Diagnostic and Prognostic Value of 11C-Methionine PET for Nonenhancing Gliomas

Abstract

Background and purpose: Noninvasive radiologic evaluation of glioma can facilitate correct diagnosis and detection of malignant transformation. Although positron-emission tomography is considered valuable in the care of patients with gliomas, (18)F-fluorodeoxyglucose and (11)C-methionine have reportedly shown ambiguous results in terms of grading and prognostication. The present study compared the diagnostic and prognostic capabilities of diffusion tensor imaging, FDG, and (11)C-methionine PET in nonenhancing gliomas.

Materials and methods: Thirty-five consecutive newly diagnosed, histologically confirmed nonenhancing gliomas that underwent both FDG and (11)C-methionine PET were retrospectively investigated (23 grade II and 12 grade III gliomas). Apparent diffusion coefficient, fractional anisotropy, and tumor-to-normal tissue ratios of both FDG and (11)C-methionine PET were compared between grade II and III gliomas. Prognostic values of these parameters were also tested by using progression-free survival.

Results: Grade III gliomas showed significantly higher average tumor-to-normal tissue and maximum tumor2-to-normal tissue than grade II gliomas in (11)C-methionine (P = .013, P = .0017, respectively), but not in FDG-PET imaging. There was no significant difference in average ADC, minimum ADC, average fractional anisotropy, and maximum fractional anisotropy. (11)C-methionine PET maximum tumor-to-normal tissue ratio of 2.0 was most suitable for detecting grade III gliomas among nonenhancing gliomas (sensitivity, 83.3%; specificity, 73.9%). Among patients not receiving any adjuvant therapy, median progression-free survival was 64.2 ± 7.2 months in patients with maximum tumor-to-normal tissue ratio of <2.0 for (11)C-methionine PET and 18.6 ± 6.9 months in patients with maximum tumor-to-normal tissue ratio of >2.0 (P = .0044).

Conclusions: (11)C-methionine PET holds promise for World Health Organization grading and could offer a prognostic imaging biomarker for nonenhancing gliomas.

Fig 1.

Schematic overview of the image-analysis process. High-intensity lesions on T2/FLAIR imaging were semiautomatically segmented in 3D by an image-intensity threshold. The segmented voxels of interest were applied to registered PET images and ADC fractional anisotropy maps, followed by calculation of the average and maximum or minimal values of each parameter within the VOI.

Fig 2.

T/N_ave (A) and T/N_max (B) of FDG-PET in patients with grade II and III gliomas. No significant difference was seen between grade II and III gliomas (P = .11 and 0.72, respectively).

Fig 3.

T/N_ave (A) and T/N_max (B) of MET PET in patients with grade II and III gliomas. Grade III gliomas show significantly higher T/N_ave and T/N_max than grade II gliomas (P = .013, P = .0017, respectively).

Fig 4.

Receiver operating characteristic analysis of MET PET and FDG-PET T/N_max was used for discriminating grade III from grade II gliomas. The area under the curve was 0.83, and the sensitivity and specificity were 83.3% and 73.9% at a cutoff T/N_max of 2.0 for MET PET. FDG-PET was unable to discriminate grade III glioma from grade II.

Fig 5.

Progression-free survival analysis by tumor grade and MET PET results. Progression-free survival is shown in Kaplan-Meier curves according to tumor grade (A). Patients with T/N_max of <2.0 showed prolonged progression-free survival compared with those with T/N_max of >2.0 (B, P = .006). This was also true when the analysis was restricted to patients who received no adjuvant therapy (C, P = .004).