422P - Prognostic value of pyrosequencing of MGMT promoter methylation in patients with glioblastoma treated by temozolomide and hypofractionated high-dos...

Date 28 September 2014
Event ESMO 2014
Session Poster Display session
Topics Anticancer Agents
Central Nervous System Malignancies
Personalised/Precision Medicine
Surgical Oncology
Basic Principles in the Management and Treatment (of cancer)
Biological Therapy
Radiation Oncology
Presenter Toshihiko Iuchi
Citation Annals of Oncology (2014) 25 (suppl_4): iv137-iv145. 10.1093/annonc/mdu330
Authors T. Iuchi1, S. Yokoi2, M. Ohira3, R. Shimada4, Y. Hasegawa1, K. Kawasaki1, T. Sakaida1
  • 1Neurological Surgery, Chiba Cancer Center, 2608717 - Chiba/JP
  • 2Gene Diagnosis, Chiba Cancer Center, 2608717 - Chiba/JP
  • 3Cancer Genomics, Chiba Cancer Center Research Institute, 2608717 - Chiba/JP
  • 4Market Development Ls Marketing, QIAGEN K.K., 1040054 - Tokyo/JP



Methylation of the O6-methlyguanine-DNA methyltransferase (MGMT) gene promoter is a key prognostic factor for patients with glioblastoma (GBM) treated by temozolomide (TMZ). We evaluated the significance of pyrosequencing (PSQ) assay of this gene on prediction of tumor control and survival of patients, and compared it with methylation specific PCR (MSP).


Newly diagnosed and pathologically confirmed 47 GBMs treated by TMZ and hypofractionated high-dose IMRT (8.5Gy x 8Fx) were enrolled. Methylation levels of 4 CpGs (76-79) of MGMT were assessed by PSQ. A double normal distribution model was fitted to the data to define the optimal threshold of methylation level at each CpG site. Differences in progression-free survival (PFS) and overall survival (OS) of patients between methlylated (Met) and unmethylated (UnMet) cases were evaluated at each CpG or combination of CpGs, and the results were compared with those by MSP.


There was no discrepancy between PSQ and MSP in diagnosis of methylation status, when methylation was observed at all or none of 4 CpG sites by PSQ. However, in cases only 1 or 2 CpGs exhibited methylation by PSQ, 81%(13/16) of the cases were diagnosed as UnMet by MSP. PFS of Met patients was significantly longer than UnMet ones regardless of the method of assessment. Hazard ratios (HRs) at CpG 76, 77, 78, and 79 were 7.24 (95%CI: 2.68-25.40), not available (no recurrence in Met cases), 9.73 (2.05-174.24), and 7.06 (2.94-19.69), respectively, and were better than 2.97 (1.19-9.02) in classification by MSP. These tendencies were also observed in OS: HRs at CpG 76, 77, 78, and 79 were 4.20 (1.56-14.59), 3.67 (1.07-23.05), 3.75 (1.09-23.50), and 2.81 (1.19-7.33), respectively, and were better than 2.53 (1.02-7.25) by MSP. As for combination of CpGs, patients with at least 1 or 2 Met CpGs demonstrated better PFS (HR: 7.24, 95%CI: 2.68-25.40) and OS (HR: 4.20, 1.56-14.59) than those with no Met CpG.


PSQ was feasible to assess MGMT methylation, because quantitative evaluation was appropriate to standardized assessment, and showed more significant prognostic value (especially in multiple CpGs) than MSP.


T. Iuchi: "MGMT Pyro Kits" used in this study were provided by Qiagen K.K.All other authors have declared no conflicts of interest.