2519 - Mapping Symptomatic Radionecrosis after Stereotactic Radiosurgery with Functional Magnetic Resonance Imaging

C. Gui¹, M. Jenabi², N. P. Mankuzhy¹, L. R. G. Pike¹, A. Ballangrud², K. Peck², A. Holodny³, and B. S. Imber¹; ¹Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, NY, ²Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, NY, ³Department of Radiology, Memorial Sloan Kettering Cancer Center, New York, NY

Purpose/Objective(s): Stereotactic radiosurgery (SRS) for brain metastases (BM) avoids the cognitive toxicity of whole brain radiotherapy. However, radionecrosis (RN) can diminish quality of life. Symptoms of RN are assumed to be dictated by the anatomic site involved, but eloquent areas of the brain are not accurately localizable with anatomic imaging alone. Functional magnetic resonance imaging (fMRI) is used in neurosurgical planning to localize eloquent areas but is seldom applied to SRS. We performed the first analysis assessing whether symptoms of RN after SRS for BM are associated with nearby functional anatomy. Materials/

Methods: From 2011 to 2022, patients with an fMRI within 3 months of SRS for BM were included. For each patient, one index BM was identified as the lesion closest to functional areas, including the primary motor, language, and visual cortices on fMRI. Outcomes included RN and focally symptomatic RN (FSRN). RN was defined as progression of a contrast-enhancing lesion at the site of the irradiated index BM, which demonstrated no viable disease after resection, stability with no intervention, or stability with corticosteroids alone. FSRN was defined as new or worsening focal neurologic symptoms associated with the nearest functional area to the RN lesion on fMRI. Predictors of RN and FSRN were modeled with univariable competing risk analysis, with death as a competing risk.

Results: This study included 94 patients with 94 index BM. Most received SRS postoperatively (n = 77, 82%). The median presurgical BM diameter was 2.4 cm (range 0.9-5.5). The nearest functional area was the hand, foot, or face motor cortex in 72 patients (77%), Broca’s or Wernicke’s area in 18 patients (19%), and the visual cortex in 2 patients (2%). Common prescriptions included 30 Gy in 5 fractions (n = 53, 56%) and 27 Gy in 3 fractions (n = 12, 13%). Median overall survival after SRS was 26 months. RN and FSRN estimates at 24 months after SRS were 18% (95% CI 10-27%) and 11% (95% CI 5-18%), respectively. Lung histology (n = 38, 40%) predicted RN (HR 3.3, 95% CI 1.3-8.2, p = 0.01). Prior SRS to the same site (n = 7, 7%) predicted RN (HR 6.9, 95% CI 2.5-19, p = 0.0002) and FSRN (HR = 5.4, 95% CI 1.4-21, p = 0.02). Among 20 patients with RN, 4 were asymptomatic, 4 had neurologic symptoms unrelated to the site of RN, and 12 had FSRN. Among patients with FSRN, 9 had weakness or focal seizures, 2 had speech deficits, and 1 had a visual field deficit, all consistent with the site of RN and nearest functional area on fMRI. The index BM directly overlapped with the nearest functional area in 4 of 12 patients with FSRN (33%) and 13 of 82 patients without FSRN (16%).

Conclusion: This is the first analysis showing that symptoms of RN after SRS to BM are associated with functional anatomy defined by fMRI. fMRI may provide major clinical value in SRS planning, especially in cases of reirradiation. The value of dose reduction to functional areas should be investigated.