TB Research

A blood-based Xpert host response assay for diagnosis and differentiation of active and latent tuberculosis in children and adolescents: A multi-center, prospective study.

Lin Sun, Min Fang, Yu Chen, Weiwei Jiao, Tingting Zou, Xiaoru Long, Yuqing Wang, Tingting Jiang, et al. (23 authors)

The Journal of infection · 2026-02

Abstract

OBJECTIVES: Non-sputum based, child-friendly triage tests are urgently needed to achieve accurate diagnosis and monitoring of tuberculosis (TB) in children. We aimed to assess an Xpert MTB Host Response (MTB-HR) assay, which provides a TB score based on the mRNA expression level of three host genes, for diagnosis and differentiation TB in children and adolescents.

METHODS: The multicenter, prospective study was conducted in four provinces of China among children and adolescents who were admitted to hospitals for TB or latent tuberculosis infection (LTBI) screening. Subjects were included in the evaluation of Xpert-MTB-HR from February 2020 to December 2021. Baselines of TB scores were analyzed in healthy children and adolescents with various ages. Accuracy was evaluated in subjects with various TB status, including microbiological data, disease severity and age.

RESULTS: Based on a composite clinical reference standard, among 780 patients enrolled, 403, 41, 109, and 227 were diagnosed as TB, LTBI, non-TB infectious diseases (DC) and healthy controls (HC), respectively. The mean TB scores decreased from 4 (IQR, 3·54-4·46) in infants to 0·91 (IQR, -0·23-2·04) in adolescents aged 17-18 years old. Using the composite clinical reference standard, the area under curves (AUCs) of the MTB-HR assay in discriminating ATB from HC, LTBI, and DC were 0·786 (95% CI, 0·749-0·823), 0·652 (95% CI, 0·559-0·744) and 0·771 (95% CI, 0·718-0·823), respectively.The optimal cutoff value was less than or equal to 2·675, resulting a sensitivity of 75·9% (95% CI, 67·0%-80·9%) and specificity of 70·5% (95%CI, 60·3%-76·7%) in TB diagnosis. The MTB-HR assay showed better auxiliary effect for diagnosis of ATB in children younger than five years of age (AUC, 0·885, 95% CI, 0·824-0·945, P=0·0006). The mean TB scores elevated at one month (P=0·0009) and three months (P=0·0061) after anti-TB treatment initiation.

CONCLUSIONS: The MTB-HR assay showed potential for ATB diagnosis and treatment monitoring in children and adolescents, especially in ages under five years old.

MeSH terms

  • Humans
  • Adolescent
  • Child
  • Prospective Studies
  • Male
  • Female
  • Latent Tuberculosis
  • Child, Preschool
  • Infant
  • China
  • Mycobacterium tuberculosis
  • Sensitivity and Specificity
  • Molecular Diagnostic Techniques
  • Tuberculosis
  • Diagnosis, Differential