Diagnostic potential of combining plasma biomarkers of tissue damage and inflammation in pediatric TB

Abstract

IntroductionImmune-based diagnostic tests for tuberculosis (TB) have suboptimal sensitivity in children and cannot differentiate between latent infection (LTBI) and active disease. This study evaluated the diagnostic potential of a broad range of biomarkers of tissue damage and inflammation in unstimulated plasma in children. MethodsWe analyzed 17 biomarkers in 15 non–M. tuberculosis (MTB)-infected controls and 33 children with TB infection (LTBI, n = 8; probable TB, n = 19; confirmed TB, n = 6). Biomarker concentrations were measured using a Luminex magnetic bead–based platform and multiplex sandwich immunoassays. Concentrations, correlations and diagnostic accuracy assessments were conducted among patient groups. ResultsConfirmed TB cases had significantly higher concentrations of IFN-γ and IL-2 and higher IFN-γ/MCP-1 and IL-2/MCP-1 ratios compared to LTBI and non–MTB-infected children. Among children with confirmed TB, there was a strong correlation between IFN-γ and IL-10 (r = 0.95; p < 0.001) and a significant correlation between IL-2 and IL-1ra (r = 0.92), IL-21 (r = 0.91), MCP-3 (r = 0.84), and MMP-1 (r = 0.85). The IFN-γ/MCP-1 ratio was the most accurate biomarker combination for differentiating between MTB-infected and non–MTB-infected children (AUC, 0.82; sensitivity, 87.9%; specificity, 66.6%; p < 0.001) and between active TB and non–MTB-infected children (AUC 0.82; sensitivity 88.0%; specificity 60.0%; p < 0.001). None of the biomarkers investigated were able to discriminate between LTBI and active TB. ConclusionOur data suggest that combining the analyses of multiple biomarkers in plasma has the potential to enhance diagnosis of TB in children and, thus, warrants additional investigation. In particular, the diagnostic potential of IFN-γ/MCP-1 ratios should be further explored in larger pediatric cohorts.