Glioblastoma: from volumetric analysis to molecular predictors

BACKGROUND:\r\nDespite decades of therapeutic and molecular refinements, the prognosis of patients with glioblastoma (GBM) still remains unfavorable. Integrative clinical studies allow a better understanding of the natural evolution of GBM. To assess independent predictors of overall survival (OS) and progression free survival (PFS) clinical, surgical, molecular and radiological variables were evaluated. A novel preoperative volumetric magnetic resonance imaging (MRI) index for tumor prognosis in GBM patients was investigated.\r\nMETHODS:\r\nA cohort of 195 cases of patients operated for newly GBM were analyzed. Extent of tumoral resection (EOR), tumor growth pattern, expressed by preoperative volumetric ΔT1-T2 MRI index, molecular markers such as O6-methylguanine-DNA methyltransferase (MGMT) methylation and isocitrate dehydrogenase 1/2 (IDH1/2) mutation, were analyzed. Analysis of survival was done using Cox-proportional hazard models.\r\nRESULTS:\r\nThe 1-, 2- years estimated OS and PFS rate for the whole population were 61% and 27%, 38% and 17%, respectively. A better survival rate, both in terms of survival and tumor progression, was observed in patient with higher EOR (p=0.000), younger age (p=0.000), MGMT methylation status (p=0.001) and lower preoperative ΔT1-T2 MRI index (p=0.004). Regarding the tumor growth pattern a cut-off value of 0.75 was found to discriminate patient with different prognosis. Patients with a preoperative ΔT1-T2 MRI index <0.75 had a 1-year estimated OS of 67%, otherwise patients with a preoperative ΔT1-T2 MRI index >0.75 hada 1-year estimated OS of 34%.\r\nCONCLUSIONS:\r\nIn this investigation longer survival is associated with younger age, EOR, promoter methylation of MGMT and preoperative tumor volumetric features expressed by ΔT1-T2 MRI index. The preoperative ΔT1-T2 MRI index could be a promising prognostic factor potentially useful in GBM management. Future investigations based on multiparametric MRI data and next generation sequences analysis, may better clarify this result.