RESUMEN
BACKGROUND: Ependymal tumors are the third most common group of brain tumors in children, accounting for about 10% of all primary brain neoplasms. According to the current WHO classification, they comprise four entities with the most frequent ependymoma and anaplastic ependymoma. The most of pediatric tumors are located within the posterior fossa, with a tendency to infiltrate the vital brain structures. This limits surgical resection and poses a considerable clinical problem. Moreover, there are no appropriate outcome prognostic factors besides the extent of surgical resection. Despite definition of molecular subgroups, the majority of childhood ependymomas present a balanced genome, which makes it difficult to establish molecular prognostic factors. METHODS: The purpose of our study was to explore whether miRNA expression could be used as prognostic markers in pediatric infratentorial ependymomas. We also performed a mRNA expression pattern analysis of NELL2 and LAMA2 genes, with immunohistochemical illustrations of representative cases. The miRNA and mRNA expression was measured in 53 pediatric infratentorial ependymomas using a real-time quantitative PCR. RESULTS: Three miRNAs were shown to efficiently differentiate between grade II and III ependymomas: miR-17-5p, miR-19a-3p, and miR-106b-5p. Survival analysis showed that the probabilities of overall (p = 0.036) and event-free survival (p = 0.002) were reduced with higher than median miRNA expression levels of miR-17-5p. Using multivariate analysis adjusted for patient's age, sex, tumor grade and localization, we showed statistically significant associations with event-free survival (p = 0004) and borderline statistical significance with overall survival (p = 0.057) for miR-17-5p. Correlation analysis of miR-19a, miR-17-5p, miR-106b revealed that their expression levels were significantly correlated with EZH2 expression, suggested marker of PFA ependymomas. Furthermore, lower expression level of LAMA2 mRNA was shown to be associated with an increased risk of death in covariate-adjusted analyses. CONCLUSIONS: Our data provide a better understanding of pediatric ependymoma and suggests the presence of plausible molecular biomarkers connected with the outcome.
