SSTR PET/CT for skull base low-grade meningioma: a critical tool for accurate gross tumor volume delineation in radiotherapy?

Fuchs, Frederik; Marschner, Sebastian N.; Hofmaier, Jan; Rottler, Maya; Hadi, Indra; Maier, Sebastian H.; Greve, Tobias; Holzgreve, Adrien; Albert, Nathalie L.; Bodensohn, Raphael; Belka, Claus ; Niyazi, Maximilian; Walter, Franziska (2025): SSTR PET/CT for skull base low-grade meningioma: a critical tool for accurate gross tumor volume delineation in radiotherapy? Radiation Oncology, 20: 142. ISSN 1748-717X

Abstract

Background

Precise delineation of gross tumor volume (GTV) is fundamental for effective radiation therapy in low-grade skull base meningiomas. Magnetic resonance imaging (MRI) serves as the primary imaging tool but may not fully represent tumor extent. This study investigates the additional value of incorporating Somatostatin receptor (SSTR)-directed PET/CT in radiation therapy planning.

Methods

A retrospective analysis was conducted with four experienced radiation oncologists contouring GTVs for skull base meningiomas using MRI alone (GTV_MRI), PET/CT alone (GTV_PET/CT), and both modalities combined (GTV_ALL). Consensus ground truth volumes were generated for each modality through a STAPLE algorithm. Agreement between modalities, excluding observer variability, was assessed using statistical metrics including Dice Similarity Coefficient (DSC), Jaccard Index (JCI), Hausdorff distance (HD95), Geographical Miss Index (GMI), sensitivity, and kappa statistics.

Results

The study included 25 patients (15 females, 10 males; median age 56 years (range: 23–74 years), with 96% achieving local control post-radiotherapy over a median follow-up of 64 months (range: 28–135 months). Substantial interobserver agreement was observed, with median kappa values of 0.74 for GTV_MRI, 0.68 for GTV_PET/CT, and 0.77 for GTV_ALL. Median consensus volumes were 6.65 cc (MRI STAPLE ), 7.21 cc (PET STAPLE ), and 6.73 cc (ALL STAPLE ). The median GMI for MRI STAPLE compared to ALL STAPLE was 0.18 (IQR: 0.11–0.39), and 0.21 (IQR: 0.15–0.28) for PET STAPLE compared to ALL STAPLE . The DSC indicated the lowest concordance between MRI STAPLE and PET STAPLE with a median of 0.75 (IQR: 0.59–0.82), followed by PET STAPLE versus ALL STAPLE with a median DSC of 0.84 (IQR: 0.79–0.89), and MRI STAPLE versus ALL STAPLE with a median DSC of 0.89 (IQR: 0.76–0.92). The integration of PET/CT with MRI significantly enhanced concordance metrics.

Conclusion

Combining MRI and PET/CT improves GTV delineation in low-grade skull base meningiomas, as PET/CT can reveal regions missed by MRI, which may slightly underestimate tumor size. This multimodal imaging approach enhances consensus and supports its role in radiotherapy planning. Standardized protocols and technical integration remain key future goals.

• Dokumententyp: Artikel (Klinikum der LMU)
• Organisationseinheit (Fakultäten): 07 Medizin > Klinikum der LMU München > Neurochirurgische Klinik und Poliklinik; 07 Medizin > Klinikum der LMU München > Klinik und Poliklinik für Nuklearmedizin; 07 Medizin > Klinikum der LMU München > Klinik und Poliklinik für Strahlentherapie und Radioonkologie
• DFG-Fachsystematik der Wissenschaftsbereiche: Lebenswissenschaften
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 545058105
• DFG: Gefördert durch die Deutsche Forschungsgemeinschaft (DFG) - 491502892