Feasibility of different meningioma delineation approaches on [18F]SiTATE PET/CT imaging

Somatostatin receptor (SSTR)-targeted PET is valuable for meningioma imaging due to high SSTR expression. [18F]SiTATE, a novel tracer, is not only promising for imaging neuroendocrine tumors but also for meningiomas. Standardized delineation methods on [18F]SiTATE PET are lacking. This study correlates CT-based volumes with PET-based delineation approaches to identify a threshold for standardized [18F]SiTATE PET volume assessment. Patients with well-delineated, extraosseous meningioma on CT (≥ 1mL) who underwent [18F]SiTATE PET/CT were included. Volumes were assessed on contrast-enhanced CT and correlated with PET-based delineation approaches: (I) fixed SUV threshold, (II) isocontour thresholding relative to SUVmax (SUV%), and thresholds relative to (III) bone marrow (SUVBM), (IV) parotid gland (SUVparotis) and (V) pituitary gland (SUVsella). 19 meningiomas in 17 PET/CT scans (16 patients) were included. A fixed SUV of 4.0 (r = 0.783, p < 0.001) showed good correlation with CT volumes without skewed distribution on Bland-Altman-Plot analysis. Using isocontour-based thresholds, 45% SUVmax (r = 0.496, p = 0.031) showed the highest concordance. Best reference-based approaches were achieved by 150% SUVBM (r = 0.859, p < 0.001), 250% SUVparotis (r = 0.460, p = 0.047) and 70% SUVsella (r = 0.819, p < 0.001). However, background-based approaches showed a trend towards overestimation of PET-volumes in larger meningiomas as assessed on Bland-Altman-Plot analyses. Uptake intensities of reference tissues (SUVBM, SUVparotis and SUVsella) were not inter-correlated (p > 0.05 each). A fixed SUV threshold of 4.0 showed strong agreement with CT-based volumes in well-delineated, extraosseous meningiomas which offers a simple, clinically applicable method without technical requirements. Reference tissue-based methods showed similar correlations but tended to overestimate volumes in larger lesions.