Accurate Cancer Diagnosis and Treatment Monitoring through Multiplexed Profiling of Protein Markers on Small Extracellular Vesicles

The detection of small extracellular vesicles (sEVs) is currently a pivotal liquid biopsy approach for noninvasive cancer diagnosis. However, the lack of adequate specificity and sensitivity, as well as labor-intensive purification and analysis procedures, present challenges in isolating and profiling sEVs. Here, we present a protein-specific enzymatic optical reporter deposition-based liquid biopsy assay for the rapid and efficient capture and ultrasensitive detection of sEVs using a minimal volume of initial biofluids (10 μL). Biotin aptamers were employed to label sEV proteins for peroxidase conjugation, catalyzing the conversion of fluorescein tyramine into highly reactive free radicals. Efficient signal conversion was achieved by depositing nanoheterolayers composed of covalent tyraminated complexes onto sEV surfaces. The present method offers a detection limit of 6.4 × 10³ particles mL^-1 with a linear range of 10⁴-10¹⁰ particles mL^-1 for sEVs. Two machine learning algorithms, principal coordinates analysis and principal component analysis, were subsequently applied for dimensionality reduction. In a clinical cohort of 84 patients, including 6 cancer types and noncancer cases, the assay achieved an overall accuracy of 100% (95% confidence interval) in distinguishing between cancer and noncancer controls and 96% in classifying cancer types. As drugs are frequently administered to patients to modulate the activity of tumor cells, we investigated the efficacy of this strategy in treatment monitoring, achieving an overall accuracy of 100%. This strategy demonstrates a cost-effective, rapid, and low sample volume consumption approach that holds significant potential for precise cancer diagnosis and auxiliary assessment of drug response in clinical settings.