Multiple positron emission tomography tracers for use in the classification of gliomas according to the 2016 World Health Organization criteria

Abstract Background The molecular diagnosis of gliomas such as isocitrate dehydrogenase (IDH) status (wild-type [wt] or mutation [mut]) is especially important in the 2016 World Health Organization (WHO) classification. Positron emission tomography (PET) has afforded molecular and metabolic diagnostic imaging. The present study aimed to define the interrelationship between the 2016 WHO classification of gliomas and the integrated data from PET images using multiple tracers, including 18F-fluorodeoxyglucose (18F-FDG), 11C-methionine (11C-MET), 18F-fluorothymidine (18F-FLT), and 18F-fluoromisonidazole (18F-FMISO). Methods This retrospective, single-center study comprised 113 patients with newly diagnosed glioma based on the 2016 WHO criteria. Patients were divided into 4 glioma subtypes (Mut, Codel, Wt, and glioblastoma multiforme [GBM]). Tumor standardized uptake value (SUV) divided by mean normal cortical SUV (tumor–normal tissue ratio [TNR]) was calculated for 18F-FDG, 11C-MET, and 18F-FLT. Tumor–blood SUV ratio (TBR) was calculated for 18F-FMISO. To assess the diagnostic accuracy of PET tracers in distinguishing glioma subtypes, a comparative analysis of TNRs and TBR as well as the metabolic tumor volume (MTV) were calculated by Scheffe's multiple comparison procedure for each PET tracer following the Kruskal–Wallis test. Results The differences in mean 18F-FLT TNR and 18F-FMISO TBR were significant between GBM and other glioma subtypes (P < .001). Regarding the comparison between Gd-T1WI volumes and 18F-FLT MTVs or 18F-FMISO MTVs, we identified significant differences between Wt and Mut or Codel (P < .01). Conclusion Combined administration of 4 PET tracers might aid in the preoperative differential diagnosis of gliomas according to the 2016 WHO criteria.

According to the 2007 World Health Organization (WHO) grading criteria, gliomas, the most common primary brain tumors, comprise a heterogeneous group of histological subtypes based on cellular alterations related to tumor aggressiveness. 1 Additionally, the 2016 WHO classification of Central Nervous System (CNS) tumors includes molecular genetic profiles for the subclassification of gliomas. 2 Mutations in coding sequences of isocitrate dehydrogenase (IDH) 1 and IDH2 and chromosome 1p

Multiple positron emission tomography tracers for use in the classification of gliomas according to the 2016 World Health Organization criteria
and 19q (1p19q) codeletion are essential for the diagnosis of gliomas reclassified as astrocytic and oligodendroglial tumors. 3,4 Surgical specimens are indispensable for the definitive molecular pathological diagnosis according to the 2016 WHO criteria. However, in some patients, glioma localization hinders sample collection for pathological assessment and preoperative methods that can predict the glioma genotype are necessary for determining treatment strategies.
Diagnostic imaging of gliomas achieved through various methods has significantly advanced over recent years. Magnetic resonance imaging (MRI), the most commonly used method to collect information on tumor morphology, cannot by itself determine the definitive diagnosis. Conversely, positron emission tomography (PET) has facilitated the establishment of noninvasive metabolic and molecular imaging methods for CNS tumors. Importantly, various molecular processes can be visualized using specific PET tracers. Among these, 18 F-fluorodeoxyglucose ( 18 F-FDG), the most frequently used radiotracer, is a glucose analog whose metabolism involves glucose transporter and hexokinase activity. Additionally, several other PET tracers for CNS tumors have been developed based on the key roles of certain amino acids such as 11 C-methionine ( 11 C-MET), 5 18 F-fluoroethyltyrosine ( 18 F-FET), 6 and 18 F-fluorodopa ( 18 F-FDOPA). 7 11 C-MET is used to evaluate protein synthesis and cell proliferation in gliomas and to detect malignant transformation. 8 18 F-fluorothymidine ( 18 F-FLT) is a radiolabeled thymidine analog used to predict tumor progression, 9 and provides a low background, facilitating tumor detection. 10 Malignant tumors are characterized by a hypoxic tissue environment which may drive peripheral tumor growth and is associated with tumor progression. One of the most widely used PET tracers for molecular imaging of hypoxia is 18 F-fluoromisonidazole ( 18 F-FMISO). 11 We previously reported the characteristics of gliomas based on the 2007 WHO criteria using 18 F-FDG, 11 C-MET, 18

Patients
This retrospective, single-center study complied with the precepts established by the Declaration of Helsinki and was approved by the Kagawa University Faculty of Medicine Human Subjects Ethics Committee (no. 2019-027). 18 F-FDG, 11 C-MET, 18 F-FLT, and 18 F-FMISO were approved for use as PET tracers by the Kagawa University Faculty of Medicine Human Subjects Ethics Committee, and an informed written consent was obtained from all participants.

