Postoperative stereotactic radiosurgery for patients with resected brain metastases: a volumetric analysis.

ABSTRACT

PURPOSE: Postoperative stereotactic radiosurgery (SRS) is increasingly utilized following resection of brain metastases (BM); however, there are no volumetric data guiding dose selection. We performed a volumetric analysis to guide cavity SRS dosing for resected BM. METHODS: 83 consecutive patients with gross total resection who underwent postoperative SRS to 90 cavities were identified. The 12 Gy isodose lines (V12total) along with the volume of brain parenchyma receiving 12 Gy excluding cavity fluid, ventricular fluid, and calvarium (V12parenchyma) were contoured. Local recurrence (LR) and radionecrosis (RN) were calculated using cumulative incidence rates. Multivariate analysis (MVA) and cutpoint analysis were conducted. RESULTS: Median follow-up was 12.3 months; median dose was 16 Gy. 1- and 2-year cumulative incidence rates of LR were 7.9% and 11.0%. Radiation dose [hazard ratio (HR) 2.04, p = 0.002] was significantly associated with time to LR on MVA. 1- and 2-year cumulative incidence rates of RN were 2.6% and 5.5% respectively. MVA demonstrated increased risk of RN with a larger V12parenchyma (HR 1.46, p = 0.0496). Cavities ≤ 10 cc showed a low 2-year RN risk (4.3%), but had a modest LR risk (13.9%). A radiation dose ≥ 18 Gy significantly improved LC (HR 4.79, p = 0.01). CONCLUSIONS: V12parenchyma should be examined in postoperative SRS to assess RN risk. Cavities > 10 cc treated with 16 Gy achieved excellent LC and minimal RN at 2 years. Cavities ≤ 10 cc may be better treated with a dose ≥ 18 Gy to significantly improve LC given the low RN rate observed with 16 Gy.