Diagnostic value of PET/CT with ¹¹C-methionine (MET) and ¹⁸F-fluorothymidine (FLT) in newly diagnosed glioma based on the 2016 WHO classification

The updated 2016 edition of the World Health Organization (WHO) Classification of Tumours of the Central Nervous System (CNS) uses molecular parameters and histology to define the main tumours categories for the first time [1]. This represents a shift from the traditional way of using neuropathological diagnoses primarily based on microscopic features, to using biologically oriented diagnoses. Among several molecular features, isocitrate dehydrogenase (IDH) mutation and chromosome 1p and 19q (1p/19q) codeletion status are the genetic alterations that have had a significant impact on the new tumour classification [2]. Several studies have shown that IDH1 mutation is a strong prognostic maker, and this may well be the most upstream genetic event in the tumourigenesis and may drive other genetic changes in tumour cells [3, 4].

Positron emission tomography (PET) with l-[methyl-¹¹C]-methionine (MET) has been widely used as an imaging tool for brain tumour detection, tumour grading and prediction of prognosis in patients with gliomas [5]. Recent studies have examined MET uptake for differentiating glioma based on the 2016 WHO classification especially in relation to IDH1 mutation and 1p/19q codeletion status. In general, MET uptakes in IDH1-wildtype gliomas were significantly higher compared with those in IDH1-mutant gliomas [6‐8]. However, several amino acid PET studies showed paradoxically higher tracer uptakes in IDH-mutant gliomas, especially with oligodendroglial components, compared with counterpart IDH1-wildtype tumours [8‐10]. These results suggest that amino acid PET studies are not fully consistent or accurate for differentiating glioma at diagnosis based on the 2016 WHO classification.

3′-deoxy-3′-[¹⁸F]fluorothymidine (FLT), a fluorinated thymidine analogue, has emerged as a promising PET tracer for evaluating tumour proliferating activity in various malignant tumours. FLT allows the direct measurement of cellular thymidine kinase-1 (TK1), which has been reported to be proportional to the proliferation activity of a tumour [11, 12]. As FLT uptake in normal brain tissue is very low, FLT-PET provides a low-background brain image and thus is considered useful for imaging brain tumours. FLT-PET has been found to be useful for assessing the proliferative activity of gliomas in vivo [13‐15]. However, several studies have shown that the major portion of FLT uptake is dependent on the influx through the disrupted blood-brain barrier (BBB) and non-enhancing tumours with an intact BBB showed limited transport of FLT [15‐18].

Previously, we reported that FLT-PET was superior to MET-PET in non-invasive tumour grading and assessment of proliferative activity in gliomas of different grades [13]. The study was conducted based on the 2007 WHO classification, and no study has evaluated the usefulness of FLT-PET for glioma differentiation based on the revised 2016 WHO classification. In the present study, we retrospectively analysed MET-PET/CT and FLT-PET/CT in newly diagnosed gliomas for differentiating glioma according to the 2016 WHO classification especially in relation to IDH1 mutation status.

IDH1 mutation status is the genetic alteration with the most significant impact on the updated 2016 edition of the WHO Classification of Tumours of the CNS [1, 2]. Patients with IDH1-mutant astrocytomas have a better overall prognosis compared with those with IDH1-wildtype astrocytomas, even after controlling for histologic grade [1, 3]. Given their distinct molecular origins, the optimal treatment strategies for IDH1-mutant versus IDH1-wildtype tumours should be different. Recently, several studies have reported that response and benefit to treatment differ depending on the IHD1 mutation status [19‐21]. Based on these studies, preoperative prediction of genotypes, especially the IDH1 mutation status, is essential to planning tailored treatment strategies including surgical resection and postoperative adjuvant therapy. This is especially the case with WHO grade II and III gliomas [20]. On the other hand, the great majority of WHO grade IV astrocytomas falls into the IDH1-wildtype category (more than 90% of cases) [1], which corresponds most frequently to the clinically defined primary GBM resulting in poor prognosis even with intensive treatments.

Based on the 2016 WHO classification, MET-PET studies in patients with newly diagnosed gliomas have reported a correlation between MET uptake and IDH1 mutation status, showing that IDH1-wildtype gliomas had significantly higher MET uptake than IDH1-mutant gliomas [6‐8]. In the present study, the MET uptakes in IDH1-wildtype tumours were also significantly higher than that in IDH1-mutant tumours in all gliomas, but this was not the case among grade II and III gliomas. Although the sensitivity for the differential diagnosis was high (88.5 %), the specificity was low (51.7 %) using MET-PET/CT. Kim et al. reported that among grade II and III gliomas, IDH1-mutant and 1p/19q-codeleted oligodendrogliomas were more likely to exhibit higher MET uptake, even paradoxically high, compared with the counterpart IDH1-wildtype astrocytomas [8]. They cautioned that MET uptake for glioma grading according to the 2016 WHO classification was more consistent and accurate for IDH1-wildtype tumours than for IDH1-mutant tumours. Previous studies have shown that MET uptake in gliomas with an oligodendroglial component is higher than in astrocytomas, even in low-grade gliomas [13, 22, 23]. High MET uptakes in IDH1-mutant oligodendrogliomas may cause low specificity for the differential diagnosis. In fact, AOs (median 5.07, IQR 4.72–7.21) showed increased MET uptake as high as those in GBMs (median 6.19, IQR 4.63–7.41), in which most of them were IDH1-wild type tumours in the present study. Several amino acid PET studies showed paradoxically higher tracer uptakes in IDH-mutant gliomas and the inability to predict the IDH1 mutation status among grade II and III gliomas [9, 10]. These results suggest that the use of amino acid PET for differentiating glioma at diagnosis may be useful but is subject to controversy. Recently, several studies have shown the usefulness of amino acid PET with dynamic analysis for glioma differentiation based on the 2016 WHO classification [10, 24].