Importance of the Study
This is the first study examining the relationship between the 2016 WHO glioma classification and glioma classification based on multiple PET tracers to evaluate different metabolic pathways, including glucose, amino acid, and nucleic acid metabolism, and the presence of hypoxic regions. The differences in mean 18 F-FLT TNR and 18 F-FMISO TBR were significant between GBM and other glioma subtypes. The differences in mean 11 C-MET TNR were significant between GBM and Mut or Wt. There were significant differences in the MTV of 18 F-FLT between GBM and Mut or Codel. A comparison between Gd-T1WI volume and the MTV of 11 C-MET was significant between GBM and Codel or Wt. A comparison between Gd-T1WI volume and the MTV of 18 F-FLT or 18 F-FMISO revealed significant differences between Wt and Mut or Codel. We suggest that multiple PET tracers using 18 F-FDG, 11 C-MET, 18 F-FLT, and 18 F-FMISO are useful for preoperative differential diagnosis of gliomas.

Histopathological and Molecular Analyses
To reclassify the study cohort according to the 2016 WHO classification, the study patients were evaluated for IDH-mut and 1p19q codeletion. For IDH-mut status, IDH1 R132H protein expression was determined by immunohistochemistry using a monoclonal antibody (clone H09, 1:50; Dianova, Germany). In cases where immunostaining was not possible, IDH1 (R132) and IDH2 (R172) were directly sequenced using the Sanger method. The 1p19q codeletion status was analyzed by fluorescence in situ hybridization with locus-specific probes for 1p36 and 19q13.
PET studies were performed using a Biograph mCT PET/ CT scanner (Siemens Medical Solutions Knoxville, TN, USA). PET scans were acquired in the three-dimensional model, and PET images were reconstructed as described in our previous study (the simultaneous acquisition of 51 transverse images per field of view [FOV], with an intersection spacing of 3 mm, for a total axial FOV of 15 cm). 10 PET radiotracers were produced using an HM-18 cyclotron (Sumitomo Heavy Industries, Tokyo, Japan). The radiochemical purity of 11 C-MET, 12 18 F-FLT, 13 and 18 F-FMISO 14 were >95%. Transmission and regional emission images of the brain were obtained as described in our previous study. 10 Fasting was initiated 6 h before all PET studies, and the examination schedule was as follows: MRI, including contrast examination, was performed on day 1, 18 F-FMISO was performed on day 2, 18 F-FLT was performed on day 3, and 11 C-MET was performed on the morning of day 4, followed by 18 F-FDG during the afternoon of day 4.

Image Analyses
The uptake of 18 F-FDG, 11 C-MET, and 18 F-FLT in brain tumors were semiquantitatively assessed by obtaining the standardized uptake values (SUVs). A region of interest around the hottest portion of each lesion was manually set by an observer. The maximum SUV (SUV max ) was considered as the representative value for each tumor. The maximum tumor-to-normal ratio (TNR) was determined by dividing the tumor SUV max by the mean SUV of the normal brain parenchyma (usually contralateral normal cerebral tissue excluding the ventricles). The uptake of 18 F-FMISO in the brain tumor was semiquantitatively assessed by evaluating the SUV max . The 18 F-FMISO PET images were converted into average venous blood concentration of 18 F-FMISO to obtain the tumor-to-blood ratios (TBRs), allowing for a three-dimensional pixel-by-pixel calculation of the maximum TBR for SUV max . The tumor volumes were measured by performing a three-dimensional, threshold-based, volume-ofinterest analysis of the hyperintensity on fluid-attenuated inversion recovery (FLAIR) images, hyperintensity on diffusion-weighted images (DWI), and contrast-enhanced lesions on gadolinium-enhanced T1-weighted images (Gd-T1WI). For PET studies, the cutoff values of 1.1 on the 18 F-FDG TNR, 1.3 on the 11 C-MET TNR, 1.3 on the 18 F-FLT TNR, and 1.2 on the 18 F-FMISO TBR were used to determine the metabolic tumor volume (MTV). 8,15 The PET and MRI datasets were transferred to a Linux workstation, and coregistration of 18 F-FDG/ 11 C-MET/ 18 F-FLT/ 18 F-FMISO/MRI was performed using Dr. View/Linux, version R2.5 (AJS, Tokyo, Japan). Before the histopathological and molecular diagnoses, 2 radiologists (Y. Y. and Y. N.) analyzed the data to lower the risk of observer bias to the maximum extent possible.

