Cancer Medicine · 10 authors, 9 centres
This summary was generated by AI from a single paper. It has not been reviewed by a clinician and is not clinical advice. Verify against the source before acting on it.
This study used large-scale targeted proteomics of plasma extracellular vesicles (EVs) to identify two novel subtypes of colorectal cancer (CRC) with significantly different overall survival. Alpha-1-acid glycoprotein 1 (ORM1) was identified as a key protein strongly associated with poor prognosis, independent of TNM staging. Liquid biopsy assessment of EV proteins may provide valuable prognostic information to complement standard clinical staging in CRC.
**Background:** Colorectal cancer (CRC) is the third most common cancer and second leading cause of cancer-related death worldwide. While TNM classification remains the gold standard for staging, molecular subtyping is increasingly important for personalized therapy. Liquid biopsy using extracellular vesicles (EVs) offers a less invasive alternative to tissue biopsies for molecular analysis. This study aimed to identify EV protein biomarkers and CRC subtypes through large-scale targeted proteomics.
**Methods:** The study analyzed two independent cohorts. Cohort 1 included 59 CRC patients and 59 healthy controls (HC). Cohort 2 comprised 150 CRC patients of all disease stages. Plasma EVs were isolated and analyzed using a large-scale targeted proteomics approach (selected reaction monitoring, SRM) targeting 457 proteins (1344 peptides). Stable isotope-labeled (SIL) peptides were synthesized using the MS-QBiC method. In cohort 1, the Mann-Whitney-Wilcoxon test and random forest analysis were used to select candidate biomarkers. In cohort 2, consensus clustering was applied to identify patient subtypes, and Kaplan-Meier survival analysis with Cox regression was performed to assess prognostic significance.
**Key Results:** In cohort 1, 99 proteins were identified as CRC-associated biomarkers. In cohort 2, consensus clustering revealed two stable clusters (k=2) with no apparent relationship to TNM staging. Cluster 1 patients had significantly shorter overall survival than cluster 2 patients (p=0.017). Twelve proteins were significantly upregulated in cluster 1 (fold change >2, p<0.05), including ORM1, SERPINA1, LRG1, C3, and FN1. Kaplan-Meier analysis showed that ORM1, SERPINA1, and LRG1 levels above the median were associated with worse OS, with ORM1 showing the most significant difference (p<0.003). Multivariate Cox regression identified TNM stage and ORM1 as independently associated with shorter OS (p<0.01). Among stage IV patients specifically, ORM1 remained a significant predictor of shorter OS. Gene Ontology analysis revealed that 8 of the 12 upregulated proteins shared terms including acute-phase response, inflammatory response, neutrophil degranulation, and platelet degranulation. Protein-protein interaction and correlation analyses confirmed strong functional associations among these proteins, with ORM1 showing strong correlations with LRG1 (r=0.80) and SERPINA1 (r=0.72).
**Clinical Implications:** This study demonstrates that large-scale targeted proteomics of plasma EVs can identify clinically meaningful CRC subtypes with prognostic significance independent of TNM staging. The two identified subtypes may help distinguish early-stage patients at risk of progression to more serious disease. ORM1, an acute-phase protein with immunomodulatory properties, emerged as a strong independent prognostic biomarker. The findings suggest that EV protein profiling could complement TNM classification for treatment planning, particularly for stage IV patients where curative resection rates differed between clusters. The potential for non-invasive detection of ORM1 in urine EVs is noted as a future direction. Further studies with larger cohorts and mechanistic investigations are needed to validate these findings and elucidate the biological role of ORM1 and associated proteins in CRC progression.