The molecular classification of astrocytic tumors.

Aim: This study will explore the genetic and epigenetic alterations in astrocytomas, and identify the critical molecular signatures and signaling pathways for prognosis assessment by multiplatform comprehensive analysis.

Method: We performed integration analyses of incorporating DNA methylation, mRNA expression, microRNA expression, and long non-coding RNA (lncRNA) expression in 33 astrocytic tumor tissues and 9 non-tumor brain tissues.

Result: We observed that 11,795 DNA methylation sites, 3,627 genes, 136 microRNAs, and 3,334 lncRNAs were significantly differential between tumors and non-tumor brain tissues, and the filtered signatures through comprehensive analysis were significantly enriched in calcium signaling pathway. Furthermore, four signatures involved in calcium signaling pathway and age could contribute to predicting the patients' overall survival. Additionally, we identified differentially expressed signatures between IDH-mutated and IDH wild-type astrocytic tumors, and complement and coagulation cascades pathway was the most significant pathway in functional enrichment analysis using multiplatform data. The IDH wild-type astrocytomas were divided into two subtypes by Cluster of Cluster (CoC) analysis, one of which was enriched for astrocytomas overexpressed in chemokine signaling pathway.

Conclusion: The calcium signaling pathway played a key role in astrocytoma tumorigenesis and prognosis. IDH mutation was a vital biomarker, and resulted in the change of expression level in complement and coagulation cascades pathway. The chemokine signaling pathway could characterize subtypes of IDH wild-type astrocytomas.