Advanced Technologies for the Diagnosis of Pulmonary Tuberculosis Using Exhaled Breath Samples: A Systematic Scoping...

Non-invasive methods based on exhaled breath samples have emerged as promising alternatives for the diagnosis of pulmonary tuberculosis, especially in settings where sputum collection is challenging and laboratory infrastructure is limited. We aimed to map the evidence on advanced technologies available for detecting pulmonary tuberculosis using alveolar breath, bioaerosols, or exhaled breath condensate.

A comprehensive scoping review following international guidelines was performed (OSF: 10.17605/OSF.IO/TRJ7W). Searches were conducted in PubMed, Scopus and Web of Science (June 2025). Primary studies evaluating physical or chemical diagnostic methods based on biological exhaled breath samples for detecting pulmonary tuberculosis in symptomatic patients were included. Studies' metadata, including method accuracy metrics, were extracted for diagnosis performance comparison. Evidence was synthesised using descriptive statistical and qualitative methods.

Overall, 19 studies published between 2010 and 2025 were included, mostly conducted in high tuberculosis-burden countries such as South Africa (16%), China (11%) and Paraguay (11%). The most evaluated techniques were electronic nose (e-nose) (42%), gas chromatography coupled with mass spectrometry (GC-MS/MS-based) (21%) and machine learning applied to mass spectrometry (ML-MS) approaches (11%). Important heterogeneity in studies' design and analytical approaches was observed. All included studies used non-invasive samples and were classified as intended for screening purposes; none aimed for confirmatory diagnosis. Among the included studies, 10 (53%) were classified as near point-of-care approaches. Of these, 8 (80%) met the minimum sensitivity requirement for screening tests, although only one also achieved the corresponding specificity threshold. The highest diagnostic performance reported showed a sensitivity of 98.5% (95% CI: 92.1-100) and a specificity of 100% (95% CI: 93.5-100).

Although research on exhaled breath-based technologies for pulmonary tuberculosis diagnosis is advancing, with some alternatives presenting moderate-to-high performance, their implementation still requires standardised validation, methodological improvements and field testing in real-world settings.