Statistical Analysis
The relationship of glioma subtypes with the volume on FLAIR, Gd-T1WI, and DWI, mean TNRs on 18 F-FDG, 11 C-MET, and 18 F-FLT, mean TBR on 18 F-FMISO, MTV on 4 PET studies were examined. To assess the diagnostic accuracy of PET tracers in distinguishing glioma subtypes, a comparative analysis of TNRs and TBR as well as the MTV were calculated by Scheffe's multiple comparison procedure of each PET tracer following the Kruskal-Wallis test. All parametric data were expressed as averages with standard deviation. Differences were considered statistically significant at a P value of <.05. The cutoff values for volume on FLAIR,   Figure 1 shows the correlation of glioma subtypes with the 18 F-FDG, 11 18 F-FDG TNRs between GBM and Mut were statistically significant (P = .027) ( Figure 1A). The mean 11 C-MET TNRs for Mut, Codel, Wt, and GBM were 3.32 ± 1.64, 4.74 ± 1.98, 3.79 ± 1.54, and 6.27 ± 2.66, respectively. The differences in mean 11 C-MET TNRs were significant between GBM and Mut (P < .001) and GBM and Wt (P = .006) ( Figure 1B). The cutoff value of 11

Correlation Among Glioma Subtypes with the Comparison Between MTVs of 4 PET Tracers and DWI Volumes.
The 11 C-MET, 18 F-FLT, and 18 F-FMISO MTVs were larger than the DWI volumes. 18 F-FDG MTVs were similar or slightly lesser than the DWI volumes. For comparison between the volumes of DWI and MTVs of 18 F-FDG, 11 C-MET, or 18 F-FLT tracers, there were no significant differences among the glioma subtypes ( Figure 3I, J, and K). Comparison between MTV of 18 F-FMISO and the volume of DWI indicated significant differences between Mut (0.34 ± 0.39) and Wt (1.45 ± 1.10, P = .046) or GBM (1.30 ± 0.69, P = .001; Figure 3L;Supplementary Table 3)

Neuro-Oncology Advances
a high 11 C-MET accumulation, it was possible to distinguish Mut from Codel using the 11  The cutoff of TNR for 11 C-MET (4.327) and 18 F-FLT (7.563) and TBR for 18 F-FMISO (1.612) could distinguish between Wt and GBM. Considering these results, case D was diagnosed as GBM ( Figure 4D).