3′-Deoxy-3′-[¹⁸F]fluorothymidine (FLT), a fluorinated thymidine analogue, has emerged as a promising PET tracer for evaluating tumour proliferating activity in various brain tumours. FLT is phosphorylated by TK1, a principle enzyme in the salvage pathway of DNA synthesis, and trapped inside cells. Phosphorylated FLT is resistant to degradation and suitable for imaging with PET. The application of FLT phosphorylation as a marker of cell proliferation is based on the assumption that cellular FLT trapping is a representation of thymidine incorporation into DNA [11, 12]. FLT-PET has been found useful for non-invasive grading and assessment of proliferative activity especially in newly diagnosed gliomas [13‐15]. Moreover, FLT-PET can provide valuable information regarding treatment response and patient prognosis and regarding the recognition of tumour recurrence in gliomas [25, 26]. In our previous study, FLT-PET was found to be likely superior to MET-PET in tumour grading and assessment of proliferative activity in newly diagnosed gliomas of different grades [13]. In the present study, MET-PET/CT could only discriminate DA from other histological types of gliomas, i.e. DA and AA, DA and AO, and DA and GBM. On the other hand, FLT-PET/CT can distinguish all different types of gliomas except AA and AO based on the 2016 WHO classification. Moreover, both PET tracers could distinguish grade II gliomas from higher-grade gliomas, but the differentiation between grade III and IV gliomas, two groups with different prognosis and management [27], was able to achieve only with FLT-PET/CT.

This is the first study to demonstrate a significant correlation between FLT uptake values and IDH1 mutation status in newly diagnosed gliomas. FLT uptakes in IDH1-wildtype tumours were significantly higher than those in IDH1-mutant tumours. ROC analysis showed that FLT-PET/CT can distinguish the IDH1 mutation status more accurately than MET-PET/CT. Histologically, the prognostic differences between DA and AA were significant [28]. However, recent studies have shown that the prognostic difference between IHD-mutant DAs and IHD-mutant AAs are not as marked [29], suggesting that IDH status might be more important than tumour grade. In the present study, more than 50% of the gliomas were GBMs, and almost all GBMs were IDH1-wildtype (43/45 cases). PET tracer uptake in GBMs is usually higher than that in other grades of gliomas, and this may influence the results in relation to IDH1 mutation status. Therefore, we conducted further examination of the uptake values in 36 grade II and III gliomas. It should be noted that FLT-PET/CT was able to distinguish the IDH1-mutant tumours from wildtype tumours not only in all gliomas, but also in this specific population, but this was not the case for MET-PET/CT. Although the sensitivity for differential diagnosis of IDH1 mutation status in the ROC analysis was high (92.3%) with FLT-PET, the specificity was not satisfactory (75.9%). This number means that one in 4 gliomas with low FLT uptake below the cutoff value (T/N ratio of 6.74) was an IHD1-wildtype tumour that could harbour an aggressive nature in the tumour. Our previous study showed that FLT uptake in tumours without an observed contrast enhancement effect on MR images was low even if the tumours were high-grade gliomas [15]. Later studies, including from our laboratory, have shown that the major portion of FLT uptake is due to increased transport and influx through the disrupted blood-brain barrier (BBB) [16, 17]. Non-enhancing tumours with an intact BBB showed limited transport of FLT, and tumour malignancy and cell proliferation activity cannot be adequately assessed by FLT-PET [18]. In the present study, FLT-PET/CT was able to distinguish the IDH1-mutant tumours from wildtype tumours statistically. However, the number of non-enhancing tumours was limited (n = 13), especially the IDH1-wildtype tumours (n = 3) and the result was not conclusive. The issue regarding the usefulness of FLT-PET/CT for differentiating IDH1 mutation status in non-enhancing tumours should be addressed with more cases in the future.

The present study had several limitations. First, there was selection bias with respect to the relatively small number of patients with low-grade glioma. Patients with low tracer uptakes were suspected of having low-grade gliomas and considered less likely candidates for surgery, especially when they had few or no symptoms. Second, patients with IHD1-wildtype DA (n = 1) and IHD1-mutant GBMs (n = 2) were excluded from the statistical analysis because of the small number of cases. We only compared the tracer uptakes in grade II and III gliomas with IDH1 mutation and grade III and IV gliomas without IDH1 mutation. Incidences of IHD1-mutant GBM (less than 10%) and IHD1-wildtype DA are rare in the general population [1, 30]. Third, we have not evaluated the tracer uptakes in relation to IDH1 mutation status within the same WHO malignancy grade due to too few datasets available for a meaningful analysis. Patients harbouring the IDH1 mutation had a longer overall survival than those without mutation within AA as well as GBM [31]. Finally, in addition to the retrospective nature of the present study, no follow-up data were evaluated in the present study, and the findings need to be analysed in relation to prognosis with a long flow-up period.