Identifying Novel Tuberculosis Endotypes Based on Gene Expression Data

An analysis of transcriptome and protein data from tuberculosis (TB) patients supported biologically characteristic TB endotypes that mediate clinical outcomes, according to the results of a preclinical study published in the European Respiratory Journal. The current study involved unbiased clustering to elucidate distinct TB endotypes with classifiable gene expression patterns and clinical results.

To identify TB endotypes, researchers conducted a systematic review and meta-analysis to find microarray gene expression data derived from tuberculosis patients for analysis. Researchers formed a discovery cohort that included 7 studies with data on 435 patients with tuberculosis and 533 asymptomatic individuals as controls. The researchers retained 1 microarray cohort with longitudinal clinical outcomes for purposes of validation, as well as 2 RNA-seq cohorts. The investigators analyzed a separate cohort of tuberculosis patients with functional immune responses to elucidate stimulated vs unstimulated immune responses.

Analysis of the discovery cohort resulted in the identification of 2 tuberculosis endotypes. Endotype A involved higher expression of genes related to inflammation and immunity and lower metabolism and proliferation, whereas endotype B exhibited higher activity of metabolism and proliferation pathways. The independent RNA-seq validation cohort represented 118 patients with tuberculosis and 179 asymptomatic individuals and confirmed results from the discovery cohort.

For endotype A, gene expression signatures for treatment failure were higher in the discovery cohort. Endotype A patients exhibited slower time to culture conversion and a decreased cure rate on further testing. The authors noted that these endotypes mirror functional immunity, which is buoyed by the observation that TB patients with a hyperinflammatory endotype exhibit decreased responsive cytokine production on stimulation.

TB is not a monomorphic disease, with protean host response involving multiple distinct molecular pathways and pathologies (ie, endotypes), the authors noted. Host, pathogen, and environmental factors all play a role in the pathogenesis of TB, although limited data regarding such interactions are available. Variables contributing to differing endotypes could include malnutrition, HIV status, helminths, tobacco use, or fuel exposure.

“These observations suggest that different endotypes display responses that are likely to have clinical and pathologic relevance and provides the basis for studies to evaluate endotype-specific host-directed therapies,” stated the authors.

Reference

DiNardo AR, Gandhi T, Heyckendorf J, et al. Gene expression signatures identify biologically and clinically distinct tuberculosis endotypes. Eur Respir J. Published online February 15, 2022. doi:10.1183/13993003.02263-2021