The current state of tuberculosis vaccines

Abstract

Tuberculosis continues to persist despite widespread use of BCG, the only licensed vaccine to prevent TB. BCG's limited efficacy coupled with the emergence of drug-resistant strains of Mycobacterium tuberculosis emphasizes the need for a more effective vaccine for combatting this disease. However, the development of a TB vaccine is hindered by the lack of immune correlates, suboptimal animal models, and limited funding. An adolescent/adult vaccine would have the greatest public health impact, but effective delivery of such a vaccine will require a better understanding of global TB epidemiology, improved infrastructure, and engagement of public health leaders and global manufacturers. Here we discuss the current state of tuberculosis vaccine research and development, including our understanding of the underlying immunology as well as the challenges and opportunities that may hinder or facilitate the development of a new and efficacious vaccine.

Introduction

The global incidence rate of tuberculosis (TB) has been slowly diminishing over the past few years.¹ Despite the progress that has been made, Mycobacterium tuberculosis (Mtb) is second only to HIV in deaths caused by a single pathogen worldwide.¹ TB continues to persist even though Bacille Calmette-Guérin (BCG), an attenuated strain of Mycobacterium bovis and the only licensed vaccine for TB, is one of the most widely administered vaccines in the world.^1,2 In 2011 there were an estimated 8.7 million new cases of TB and 1.4 million deaths due to TB.¹ Furthermore, TB is the leading cause of death for people living with HIV.³ Multidrug-resistant (MDR), extensively drug-resistant (XDR) and “totally” drug-resistant (TDR) strains of Mtb have emerged, making the difficult task of treating infected individuals even harder.^1,4-9

There are several ways in which a vaccine for TB could be effective at controlling disease. The most obvious of these would be to prevent initial infection. However, because only ~10% of infected individuals develop active disease and many of those do so years or even decades after infection, a vaccine to prevent active disease in those already infected would have the greatest impact in the shortest time on global health.¹⁰ In addition, therapeutic vaccines designed to shorten the course of chemotherapy for active TB or to increase the efficacy of MDR/XDR/TDR treatment could help to increase adherence to treatment regimens, decrease the high public health and personal patient costs associated with treatment, and help prevent emergence of further drug-resistance. While these opportunities exist, the path forward is not entirely clear and there are challenges that hinder vaccine development.

BCG was isolated nearly a century ago, and has spread worldwide through laboratory transfer, diversifying through separate culture conditions and manufacturing processes. There are now recognized to be at least three major lineages with significant genetic disparity, which may at least partially account for the extreme variability in efficacy (0–80%) that is reported.^11-14 This clinical variability could also be influenced by exposure to non-tuberculous mycobacteria, force of infection with Mtb, population genetics, and the impact of poverty and nutrition.^15-17 BCG is a complex organism (~4,000 genes) that includes proteins, lipids, and glycolipids capable of stimulating and modifying immune responses.¹⁸ While BCG has some protective efficacy on certain forms of childhood TB, the vaccine is not recommended for use in HIV-infected infants due to safety concerns.¹⁹ It is clear that a safe and more effective vaccine is needed to protect HIV-positive infants as well as to control the most common form of the disease, which is adult pulmonary TB.

TB Immunology

Most people exposed to Mtb fail to develop disease.¹⁰ Disruption of immune function, either by drugs,²⁰ therapeutic monoclonal antibodies,^21,22 or other disease,³ increases the likelihood of active TB disease, suggesting that the immune system is responsible for the control of the infection and prevention of active disease in the majority of infected individuals. However, generation of host immune responses to fight TB is not without risk. It has become clear that the pathology associated with tuberculosis is at least in part due to host inflammatory immune responses to the pathogen, which can result in damage to the infected tissues.^23-25 While this may be acceptable collateral damage for most infections, in the setting of the lungs this response can be highly detrimental to the host. Identification of vaccines that can selectively improve immune responses to Mtb infection and limit lung damage will be incredibly challenging.

