Search

Circulating exosomes in the blood are promising tools for biomarker discovery in cancer. Due to their heterogeneity, different isolation methods may enrich distinct exosome cargos generating different omic profiles. In this study, we evaluated the effects of plasma exosome isolation methods on detectable multi-omic profiles in patients with non-small cell lung cancer (NSCLC), castration-resistant prostate cancer (CRPC), and healthy controls, and developed an algorithm to quantify exosome enrichment. Plasma exosomes were isolated from CRPC (n = 10), NSCLC (n = 14), and healthy controls (n = 10) using three different methods: size exclusion chromatography (SEC), lectin binding, and T-cell immunoglobulin domain and mucin domain-containing protein 4 (TIM4) binding. Molecular profiles were determined by mass spectrometry of extracted exosome fractions. Enrichment analysis of uniquely detected molecules was performed for each method with MetaboAnalyst. The exosome enrichment index (EEI) scores methods based on top differential molecules between patient groups. The lipidomic analysis detected 949 lipids using exosomes from SEC, followed by 246 from lectin binding and 226 from TIM4 binding. The detectable metabolites showed SEC identifying 191 while lectin binding and TIM4 binding identified 100 and 107, respectively. When comparing uniquely detected molecules, different methods showed preferential enrichment of different sets of molecules with SEC enriching the greatest diversity. Compared to controls, SEC identified 28 lipids showing significant difference in NSCLC, while only 1 metabolite in NSCLC and 5 metabolites in CRPC were considered statistically significant (FDR < 0.1). Neither lectin-binding- nor TIM4-binding-derived exosome lipids or metabolites demonstrated significant differences between patient groups. We observed the highest EEI from SEC in lipids (NSCLC: 871.33) which was also noted in metabolites. These results support that the size exclusion method of exosome extraction implemented by SBI captures more heterogeneous exosome populations. In contrast, lectin-binding and TIM4-binding methods bind surface glycans or phosphatidylserine moieties of the exosomes. Overall, these findings suggest that specific isolation methods select subpopulations which may significantly impact cancer biomarker discovery.

Advancements in the clinical practice of non-small cell lung cancer (NSCLC) are shifting treatment paradigms towards increasingly personalized approaches. Liquid biopsies using various circulating analytes provide minimally invasive methods of sampling the molecular content within tumor cells. Plasma-derived circulating tumor DNA (ctDNA), the tumor-derived component of cell-free DNA (cfDNA), is the most extensively studied analyte and has a growing list of applications in the clinical management of NSCLC. As an alternative to tumor genotyping, the assessment of oncogenic driver alterations by ctDNA has become an accepted companion diagnostic via both single-gene polymerase chain reactions (PCR) and next-generation sequencing (NGS) for advanced NSCLC. ctDNA technologies have also shown the ability to detect the emerging mechanisms of acquired resistance that evolve after targeted therapy. Furthermore, the detection of minimal residual disease (MRD) by ctDNA for patients with NSCLC after curative-intent treatment may serve as a prognostic and potentially predictive biomarker for recurrence and response to therapy, respectively. Finally, ctDNA analysis via mutational, methylation, and/or fragmentation multi-omic profiling offers the potential for improving early lung cancer detection. In this review, we discuss the role of ctDNA in each of these capacities, namely, for molecular profiling, treatment response monitoring, MRD detection, and early cancer detection of NSCLC.