TB vaccines: the paradigms they are a-shifting

Traditional roadblocks to effective TB control (e.g., lack of attention to the disease, lack of resources, antiquated diagnostic tests, lack of patient compliance with therapy and a huge reservoir of persistently infected individuals) have been compounded by the terrible synergy between TB and HIV, and the development of almost untreatable multi- and extensively drug-resistant strains of Mycobacterium tuberculosis[1]. Case-finding and chemotherapy employing the Directly Observed Treatment, Short-course strategy, and several promising new TB drugs currently in the pipeline, will undoubtedly continue to have a significant impact on TB incidence [2,3]. However, it is clear that an effective vaccine is essential if TB is to be controlled as rapidly as possible [4–6]. The urgency to license a new TB vaccine has driven a worldwide effort over the past 20 years involving basic scientists, public and private funding agencies, regulatory bodies, and disease control experts in high-burden countries [7]. That effort has been shaped at various stages by several paradigms that reflected the current understanding of the type of vaccine that would be needed, the manner in which vaccines should be evaluated for efficacy in animal models, the process by which preclinical development of vaccines would occur, the design of clinical trials, and the public and private organizations that would participate in the process. Almost without exception, those paradigms have shifted as the understanding of vaccine-induced protection has improved, an appreciation for the biological relevance of certain animal models of TB has grown, new strategies for organizing and funding preclinical development and human trials have emerged, and the realities of testing and using TB vaccines in diverse high-burden populations have become apparent.

Variety is the spice of life

In general, the induction of protective cell-mediated immunity is much more challenging than inducing a protective antibody response, a fact which has complicated the search for a better TB vaccine. The current vaccine, bacillus Calmette–Guérin (BCG), is a living, attenuated strain (actually several different strains) of Mycobacterium bovis, which has served as a model for new TB vaccines and a gold standard for their evaluation in animal models [8]. BCG has many characteristics that recommend it as a platform for new TB vaccines, and several of the promising new vaccines utilize recombinant BCG strains overexpressing the genes for several mycobacterial antigens [9,10]. However, the fact that BCG does not seem to protect against adult pulmonary TB in some high-burden settings has stimulated the search for other types of TB vaccines. The traditional approach of using purified components of the bacterium together with an effective adjuvant has been modernized by cloning the genes for target antigens and either expressing those genes in an expression vector (e.g., vaccinia virus and adenovirus) or using the DNA directly [11,12]. In fact, the variety of TB vaccine types that have been taken into the clinic in the past few years is impressive and reflects the realization that the unknown factors that impair BCG efficacy may also render less effective new TB vaccines that depend upon BCG as a delivery vehicle and are used as ‘stand-alone’ vaccines [13].

One size fits all

While the TB control community would be ecstatic with any effective new TB vaccine, it has become clear that different vaccines will be required to prevent different forms of the disease in distinct populations. We can no longer talk about a search for a new TB vaccine, but rather for several new TB vaccines with fundamentally different indications. Thus, vaccines are needed to protect against primary, pulmonary TB in immunologically naive hosts, to reduce the risk of reactivation disease in a postexposure situation, or to be used as an adjunct to therapy in patients infected with multidrug-resistant strains [14,15]. One of the biggest challenges for TB vaccine development is a vaccine which works and is safe in HIV-infected individuals. It is thought that a TB vaccine might work in HIV-positive individuals with a normal CD4⁺ T-cell count who are receiving effective antiretroviral therapy. However, it is likely that BCG and similarly live-attenuated bacterial vaccines will not be used in HIV-positive individuals due to the risk of vaccinosis [16]. While the reliable testing and determination of the HIV status of potential vaccinees would allow the identification of those individuals in whom a living, BCG-like vaccine could be applied safely, conditions in countries with a very high rate of TB and HIV coinfection currently preclude such testing. Therefore, the same, completely safe vaccine would have to be used in all vaccinees. Such a vaccine will probably be a nonviable antigenic preparation with an adjuvant, or a stable, highly attenuated, nonreplicating mycobacterial strain [17]. While the prevention of TB in HIV-positive individuals is obviously a high priority for public-health officials worldwide, it would be a shame if the HIV-negative segment of the population was denied access to a living attenuated vaccine, which may well prove to be more effective.

