Background: The utility of O6-methylguanine-DNA methyltransferase (MGMT) gene promoter methylation status as a prognostic marker in patients with glioblastoma (GBM) has been established. However, the number of CpG sites that must be methylated to cause transcriptional silencing remains unclear, and no significant consensus exists on the optimal method of assessing MGMT methylation. We developed a new high-performance liquid chromatography (HPLC) method that enables accurate analysis of DNA methylation levels using long PCR products. In the present study, we analyzed the MGMT methylation status of 28 isocitrate dehydrogenase-wild-type GBMs treated with temozolomide using ion-exchange HPLC and set the optimal cutoff values.
Results: We designed three primers for separate regions (regions 1-3) that had 21 to 38 CpGs for PCR and validated the MGMT promoter methylation status using frozen samples. There was a strong correlation between HPLC and bisulfite sequencing results (R = 0.794). The optimal cutoff values for MGMT methylation in HPLC were determined to allow differentiation of patient prognosis by receiver operating characteristic curve analysis. The cutoff values were 34.15% for region 1, 8.84% for region 2, and 36.72% for region 3. Kaplan-Meyer curve analysis estimated that the most differentiated prognosis was enabled in the setting of 8.84% methylation of MGMT in region 2. Progression-free survival and overall survival were significantly longer for patients in this setting of region 2 methylation (p = 0.00365 and p = 0.00258, respectively).
Conclusions: The combination of our HPLC method and the original primer setting provides a new standard method for determination of MGMT methylation status in patients with GBM and is useful for refining MGMT-based drug selection.
Keywords: DNA methylation; Glioblastoma; HPLC; MGMT; Prognostic factor.