How can immunology contribute to the control of tuberculosis?

Tuberculosis presents a significant health threat to the world population, with 8 million new cases of disease and 2 million deaths per year. The aetiological agent, Mycobacterium tuberculosis, remains well controlled in most of the 2 billion infected people worldwide, and disease develops in only a minority. Typically, disease develops years, or even decades, following initial infection, due an exogenous insult, intrinsic genetic susceptibility or a combination of both. The currently used vaccine, BCG, fails to provide complete protection against the most common form of the disease, namely pulmonary tuberculosis in adults. Understanding the mechanisms that are induced, neglected or inhibited by vaccination with BCG or natural infection with M. tuberculosis provides guidelines for the rational design of a more efficacious vaccine. T cells are responsible for protection against tuberculosis, and hence represent the main targets of any vaccination strategy. The entire T-cell compartment, consisting of conventional CD4 and CD8 T cells as well as non-conventional T cells, such as CD1-restricted T cells and γδ T cells, participates in optimal protection. TH1 cytokines, notably interferon-γ, are critical but insufficient for protection, which also depends on cytolytic mechanisms. Currently, two main vaccination strategies are being pursued. The first strategy uses subunit vaccines in the form of protein–adjuvant formulations, naked DNA, or recombinant bacterial or viral carriers that express defined antigens. Promising results have been obtained, but so far no vaccine candidate tested in animal models has proven to be better than BCG. The second strategy, comprising viable mycobacterial vaccines, either attenuated viable M. tuberculosis or BCG, or recombinant BCG overexpressing certain antigens or immumodulators, is also being pursued and shows promise. At present, it is unclear which vaccine candidate is most suitable to tackle the problem, and it is possible that a combination of different vaccines administered in a heterologous prime-boost schedule would prove best. Different vaccination strategies should be pursued and tested in standardized animal models of mice, guinea pigs and non-human primates. As human trials are being initiated, the identification of correlates and surrogates of protection will gain importance, as so far no single marker bears the characteristic signature of protection. With the information from the genome of M. tuberculosis, as well as the deeper insights into the immune response in protection and disease, prospects of developing a successful vaccine are improving.