Multifunctional Analysis of CD4⁺ T-Cell Response as Immune-Based Model for Tuberculosis Detection

Figure 2

Analysis of cytokine production by CD4⁺ cells at the single-cell level. (a) ROC curve (plotting sensitivity versus 1 − specificity) to discriminate infected (active TB and LTBI) from uninfected patients. The area under curve (AUC) was 0.9722. (b) Analysis of activated M. tuberculosis-specific CD4⁺ T cells producing any of the 3 cytokines (IFN-γ, IL-2, or TNF-α), using a cut-off to score responses as either positive or negative. The subjects were considered as positive (black) whether the frequency of CD4⁺ T cells was >0.45% and negative (white) when the frequency was <0.45%. Horizontal bars represent the median values and horizontal dashed line indicates the cut-off of 0.45%. ((c), (d), and (e)) Frequency of “total IFN-γ⁺,” “total IL-2⁺,” and “total TNF-α⁺” Mtb-specific CD4⁺ T cells in active TB patients (), in LTBI patients (), and in healthy controls () is shown. Horizontal bars represent the median values. Statistical analysis was performed using Kruskal-Wallis ANOVA with Dunn’s posttest comparison and significant differences are indicated by asterisks ().

(a)

(b)

(c)

(d)

(e)