Identification of clinical prognosis features and significant DNA methylation regulation in pineoblastoma.

ABSTRACT

BACKGROUND: Pineoblastoma (PB) represents a great challenge for clinical management due to lack of a specific therapeutic regimen. This study aims to identify relevant prognostic factors and potential treatment targets by mining public databases. METHODS: The clinical characteristics and survival data of PB patients were obtained from the SEER database between 2000 and 2019 for Cox regression analysis and nomogram construction. The PB's DNA methylation data was acquired from two GEO datasets, GSE133801 and GSE215240, for bioinformatics analysis. RESULTS: Of 383 PB patients, Cox univariate analysis unveiled that male gender (p = 0.017), age younger than 3 years at diagnosis (p < 0.001) and absence of radiotherapy (p < 0.001) correlated with poorer overall survival (OS), the subsequent multivariate analysis confirmed sex (p = 0.036), age (p < 0.001) and radiotherapy (p = 0.005) as independent prognostic factors for OS. A nomogram showed robust predictive accuracy as evidenced by AUC values (1-year OS 0.774, 3-year OS 0.692, 5-year OS 0.643). DNA methylation analysis observed tumor hypomethylation, notably in promoter regions. Later, the GO enrichment analysis of aberrantly methylated genes indicated associations with embryonic organ development, cellular membrane composition and DNA-binding transcription, while KEGG analysis revealed enrichment in tumor-associated MAPK, calcium and RAS signaling pathways. CONCLUSIONS: The prognosis of PB is closely associated with sex, age and receipt of radiotherapy, potentially linked to aberrations in the RAS and MAPK signaling pathways. The individual case suggests that dasatinib and trametinib are potential targeted therapies for improving PB prognosis.