One of the major challenges in developing an improved vaccine is the lack of known immune correlates for either protective immunity or control of infection. While immune responses to BCG can be detected, it is unclear which responses are responsible for the limited efficacy observed with this vaccine. Fortunately, there are immunological clues in humans and animals that provide some insight into what type of immune responses may be necessary for combating this disease. Although IFN-γ secretion does not appear to be correlated with TB prevention,^26,27 Th1-biased immune responses, characterized by IFN-γ and TNF secretion by T cells, are believed to be pivotal for the control of infection. Clinical observations reveal that patients with genetic defects in IL-12 or IFN-γ signaling pathways, which are critical for development of Th1 immune responses, are more susceptible to mycobacterial infections, including disseminated infection with BCG following vaccination.^28-31 In addition, anti-TNF treatment for autoimmunity also leads to an increased likelihood of active TB disease in those already latently infected with Mtb.^21,22 Furthermore, CD4+ T cells appear to be important for control of Mtb infection, as HIV/Mtb coinfection greatly increases the risk for development of active disease.^3,32,33 Animal models, while imperfect, have also provided insight into immunological mechanisms of control. Mice are useful for examining specific pathways that can be either blocked or knocked out. The mouse model has supported the clinical observations of the importance of Th1 immunity.³⁴ Additionally, CD8 depletion in nonhuman primates (NHPs) results in a loss of control of infection.³⁵ Together, the data suggest that both CD4+ and CD8+ T cells are helpful in controlling infection and that IFN-γ and TNF are necessary for control. However, BCG vaccination readily induces Th1 immunity, yet has failed in many settings and individuals to provide protection from pulmonary disease when given in infancy or when used to revaccinate adolescents.³⁶ These data suggest that either BCG is directing responses against epitopes that are not beneficial for control of Mtb or that the secretion of IFN-γ and TNF alone is not sufficient for disease prevention. One interesting theory proposes that the Th1 immune response stimulated by BCG is deficient, possibly corrupted by the bacteria to avoid immunological control.³⁷ What is clear is that novel vaccine approaches are needed to combat this disease.

Vaccine Platforms

Since BCG was first developed, there have been numerous advances in immunology, allowing for the design of more sophisticated and safer vaccines. New vaccine platforms are under development, including modification of the existing vaccine, BCG, as well as development of modern platforms such as recombinant proteins, novel adjuvants, recombinant viruses, and DNA. These approaches may be able to elicit a more appropriate immune response or direct the response against more suitable targets for the control or prevention of TB.

Progress toward the development of TB vaccines has been met with some disappointment. The most advanced clinical candidate vaccine is the MVA85A vaccine, which was first developed at Oxford University. This vaccine is an attenuated modified vaccinia Ankara (MVA) that expresses the Mtb antigen (Ag) 85A and was found to be safe and immunogenic in humans, generating CD4+ T cell responses to the encoded Ag85A.^38,39 This candidate entered a phase IIB efficacy trial in 2009. Despite indications of efficacy in preclinical models, and generation of modest CD4/TH1 responses, MVA85A was found to have no statistically significant impact on efficacy in BCG-vaccinated infants.^40-42 Another recent disappointment was seen with the vaccine candidate, AERAS-422 - a recombinant BCG vaccine that overexpressed three Mtb antigens (Ag85A, Ag85B, and Rv3407) and expressed perfringolysin O from C. perfringens to allow antigens to escape the endosomal compartment to enhance presentation of antigens to CD8+ T cells.^43-45 While AERAS-422 was found to be safe and immunogenic in animal models, further clinical development has been stopped as a consequence of a safety signal seen in the first human safety trial; administration at high dose to young adults was followed 60 or more days later by shingles in some individuals.^46,47

The Path Forward

HIV research and funding has driven the development of better tools for the stimulation and detection of Th1 immune responses as measured by IFN-γ, IL-2, and TNF. However, reagents for measuring other responses are often not as reliable or robust. It is incredibly important for the TB vaccine field, and the larger field of immunology, to expand our understanding of not only other types of responses but also of other cytokines and functions that may accompany the traditional Th1 response. For example, IL-17 and IL-22 have both been shown to play a role in protective immune responses against TB^48-51 and their inclusion in complex immunological assays may improve our knowledge of the disease and the key immune responses in TB. In general, the breadth of the measurement of immune responses must be increased. This may be partially accomplished through the incorporation of gene arrays into the analysis of immune responses to vaccination to capture both innate and adaptive immune signatures missed by standard immune assays. Furthermore, while antibodies have largely been dismissed in the TB vaccine field, there are some data to suggest that they may play a role in protection, although these observations are controversial.^52,53 In order to evaluate the different immune responses that are possible, the vaccine development pipelines must be augmented with novel platforms with the capacity to stimulate unique immune signatures. DNA vaccines offer an exciting opportunity in this area because of the ability to encode or co-deliver molecular adjuvants.^54,55

