Clinical trials have demonstrated a significantly prolonged median survival of GBM patients treated with TMZ associated to radiotherapy after surgical resection
[18–20]. In addition, a strong correlation between the MGMT gene promoter methylation status and the TMZ treatment effect and outcome was shown by Hegi et al.
, confirming previous results which correlated MGMT inactivation and clinical response to alkylating agents
[16, 21]. Criniere et al.
 showed that while the MGMT promoter methylation status had no impact on the OS of GBM patients treated with alkylating agents, it did have an impact on those treated with chemo-radiotherapy, suggesting that the prognostic impact of this methylation is dependent on therapeutic modalities. In fact, Weller et al.
 analyzed patients with GBM treated with radiotherapy, with chemotherapy and with both radiotherapy and chemotherapy and showed that the methylation status was a significant prognostic factor for OS and PFS only in the subgroup of patients treated with radiotherapy and concurrent TMZ. These results have been recently confirmed by Cao et al.
, Minniti et al.
 and Kim et al.
. Recently, the methylation status was analyzed in elderly patients with GBM and was found to be strictly correlated with the pattern of, and time until, GBM recurrence
, but not with its evolution after recurrence
. Rivera et al.
 demonstrated that MGMT promoter methylation was a predictor of survival in GBM treated exclusively with radiotherapy. In addition, a randomized phase III trial comparing standard adjuvant TMZ with a dose-dense schedule in newly diagnosed GBM confirmed the prognostic significance of MGMT methylation in GBM
. However, previous studies considered promoter methylation of the MGMT gene not to be a reliable prognostic factor of responsiveness to alkylating agents in glioblastomas
. Recent studies have also questioned the role of promoter methylation status of MGMT in GBM. Yachi et al.
 failed to establish this correlation in a larger number of patients and showed that neither MSP-MGMT methylation nor immunohistochemical MGMT expression had prognostic implications in GBM patients. Similarly, Tang et al.
 did not find a correlation between progression-free survival and MGMT promoter methylation in chinese patients. In this context, we conducted a retrospective study in a homogeneous series of patients diagnosed with GBM treated with radiotherapy and concurrent temozolomide using MSP, which has been proposed as the most convenient technique in clinical routine diagnostics
. We found that the percentage of methylation of the MGMT promoter to be higher than that reported by Hegi et al.
 but similar to those reported by other authors
[22, 34]. The MGMT methylation status was clearly confirmed as an independent prognostic factor of GBM PFS and OS. In addition, patients with a methylated MGMT promoter status had higher response rates to TMZ and radiation therapy compared to those with a non-methylated status, so that MGMT methylation was a predictive factor for radiologic response
On the other hand, the association between MGMT protein expression and promoter methylation in vivo has been widely discussed. Several studies have reported a significant association of high MGMT expression and poor prognosis of patients
[7–10, 35–38]. We analyzed the expression of MGMT by immunochemistry using a digital quantitative method to avoid observer variability
. Our study clearly showed no correlation between MGMT expression and MGMT promoter methylation, supporting the finding observed by Rodriguez et al.
 and Uno et al.
, or overall survival or radiological response coinciding with those findings of Preusser et al.
. The regulation of the MGMT gene is a complex phenomenon in which promoter hypermethylation is one of the factors that influence the final expression of the protein. In this context, the recent demonstration that there is discordance between MGMT promoter methylation and levels of MGMT mRNA expression suggests that other mechanisms may regulate the expression of this enzyme
. It also possibly suggests that promoter methylation and expression alone are not sufficient to provide information on the expected clinical course in patients with malignant glioma who receive chemotherapy with alkylating agents. In fact, a comprehensive study undertaken by sequencing the MGMT gene promoter has shown a strong correlation between the methylation site and treatment response
CD133, a five-transmembrane cell surface protein found in human stem cells from various sources including the central nervous system
, has been proposed to detect GBM CSCs. Considering the inherent resistance of CSCs to chemotherapy and radiotherapy
[14, 15], it has been hypothesized that the clinical outcome will be inversely related to the presence of CSC-marker-positive cells. However, the impact of the presence of CSC in the clinical progression of tumors is unknown. Few studies exploring the prognostic value of CD133 expression as a marker of CSC in GBM have been undertaken; most were heterogeneous and used different methods (QRT-PCR, immunohistochemistry, FACS) so that comparisons are difficult. Intense CD133 expression was detected in high-grade oligodendroglial tumors
 and in grade II-IV gliomas
, both with poor prognoses. Similar results were founded in grade IV gliomas where CD133 expression was related to OS and PFS
. Murat et al.
 provided evidences that the glioblastoma stem cell phenotype (including CD133 expression) correlated with chemoradiotherapy resistance and patient survival. In a study with very homogeneous samples of GBM, high expression of CD133 was found to be an unfavorable prognostic factor
. In addition, an in vitro study using CSC obtained from 44 GBM patients to evaluate CD133/Ki67 expression by immunohistochemical analysis, concluded that CD133+ was correlated to survival
. However, the same authors demonstrated that high expression of CD133 was associated with a better prognosis when detected by FASCcan
. Similar results were previously found by Joo et al.
 who described CD133 as a favourable prognostic factor for GBM. In this context, our study clearly demonstrates that CD133 has no implication in the prognoses of GBM patients supporting similar results of Kim et al.
. Finally, the prognostic impacts of variations of MGMT promoter methylation or MGMT and CD133 protein expression after treatment, are not known. A previous study determined that MGMT methylation status has no prognostic value after GBM recurrence
. Our preliminary study in samples obtained before and after treatment from only ten patients with recurrent GBM suggested that there are no differences between the analyzed variables although the low number of patients does not allow to obtain statistically significant conclusions.