Two immune response types found in TB-exposed healthcare workers

Tuberculosis exposure is common among healthcare workers in high-burden settings, yet some people repeatedly exposed to Mycobacterium tuberculosis (Mtb) do not show standard signs of infection. In a study led by corresponding author Muki Shey, researchers screened HIV-uninfected healthcare workers who had worked in high TB exposure healthcare facilities for over 5 years using the interferon-gamma release assay (IGRA) and the tuberculin skin test (TST). From that screening they defined Resisters (TST<10mm and IGRA<0.35 IU/mL, n=129) and people with latent TB infection (LTBI; TST≥10mm and IGRA≥0.35 IU/mL, n=145). For deeper study they selected 'extreme resisters' (TST=0mm and IGRA<0.2 IU/mL; n=26) and 'extreme LTBI' (TST≥15mm and IGRA≥1 IU/mL; n=24). Blood samples were collected and peripheral blood mononuclear cells (PBMC) were isolated for laboratory tests designed to probe immune responses to live Mtb. The goal was to find whether people who appear uninfected by standard tests might nonetheless show hidden, Mtb-specific immune activity that could explain their apparent resistance.

To compare immune activity, the investigators cultured PBMC with live H37Rv Mtb for 18 hours, then collected culture supernatants to measure 65 secreted analytes by Luminex. They also evaluated cell-associated cytokine expression by CD4 T cells and monocytes using flow cytometry. Using feature selection in R, the team identified four cytokines—IL-17A, MCP-1, IL-8, and MDC—that together distinguished extreme Resisters from extreme LTBI with an area-under-the-curve (AUC) of 0.67 (0.54-0.85). When the researchers examined just the extreme Resisters, hierarchical clustering split them into two main groups, labeled Resister_c1 and Resister_c2. Comparing those clusters showed 37 cytokines that were significantly higher and 15 cytokines that were lower in Resister_c2 compared with Resister_c1. A focused set of five cytokines—TRAIL, MIP-1β, Fractalkine, GRO-α and IL-1α—separated these two Resister clusters perfectly in this dataset, with an AUC of 1 (1-1). Notably, CD4 T cell and monocyte responses to Mtb did not differ significantly between the two clusters.

The findings suggest that apparent resistance to Mtb exposure is not a single uniform state. Instead, extreme Resisters split into two distinct cytokine-response patterns when their PBMC are challenged with live H37Rv Mtb, indicating that some people who test negative by IGRA and TST may nonetheless have detectable, Mtb-specific immune profiles. The identification of small sets of cytokines that classify extreme Resisters from extreme LTBI, and that perfectly separate the two Resister clusters in this study, points to the possibility of developing more sensitive biomarker approaches. Such biomarkers could help discriminate true resistance from alternative immune responses that evade standard tests. These results are based on a selected subset of participants and specific laboratory conditions, so further study will be needed to confirm these cytokine signatures in larger and different groups and to understand whether they predict long-term protection from infection or disease.