Plasma metabolomics panels show high discriminating power for lung, colorectal, and ovarian cancers

In this cohort-based metabolomics biomarker discovery effort, investigators analyzed plasma samples from 615 patients with lung, ovarian, or colorectal cancer at diagnosis and 95 non-cancerous control subjects. The primary aim was to identify specific panels of metabolites with high discriminating power, with secondary goals of evaluating potential roles in diagnosis and monitoring disease progression.

The main result was a combined ROC AUC value up to 0.95 for the metabolite panels in discriminating cancer patients from non-cancerous controls. No additional effect sizes, p-values, confidence intervals, or absolute event numbers were provided for this or other outcomes.

Safety and tolerability were not assessed, and no adverse events, serious adverse events, or study discontinuations were reported, consistent with the observational design and use of banked plasma.

Key limitations include the cross-sectional case-control design, which limits inference about causality or temporal changes, and the absence of external validation cohorts, standardized cutoffs, and performance metrics such as sensitivity, specificity, or PPV/NPV. Generalizability to other populations, settings, and cancer stages is uncertain.

While the authors suggest potential for improved diagnosis and monitoring, these findings are preliminary. Clinical use would require rigorous validation, comparison with existing diagnostics, assessment in real-world settings, and demonstration of incremental value over current standards.