Discriminating MGMT promoter methylation status in patients with glioblastoma employing amide proton transfer-weighted MRI metrics

作者:Jiang, Shanshan*; Rui, Qihong; Wang, Yu; Heo, Hye-Young; Zou, Tianyu; Yu, Hao; Zhang, Yi; Wang, Xianlong; Du, Yongxing; Wen, Xinrui; Chen, Fangyao; Wang, Jihong; Eberhart, Charles G.; Zhou, Jinyuan; Wen, Zhibo*
来源:European Radiology, 2018, 28(5): 2115-2123.
DOI:10.1007/s00330-017-5182-4

摘要

To explore the feasibility of using amide proton transfer-weighted (APTw) MRI metrics as surrogate biomarkers to identify the O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation status in glioblastoma (GBM). @@@ Eighteen newly diagnosed GBM patients, who were previously scanned at 3T and had a confirmed MGMT methylation status, were retrospectively analysed. For each case, a histogram analysis in the tumour mass was performed to evaluate several quantitative APTw MRI metrics. The Mann-Whitney test was used to evaluate the difference in APTw parameters between MGMT methylated and unmethylated GBMs, and the receiver-operator-characteristic analysis was further used to assess diagnostic performance. @@@ Ten GBMs were found to harbour a methylated MGMT promoter, and eight GBMs were unmethylated. The mean, variance, 50th percentile, 90th percentile and Width(10-90) APTw values were significantly higher in the MGMT unmethylated GBMs than in the MGMT methylated GBMs, with areas under the receiver-operator-characteristic curves of 0.825, 0.837, 0.850, 0856 and 0.763, respectively, for the discrimination of MGMT promoter methylation status. @@@ APTw signal metrics have the potential to serve as valuable imaging biomarkers for identifying MGMT methylation status in the GBM population. @@@ aEuro cent APTw-MRI is applied to predict MGMT promoter methylation status in GBMs. @@@ aEuro cent GBMs with unmethylated MGMT promoter present higher APTw-MRI than methylated GBMs. @@@ aEuro cent Multiple APTw histogram metrics can identify MGMT methylation status. @@@ aEuro cent Mean APTw values showed the highest diagnostic accuracy (AUC = 0.825).