Emerging tests show promise for telling latent tuberculosis apart from active disease

A systematic review published in BMC Infectious Diseases has evaluated whether new diagnostic approaches can reliably distinguish latent tuberculosis infection from active tuberculosis, a distinction that current immunological tests cannot make with confidence.

Latent tuberculosis infection affects approximately one quarter of the global population and represents a major reservoir for future active tuberculosis. The existing immunological diagnostics, including the tuberculin skin test and interferon-gamma release assays, are unable to reliably distinguish latent infection from active disease, which limits their clinical utility for targeted prevention strategies.

The review, led by Sofia Kostoudi and Robert J H Hammond, was conducted in accordance with PRISMA 2020 guidelines. In its initial phase, the authors searched PubMed and Scopus from January 2005 to May 2022 for primary studies evaluating diagnostic tests intended to differentiate latent tuberculosis infection from active tuberculosis. A second phase repeated a similar search from May 2022 to 2026 to capture more up-to-date results.

Eligible studies included immunocompetent adults and reported diagnostic performance metrics. The review excluded studies involving immunocompromised populations, children, pregnancy, animal models, or secondary literature. Risk of bias was assessed using QUADAS-2, and certainty of evidence was assessed narratively using GRADE. Sensitivity and specificity were extracted where available.

Fifteen studies comprising 3,967 participants met the inclusion criteria. The diagnostic approaches fell into two categories: immunological assays, with eleven studies, and transcriptional profiling, with four studies.

Four approaches demonstrated promising diagnostic performance. A T-SPOT machine-based learning algorithm using a cforest model achieved a sensitivity of 92.80% and a specificity of 89.86%. A multi-antigen ELISA-based IgG logistic regression model reached a sensitivity of 93.33% and a specificity of 93.10%, and a composite T-SPOT.TB diagnostic model showed a sensitivity of 93.21% and a specificity of 90.00%.

The most striking result came from microRNA-based diagnostic biomarkers. The microRNAs let-7d-5p and miR-140-5p, in combination, established a sensitivity of 100% and a specificity of 100% in an internal test set, although the review notes that no independent validation cohort was available for this approach.

Despite these encouraging figures, the authors report considerable heterogeneity in reference standards and biomarker thresholds across the included studies. They conclude that several emerging diagnostics demonstrate promising accuracy for distinguishing latent tuberculosis infection from active tuberculosis in immunocompetent adults, but that heterogeneity in study design and reference standards limits immediate clinical translation. Future research, they write, should prioritize standardization, validation in high-burden settings, and evaluation in immunologically diverse populations.

Subject of Research: Systematic review of emerging diagnostics for distinguishing latent tuberculosis infection from active tuberculosis

Article Title: What is the future of latent tuberculosis diagnostics? A PRISMA-compliant systematic review

Article References: Kostoudi, S., & Hammond, R. J. H. (2026). What is the future of latent tuberculosis diagnostics? A PRISMA-compliant systematic review. BMC Infectious Diseases. https://doi.org/10.1186/s12879-026-14570-2

Image Credits: AI Generated

DOI: 10.1186/s12879-026-14570-2

Keywords: latent tuberculosis infection, active tuberculosis, diagnostics, systematic review, PRISMA, interferon-gamma release assays, tuberculin skin test, microRNA biomarkers, T-SPOT, machine learning, sensitivity and specificity, QUADAS-2