Clinical Outcomes of Single-Isocenter Versus Multiple-Isocenter Stereotactic Radiosurgery Techniques for Multiple Brain Metastases.

Abstract

PURPOSE/OBJECTIVE(S)\nThere are multiple technical approaches for stereotactic radiosurgery (SRS) delivery to multiple brain metastases, including single-isocenter (SIT) and multiple-isocenter (MIT) techniques. SIT is increasing employed to streamline treatment planning and delivery in the setting of multiple brain metastases such that a single radiosurgery plan is generated to treat all metastases simultaneously, rather than generating and delivering separate plans for each metastasis. In this study, we compare clinical outcomes using a single-isocenter versus multiple-isocenter approaches for treating multiple brain metastases amenable to single fraction radiosurgery.\n\n\nMATERIALS/METHODS\nThis is a single-institution, retrospective review of patients treated from 2017-2019 with intact brain metastases treated using a SIT or MIT with linear accelerator-based SRS. The typical prescription dose was 18 Gy in 1 fraction prescribed to the 80-90% isodose line. Time from end of treatment to local failure or radionecrosis (RN) for each lesion was recorded. On the per-lesion basis, local control (LC) and freedom-from-radionecrosis (FFR), were estimated with Kaplan Meier method and compared between SIT and MIT group via log-rank test. Evidence of RN was defined as radiographic evidence of cerebral edema surrounding a treated lesion with associated clinical symptoms. Multivariable analysis using Cox proportional-hazard regression was used to establish potential predictors for RN, and significance assessed by Likelihood ratio test.\n\n\nRESULTS\n104 patients with a total of 437 metastases (80% supratentorial and 20% infratentorial) were included. 232 lesions were treated with a SIT approach, ranging from 2-13 (median 5) intact metastases treated per treatment course. Median radiographic follow up time was 6.4 months for SIT lesions and 10.7 months for MIT lesions. Primary malignancies included non-small cell lung cancer (n\u202f=\u202f41), breast cancer (n\u202f=\u202f33), melanoma (n\u202f=\u202f14), gastrointestinal cancer (n\u202f=\u202f6), head and neck cancer (n\u202f=\u202f5), and kidney cancer (n\u202f=\u202f5). On a per-lesion basis, 6-month LC was 97% for SIT and 99% for MIT (P\u202f=\u202f0.12). On a per-lesion basis, 6-month FFR was 96% for SIT and 96% for MIT (P\u202f=\u202f0.81). Fourteen patients developed symptomatic RN leading to steroid treatment in 7 patients, laser interstitial thermal therapy in 2 patients, anti-VEGF therapy in 3 patients, and continued observation in 2 patients. The only multivariable predictor for developing RN was increasing PTV size (HR 1.76, P < 0.01).\n\n\nCONCLUSION\nThese data indicate high rates of LC and low rates of RN for both MIT and SIT for SRS. Efficient and effective treatment planning and delivery for the treatment of multiple metastases with a single isocenter improves the clinical feasibility of delivering initial and subsequent radiosurgery courses for patients, allowing for optimal integration with systemic therapy in patients with metastatic disease.\n\n\nAUTHOR DISCLOSURE\nJ. Deng: Employee; UCLA.M.K. Li: None. R.R. Savjani: Research Grant; ASTRO-Varian Research Training Fellowship. F. Chu: None. S.E. Tenn: Research Grant; Brain Lab. Honoraria; Brain Lab. C. Lee: Stock; Varian Medical Systems. N. Agazaryan: Honoraria; Brainlab and Novalis Accreditation. Consultant; Brainlab and Novalis Accreditation. expert panel member; Brainlab and Novalis Accreditation. I. Yang: Consultant; Baxter, Brain Lab. R. Everson: None. W. Kim: None. N. Pouratian: Research Grant and Speaker; Brain Lab. A.U. Kishan: None. R.K. Chin: None. M.L. Steinberg: None. T.B. Kaprealian: Research Grant and Honoraria; Brain Lab. J.V. Hegde: Research Grant; Soylent.