11C-methionine positron emission tomography for target delineation of recurrent glioblastoma in re-irradiation planning

Aim To define the optimal margin on MRI scans in the re-radiation planning of recurrent glioblastoma using methionine positron emission tomography (MET-PET). Background It would be very useful if the optimal margin on MRI to cover the uptake area on MET-PET is known. Materials and Methods CT, MRI, and MET-PET were performed separately over the course of 2 weeks. Among the MRI scans, we used the contrast-enhanced T1-weighted images (Gd-MRI) and T2-weighted images (T2-MRI). The Gd-MRI-based clinical target volume (CTV) (CTV-Gd) and the T2-MRI-based CTV (CTV-T2) were defined as the contrast-enhanced area on Gd-MRI and the high intensity area on T2-MRI, respectively. We defined CTV x mm (x=5, 10, 15, 20) as x mm outside the CTV. MET-PET-based CTV (CTV-MPET) was defined as the area of accumulation of MET-PET. We calculated the sensitivity and specificity of CTV-Gd and CTV-T2 following comparison with CTV-MPET, which served as the gold standard in this study. Results The sensitivity of CTV-T2 5mm (98%) was significantly higher than CTV-T2 (87%), and there was no significant difference in the sensitivity between CTV-T2 5mm and CTV T2 10, 15, or 20mm. The sensitivity of CTV-Gd 20mm (97%) was lower than that of CTV-T2 5mm (98%). Conclusions A margin of at least 5mm around the high intensity area on T2-MRI is necessary in the target volume delineation of recurrent glioblastoma for the coverage of MET-PET findings in re-radiation therapy planning.