Background: Breast cancer (BC) is the second leading cause of cancer-related deaths among women, primarily due to metastases to other organs rather than the primary tumor.
Methods: In this study, a comprehensive analysis of plasma proteomics and metabolomics was conducted on a cohort of 51 BC patients. Potential biomarkers were screened by the Least Absolute Shrinkage and Selection Operator (LASSO) regression and Random Forest algorithm. Additionally, enzyme-linked immunosorbent assay (ELISA) kits and untargeted metabolomics were utilized to validate the prognostic biomarkers in an independent cohort.
Results: In the study, extracellular matrix (ECM)-related functional enrichments were observed to be enriched in BC cases with bone metastases. Proteins dysregulated in retinol metabolism in liver metastases and leukocyte transendothelial migration in lung metastases were also identified. Machine learning models identified specific biomarker panels for each metastasis type, achieving high diagnostic accuracy with area under the curve (AUC) of 0.955 for bone, 0.941 for liver, and 0.989 for lung metastases.
Conclusions: For bone metastasis, biomarkers such as leucyl-tryptophan, LysoPC(P-16:0/0:0), FN1, and HSPG2 have been validated. dUDP, LPE(18:1/0:0), and aspartylphenylalanine have been confirmed for liver metastasis. For lung metastasis, dUDP, testosterone sulfate, and PE(14:0/20:5) have been established.
Keywords: biomarkers; breast cancer; metabolomics; metastasis; proteomics.
© 2024 Wiley‐VCH GmbH.