7 Mullerpattan JB, Udwadia ZZ, Banka RA, Ganatra SR, Udwadia ZF.
of tuberculosis. Tuberculosis (Edinb) 2016; 101: 164–73.
Catastrophic costs of treating drug resistant TB patients in a tertiary
care hospital in India. Indian J Tuberc 2019; 66: 87–91. 30 von Reyn CF, Lahey T, Arbeit RD, et al. Safety and immunogenicity
of an inactivated whole cell tuberculosis vaccine booster in adults
8 Bloemberg GV, Keller PM, Stucki D, et al. Acquired resistance to
primed with BCG: a randomized, controlled trial of DAR-901.
bedaquiline and delamanid in therapy for tuberculosis.
PLoS One 2017; 12: e0175215.
N Engl J Med 2015; 373: 1986–88.
31 Vilaplana C, Montané E, Pinto S, et al. Double-blind, randomized,
9 Heimbeck J. Incidence of tuberculosis in young adult women,
placebo-controlled phase I clinical trial of the therapeutical
with special reference to unemployment. Brit J Tuberc 1938;
antituberculous vaccine RUTI. Vaccine 2010; 28: 1106–16.
32: 154–56.
32 Porcelli S, Morita CT, Brenner MB. CDlb restricts the response of
10 Andrews JR, Noubary F, Walensky RP, Cerda R, Losina E,
human CD4−8−T lymphocytes to a microbial antigen. Nature 1992;
Horsburgh CR. Risk of Progression to active tuberculosis following
360: 593–7.
reinfection with Mycobacterium tuberculosis. Clin Infect Dis 2012;
54: 784–91. 33 Beckman EM, Porcelli SA, Morita CT, Behar SM, Furlong ST,
Brenner MB. Recognition of a lipid antigen by CD1-restricted
11 Mahan CS, Zalwango S, Thiel BA, et al. Innate and adaptive
αβ+ T cells. Nature 1994; 372: 691–4.
immune responses during acute M tuberculosis infection in adult
household contacts in Kampala, Uganda. Am J Trop Med Hyg 2012; 34 Van Rhijn I, Moody DB. CD1 and mycobacterial lipids activate
86: 690–97. human T cells. Immunol Rev 2015; 264: 138–53.