To boost or not to boost – that is the question

The history of BCG use as a prophylactic vaccine administered once to immunologically naive individuals (newborns or infants in most countries) has had a major influence on the perception about how new TB vaccines would be used. Since dogma holds that a single dose of live-attenuated vaccines such as BCG will induce lifelong immunity, little consideration was given to the possible benefit of booster doses. This bias was compounded by the observation that repeated vaccination with BCG itself appeared to afford no added protection in countries that employed a boosting strategy [18]. However, as the possibility of testing new TB vaccines in humans became a reality, it was apparent that these vaccines would be tested by necessity in individuals who had already been vaccinated with BCG. This fact drove a major change in the strategy of TB vaccine development and testing. The concept of enhancing immunity induced by priming with BCG or some recombinant version of BCG, and subsequently boosting with a protein or virally vectored vaccine has become a major new paradigm in TB vaccination. In fact, several of the novel TB vaccine candidates currently in human trials are intended to be used in a prime–boost strategy [7,19].

When is a model not a model?

Since there are no validated immunological correlates of vaccine-induced resistance in TB, any new TB vaccine must be tested for protective efficacy in an animal model. Historically, much of the testing was conducted in mice infected by injection (often intravenously) with massive doses of virulent mycobacteria. Nearly 20 years ago, it was recognized that one of the bottlenecks in the TB vaccine development pipeline was the availability of laboratories with the biohazard facilities and expertise to test the protective efficacy of large numbers of novel vaccines. That bottleneck was effectively circumvented by the development of a US NIH-funded contract to test vaccines without cost for any investigator in the world. Since the inception of that testing service and a similar vaccine screening program funded by the EU TB Vaccine Cluster, hundreds of vaccine candidates have been tested in mice and guinea pigs [20,21]. During that time, four important paradigm shifts have occurred. First, the use of low-dose aerosol infection has become the standard for vaccine evaluation in animals. Second, the unique contributions of the guinea pig as a model of critical aspects of the disease (e.g., human-like granulomas) have been recognized. Third, the value of experimental readouts of disease prevention (e.g., long-term survival and reduced histopathology), in addition to control of the infection (e.g., reduced bacterial loads in lungs and spleens), has been appreciated. Fourth, while the vast majority of new TB vaccine candidates have been evaluated prophylactically in immunologically naive animals, there is a shift toward the development of improved models in which prime–boost strategies and postexposure regimens can be tested [22–24]. Ironically, one of the most dramatic paradigm shifts in the use of animal models for TB vaccine testing would be to eliminate the reliance on protection data from models in the selection of vaccines for human trials. Proponents of such a shift conclude that existing models are imperfect and may not predict the performance of vaccines in humans and, therefore, support a more empirical approach.

Bridging the gap between bench & bedside

Nearly 10 years ago, Carol Nacy and others recognized that another important bottleneck in the TB vaccine development pipeline was the gap between the basic scientists developing new vaccines at the bench, and the vaccine companies that would ultimately produce and market the new vaccines. The scientists were not in a position to take on the challenges of preclinical development, including issues such as manufacturing test lots in bulk under Good Laboratory Practice conditions, interacting with regulatory agencies, negotiating with private-sector partners, and so forth. On the other hand, companies would not come to the table until a new vaccine began to look like a product. Nacy conceived of a not-for-profit foundation – the Sequella (now Aeras) Global TB Vaccine Foundation – which would serve as a bridge from bench science to the commercialization of TB vaccines. With generous grants from the Bill and Melinda Gates Foundation, Aeras has taken a portfolio of promising TB vaccines into Phase I and II clinical testing, and is developing field trial sites for Phase III trials [7]. Major contributions from philanthropic organizations and the advent of Product Development Partnerships, such as Aeras represent important paradigm shifts in the strategy for moving promising new TB vaccines from the bench to the bedside.

Future paradigm shifts

Amazing progress has been made in the development and testing of new TB vaccines, and it is likely that one or more will be licensed in the next 8–10 years. However, there are some current paradigms that may need to shift in the next few years in order to ensure that the pipeline of vaccine candidates remains full and flowing. First, while several different types of TB vaccines are currently under investigation, most rely upon the same, very small number of, mycobacterial antigens (e.g., Ag85, ESAT-6 and Mtb10.4, among others). Unless there is a shift in favor of new antigen discovery, the potential failure of those vaccines/antigens will hinder the development of second-generation vaccines. Second, the paradigm of TB vaccination to prevent new clinical cases (and reduce transmission) in recently infected individuals ignores the even greater need to use vaccination to prevent reactivation disease in 2 billion persistently infected individuals. The evaluation of such vaccines will require the development of valid animal models of persistence and reactivation. Finally, while a mass vaccination paradigm using the same TB vaccine for everyone is obviously highly desirable from the public health programmatic perspective, it overlooks the potential benefit of targeting specific subpopulations (e.g., HIV-positive and -negative individuals) with different TB vaccines in an attempt to provide optimum protection for everyone.

Financial & competing interests disclosure

David McMurray is a member of the Board of Directors of the Aeras Global TB Vaccine Foundation. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.