Exploring the role of IL-6 in the risk of developing TB disease

The role played by universal IL-6 responses in determining human tuberculosis (TB) disease risk is unknown. We used single nucleotide polymorphisms (SNPs) in and near the IL-6 receptor (IL6R) gene, including the non-synonymous variant, rs2228145, for which the C allele contributes to reduced classical (cis) IL-6 signalling activity, to test the hypothesis that altered IL-6 signalling is causally associated with the risk of developing TB disease. 

Methods: We identified 17 genome wide association studies (GWAS) of TB disease. We performed multiancestry, two-sample Mendelian randomisation (MR) analyses to estimate the causal effect of reduced IL-6 signalling on TB disease. We used rs2228145-C as a genetic instrument, weighted on CRP reduction as a measure of the effect on IL-6 signalling, and also ancestry-specific, multiple SNP approach using IL6R plasma protein as an exposure.

Findings: From 17 GWAS we collated 19,302 cases of TB disease and 1,019,821 controls across multiple ancestries. Multi-ancestry MR analyses revealed the rs2228145-C allele decreased the odds of TB disease with OR of 0.52 (95% CI 0.39 - 0.69, p = 6.8 x 10-6) for each natural log CRP decrease, showing only weak heterogeneity (I2 = 0.315, p = 0.11). Ancestry-specific, multiple SNP MR using increase in IL6R plasma protein as an exposure revealed a similar reduced TB disease risk (OR 0.94; 95% CI 0.930.96), p = 2.4 x 10-10). The protective effects with rs2228145-C were comparable in size and direction to those in critical COVID-19, Crohn’s disease and rheumatoid arthritis in which IL-6 antagonism has beneficial therapeutic effects.

Interpretation: Our findings propose a causal relationship between reduced IL-6 signalling and lower risk of TB disease, akin to that seen in other IL-6 mediated diseases. This suggests IL-6 antagonists do not increase the risk of TB disease but rather should be investigated as therapeutic adjuncts in its treatment.