MGMT promoter methylation status for glioblastoma: defining the clinically relevant cut-off value for pyrosequencing

MGMT promoter methylation is a key predictive biomarker for response to alkylating agents in glioblastoma. However, there is no consensus regarding optimal analytical method or cut-off. This study aimed to define a clinically relevant survival-based cut-off value for MGMT using a standardized pyrosequencing assay.

Patients from five Swedish university hospitals with MGMT promoter methylation status analyzed using the Therascreen MGMT Pyro Kit, investigating CpGs 76-79, were identified. Glioblastoma patients treated with radiotherapy and concomitant temozolomide were selected from the Swedish CNS Tumor Registry. Quantitative MGMT status, both mean value and percentage of methylation for each individual CpG, was analyzed using an unsupervised bimodal normal mixture model and a survival-informed approach adjusted for established prognostic factors.

A total of 451 patients were included. The unsupervised model identified a cut-off at ≥ 11% for methylated MGMT, whereas the supervised, survival-informed analysis defined a cut-off for unmethylated tumors at ≤ 8%. A small intermediate gray zone (> 8% to < 11%) was observed. No individual CpG provided superior predictive performance compared to the mean of all CpGs.

A clinically relevant, survival informed three-tier classification of MGMT promoter methylation is proposed for the pyrosequencing kit analyzing CpGs 76-79 in glioblastoma. A threshold of ≤ 8% identifies a subgroup with truly unmethylated tumors, while ≥ 11% defines a clearly methylated group. The intermediate range represents a gray zone. These cut-offs support individualized therapy by improving clinically relevant stratification and facilitate future clinical trials focusing on patients with truly unmethylated MGMT.