Using immune clusters for classifying Mycobacterium tuberculosis infection.

BACKGROUND

The differential diagnosis between active tuberculosis (ATB) and latent tuberculosis infection (LTBI) is still a challenge worldwide.

METHODS

Immune indicators involved in innate, humoral, and cellular immune cells, as well as antigen-specific cells were simultaneously assessed in patients with ATB and LTBI.

RESULTS

Of 54 immune indicators, no indicator could distinguish ATB from LTBI, likely due to an obvious heterogeneity of immune indicators noticed in ATB patients. Cluster analysis of ATB patients identified three immune clusters with different severity. Cluster 1 (42.1 %) was a ''Treg/Th1/Tfh unbalance type" cluster, whereas cluster 2 (42.1 %) was an "effector type'' cluster, and cluster 3 was a ''inhibition type'' cluster (15.8 %) which showed the highest severity. A prediction model based on immune indicators was established and showed potential in classifying Mycobacterium tuberculosis infection.

CONCLUSIONS

We depicted the immune landscape of patients with ATB and LTBI. Three immune subtypes were identified in ATB patients with different severity.