Improvements are also needed in other areas. In the absence of an immune correlate, assays that can help to screen for anti-mycobacterial activity may be useful for screening vaccine candidates. The field of malaria vaccines has benefitted from a human challenge model.^56-58 While a human challenge model using virulent Mtb is unrealistic, suitably attenuated and marked Mtb or other mycobacteria such as BCG may be suitable surrogates. Exciting work has been performed in developing a BCG, skin-based human challenge assay that, in time, may prove to facilitate TB vaccine selection.^59,60 There is also a desperate need for more predictive animal challenge models for TB. Current models rely on the deposition of large amounts of bacteria into the lungs of vaccinated animals. This large number of bacteria is not physiologically realistic and may well overwhelm a protective immune response that may be present. A standardized natural transmission model would be incredibly useful, particularly for measuring responses that may block transmission of Mtb. One model that seeks to address this issue is the Riley guinea pig model, which was developed decades ago and exposes animals to air ventilated from hospital TB wards.⁶¹ Efforts are currently underway to re-establish this model for evaluating vaccines and immune responses that may prevent transmission.⁶¹ There has also been movement toward low-dose challenge, similar to what has been done in the HIV vaccine field. Researchers at the University of Pittsburgh have been developing a low-dose challenge model using cynomolgus macaques that results in latent infection in approximately 50% of the animals.⁶² However, even with low-dose challenge, the number of bacteria that is delivered likely exceeds what is seen during a natural human infection. Furthermore, the expense of challenge models with Mtb, which require an ABSL-3 or greater facility, is substantial. Costs increase dramatically when using nonhuman primates, where a single vaccine challenge study can cost over a million US dollars, yet these animals will likely continue to be important tools in developing predictors of human responses.

In addition to these substantial research challenges, there will be equally or even more daunting challenges to meet in ensuring that an improved vaccine to prevent pulmonary TB will be affordable, adopted, and adequately deployed, particularly in high burden settings. A recent study by Knight and colleagues at the London School of Hygiene and Tropical Medicine demonstrated that a 60% efficacious vaccine with a ten year duration of protection, administered to only 20% of the adolescents and adults in mass campaigns every ten years, could be cost-effective in the 22 TB high burden countries and avert on the order of 50 million cases and 5 million deaths from TB in the first 25 y of its use.⁶³ Much more work needs to be done to inform and persuade funders, donors, the public health community, and global and country-level decision makers that an adolescent/adult TB vaccine, once available, can and must be adopted and delivered in a strategic manner to maximize public health impact. The necessary information will include more precise assessments of vaccine costs based on specific candidates, and accurate estimations of both market size and potential public health impact on a global and on a country-by-country basis. Effective vaccine delivery will also require improved knowledge of TB epidemiology in many parts of the world to ensure appropriate vaccination strategies are planned and adequate infrastructure created for vaccine delivery. Such information is essential to ensure engagement of an appropriate number and global distribution of manufacturers so that there will be adequate vaccine supply to meet demand.

Despite the formidable challenges ahead, there are reasons to be optimistic. The global pipeline of vaccines in clinical trials is relatively robust compared with any previous period in history, and includes protein/adjuvant combinations, recombinant BCGs, recombinant viral vectors, attenuated Mtb strains, and mycobacterial extracts (Fig. 1). Furthermore 2nd generation approaches, including nucleic acid-based (DNA and RNA) vaccines are being evaluated preclinically by Aeras, the Tuberculosis Vaccine Initiative, academic and pharmaceutical partners, and others. However, it is becoming increasingly clear that, given the lack of immune correlates and the current state of funding, additional approaches and increased collaboration are needed to speed the development of a new and effective vaccine for TB. There has been considerable movement in the TB field toward cooperation and sharing of resources, within the field and with HIV research organizations, such as the HVTN (HIV Vaccine Trials Network) and IMPAACT networks supported by NIAID, NIH, and IAVI (International AIDS Vaccine Initiative), in an effort to reduce the need for new clinical trial sites and to minimize redundancies. These efforts will enable both fields to move forward in a more economical manner and allow increased knowledge sharing between the organizations. The coming decade should be a time of extraordinary progress.

Figure 1. Global TB vaccine pipeline. TB vaccine candidates currently in clinical trials include viral vectors (yellow), recombinant BCG (green), protein/adjuvant combinations (purple), attenuated M.tb (red) and non-BCG mycobacteria or mycobacterial extracts (orange). Trials sponsored at least in part by Aeras are indicated.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.