LAUSR

IDH-wild-type and GBM IDH-mutated. Frequently, some entities, different from GBM, demonstrate EGFR amplification, TERT+, mutation of P53, PTEN and NF1. These diseases are characterized by a particularly poor glioblastoma-like prognosis, even in the absence of micro vascular proliferation or necrosis. Moreover, the Cancer Genome Atlas (TCGA) study [2] has already proposed a GBM subclassification based on gene-level expression profiles. The authors identified four molecular subgroups of GBM: classical, proneural, neural and mesenchymal. From this report EGFR amplification is a specific feature of classical GBM subtype. Up to now all these biologically different entities have been treated with a single treatment protocol and this choice can only obtain suboptimal results both in terms of disease free and overall survival. The different response of these entities to the same treatment has already been demonstrated, at least once, in the case of Bevacizumab as first line treatment with radiotherapy and temozolomide (Avaglio protocol). Thus, the main objective of the modern neuro-oncology should be to increase the knowledge of both prognostic and predictive factors. We read with great interest the article by Tini et al. [3] pertaining the challenging quest for predictive factors. In this study the authors conclude that the considerable incidence of marginal and distant patterns of recurrence in GBM patients with high-EGFR expression might be relevant for the contouring of radiotherapy target volumes. More, the effect of EGFR expression on the response to different temozolomide schedules can be outlined, as showed in the published retrospective experience where patients with higher EGFR expression could have better outcome when TMZ was used concomitant to and after RT with a metronomic schedule [4]. To the Editor,