ESTRO 2025 - Abstract Book

S707

Clinical - CNS

ESTRO 2025

2775

Mini-Oral Estimating Radiation Necrosis Risk following Single-fraction Stereotactic Radiosurgery in the Modern Era: An In-Depth Dose-Volume Analysis Sreenija Yarlagadda 1 , Yanjia Zhang 2 , D Jay J. Wieczorek 1,3 , Ranjini Tolakanahalli 1,3 , Anshul Saxena 2 , Tatiana Bejarano 1 , Yongsook C. Lee 1,3 , Haley Appel 1 , Matthew D. Hall 1,3 , Robert H. Press 1,3 , Michael W. McDermott 4 , Alonso N. Gutierrez 1,3 , Minesh P. Mehta 1,3 , Rupesh Kotecha 1,3 1 Radiation Oncology, Miami Cancer Institute, Baptist Health South Florida, Miami, USA. 2 Biostatistics, TD – Artificial Intelligence and Machine Learning, Miami Cancer Institute, Baptist Health South Florida, Miami, USA. 3 Radiation Oncology, Herbert Wertheim College of Medicine, Florida International University, Miami, USA. 4 Neurosurgery, Miami Neuroscience Institute, Baptist Health South Florida, Miami, USA Purpose/Objective: Dose-volume constraints for stereotactic radiosurgery (SRS) rely on dose to whole-brain to minimize the risk of radiation necrosis (RN). However, these constraints should ideally focus on the dose delivered to brain tissue immediately surrounding the targeted metastasis. We aim to define single-fraction SRS locoregional constraints for small brain metastases (SBM; ≤ 2 cm) focusing on actual brain treated. Material/Methods: Consecutive patients with intact SBM treated with SRS to 20-24 Gy between January 2017 and July 2022 were included. The locoregional at-risk normal brain (defined as a 1 cm volume around each lesion excluding bone and ventricles) and volumetric dose (V 8 -V 22 ) to this zone were recorded on a per lesion basis. Fine-Gray competing risk was used to compute the cumulative incidence of RN (CTCAEv5), with death as competing risk. Univariate logistic regression models were used to estimate the relationship between RN and per lesion volumetric dose. Predictive performance was assessed by AUC, McFadden R 2 , and Brier Score to identify the best metric. Results: 1503 SBMs in 235 patients (358 SRS courses) that received SRS were analyzed. The median prescription dose was 24 Gy, median lesion diameter was 0.45 cm (Interquartile range [IQR]: 0.3-0.7 cm), median volume was 0.03 cc (IQR: 0.01-0.14 cc), and a median of 9 lesions (IQR: 5-14) were treated per SRS course. With a median follow-up of 12 months, 52 (3.5%) RN events were observed in 37 patients. The cumulative incidences of RN at 6-months, 1-year, and 2-year were 1.15% (95% CI: 0.69 – 1.9%), 2.24% (95% CI: 1.64 – 3.57%), and 5.42% (95% CI: 3.95 – 7.42%). Locoregional V 15 Gy was the best predictor of any grade RN and G2+ RN followed by V 18 Gy . V 15 Gy cutoffs of 2.52cc, 4.55cc, 6.06cc; 4.65cc, 6.36cc, 7.68cc were associated with <5%, <7.5%, and <10% risk of any grade RN and G2+ RN, respectively. V 18 Gy cutoffs of 1.92cc, 3.54cc, 4.65cc; 3.54cc, 4.95cc, 5.96cc were associated with <5%, <7.5%, and <10% risk of any grade RN and G2+ RN, respectively (Table 1). The probability of symptomatic RN with V 15 Gy < 5cc was 5.3% (95% CI: 2.8-9.9%), V 18 Gy < 4cc was 5.5% (95% CI: 2.8-10.6%) (Figure 1).