Recent advances towards tuberculosis control: vaccines and biomarkers

Abstract

Of all infectious diseases, tuberculosis (TB) remains one of the most important causes of morbidity and mortality. Recent advances in understanding the biology of Mycobacterium tuberculosis (Mtb) infection and the immune response of the infected host have led to the development of several new vaccines, a number of which are already undergoing clinical trials. These include pre-exposure prime vaccines, which could replace bacille Calmette-Guérin (BCG), and pre-exposure booster vaccines given in addition to BCG. Infants are the target population of these two types of vaccines. In addition, several postexposure vaccines given during adolescence or adult life, in addition to BCG as a priming vaccine during infancy, are undergoing clinical testing. Therapeutic vaccines are currently being assessed for their potential to cure active TB as an adjunct to chemotherapy. BCG replacement vaccines are viable recombinant BCG or double-deletion mutants of Mtb. All booster vaccines are composed of one or several antigens, either expressed by viral vectors or formulated with adjuvants. Therapeutic vaccines are killed mycobacterial preparations. Finally, multivariate biomarkers and biosignatures are being generated from high-throughput data with the aim of providing better diagnostic tools to specifically determine TB progression. Here, we provide a technical overview of these recent developments as well of the relevant computational approaches and highlight the obstacles that still need to be overcome.

Keywords: bacille Calmette-Guérin; biomarker; biosignature; tuberculosis; vaccine.

Figure 1

Overview of different vaccine types. Advanced vaccine candidates for different target populations, with stage of vaccine administration.

Figure 2

Classification of serum samples (patients with TB and control subjects) using independent predictors. Combining a predictor with an error rate of 22% with a correlated predictor results in a higher joined error rate (13%) than combining the same predictor with an inferior predictor which is not as strongly correlated (joined error rate 8%). Data from Ref. .

Figure 3

Number of variables selected as predictors versus error rate in a classification task. The error rate decreases with the number of variables that have been chosen for classification. However, adding more than 25 variables to the biosignature does not result in a further improvement of the error rate. Figure from Ref. .