Correlation of epigenetic markers with glioblastoma patients’ survival and molecular subtypes: insights from methylation-specific MLPA analysis

Objectives Glioblastoma (GB) is a heterogeneous group of tumors with poor patient’s outcome. The epigenetic markers could influence patient’s outcome and add more precision in sub-typing. Methods The study aims to determine a correlation between the methylation status of 65 genes and the overall survival (OS) in 50 GB Tunisian patients using methylation specific-multiplex ligation-dependent probe amplification (MS-MLPA) technique. Results OS was significantly longer for patients with tumors harboring unmethylated tumor suppressor genes ATM , BRCA1 , BRCA2 , TP53 and TP73 and the DNA repairing gene MSH6 . Furthermore, simultaneous unmethylation of ATM , BRCA2 , CD44 and VHL genes was associated with GB presenting normal EGFR status. This subgroup exhibits a better prognosis (OS=15 months, p=0.005). Our results also showed that, combined methylation of TP73 , THBS1 , GSTP1 and ESR1 genes in amplified EGFR GB subtype resulted in a poor prognosis group (OS=3 months, p=0.041). Simultaneous methylation of HLTF and SFRP5 genes was associated with wild-type IDH1 GB defining therefore a very poor prognosis group (OS=1 month, p=0.026). Besides, BRCA1 methylation in wild-type IDH1 group was significantly associated with poor prognosis (OS=6 months, p=0.002). Interestingly, RBM14 and PCCA genes were co-methylated in 80 % of prolonged survival cases (GB+). In absence of EGFR amplification and with unmethylation of BRCA1 , BRCA2 , ATM , VHL , CD44 , HLTF and SFRP5 genes, GB+ tumors seemed to respond better to treatment, avoiding the relapse and conferring prolonged survival (>36 months). Conclusion Combining genes methylation status of GB with EGFR and IDH1 profiles will refine subtyping, predict patient’s outcomes and guide personalized therapy.