Introduction/Purpose Glioblastoma multiforme is the most common and aggressive type of brain tumor, with a median survival time of 15 months despite treatment. We propose the use of Y-90 endovascular radiosurgery (ER) to increase treatment efficacy while reducing the neurotoxicity associated with radiotherapy in glioblastoma. To that end, this study aimed to evaluate the safety and feasibility of Y-90 ER in the treatment of glioma in a canine model.

Materials and Methods Three healthy research dogs (R1-3) and five client-owned dogs on anti-epileptics with spontaneous, intra-axial brain masses (P1-5) received unilateral Y-90 glass microsphere infusions in either the PCA (R1), MCA (R2), or ICA (R3, P1-5), followed by quantitative Y-90 PET/CT. R1-3 had neurological exams, as clinically indicated, and a 4-week post-ER MRI. P1-4 had serial neurological exams and 1-, 3-, and 6-month MRIs. Due to a small sample size, only descriptive statistics are reported.

Results R2-3 developed transient neurologic defects consistent with the treated side, which resolved in 13 days. R1 had no post-procedure neurologic deficits. The treated hemisphere in R1-3 received a maximum of 378±121Gy (x̄±σ) of radiation, with maximum dose twice as high in more distal deliveries (PCA, MCA). MRIs at 1 month were normal without atrophy or microinfarction.

All dogs except P2 were on corticosteroids and seizure-free prior to treatment. P1-3 had transient post-procedure neurologic deficits which resolved in 20–33 days, while P4 had no neurologic deficits. P5 passed away 12-hours post-ER. The masses received 46.2±11.7Gy of radiation with 19.3±12.2% of the mass volume receiving >70Gy. MRIs at 1-month post-ER showed decreased mass size in all four dogs: by 69% and 59% in P1 and P2 on post-contrast MRI and by 24% and 26% in P3 and P4 on FLAIR MRI. The average number of spheres per cm³ of lesion, calculated using the measured activity per cm³ on PET/CT and expected activity per bead, ranged from 1490 (P2) to 5280 (P1).

At 53 days, P2’s seizures returned with tumor enlargement, and he was euthanized 5 months post-ER. P1 remained asymptomatic until her 6-month visit, when left rear limb proprioceptive delay was observed. At her 12-month follow-up, left thoracic limb proprioceptive delay was also observed, though this did not suggest significant change in the known right cerebral cortex lesion. P3 remained asymptomatic with stable disease at his 6-month visit. P4 developed medically manageable seizure activity and a unilateral menace deficit with no mass growth on MRI at 3 months post-ER, though at 6 months mass volume began trending towards pre-treatment size.

Conclusion Y-90 ER in the canine brain is technically feasible and caused no permanent neurologic deficit despite >400Gy of radiation to critical brain structures. Four of five patient dogs had favorable dosimetric, radiologic, and clinical outcomes, all outliving the 63-day mean survival time associated with their original diagnosis and symptomatic treatment. Long-term outcomes, histopathology, and a larger sample size are needed to better understand brain Y-90 ER viability.

Disclosures S. Manupipatpong: None. A. Pasciak: None. F. Hui: None. R. Krimins: None. L. Gainsburg: None. M. Dreher: 5; C; BTG/Boston Scientific. D. Kraitchman: None. C. Weiss: 1; C; BTG/Boston Scientific. 2; C; BTG/Boston Scientific.