Surgical Cavities Exhibit Substantial Remodeling Post Resection of Brain Metastases: Implications for Stereotactic Radiosurgery (SRS)

Abstract

SRS is commonly employed following resection of brain metastases to reduce the risk of local failure. It is unclear when the planning MRI should be performed in order to allow for stabilization of the shape, size, and position of the resection cavity. This quality assurance study quantifies changes in the resection cavity over time by comparing target volumes created on an immediate post-operative MRI with those created on a planning MRI performed two or more weeks after resection. Patients with brain metastases treated with post-operative SRS at our institution over a recent nine-month period were identified. The immediate pre- and post-op MRIs were fused to the planning MRI, typically performed 2 to 3 weeks post-op. The initial tumor volume was delineated on the pre-operative MRI, and defined as cystic or solid. On the immediate post-op and planning MRIs, a post-op cavity CTV was created in accordance with Consensus Contouring Guidelines. The CTV was uniformly expanded 2mm to yield a PTV. We calculated and compared the mean and median post-op and planning volumes, the absolute difference between post-op and planning volumes for each patient, and the percentage difference in volumes relative to the post-op MRI. After overlaying the post-op and planning PTVs, volumes were created to quantify the amount of tissue that would have been either unnecessarily included in, or excluded from, the PTV, had the post-op MRI been used for planning purposes. Twenty-one patients with a total of 23 resection cavities were identified. The median pre-operative tumor volume was 14.3 mL (range 3.1-57.0 mL). The median time from resection to post-op MRI was 1 day (0-2), and to planning MRI was 20 days (13-42). The median post-op CTV was 20.4 mL (3.2-34.1) and PTV was 34.6 mL (6.9-53.9), while the median planning CTV was 20.7 mL (2.7-46.0) and PTV was 33.1 mL (6.0-67.3). In 17 of 23 cavities (74%) the post-op PTV was larger than the planning PTV. The median absolute PTV difference was 5.9 mL (0.8-31.3), or 19.9% (2.9-87.3) of the post-op PTV. In the case of tumors defined as solid (n=12), the median relative volume change was 33.4%, while in cystic tumors (n=11) it was 19.3% (NSS). The mean volume of tissue included in the post-op PTV and not the planning PTV was 9.3 mL (1.2-24.5), while the mean volume of tissue not included in the post-op PTV but included in the planning PTV was 8.1 mL (1.3-32.5); these results represent a 30% and 27% relative difference in volume, respectively, referred to the post-op PTV. SRS is now regularly employed in the treatment of resected brain metastases and a rigorous treatment planning protocol is essential. As the resection bed undergoes substantial remodeling in the weeks following resection, target volumes should be based on a planning MRI, performed shortly before SRS, and not the immediate post-op study.