8 www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5

 Review

35 Busch M, Herzmann C, Kallert S, et al. Lipoarabinomannan- 58 Nangpal P, Bahal RK, Tyagi AK. Boosting with recombinant MVA
responsive polycytotoxic T cells are associated with protection in expressing M tuberculosis α-crystallin antigen augments the
human tuberculosis. Am J Respir Crit Care Med 2016; 194: 345–55. protection imparted by BCG against tuberculosis in guinea pigs.
36 Spencer CT, Abate G, Blazevic A, Hoft DF. Only a subset of Sci Rep 2017; 7: 17286.
phosphoantigen-responsive γ9δ2 T cells mediate protective 59 Leung-Theung-Long S, Gouanvic M, Coupet C-A, et al. A novel
tuberculosis immunity. J Immunol 2008; 181: 4471–84. MVA-based multiphasic vaccine for prevention or treatment of
37 Commandeur S, van Meijgaarden KE, Prins C, et al. An unbiased tuberculosis induces broad and multifunctional cell-mediated
genome-wide Mycobacterium tuberculosis gene expression approach to immunity in mice and primates. PLoS One 2015; 10: e0143552.
discover antigens targeted by human T cells expressed during 60 TuBerculosis Vaccine Initiative. Accelerating TB vaccine research
pulmonary infection. J Immunol 2013; 190: 1659–71. & development through partnership, TuBerculosis Vaccine
38 Coppola M, van Meijgaarden KE, Franken KLMC, et al. New genome- Initiative annual report 2014. Lelystad: TuBerculosis Vaccine
wide algorithm identifies novel in-vivo expressed Mycobacterium Initiative, 2014.
tuberculosis antigens inducing human T-cell responses with classical 61 TuBerculosis Vaccine Initiative. Pipeline of Vaccines. http://www.
and unconventional cytokine profiles. Sci Rep 2016; 6: 37793. tbvi.eu/what-we-do/pipeline-of-vaccines/ (accessed Dec 15, 2019).
39 Aagaard C, Hoang T, Dietrich J, et al. A multistage tuberculosis 62 Kaushal D, Foreman TW, Gautam US, et al. Mucosal vaccination
vaccine that confers efficient protection before and after exposure. with attenuated Mycobacterium tuberculosis induces strong central
Nat Med 2011; 17: 189–94. memory responses and protects against tuberculosis. Nat Commun
40 Boisson-Dupuis S, Bustamante J, El-Baghdadi J, et al. Inherited and 2015; 6: 8533.
acquired immunodeficiencies underlying tuberculosis in childhood. 63 Marcus SA, Steinberg H, Talaat AM. Protection by novel vaccine
Immunol Rev 2015; 264: 103–20. candidates, Mycobacterium tuberculosis ΔmosR and ΔechA7, against
41 Lewinsohn DA, Lewinsohn DM, Scriba TJ. Polyfunctional challenge with a Mycobacterium tuberculosis Beijing strain. Vaccine
CD4+ T cells as targets for tuberculosis vaccination. Front Immunol 2015; 33: 5633–39.
2017; 8: 1262. 64 Nieuwenhuizen NE, Kaufmann SHE. Next-Generation Vaccines
42 Coscolla M, Copin R, Sutherland J, et al. M. tuberculosis T cell epitope Based on Bacille Calmette-Guérin. Front Immunol 2018; 9: 121.
analysis reveals paucity of antigenic variation and identifies rare 65 Hansen SG, Zak DE, Xu G, et al. Prevention of tuberculosis in
variable TB antigens. Cell Host Microbe 2015; 18: 538–48. rhesus macaques by a cytomegalovirus-based vaccine. Nat Med
43 Godfrey DI, Uldrich AP, McCluskey J, Rossjohn J, Moody DB. 2018; 24: 130–43.
The burgeoning family of unconventional T cells. Nat Immunol 2015; 66 Lewinsohn DA, Swarbrick GM, Park B, et al. Comprehensive
16: 1114–23. definition of human immunodominant CD8 antigens in
44 Achkar JM, Chan J, Casadevall A. B cells and antibodies in the tuberculosis. NPJ Vaccines (Basel) 2017; 2: pii 8.
defense against Mycobacterium tuberculosis infection. Immunol Rev 67 Joosten SA, van Meijgaarden KE, van Weeren PC, et al.
2015; 264: 167–81. Mycobacterium tuberculosis peptides presented by HLA-E molecules
45 Day CL, Tameris M, Mansoor N, et al. Induction and regulation of are targets for human CD8+ T-cells with cytotoxic as well as
T-cell immunity by the novel tuberculosis vaccine M72/AS01 in regulatory activity. PLoS Pathog 2010; 6: e1000782.
South African adults. Am J Res Crit Care Med 2013; 188: 492–502. 68 Lindestam Arlehamn CS, Lewinsohn D, Sette A, Lewinsohn D.
46 Penn-Nicholson A, Tameris M, Smit E, et al. Safety and Antigens for CD4 and CD8 T cells in tuberculosis.
