Stereotactic Radiosurgery for Resected Brain Metastases: Single-Institutional Experience of over 500 Cavities

Abstract

While post-operative stereotactic radiosurgery (SRS) is the standard of care for resected brain metastases, SRS techniques are not standardized. We report patterns of failure and toxicity outcomes from our institutional experience with post-resection SRS. This IRB-approved retrospective review included patients with resected brain metastases treated from 2007-2018 with cavity SRS within 70 days of resection. Brain failure was defined as local (LF) if within the surgical cavity involved with tumor pre-resection, distant (DF) if a new parenchymal tumor, or leptomeningeal (LMD) for new nodular or classical leptomeningeal enhancement. The cumulative incidences of failure and adverse radiation effect (ARE) were analyzed with death and whole brain radiation therapy (WBRT) as competing risks. Overall survival (OS) was analyzed with the Kaplan Meier method. In total, 442 patients with 501 brain metastases were included. Median age of the cohort was 62 years (interquartile range [IQR] 54-69 years). Primary histologies included cancers of the breast (18%), lung (34%), melanoma (11%), kidney (6%), gastrointestinal tract (12%), and other (19%). Prior to resection, 54% of patients had more than 1 brain metastases (median 2, IQR 1-3), 9% had prior WBRT, and 18% had prior resection or SRS to other brain metastases. Median diameter and distance from pia of lesions were 3 cm (IQR 2.2-4 cm) and 0 cm (IQR 0-0.6 cm), respectively, with 24% lesions located below the tentorium and 12% having prior resection or SRS. The majority of lesions (89%) were gross totally resected. Lesions were treated with SRS at median 19 days (IQR 14-27 days) after resection using median dose of 24 Gy (range 12-37 Gy) in median 3 fractions (range 1-5 fractions), with median single fraction equivalent dose of 20 Gy (range 12-33 Gy). The majority of cavities (76%) were treated with a margin (1-3 mm), with 65% of deep tumors (>1 cm from pia) having had the surgical corridor covered. Median planning treatment volume was 14 cm3 (IQR 9-22 cm3). Median follow up after SRS was 9 months (IQR 3-19 months). Median OS after SRS was 13 months (95% confidence interval [CI], 12-15 months). The 12-month failure rates were: LF 7% (95% CI, 5-10%), DF 37% (95% CI, 33-42%), and LMD 13% (95% CI, 10-16%). Of the LMD failures, 54% were classical and 46% were nodular. The 12-month ARE rate was 6% (95% CI 4-8%), with 50% requiring steroids or bevacizumab, and the other 50% requiring re-resection. We report here the largest clinical experience of post-operative SRS for resected brain metastasis, showing its safety and efficacy for controlling brain metastasis progression. Future research to further refine SRS technique and improve outcomes is warranted.