Influence of molecular classification in anaplastic glioma for determining outcome and future approach to management

Introduction Assess survival of patients with anaplastic glioma (AG) and the relationship to molecular subtype. Methods Patients with AG managed with IMRT between 2008 and 2014 were entered into a prospective database assessing relapse-free survival (RFS) and overall survival (OS). Isocitrate dehydrogenase (IDH) mutations were assessed prospectively from 2011, and subsequent testing of historical patients allowing categorisation under WHO 2016 classification as anaplastic astrocytoma IDH wild type (AAwt), anaplastic astrocytoma IDH mutated (AAmut), anaplastic oligodendroglioma (AOD) or other glial tumour (OTH). Kaplan-Meier estimates of survival distribution were calculated for the primary endpoint of overall survival and Log-rank test used to determine associated factors. Results One hundred and fifty-six patients were included with median follow-up for survivors of 4.7 years. Fifty-six per cent were managed after initial diagnosis, whilst 18% received IMRT at second or later relapse. Seventy-three per cent had temozolomide as part of initial therapy. A total of 118 or 75% of patients had IDH mutated glioma, of which 61 were AOD and 57 AAmut. There were 68 relapses and 52 deaths for a 6yrRFS of 51.2% and 6yrOS of 62.5%. AAwt was associated with worse survival (P < 0.001); and delay of RT until second or later relapse (P = 0.03). Within the 118 patients with IDH mutated tumours, 6yrOS for AOD and AAmut were 90.0% and 62.5%, respectively (P = 0.003). Also two or more craniotomies (P < 0.001), delayed RT (P = 0.006) and age <40 years (P = 0.022) were associated with worse survival on univariate analysis but only AAmut subtype and number of craniotomies on multivariate analysis. Conclusion Within AG, molecular classification predicts for survival, and should influence current decision-making.