Genetic Susceptibility to Tuberculosis and the Utility of Polygenic Scores in Population Stratification.

Tuberculosis (TB) is one of the leading infectious causes of mortality worldwide. Although a significant proportion of the population (up to 36%, depending on the region) is infected with the latent form of TB, only about one in ten of these people will develop an active form of the disease in their lifetime. This is due to a complex interaction between the host's genetic predisposition and environment. However, the genetic determinants of TB are not well established and have been insufficiently explored in previous genome-wide association studies (GWAS) with sparse and incongruent results. We reviewed recent evidence on host genetic susceptibility to TB, highlighting population-specific characteristics, host-pathogen coevolution, and the limitations of conventional GWAS approaches in terms of clinical and genetic heterogeneity. While rare variants with high penetrance, such as TYK2 P1104A, lead to monogenic susceptibility, most heritable risk results from the cumulative effect of numerous common variants. This cumulative effect may be summarized using polygenic risk scores (PRSs). Although their use has been proven for non-communicable diseases, PRSs are not applied to infectious disease susceptibility. To date, no PRS model for susceptibility to tuberculosis has been consistently validated. The development of PRSs for TB susceptibility is limited by phenotypic heterogeneity, population structure, and co-adaptation between host and pathogen. Another major challenge is to take into account the considerable influence of environmental factors. This difficulty in modeling environmental influences probably explains the current lack of a clinically applicable PRS for TB susceptibility. However, taking these caveats into account, polygenic models could improve risk stratification at the individual level compared to single-variant association and allow for earlier targeted treatment and prophylaxis.