Antigen-specific chemokine profiles as biomarkers for detecting Mycobacterium tuberculosis infection

Background Latent tuberculosis (TB) infection can progress to active TB, which perpetuates community transmission that undermines global TB control efforts. Clinically, interferon-γ release assays (IGRAs) are commonly used for active TB case detection. However, low IGRA sensitivity rates lead to false-negative results for a high proportion of active TB cases, thus highlighting IGRA ineffectiveness in differentiating MTB-infected individuals from healthy individuals. Methods Participants enrolled at Beijing Chest Hospital from May 2020-April 2022 were assigned to healthy control (HC), LTBI, IGRA-positive TB, and IGRA-negative TB groups. Screening cohort MTB antigen-specific blood plasma chemokine concentrations were measured using Luminex xMAP assays then were verified via testing of validation cohort samples. Results A total of 302 individuals meeting study inclusion criteria were assigned to screening and validation cohorts. Testing revealed significant differences in blood plasma levels of CXCL9, CXCL10, CXCL16, CXCL21, CCL1, CCL19, CCL27, TNF-α, and IL-4 between IGRA-negative TB and HC groups. Levels of CXCL9, CXCL10, IL-2, and CCL8 biomarkers were predictive for active TB, as reflected by AUC values of ≥0.9. CXCL9-based enzyme-linked immunosorbent assay sensitivity and specificity rates were 95.9% (95%CI: 91.7-98.3) and 100.0% (92.7-100.0), respectively. Statistically similar AUC values were obtained for CXCL9 and CXCL9-CXCL10 assays, thus demonstrating that combined analysis of CXCL10 and CXCL9 levels did not improve active TB diagnostic performance. Conclusion The MTB antigen stimulation-based CXCL9 assay may compensate for low IGRA diagnostic accuracy when used to diagnose IGRA-negative active TB cases and thus is an accurate and sensitive alternative to IGRAs for detecting MTB infection.