A liquid biopsy algorithm exploiting local exosome heterogeneity

Abstract

We developed a novel cancer detection method that leverages changes in local exosome heterogeneity to regulate bulk-scale heterogeneity in body fluids. Using an automated, reversible isolation system and downstream immunoassays, we investigated exosome subclasses exhibiting marker composition changes associated with cancer. We introduced an index, R, to measure local heterogeneity changes, quantifying the proportion of a third marker within double-marker-positive subclasses. This method, which employs pan-exosome tetraspanins CD9, CD63, and CD81, revealed distinct heterogeneity differences between normal and cancer samples. These differences were consistently demonstrated in exosome samples derived from various cell lines and human sera. Notably, in clinical samples, R values distributed samples from healthy donors and cancer patients into unique patterns based on the third marker within specific subclasses. However, across subclasses, R value changes were significantly smaller in healthy samples than in cancer samples. Sorting patterns were simulated using exosomes derived from immune cells, indicating an immune response contribution to the observed signals. These findings highlight the method's potential for identifying biomarkers for specific cancer diagnoses and multi-cancer screening.