immunogenicity of the novel tuberculosis vaccine ID93 + GLA-SE Cold Spring Harb Perspect Med 2014; 4: a018465.
in BCG-vaccinated healthy adults in South Africa: a randomised, 69 Zhu B, Dockrell HM, Ottenhoff THM, Evans TG, Zhang Y.
double-blind, placebo-controlled phase 1 trial. Lancet Respir Med 2018; Tuberculosis vaccines: opportunities and challenges: novel TB
6: 287–98. vaccines and evaluation. Respirology 2018 ; 23: 359–68.
47 Luabeya AKK, Kagina BMN, Tameris MD, et al. First-in-human trial 70 Bhatt K, Verma S, Ellner JJ, Salgame P. Quest for correlates of
of the post-exposure tuberculosis vaccine H56:IC31 in Mycobacterium protection against tuberculosis. Clin Vaccine Immunol 2015;
tuberculosis infected and non-infected healthy adults. Vaccine 2015; 22: 258–66.
33: 4130–40. 71 De Libero G, Singhal A, Lepore M, Mori L. Nonclassical T cells
48 Tkachuk AP, Gushchin VA, Potapov VD, Demidenko AV, Lunin VG, and their antigens in tuberculosis. Cold Spring Harb Perspect Med
Gintsburg AL. Multi-subunit BCG booster vaccine GamTBvac: 2014; 4: a018473.
assessment of immunogenicity and protective efficacy in murine and 72 Caccamo N, Pietra G, Sullivan LC, et al. Human CD8 T
guinea pig TB models. PLoS One 2017; 12: e0176784. lymphocytes recognize Mycobacterium tuberculosis antigens
49 Tait DR, Hatherhill M, Van Der Meeren O, et al. Final analysis of a presented by HLA-E during active tuberculosis and express type 2
trial of M72/AS01E vaccine to prevent tuberculosis. N Engl J Med cytokines. Eur J Immunol 2015; 45: 1069–81.
2019; 381: 2429–39. 73 Montamat-Sicotte DJ, Millington KA, Willcox CR, et al. A mycolic
50 Bloom BR. New promise for vaccines against tuberculosis. acid-specific CD1-restricted T cell population contributes to acute
N Engl J Med 2018; 379: 1672–74. and memory immune responses in human tuberculosis infection.
51 Ahsan MJ. Recent advances in the development of vaccines for J Clin Invest 2011; 121: 2493–503.
tuberculosis. Ther Adv Vaccines 2015; 3: 66–75. 74 Gold MC, Napier RJ, Lewinsohn DM. MR1-restricted mucosal
52 Smaill F, Xing Z. Human type 5 adenovirus-based tuberculosis associated invariant T (MAIT) cells in the immune response to
vaccine: is the respiratory route of delivery the future? Mycobacterium tuberculosis. Immunol Rev 2015; 264: 154–66.
Expert Rev Vaccines 2014; 13: 927–30. 75 Meierovics A, Yankelevich W-JC, Cowley SC. MAIT cells are critical
53 Pinschewer DD. Virally vectored vaccine delivery: medical needs, for optimal mucosal immune responses during in vivo pulmonary
mechanisms, advantages and challenges. Swiss Med Wkly 2017; bacterial infection. Proc Nat Acad Sci USA 2013; 110: e3119–28.
147: w14465. 76 Huang D, Chen CY, Zhang M, et al. Clonal immune
54 Ellis RD, Hatherill M, Tait D, et al. Innovative clinical trial designs to responses of Mycobacterium-specific γδ T cells in tuberculous and
rationalize TB vaccine development. Tuberculosis (Edinb) 2015; non-tuberculous tissues during M. tuberculosis infection. PLoS One
95: 352–57. 2012; 7: e30631.
55 Hesseling AC, Walzl G, Enarson DA, et al. Baseline sputum time 77 Kawahara JY, Irvine EB, Alter G. A case for antibodies as
to detection predicts month two culture conversion and relapse in mechanistic correlates of immunity in tuberculosis. Front Immunol
non-HIV-infected patients. Int J Tuberc Lung Dis 2010; 14: 560–70.
 2019; 10: 996.
56 Jeyanathan M, Thanthrige-Don N, Afkhami S, et al. Novel 78 Garcia-Contreras L, Wong Y-L, Muttil P, et al. Immunization by a
chimpanzee adenovirus-vectored respiratory mucosal tuberculosis bacterial aerosol. Proc Nat Acad Sci USA 2008; 105: 4656–60.
vaccine: overcoming local anti-human adenovirus immunity for 79 Manjaly Thomas Z-R, Satti I, Marshall JL, et al. Alternate aerosol
potent TB protection. Mucosal Immunol 2015; 8: 1373–87. and systemic immunisation with a recombinant viral vector for
57 Chen Z, Gupta T, Xu P, et al. Efficacy of parainfluenza virus 5 tuberculosis, MVA85A: a phase I randomised controlled trial.
(PIV5)-based tuberculosis vaccines in mice. Vaccine 2015; 33: 7217–24. PLoS Med 2019; 16: e1002790.

