High-purity isolation of tumor-derived extracellular vesicles via SEC-immunoaffinity integration improves diagnostic sensitivity in breast cancer

Background: Extracellular vesicles (EVs) provide a non-invasive approach for liquid biopsy due to their enriched molecular cargo and structural stability. However, EV heterogeneity and contamination of current EV isolation methods by non-vesicular extracellular particles (NVEPs), particularly lipoproteins, remain substantial obstacles to reliable biomarker analysis. Reports showing that several proposed EV biomarkers are enriched in NVEP fractions underscore the need for high-purity EV isolation. Methods: To improve plasma-derived small EV (sEV) purification, we systematically evaluated size-exclusion chromatography (SEC) resins using purified control materials representing sEVs, large EVs, lipoproteins (HDL, LDL, VLDL) and albumin. Fraction distribution patterns were mapped to predict NVEP co-elution during plasma processing. Based on these observations, SEC was combined with apoB100-based immunoaffinity depletion to reduce lipoprotein contamination. Tumor-derived EVs (tdEVs) were then enriched using immunocapture targeting breast cancer-associated surface epitopes, enabling downstream multi-omics analysis. Results: CL-6B generated sharp peaks for individual control materials but showed convergence of VLDL and a portion of LDL in fraction 6, indicating a high likelihood of overlap with sEV-associated fractions when applied to plasma. In contrast, CL-4B produced broader elution patterns, with LDL largely absent from sEV-associated fractions and VLDL distributed across fractions 5-15, reducing the degree of overlap with sEVs. These findings indicate that CL-4B provides more favorable separation characteristics for plasma sEV purification. ApoB100-mediated immunoaffinity depletion further reduced VLDL and HDL signals while maintaining sEV recovery. The resulting fractions showed improved particle-to-protein ratios, decreased ApoA1/ApoB signatures, and retention of canonical sEV markers. Breast cancer-specific immunocapture enriched tdEVs carrying elevated levels of an oncogenic miRNA signature previously reported in our cohorts. Notably, this tdEV-focused miRNA panel demonstrated improved diagnostic sensitivity compared with our earlier total-EV-based approach, suggesting that high-purity tdEV enrichment more accurately reflects tumor-derived molecular profiles. Conclusion: We developed an analytically stable EV purification workflow that combines optimized SEC with sequential immunoaffinity strategies to minimize NVEP interference while preserving sEV integrity. This high-purity, tumor-focused platform enables reliable multi-omics profiling of circulating tdEVs and supports the development of clinically applicable biomarkers for early breast cancer detection and disease monitoring.