Prognostic and microRNA profile analysis for CD44 positive expression pediatric posterior fossa ependymoma.

Background: Ependymoma is the third most common pediatric brain tumor and occurs most frequently in the posterior fossa. However, the lack of immortalized cell lines, xenografts, or animal models has significantly hindered the study of pediatric posterior fossa ependymoma (P-PF-EPN) pathogenesis. This prompted us to use clinical big data to study this rare disease.

Methods: Application of the robust rank aggregation method revealed CD44 as a reliable biomarker in P-PF-EPN. 120 P-PF-EPN samples after surgical resection were selected for Kaplan-Merier and Cox proportion hazard regression survival analysis. Immunohistochemical analysis was performed to assess CD44 expression in the tumor samples. The miRNA profile was determined using a whole-genome miRNA microarray. The expression patterns of related mRNAs, miRNAs and proteins were validated by qRT-PCR or Western blotting.

Results: CD44 was found to be an independent predictor of prognosis in survival analysis. It improved the accuracy of using LAMA2/NELL2 for classifying P-PF-EPN molecular subgroups. Fourteen miRNAs were underexpressed, and one miRNA was overexpressed in CD44-positive P-PF-EPNs. miR-543, miR-495-3p, miR-299-3p, miR-139-5p and miR-128-3p were identified to have CD44 positively co-regulated potential target oncogenes. Two PI3K-Akt signaling pathway related potential target oncogenes (VEGFA, CSF1) for miR-299-3p and miR-495-3p were validated overexpression in CD44 positive P-PF-EPNs. Abnormal activation of the PI3K-Akt pathway was confirmed in CD44-positive cases.

Conclusions: CD44 is of great clinical significance as a prognostic biomarker. The survival difference between CD44 positive and negative P-PF-EPN is determined by a complex functional miRNA-mRNA-signaling pathway regulatory network.