www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5 9

 Review

80 Manjaly Thomas Z-R, McShane H. Aerosol immunisation for TB: 92 Coleman MT, Maiello P, Tomko J, et al. Early changes by
matching route of vaccination to route of infection. 18
fluorodeoxyglucose positron emission tomography coregistered
Trans R Soc Trop Med Hyg 2015; 109: 175–81. with computed tomography predict outcome after Mycobacterium
81 Anacker RL, Brehmer W, Barclay WR, et al. Superiority of tuberculosis infection in cynomolgus macaques. Infect Immun 2014;
intravenously administered BCG and BCG cell walls in protecting 82: 2400–04.
rhesus monkeys (Macaca mulatta) against airborne tuberculosis. 93 Laddy DJ, Bonavia A, Hanekom WA, et al. Toward tuberculosis
Z Immunitatsforsch Exp Klin Immunol 1972; 143: 363–76. vaccine development: recommendations for nonhuman primate
82 Sharpe S, White A, Sarfas C, et al. Alternative BCG delivery study design. Infect Immun 2017; 86: e00776–17.
strategies improve protection against Mycobacterium tuberculosis in 94 Roestenberg M, Mo A, Kremsner PG, Yazdanbakhsh M. Controlled
non-human primates: protection associated with mycobacterial human infections: a report from the controlled human infection
antigen-specific CD4 effector memory T-cell populations. models workshop, Leiden University Medical Centre 46 May 2016.
Tuberculosis (Edinb) 2016; 101: 174–90. Vaccine 2017; 35: 7070–76.
83 Lu S. Heterologous prime-boost vaccination. Curr Opin Immunol 95 Minassian AM, Satti I, Poulton ID, Meyer J, Hill AVS, McShane H.
2009; 21: 346–51. A human challenge model for Mycobacterium tuberculosis using
84 Kaufmann SHE, Weiner J 3rd, Maertzdorf J. Accelerating Mycobacterium bovis bacille Calmette-Guerin. J Infect Dis 2012;
tuberculosis vaccine trials with diagnostic and prognostic 205: 1035–42.
biomarkers. Expert Rev Vaccines 2017; 16: 845–53. 96 UN General Assembly. Resolution A/RES/73/3 of the UN General
85 The GC6–74 consortium, Weiner J, Maertzdorf J, et al. Metabolite Assembly of 10 October 2018: political declaration of the high-level
changes in blood predict the onset of tuberculosis. Nat Commun meeting of the General Assembly on the fight against tuberculosis.
2018; 9: 5208. New York, NY: UN General Assembly, 2018.
86 Sharpe S, White A, Gleeson F, et al. Ultra low dose aerosol 97 Harris RC, Sumner T, Knight GM, White RG. Systematic review of
challenge with Mycobacterium tuberculosis leads to divergent mathematical models exploring the epidemiological impact of
outcomes in rhesus and cynomolgus macaques. Tuberculosis (Edinb) future TB vaccines. Hum Vaccin Immunother 2016; 12: 2813–32.
2016; 96: 1–12. 98 Knight GM, Griffiths UK, Sumner T, et al. Impact and
87 Dijkman K, Sombroek CC, Vervenne RAW, et al. Prevention of cost-effectiveness of new tuberculosis vaccines in low- and
tuberculosis infection and disease by local BCG in repeatedly middle-income countries. Proc Natl Acad Sci USA 2014; 111: 15520–25.
exposed rhesus macaques. Nat Med 2019; 25: 255–62. 99 Barker L, Hessel L, Walker B. Rational approach to selection and
88 McShane H, Williams A. A review of preclinical animal models clinical development of TB vaccine candidates. Tuberculosis (Edinb)
utilised for TB vaccine evaluation in the context of recent human 2012; 92 (suppl 1): S25–29.
efficacy data. Tuberculosis (Edinb) 2014; 94: 105–10. 100 Schrager LK, Chandrasekaran P, Fritzell BH, et al. WHO preferred
89 Clark S, Hall Y, Williams A. Animal models of tuberculosis: product characteristics for new vaccines against tuberculosis.
guinea pigs. Cold Spring Harb Perspect Med 2015; 5: a018572. Lancet Infect Dis 2018; 18: 828–29.
90 Scanga CA, Flynn JL. Modeling tuberculosis in nonhuman
primates. Cold Spring Harb Perspect Med 2014; 4: a018564. © 2020 Elsevier Ltd. All rights reserved.
91 Maiello P, DiFazio RM, Cadena AM, et al. Rhesus macaques are
more susceptible to progressive tuberculosis than cynomolgus
macaques: a quantitative comparison. Infect Immun 2018;
86: e00505-17.

10 www.thelancet.com/infection Published online January 31, 2020 https://doi.org/10.1016/S1473-3099(19)30625-5