Background and purpose: World Health Organization (WHO) grading of meningiomas reflects recurrence rate and prognosis. Positron emission tomography (PET) investigates metabolic activity, allowing for distinction between low- and high-grade tumors. As preoperative suspicion for malignant meningioma will influence surgical strategy in terms of timing, extent of resection, and risks taken to achieve a total resection, we systematically reviewed the literature on PET-imaging in meningiomas and relate these findings to histopathological analysis.
Methods: Searches in PubMed, EMBASE, and The Cochrane Library, from inception to September 2019, included studies of patients who had undergone surgery for a histologically verified intracranial meningioma, with a PET-scan prior to surgery and description of (semi)quantitative PET values for meningiomas from two different WHO groups. Studies comparing more than 1 patient per WHO group were included in the meta-analysis.
Results: Twenty-two studies (432 patients) were included. 18fluor-fluorodesoxyglucose (18F-FDG) PET was mostly described to differentiate benign from malignant meningiomas. Pooled data showed differences in mean (95% CI) Standardized Uptake Value (SUV) for WHO II/III compared to WHO I of 2.51 (1.36, 3.66), and in tumor-to-normal (T/N) ratio (T/N ratio) for WHO II/III versus WHO I of .42 (.12, .73).
Conclusions: We found that SUV and T/N ratio in 18F-FDG PET may be useful to noninvasively differentiate benign from malignant meningiomas. T/N ratio seems to have a high specificity for the detection of high-grade meningiomas. Other PET tracers were studied too infrequently to draw definitive conclusions. Before treatment strategies can be adapted based on 18F-FDG PET, prospective studies in larger cohorts are warranted to validate the optimal T/N ratio cutoff point.
Keywords: Meningiomas; WHO grade; meta-analysis; positron emission tomography.
© 2020 The Authors. Journal of Neuroimaging published by Wiley Periodicals LLC on behalf of American Society of Neuroimaging.