P01.053 the Role of PET- and MRI-derived [18F]FET PET Parameters in the Discrimination of Low- and High-Grade Gliomas

Gliomas are the most frequent malignant primary brain tumours in adults and can be classified into low-grade gliomas (LGG, WHO grade I-II) and high-grade gliomas (HGG, WHO grade III-IV). Overall survival rate of HGG is poor (months) when compared to LGG (years). Accurate and early diagnosis, for which histopathology is the gold standard, is essential to decide the optimal treatment. Because biopsy and resection bear certain risks, interest is shown in non-invasive diagnostic methods to guide treatment decisions. In this study, different [18F]Fluoro-Ethyl-Tyrosine ([18F]FET) positron emission tomography (PET) parameters were examined for discriminating HGG from LGG. A pre-treatment [18F]FET PET scan and an anatomical MRI were performed in 30 patients. Twenty-four PET-derived parameters, e.g. metabolic tumour volume (MTV), total lesion uptake (TLU) and tumour-to-background ratio (TBR), were calculated using various standardised uptake value (SUV) thresholds. Subsequently, an in-house developed and segmentation algorithm was used to label different tissue categories on both contrast-enhanced T1-weighted and FLAIR MR images: necrosis (N), oedema (OE), non-enhancing tumour (NET) and contrast-enhancing tumour (CET). MR-derived tumour masks were applied on the PET images to calculate 55 PET parameters. PET- and MR-derived FET PET parameters were compared between HGG and LGG. Significantly different parameters were analysed by receiver-operating characteristics (ROC) analysis, yielding area under the curve (AUC) values. Pathological diagnosis served as the reference. Significant differences between HGG and LGG were observed for 13 and 17 for the PET- and MR-derived PET parameters, respectively. The AUC values for significant PET-derived FET PET parameters ranged between 0.78 and 0.85. The lowest AUC value corresponded with TBR1.5 that is defined as the tumour-to-background ratio for the voxels exceeding 1.5 times the mean background SUV. The highest AUC value corresponded with MTV1.5 that is defined as the total volume of voxels with a SUV exceeding 1.5 times the mean background SUV. The AUC values for significant MR-derived FET PET parameters ranged between 0.71 and 0.84. The lowest AUC value corresponded with TLUN+OE+CET+NET that is defined as the total lesion uptake in a volume covering N, OE, CET and NET. The highest AUC value corresponded with TBRmax for all tumour tissue categories (N, OE, CET, NET). TBRmax is defined as the maximal tumour-to-background ratio for N, OE, CET and NET. This study illustrates that various PET- and MR-derived [18F]FET PET parameters differ significantly between HGG and LGG. Hence, PET and MRI of brain tumours may aid the pre-treatment diagnostic classification of gliomas. Further studies in large populations are needed to investigate which parameters may serve as a surrogate for biopsy.