Abstract
BACKGROUND: Tuberculosis (TB) remains a major cause of infectious disease mortality globally, with significant underdiagnosis perpetuating transmission. Tongue swab analysis has emerged as a promising non-invasive method for pulmonary TB diagnosis. This study evaluates the diagnostic accuracy of the TB-EASY quantitative PCR (qPCR) assay using tongue swab specimens. METHODS: In this prospective multicentre study, conducted across seven designated TB hospitals in China, 729 participants were included in the analysis. Tongue swabs were tested using the new TB-EASY assay from Hugobiotech, while sputum samples were analysed by Xpert MTB/RIF (Xpert), smear and culture tests. Diagnostic performance was compared to a composite microbiological reference standard (MRS). RESULTS: The TB-EASY assay demonstrated high diagnostic accuracy, with sensitivity and specificity of 89.6% and 96.2% compared to sputum Xpert, and 87.4% and 98.0% compared to the MRS. Sensitivity varied by bacterial load, ranging from 100% in high-load cases to 70.4% in very-low-load cases. The assay demonstrated robust performance in diverse epidemiological settings. CONCLUSIONS: The TB-EASY qPCR assay using tongue swabs offers a reliable, non-invasive diagnostic alternative for pulmonary TB, especially where sputum collection is challenging. Its potential for wider use in high TB burden settings warrants further validation in community-based studies. Limitations include potential overestimation of sensitivity due to the selection of symptomatic patients and the use of sputum Xpert rather than Xpert Ultra. Additionally, the performance in non-sputum-producing patients remains untested, and the cost-effectiveness should be further evaluated to assess the feasibility of its implementation.