Preoperative ¹¹C-methionine PET imaging in glioma grading efficacy and its predictive value for IDH1 gene mutation status

Objective: To assess the preoperative ¹¹C-methionine (¹¹C-MET) PET imaging in glioma grading efficacy and its predictive value for isocitrate dehydrogenase enzyme 1 (IDH1) gene mutation status. Methods: A total of 118 glioma cases (70 males, 48 females;median age 45 years, age range: 10-71 years; II grade 65 cases, III grade 34 cases, IV grade 19 cases) received ¹¹C-MET PET imaging in PET Center of Huashan Hospital from February 2012 to November 2017 were retrospectively analyzed. Lesion-based semi-quantitative analysis was conducted on the ¹¹C-MET imaging. Maximum standardized uptake value (SUV_max), peak standardized uptake value (SUV_peak), tumor-to-background ratio (TBR; SUV_max in lesion/mean standardized uptake value (SUV_mean) in normal contralateral cortex) were calculated. Independent-sample t test and one-way analysis of variance were applied to assess the differentiating efficacy of ¹¹C-MET PET imaging for different glioma groups. Based on IDH1 immunohistochemical staining results, predictive efficacy of ¹¹C-MET PET diagnostic parameters on IDH1 mutation status in glioma patients was further analyzed with receiver operating characteristic (ROC) curve analysis. Results: Low-grade glioma (LGG; grade II) group showed significant differences from high-grade glioma (HGG; grade III-IV) group in SUV_max (2.458 ± 1.100 vs 3.828 ± 1.540; t = 5.624, P < 0.01), SUV_peak (2.160 ± 0.991 vs 3.261 ± 1.319; t = 5.175, P < 0.01) and TBR (2.283 ± 0.942 vs 3.434 ± 1.395; t = 5.328, P < 0.01). SUV_max (2.458 ± 1.100, 3.591 ± 1.611 and 4.251 ± 1.343; F = 17.67, P < 0.01), SUV_peak (2.160 ± 0.991, 3.040 ± 1.335 and 3.656 ± 1.225; F = 15.48, P < 0.01) and TBR (2.283 ± 0.942, 3.010 ± 1.242 and 4.192 ± 1.358; F = 22.73, P < 0.01) were different in grade II, III and IV glioma subgroups. SUV_max, SUV_peak and TBR all showed significant differences between grade II and grade III gliomas, grade II and grade IV gliomas, and there were also statistical differences between grade III and grade IV glioma with TBR (all P < 0.01). SUV_max indicated the best single-parameter prediction performance (area under curve (AUC) = 0.808, z = 7.193, P < 0.01), while the SUV_max + SUV_peak showed the best performance (AUC = 0.852, z = 9.115, P < 0.01). In the subgroup of grade II (n = 55), TBR of patients with IDH1 gene mutation (n = 41) was lower than that of patients with IDH1 wild-types (n = 14; 2.152 ± 0.759 vs 2.793 ± 1.208; t = 2.326, P = 0.02), while TBR of those with oligodendrogenic components (n = 26) was higher than that of patients with IDH1 gene mutation only (n = 18; 2.383 ± 0.825 vs 1.854 ± 0.478; t = 2.447, P = 0.02). Conclusions: Preoperative semi-quantitative parameters (SUV_max, SUV_peak, TBR) of ¹¹C-MET brain PET imaging have satisfactory grading discrimination performance for glioma patients. SUV_max is the best predictor for IDH1 mutation as a single parameter, while SUV_max + SUV_peak showed the most optimized predictive ability. The oligodendrogenic components in glioma can increase the uptake of ¹¹C-MET, which may affect the effectiveness of ¹¹C-MET in determining glioma grade to some extent. (authors)