Comparison of 18F-FET-PET- and MRI-based target definition for re-irradiation treatment of recurrent diffuse glioma
Everard, Albert, Versnel, Daniel, Ruijters, Veerle, Tolboom, Nelleke, Philippens, Marielle, Snijders, Tom, Verhoeff, Joost, David, Szabolcs, Beijst, Casper
DOI: https://doi.org/10.1016/j.ctro.2025.100931
Clinical and translational radiation oncology 52
Abstract
Background and purpose: The target area definition for re-irradiation of recurrent glioma typically relies on T1-weighted gadolinium-enhanced MRI (T1Gd), but T1Gd fails to capture glioma infiltration. While T2 FLAIR and CTV margins can address this limitation, their extend in re-irradiation remains debated to minimize toxicity. In contrast, 18F-FET-PET imaging can better visualize infiltrative components, improving target definition. This study investigates the role of 18F-FET-PET in target definition by analyzing shape differences and gray-white matter tissue composition. Material and methods: Retrospective data from 36 patients with recurrent glioma were used. These patients underwent 18F-FET-PET, on which Biological Tumor Volume (BTVPET) was delineated, and T1Gd, on which Gross Tumor Volume (GTVMRI) and Clinical tumor volume (CTVMRI) were delineated. Lesion volume, shape (Hausdorff distance (HD) and Center of Mass Distance (CoM-D)), tissue composition and re-recurrence volume were compared between the delineations. Results: BTVPET and the non-overlapping BTVPET-GTVMRI volumes were significantly larger than GTVMRI and non-overlapping GTVMRI-BTVPET (P < 0.001), but significantly smaller than the CTVMRI (P < 0.001). The median HD and CoM-D were 19 mm and 6 mm. A correlation was found between the white matter (WM) ratio in GTVMRI and WM in non-overlapping BTVPET (slope = 0.36, R2 = 0.36;P < 0.001) and a correlation with gray matter (GM) in non-overlapping BTVPET (slope = -0.15, R2 = 0.11;P < 0.05). Recurrence after reirradiation was significantly higher within BTVPET than GTVMRI (P < 0.001). The ratio of the GTVMRI which recurred was significantly higher than the non-overlapping BTVPET (P < 0.001). Conclusion: Our results suggest that 18F-FET-PET reveals a more complete extent of glioma's infiltration, suggesting that the current practice of using isotropic CTV margins around the T1Gd-GTV should be reconsidered in favor of anisotropic CTV based on 18F-FET-PET.