血浆蛋白-转录组整合鉴定室管膜瘤中的分子特征和循环生物标志物

Background: Ependymomas (EPNs) are heterogeneous central nervous system (CNS) tumors that lack reliable blood-based biomarkers for diagnosis, prognosis, treatment monitoring, and recurrence detection. We generated the first plasma proteo-transcriptomic atlas of pediatric EPN to define circulating signatures across tumor subtypes, grades, treatment states and recurrence. Methods: Plasma from 33 EPN patients and seven controls was analyzed using quantitative label-free untargeted LC-MS/MS and integrated with 44 tumor Total RNA sequencing and six control (differentially expressed genes (DEGs); |log₂FC| ≥ 1, FDR < 0.05). Differentially expressed proteins (|log₂FC| ≥ 0.58, FDR < 0.05) were examined via GO/KEGG/Reactome enrichment, STRING protein-interaction networks, and multi-omics concordance analyses. Biomarkers correlated with Cyclin D1 positivity, 1q gain, ZFTA-RELA fusion, and CDKN2A loss. CancerMine contextualizes known oncogenic roles. Results: We identified 270 dysregulated plasma proteins that delineated a systemic pattern of ECM activation, coagulation complementation, lipid metabolic reprogramming, and humoral immune suppression. FN1 was the top marker, elevated 26.36-fold in subependymoma, 7.66-fold in supratentorial, and 6.37-fold in posterior fossa tumors, sharply decreasing post-treatment (1.64-fold), reflecting tumor dynamics. APOE (3.29-fold), APOD, APOH, APOA4, and TMEM198 (2.12-fold) were consistently elevated and decreased after therapy. Immune suppression was evident, with JCHAIN reduced to 0.32-fold, immunoglobulins (IGHD 0.07-fold), and complement proteins (C1QB, C4A/B) downregulated. IGHV3-35 was elevated pre-treatment (2.03-fold) and at recurrence, serving as a dual biomarker. LPA and SERPINF1 were recurrence-specific, and CD5L was supratentorial-specific. Transcriptomics identified 1,286 DEGs, of which 132 mapped to plasma proteins (57 upregulated, 75 downregulated). FN1, TMEM198, and IGKC showed RNA-protein concordance, whereas APOE, APOD, and JCHAIN showed secretion-driven discordance. STRING analysis revealed dense ECM and coagulation-lipid interaction hubs (PPI p < 1x10⁻¹⁶). CancerMine validated oncogenic roles of FN1, APOE/APOD, JCHAIN, C1QB, and TTR in other cancers, novel to EPN. Biomarkers were strongly correlated with aggressive genotypes. Conclusion: This plasma proteo-transcriptomic atlas revealed a robust systemic signature in pediatric EPN, defined by ECM remodeling, coagulation dominance, metabolic rewiring, and immune suppression. Key biomarkers, such as FN1, TMEM198, APOE/APOD, IGHV3-35, JCHAIN, LPA, and SERPINF1, offer strong potential for non-invasive diagnosis, stratification, therapy monitoring, and early recurrence detection, establishing plasma proteomics as a transformative precision oncology platform.