Dosimetric factors and toxicity following Gamma Knife radiosurgery for convexity meningiomas

Introduction: In patients with meningioma unsuitable for or declining observation or surgical resection, radiosurgery (SRS) is a viable treatment option. This study aimed to establish crucial dosimetric parameters, particularly for convexity meningiomas, to predict post-treatment toxicity risk when optimal planning is essential. Dose-volume distribution to normal brain was explored, alongside other dosimetric factors.

Methods: The study encompassed patients treated with SRS for convexity, parasagittal, and falcine meningiomas between 2004 and 2020, with at least six months of follow-up. Patient and tumor data, clinical outcomes, and dose-volume parameters were analyzed. The brain parenchyma volume, excluding cerebrospinal fluid, was used for V12 calculations. ROC curve analysis identified an optimal V12 volume threshold for predicting radiation-induced cerebral edema, which was then compared to toxicity incidence through logarithmic regression.

Results: The study included 68 patients with 85 lesions treated with GK. Median tumor volume was 2.83 mL, and the prescription dose was 15 Gy to the 50% IDL. Toxicity (grade 3 in 22 cases) occurred without grade 4/5 events, with cerebral edema accounting for most grade 3 events, 13 of which were symptomatic. Seven required dexamethasone treatment, and two needed surgical resection. Multivariate analysis revealed that surface area, target volume, and V10-V15 predicted grade 3 toxicity. The optimal V12 threshold was 2.65 mL for grade 3 toxicity (82% sensitivity, 71% specificity) and symptomatic edema (85% sensitivity, 75% specificity). Probability distribution analysis showed increased grade 3 toxicity from 8.1% to 50% and symptomatic edema from 4.1% to 31% for V12 values exceeding 2.65 mL.

Conclusion : This study highlights the significance of target volume, tumor surface area, and dose spillage into the cerebral cortex as predictors of peritumoral edema in meningioma patients treated with SRS. Although the multivariate model was not significant for symptomatic edema, discrete V12 thresholds are suggested for further exploration. Understanding the V12 volume threshold can aid SRS planning, particularly for convexity meningiomas in eloquent areas, to prevent symptomatic edema.