Discussion
PET uses radiotracers to achieve metabolic and molecular imaging and, in combination with MRI, can provide useful information that may improve the diagnostic accuracy of brain tumors. 16,17 One PET tracer is suitable for assessing related metabolism, but not for others. Therefore, the only approach which allows the simultaneous evaluation of various metabolites is using multiple PET tracers. Previous reports evaluating multiple PET tracers were based on systematic reviews based on the meta-analyses of published studies, and few reports evaluated multiple PET tracers used in the same patient. [18][19][20] Furthermore, no report to date has evaluated the utility of multiple PET tracers including 18 F-FMISO. The systematic reviews of published meta-analyses related to PET were based on the glioma classification according to WHO grades II, III, and IV and did not conform to the 2016 WHO classification of gliomas. This is the first report examining the interrelationship between 4 glioma subtypes based on the 2016 WHO classification and multiple PET tracers. PET tracer guidelines have been recently revised to provide joint practice guidelines and procedure standards for uniform, high-quality diagnostic accuracy imaging by the Working Group for Response Assessment in Neurooncology with PET. 17 18 F-FDG is the most commonly used PET tracer in oncology. The present study results showed that 18 F-FDG could distinguish between WHO grade II and IV gliomas, but 18 F-FDG had the lowest sensitivity and specificity among the 4 PET tracers.
Optimal quantitative thresholds and visual analysis criteria have not been established for the definitive differentiation of glioma grade based on 18 F-FDG PET alone. 21 Regarding amino acid PET tracers, especially, 11 C-MET and 18 F-FET are preferred over 18 F-FDG due to the higher sensitivity. 17 The present study revealed that the accumulation of 11 C-MET increased in parallel with higher WHO malignancy grades. Although higher 11C-MET accumulations were observed in both OD and AO in previous studies, 22 11 C-MET could not significantly differentiate between Mut and Codel at this time. Some reasons should be considered that there were relatively fewer patients with glioma subtypes other than GBM IDH-wt. Mut is composed of DA IDH-mut and AA IDH-mut, and Code is composed of OD and AO. Hence, AA and AO with higher malignancy are included in their respective subtypes; therefore, the accumulation of MET may have been high, and the difference between Mut and Codel may no longer be recognized. We would reanalyze the present study cohort with the addition of more patients and more extensive analysis between Mut and Codel in future investigations.
We previously reported that 18 F-FLT could distinguish gliomas based on the 2007 WHO classification and that the 18 F-FLT accumulation exhibited a strong correlation with the histopathologic proliferation marker Ki-67. 9 Therefore, 18 F-FLT is considered as a suitable tracer for evaluating tumor proliferation. However, careful consideration should be given to increased 18 F-FLT accumulation related to its leakage from tumor vessels in brain tumors with a disrupted blood-brain barrier (BBB), 23,24 and tumor blood flow. 25 The 2016 WHO classification of gliomas is based on the IDH mutation status. The IDH mutation status was reported to be associated with tumor proliferation and prognosis in lower-grade gliomas. 26 Takei et al. reported that 11 C-MET could be used to differentiate between AA IDHmut and AA IDH-wt and between GBM IDH-mut and GBM IDH-wt. 22 No report to date has compared the association of 18 F-FLT accumulation with tumor prognosis or IDH mutation status. In the present study, we first demonstrated that the comparison between 18 F-FLT MTV and Gd-T1WI volume could be used to distinguish between Mut and Wt. In other words, it is suggested that Wt has a wider accumulation region of 18 F-FLT than the enhancement region on Gd-T1WI compared with Mut. 18 F-FMISO is a nitroimidazole derivative that is exclusively trapped in hypoxic cells. GBM presents with necrosis and hypoxic environment, whereas lower-grade gliomas do not develop necrosis; therefore, 18 F-FMISO is more likely to accumulate in the hypoxic GBM environment. 11 The present study results also suggest that 18 F-FMISO can differentiate GBM from lower-grade gliomas. In the GBM microenvironment where hypoxia has progressed, hypoxia-inducible factor 1α (HIF1α) associated with hypoxia is activated. Most of GBM leads to upregulating HIF1α. 27 We previously reported that the accumulation of 18 F-FMISO was significantly correlated with the expression of vascular endothelial growth factor related to HIF1α. 11 Therefore, it is reasonable to assume that the accumulation of 18 F-FMISO would be high in patients with GBM. MTV of 18 F-FMISO could be distinguished GBM IDH-wt from GBM IDH-mut, but 18 F-FMISO accumulation alone cannot distinguish these subtypes. A recent report showed that not only hypoxia-related signaling pathways but also transforming growth factor β might be related to gliomas with IDH-wt. 28 The comparison between 18 F-FMISO MTV and Gd-T1WI volume or DWI volume could be used to distinguish between Mut and Wt. Gd-T1WI is related to the permeability of gadolinium, while DWI reflects on cell density. Because the 18 F-FMISO MTV evaluates a wider area the Gd-T1WI and DWI volumes, 18 F-FMISO in Wt might evaluate active tumor cell lesions, under hypoxia, and various other conditions. More evidence based on further investigation of larger cohorts is needed to confirm that 18 F-FMISO can be used to differentiate between IDH-wt and IDH-mut gliomas.
The present study has several limitations. First limitation is that the metabolism of gliomas exhibiting various molecular changes could not be evaluated using only one PET tracer. In the present study, using 4 PET tracers that could assess different metabolic pathways allowed us to classify the study patients according to the 2016 WHO glioma classification, even though not all metabolic pathways could be evaluated. Codel and Wt could not be distinguished; however, these cases can generally be discriminated by comparing 18 F-FMISO and MRI, and further examination using other tracers remains necessary. A second limitation was that few patients with Mut, Codel, and Wt were included in this study. The glioma subtypes were distributed non-normally and not homoscedastically. Therefore, it was Scheffe's multiple comparison procedure following the Kruskal-Wallis test was used for statistical analyses. The distribution can converge to a normal distribution by securing a greater number of cases; however, this will take time with a single center. The utility of multiple PET tracers in a greater number of patients across multiple institutions should be investigated.

Conclusion
This is the first study examining the relationship between glioma classification based on the 2016 WHO classification and multiple PET tracers evaluating different metabolic pathways. We suggest that all PET tracers using 18 F-FDG, 11 C-MET, 18 F-FLT, and 18 F-FMISO are useful for the Neuro-Oncology Advances preoperative differential diagnosis of gliomas according to the 2016 WHO classification.

Supplementary Data
Supplementary data are available at Neuro-Oncology Advances online.

Keywords
2016 World Health Organization classification | glioma | positron emission tomography