Prognostic value of MGMT promoter status in non-resectable glioblastoma after adjuvant therapy

Background Methylation of MGMT promoter has been identified as a favourable predictive factor of benefit from XRT/TMZ → TMZ. Patients with non-resectable glioblastoma (GBM) generally exhibit a poor prognosis, even after XRT/TMZ. Few data are available concerning the predictive value of MGMT promoter methylation in this population. Methods This is an observational retrospective study in patients with malignant brain glioma, treated between June 2008 and October 2011 and followed up until April 2012 at the Neurosurgery-Neurotraumatology Unit of the University Hospital of Parma and at the Neurosurgery Unit of IRCCS "ASMN" of Reggio Emilia, Italy. The medical records of an overall number of 174 patients with a newly diagnosed GBM were reviewed. Volumetry analysis of the lesions was performed on pre- and post-operative neuroimaging by Voxar 3D Ebit AET software. The genetic characterization was performed on paraffin embedded tissue from all resected tumours. Isolation of nucleic acids, bisulfite modification of DNA, methylation-specific PCR and sequencing analyses were done mainly on fresh tissue from biopsy withdrawals. Within 3-4 weeks after either biopsy or surgery, patients were assigned to receive XRT/TMZ → TMZ: treatment included XRT (60 Gy in 30 fractions)/TMZ (daily dose of 75 mg/m2)/TMZ (150-200 mg/m2 per day for 5 days of every 28-day cycle). Results and discussion A total of 55 consecutive patients (23 men, 22 women) fulfilled inclusion criteria consisting of age over 18 years, supratentorial histologically proven primary malignant glioma, complete determination of the MGMT methylation status, no prior history of surgery, XRT and/or chemotherapy, adequate clinical and radiological follow-up no lesser than 6 months. Twenty-three patients underwent neuronavigation needle biopsy (B Group) and thirty-two patients were operated with craniotomy for tumour resection (R Group). The pre-operative mean age was similar between groups (61.7 ± 10.7 vs 60.3 ± 11.8 years in the B and R groups respectively; p > 0.05). The B groups showed a slightly lower KPS than the R Group (82.1 ± 17.3 vs 90.3 ± 14.1 respectively; p > 0.05). The mean pre-operative volume of the tumour did not differ between groups (46.2 ± 40.2 cm3 vs 44.1 ± 33.2 cm3 in the R Group and B Group respectively; p > 0.05). The MGMT promoter was methylated in 12 patients (51.2%) of B Group and in 17 patients (53.1%) of R Group. XRT/TMZ → TMZ was accomplished in 11 patients (47.8%) of B Group and in 24 patients (75%) of R Group; in 24/29 methylated patients (82.8%) and in 11/26 unmethylated patients (42.3%). Survival analysis of methylated vs unmethylated tumours was statistically significant (Log Rank Mantel Cox: 0.019 in B Group and 0.023 in R Group). In B Group the mean overall survival (OS) of methylated patients was 11.4 months (IC 95% 6.5-16.4) vs 4.8 months (95% IC, 2.6-7.0) of unmethylated patients. In R Group the mean OS was 21.7 months (95% IC, 16.9-26.6) for methylated patients and 14.0 months (95% IC, 8.5-19.4) for unmethylated patients. At the multivariate Cox regression analysis conducted on the total population (55 patients), XRT and TMZ were found to be predictive of OS. In the R Group, KPS, XRT and TMZ correlated with a better outcome. In the B Group, XRT and MGMT promoter methylation were favourably related with OS. Conclusion MGMT promoter unmethylation has a predominant unfavourable impact on clinical outcomes even in the subpopulation of patients with non-resectable GBM. The unmethylated MGMT promoter status could be considered the main predictor of poor prognosis in biopsied GBM, due to the greater probability of patients not having benefits from adjuvant therapies and not being able to accomplish XRT/TMZ → TMZ. The frameless neuronavigation biopsy technique is safe and effective for predictive evaluation and could help in treatment decision making.