Surgical extent impacts the value of the established prognosticators in glioblastoma patients: A prospective translational study in Asia

Background While multimodal treatments have survival benefits in patients with glioblastoma, tumour volume resection still remains the most effective treatment. Methods In this prospective translational study, we analysed the clinical values of the established survival predictors in the context of extensive tumour resection (at least near-total resection) in 234 newly diagnosed glioblastoma patients. Common survival factors such as adjuvant therapy modality, O6-methylguanine DNA methyltransferase (MGMT) and isocitrate dehydrogenase 1 (IDH1) were analysed. Results The more extensive resection resulted in a favourable outcome, especially the median progression-free survival up to 11.1 months. Age at diagnosis, the initial alkylating radiochemotherapy and MGMT promoter status were correlated with survival in univariate analysis, but their independent predictive values were lost upon multivariate analysis. Multivariate analysis identified the only independent prognosticators for a longer overall survival were gross total resection (relative risk [RR] = 0.36; P = 0.003) and higher Karnofsky Performance Status score (RR = 0.41; P = 0.008), and gross-total resection (RR = 0.56; P = 0.050), higher Karnofsky Performance Status score (RR = 0.52; P = 0.028) and IDH1 mutation (RR = 0.42; P = 0.011) for a favourable progression- free survival. Interestingly, in the subgroup of patients receiving gross-total resection, neither MGMT nor IDH1 effectively predicted overall survival or progression-free survival; and also radiotherapy alone did not significantly improve the outcome though alkylating radiochemotherapy showed a strong survival advantage. Further analysis showed an additive effect of surgical extent and molecular predictor IDH1 mutation. Conclusion Surgical extent most significantly impacts the value of molecular and clinical predictors, and therefore should be taken into full account when assessing a new therapy or stratification variable in glioblastoma multiformes.