Clinical Validation of Metabolite Markers for Early Lung Cancer Detection

Non-small cell lung cancer (NSCLC), comprising 85% of lung cancers, is a leading cause of cancer mortality. Early detection enhances survival, but current screening methods are limited. This retrospective study used targeted mass spectrometry-based metabolomics on 680 plasma samples from NSCLC patients and controls (discovery cohort) and 216 samples (validation cohort). Logistic regression models with a subset of ten metabolites achieved over 90% area under the ROC curve (AUROC) for distinguishing patients from controls, including early-stage disease. Incorporating smoking history improved model performance. In the discovery cohort, AUROCs were 93.6% (all stages), 93.7% (Stage I and II), and 93.9% (Stage I). Validation confirmed the high sensitivity and specificity of the models. This study demonstrates that metabolomic biomarkers provide a minimally invasive, sensitive, and specific tool for early NSCLC detection, potentially improving screening and patient outcomes. Future studies should validate these biomarkers in diverse populations. Statement of significanceThis study identifies plasma metabolite biomarkers that enable sensitive and specific early detection of NSCLC using minimally invasive blood sampling. Achieving over 90% area under the ROC curve for early-stage patients, the findings promise to improve lung cancer screening methods and enhance early interventions and patient outcomes.