Immune reactive profiling in subjects at risk of latent tuberculosis infection with TST(+)/IGRA(-) results

Rationale: Asymptomatic subjects with positive tuberculin skin testing (TST) and negative interferon-gamma release assay (IGRA) results are often felt to have false(+) TST for latent tuberculosis infection (LTBI). We hypothesized that optimized immune profiling methods in subjects with TST(+)/IGRA(-) can identify LTBI with potential false(-) whole-blood IGRA results. Objective: Optimized IGRA and flow cytometric (FC) antigen-specific immune reactive methods were applied to seek biomarkers that differentiate subjects with TST(+)/IGRA(-) at risk of LTBI from unexposed subjects. Methods: An ELISPOT and FC assay that detects induction of CD25+CD134+ co-expression on TB-antigen-stimulated T-cells from PBMCs after 2-day incubation with co-stimulatory antibodies. Results: Ninety-five subjects with TST(+)/IGRA(-) and TST(-)/IGRA(-) results were studied, including 68 (71.6%) subjects at risk of LTBI. There were statistical differences in ELISPOT (median sfc [IQR]) and FC assay (median %CD8+ T-cells) between subjects with TST(+)/IGRA(-) and TST(-)/IGRA(-) results: ELISPOT (ESAT6/CFP10 or RD1 peptides –nil)= 52.3[8-92] vs. 17.3[2.2-58.8] (P=0.033), ELISPOT (RD1 peptides –nil)/ELISPOT (PPD-nil) ratio= 64.1[27.4-117.9] vs. 25.7[7.6-74.0] (P=0.007), and FC assay for CD8+ T-cells (RD1 peptides –nil)= 0.17[0.04-1.08] vs. 0.07[0.0-0.24] (P=0.021). Subjects with TST(+)/IGRA(-) had higher likelihood of potential TB exposure than TST(-)/IGRA(-) subjects, after adjusting for BCG vaccination (OR=5.1 [95%CI:1.3-34.2]; P=0.017). Conclusion: The strategy of optimizing immunoassay analysis is capable of detecting subset profiles that distinguish subjects with TST(+)/IGRA(-) and potential false(-) IGRA results.