Article
A quantitative real-time PCR high resolution melting (HRM) method for evaluating the O6-methylguanine-DNA-methyltransferase (MGMT) promoter methylation status in malignant gliomas is superior to conventional methylation specific PCR in predicting clinical outcome and survival in patients with malignant glioma
Search Medline for
Authors
-
Olivier J. Switzeny
- Universitätsmedizin der Johannes Gutenberg Universität, Mainz,,Institut für Toxicologie
-
Mirjam Renovanz
- Neurochirurgische Klinik
-
Clemens Sommer
- Institut für Neuropathologie
-
Markus Christmann
- Universitätsmedizin der Johannes Gutenberg Universität, Mainz,,Institut für Toxicologie
-
Alf Giese
- Neurochirurgische Klinik
-
Bernd Kaina
- Universitätsmedizin der Johannes Gutenberg Universität, Mainz,,Institut für Toxicologie
| Published: | June 2, 2015 |
|---|
Outline
Text
Objective: MGMT promoter status is an important prognostic factor in high-grade glioma patients and indispensable for decision making in neurooncological therapy. Currently, the MGMT promoter methylation is determined by methylation-specific PCR (MSP). The development of new approaches to improving accuracy of assessments of MGMT promoter methylation status is a matter of high priority. The main objective of this study was to compare the accuracy of a quantitative closed tube real-time PCR high resolution melting (HRM) approach with that of conventional MSP.
Method: First, we validated the method in 14 GBM cell lines, from which promoter methylation was quantified by MSP and HRM and correlated with the MGMT enzyme activity. Genomic DNA was isolated from 100 formalin-fixed, paraffin-embedded (FFPE) tumor samples of patients with high-grade glioma (°III, °IV) previously diagnosed by MSP and analyzed for the MGMT promoter methylation status by HRM. HRM readouts were classified into two groups upon a methylation threshold. Kaplan-Meier estimates for progression-free survival (PFS) and overall survival (OS) were calculated for both methods. The predictive value of HRM and MSP readouts was evaluated by multivariate analysis (Cox regression). A ROC curve was generated to graph sensitivity and specificity of HRM and MSP to predict PFS >12 months. Sociodemograpic and clinical parameters (age, gender, administered therapy, KPS) were retrospectively assessed using our electronic patient database.
Results: In GBM lines, a strong inverse correlation was found between the results of HRM and MGMT enzyme activity. In patient samples, Kaplan-Meier estimates for PFS (p=0.004 vs. P=0.108) and OS (p=0.001 vs. P=0.019) resulted in better group separation by HRM. For the patient's group that has been categorized as MGMT methylated by HRM, Cox regression revealed that the hazard ratio for progression of the tumor was 0.495 (95 % CI, 0.270 to 0.905). MSP could not predict a significant hazard ratio for progression. The ROC curve for HRM confirms an advantage compared to MSP. A MGMT methylation threshold with clinical importance was found.
Conclusions: HRM is a straightforward, highly accurate and cost-effective alternative to MSP for predicting clinical outcome in glioma patients. Furthermore HRM was found to be superior to MSP in predicting PFS and providing a clearer separation between MGMT methylation sub-groups. Therefore, we wish to propagate HRM as a predictive tool in glioma therapy.
