Discovery of biomarkers is crucial for the early diagnosis and precision treatment of cancer. Traditional methods (genomics and metabolomics, etc.) are low-throughput, expensive, and require complex sample processing. Surface-enhanced Raman scattering (SERS), as a simple, time- and cost-effective trace-level detection technique, offers an alternative due to its molecular fingerprint recognition capability. We developed here a three-principal-substrate SERS strategy for serum biomarker screening in cancers using breast cancer as a case study. Three Ag-Au nanorod substrates of different surface properties were selected by principal component analysis (PCA) of serum SERS spectra on them. With the three substrates, serum samples used in exploring biomarkers were screened by partial least-squares discriminant analysis (PLS-DA) and unanimity principle. Eleven candidate metabolites were indicated by the SERS spectral differences and their projection variable importance (VIP) values, supported by consistent trends observed in MALDI-TOF MS measurements. Nine of the candidates were reported as biomarkers of breast cancer, while the other two related to the metabolic derivatives and pathways relevant to breast cancer. Our results demonstrate that the three-principal-substrate SERS framework can serve as an effective screening approach for serum biomarker discovery, especially for diseases with biomarkers unknown.
Dong et al. (2026) studied this question.