Despite recent therapeutic advances, the prognosis of patients afflicted by glioblastoma remains poor, with a progression-free survival in the range of months, even with multimodal therapy including surgery, radio-and chemotherapy. Temozolomide (TMZ), an oral alkylating agent, has demonstrated activity against recurrent and newly diagnosed glioblastoma (Yung et al. 2000;Stupp et al. 2005), and is being used as the standard of care for newly diagnosed glioblastoma since 2005.The DNA repair protein O 6 -methylguanine methyltransferase (MGMT) removes O 6 -alkyl-guanine adducts from DNA through irreversible binding and degradation, thereby minimizing the DNA-damaging effects of alkylating agent chemotherapy (Wang et al. 1996;Phillips et al. 1997 Abstract Temozolomide (TMZ) is an alkylating chemotherapeutic agent that prolongs the survival of patients with glioblastoma. Clinical benefit is more prominent in patients with methylation of the O 6 -methyl-guanine DNA methyltransferase (MGMT) promoter. However, all patients eventually suffer from tumor progression because their tumors become resistant to TMZ. Here, we modeled acquired TMZ resistance in glioma cells in vitro to identify underlying molecular mechanisms. To this end, the glioma cell lines LNT-229, LN-308, and LN-18 were exposed repetitively to increasing concentrations of TMZ to induce a stable resistant phenotype (R) defined by clonogenic survival assays. The molecular mechanisms mediating acquired resistance were assessed by immunoblot, PCR, and flow cytometry. Rescue experiments were performed with siRNA-mediated candidate gene silencing. We found in LN-18 cells constitutively expressing MGMT a strong up-regulation of MGMT levels in TMZ-resistant cells. TMZ resistance in the MGMT-negative cell lines LNT-229 and LN-308 was not associated with de novo expression of MGMT. Instead, we found a down-regulation of several DNA mismatchrepair proteins in resistant LNT-229 cells. A TMZ-resistant phenotype was also achieved by silencing selected DNA mismatch repair proteins in parental LNT-229 cells. No obvious mechanism of resistance was identified in the third cell line, LN-308, except for reduced methylation of LINE-1 repetitive elements. In conclusion, we demonstrate that different molecular mechanisms may contribute to the development of acquired TMZ resistance in glioma cells, indicating the need to develop distinct strategies to overcome resistance.

J. Neurochem. (2011) 118, 449–457. Abstract The prognosis of patients afflicted by glioblastoma remains poor. Biomarkers for the disease would be desirable in order to allow for an early detection of tumor progression or to indicate rapidly growing tumor subtypes requiring more intensive therapy. In this study, we investigated whether a blood‐derived specific miRNA fingerprint can be defined in patients with glioblastoma. To this end, miRNA profiles from the blood of 20 patients with glioblastoma and 20 age‐ and sex‐matched healthy controls were compared. Of 1158 tested miRNAs, 52 were significantly deregulated, as assessed by unadjusted Student′s t‐test at an alpha level of 0.05. Of these, two candidates, miR‐128 (up‐regulated) and miR‐342‐3p (down‐regulated), remained significant after correcting for multiple testing by Benjamini–Hochberg adjustment with a p‐value of 0.025. The altered expression of these two biomarkers was confirmed in a second cohort of glioblastoma patients and healthy controls by real‐time PCR and validated for patients who had received neither radio‐ nor chemotherapy and for patients who had their glioblastomas resected more than 6 months ago. Moreover, using machine learning, a comprehensive miRNA signature was obtained that allowed for the discrimination between blood samples of glioblastoma patients and healthy controls with an accuracy of 81% [95% confidence interval (CI) 78–84%], specificity of 79% (95% CI 75–83%) and sensitivity of 83% (95% CI 71–85%). In summary, our proof‐of‐concept study demonstrates that blood‐derived glioblastoma‐associated characteristic miRNA fingerprints may be suitable biomarkers and warrant further exploration.

No abstract

clinicaltrials.gov Identifier: NCT00717210.

The results of this analysis do not justify the use of VPA or LEV for reasons other than seizure control in patients with newly diagnosed glioblastoma outside clinical trials.