MGMT methylation assessed by methyl-BEAMing technique is a prognostic and predictive biomarker in glioblastoma and metastatic colorectal cancer patients

Background: O6-methylguanine DNA methyltransferase (MGMT) silencing by promoter methylation is a common alteration found in different cancer types. This has been shown to be both a prognostic and a predictive marker of sensitivity to alkylating agent-based therapy like dacarbazine and temozolomide in glioblastoma. However in other malignancies its value remains controversial. This might be due to sampling issues, tumor heterogeneity or to the use of inadequate detection methods. In this study, we present a new assay to reliably measure MGMT methylation both in tumor and plasma samples. Material and Methods: Methylation of MGMT has been assessed by an ultra-sensitive digital PCR technique, in which a two-step PCR is followed by detection via fluorocytometer (Methyl-BEAMing). Results were compared to two other commonly used techniques (Methylation Specific PCR, MSP and pyrosequencing). Two samples datasets have been evaluated: tumors from a cohort of 98 newly diagnosed glioblastoma patients from the pre-temozolomide era, and specimens from a cohort of 68 metastatic colorectal cancer patients treated with dacarbazine in a phase II clinical trial (DETECT-01 trial, EUDRACT number 2011–002080−21). The prognostic and/or predictive value of MGMT methylation has also been evaluated. As a proof of concept, the three methods were assessed in a subset of colorectal cancer patients’ plasma derived DNA to evaluate their performance as a liquid biopsy test. Results: Methyl-BEAMing showed high reproducibility across independent experiments, as well as high sensitivity (up to 0.09% methylation detected) and specificity. In the glioblastoma cohort, Methyl-BEAMing methylated status (>50%) was associated with a decreased hazard ratio for death (HR = 0.35; p < 0.0001) compared to MSP (HR = 0.54; p = 0.006) or pyrosequencing (HR = 0.61; p = 0.059). In mCRC, tissue where tumor heterogeneity is possibly higher, both Methyl-Beaming and pyrosequecing assays provided better prediction of objective response to dacarbazine than MSP. Progression free survival was also improved in metastatic colorectal cancer with methylated status when samples were assessed with Methyl-BEAMing (p = 0.0012) or pyrosequencing (p = 0.0005). Quantitative evaluation of MGMT methylation in circulating tumor DNA was effective with Methyl-BEAMing. Conclusions: MGMT methylation testing based on BEAMing technology outperforms commonly used methods and might allow the non-invasive follow-up of patients, upon alkylating agent treatments using blood circulating DNA.