ESTRO 2021 Abstract Book

S47

ESTRO 2021

Maximum doses ≥ 40 Gy in SRS1 and SRS2 resulted in SRN rates of 16% and 21%, compared to 3.8% and 6.5% with maximum does ≤ 30 Gy (Figure 1).

Similarly, V12 > 9 cc in SRS1 and SRS2 resulted in SRN rates of 12.5% and 17.2%, compared to 7.1% and 4.2% with V12 ≤ 5 cc (Figure 2).

Prior immunotherapy was not predictive of RN or SRN. At last follow-up, intracranial disease progression was noted after 45.5% of treatments. Additional radiotherapy (24.6%) and surgery (13.4%) were often utilized in

these cases. Conclusion

We present a large, multi-institutional cohort of brain metastases treated with SRS after local or marginal failure following prior SRS. Repeat SRS generally afforded a high rate of local tumor control and a low rate of SRN. Maximum dose and V12 are predictive factors for RN. Lower V12 and maximum dose may reduce the risk of SRN, but these should not be reduced at the expense of delivering a therapeutic dose to achieve local tumor control. OC-0074 Different pre-operative glioma location patterns in patients with or without post-radiation injury A. van der Boog 1 , S. David 1 , A. Steennis 1 , J.W. Dankbaar 2 , T. Snijders 3 , P. Robe 3 , J. Verhoeff 1 1 University Medical Center Utrecht, Radiation Oncology, Utrecht, The Netherlands; 2 University Medical Center Utrecht, Radiology, Utrecht, The Netherlands; 3 University Medical Center Utrecht, Neurology and Neurosurgery, Utrecht, The Netherlands Purpose or Objective Pseudoprogression and radionecrosis develop in 20-30% of glioma patients after high-dose radiotherapy. These lesions of post-radiation injury can complicate second line treatment decisions as they are often hard to distinguish from progressive disease. The precise cause of post-radiation injury remains unknown, but damage to vascular structures is suspected to play a role. Although radiation dose and administration of chemotherapy are known risk factors, not much is known about spatial predisposing factors. Therefore, we aimed to investigate the role of pre-operative glioma location in the development and extent of post-radiation injury. Materials and Methods In a retrospective database we selected 58 of 144 patients with WHO grade II-IV supratentorial gliomas with adequate contrast enhancement that enabled delineation. They received surgery and postoperative MRI within 3 days, were treated with radiotherapy after first surgery, and had a follow-up of at least 6 months. Tumors were manually delineated on pre-operative T1 contrast-enhanced (CE-T1) MRI and used for localization and volume measurement. Post-radiation injury, defined as new enhancements on CE-T1 MRI which stabilized or decreased after at least 3 months follow-up, was delineated on the scan which showed the largest extent. After spatial normalization to stereotaxic MNI space, we performed voxel-based analysis (VBA) to compare tumor prevalence between two groups of patients: those who developed post-radiation injury versus patients without post-radiation injury. The results were further graded to aid the interpretation of results via the Harvard-Oxford cortical atlas, a white matter atlas (XTRACT) and an arterial territory map. Correction for multiple comparison was executed via family-wise error rate correction to eliminate false positive results. Results Out of 58 glioma patients, 23 developed post-radiation injury on CE-T1 MRI. Tumors located in two confluent areas in the left temporal and insular region showed subsequent post-radiation injury more frequently than tumors in other areas ( P < 0.05) (Fig. 1, Table 1). The area was predominantly located in the arterial territory of the middle cerebral artery (MCA), including the watershed area between the MCA and the posterior cerebral artery (PCA). No tumor locations were identified that were significantly associated with absence of subsequent post-radiation lesion. Contrary to expectations, preoperative tumor volume was not associated with the incidence or volume of post-radiation injury