Magnetic Levitation Derived Metabolomic Fingerprinting Enables Exploratory Discrimination of Breast Cancer Subtypes.

Breast cancer (BC) consists of heterogeneous molecular subtypes with distinct biological behavior and therapeutic response, including triple negative breast cancer (TNBC), human epidermal growth factor receptor 2 positive (HER2+), and Luminal A tumors. Current subtype classification primarily relies on tissue biopsy and molecular pathology, highlighting the need for complementary non invasive analytical approaches capable of capturing systemic biochemical differences among BC subtypes. In this exploratory study, we investigate whether magnetic levitation (MagLev) derived plasma patterning combined with untargeted metabolomics analysis can distinguish BC subtypes based on subtype specific metabolomic fingerprints. Representative plasma samples from TNBC, HER2+, and Luminal A patients were levitated in a standard MagLev system containing superparamagnetic iron oxide nanoparticles (SPIONs), generating visibly distinct levitation patterns throughout the levitation process. Individual levitated plasma layers were subsequently extracted and subjected to untargeted metabolomics analysis. Multivariate analysis demonstrated clear subtype dependent metabolic separation among the three BC subtypes. Heatmap clustering, PLS DA analysis, and variable importance profiling identified distinct metabolic signatures involving glycolysis, pentose phosphate pathway metabolism, TCA cycle activity, amino acid metabolism, and lipid remodeling. Elevated levels of phosphoglycerate, pyruvate, ribose 5 phosphate, glutamate, and 2 oxoglutarate suggested enhanced proliferative and biosynthetic metabolism, while enrichment of long chain fatty acids in Luminal A samples indicated subtype specific lipid metabolic remodeling. These findings demonstrate the feasibility of combining MagLev derived plasma organization with metabolomics analysis to generate disease specific metabolomic fingerprints and establish a foundation for future large scale validation studies.