Exhaled breath metabolomics: Clinical diagnostic applications, technological evolution and translational challenges

Breathomics, the analysis of volatile organic compounds (VOCs) and other metabolites in exhaled breath, offers a non-invasive approach to disease screening, longitudinal monitoring and assessment of treatment response. Advances in gas chromatography-mass spectrometry, real-time mass spectrometry, electronic-nose (e-nose) sensors and machine-learning workflows have markedly improved the sensitivity and diagnostic performance of breath analysis. However, these gains have been demonstrated mainly in proof-of-concept and analytical validation studies, and remain insufficiently confirmed in prospective clinical cohorts. In this Review, we examine breathomics across the translational pathway, from breath sampling and pre-analytical standardization to analytical platforms, data processing, disease applications and clinical implementation. We focus on three clinically prominent areas, cancer, chronic respiratory disease and infectious disease, and use selected systemic disorders to illustrate the limited disease specificity of individual VOCs. Current evidence supports the presence of detectable disease-associated breath signatures, but most VOC panels remain insufficiently standardized, externally validated and mechanistically interpretable for routine clinical use. Translation will require harmonized sampling protocols, rigorous analytical validation, prospective multicenter cohorts, explainable models, regulatory approval, and evidence of clinical utility and cost-effectiveness.