Comparison of preoperative versus postoperative treatment dosimetry plans of single-fraction stereotactic radiosurgery for surgically resected brain metastases.

Objective: Stereotactic radiosurgery (SRS) for operative brain metastasis (BrM) is usually administered 1 to 6 weeks after resection. Preoperative versus postoperative timing of SRS delivery related to surgery remains a critical question, as a pattern of failure is the development of leptomeningeal disease (LMD) in as many as 35% of patients who undergo postoperative SRS or the occurrence of radiation necrosis. As they await level I clinical data from ongoing trials, the authors aimed to bridge the gap by comparing postoperative with simulated preoperative single-fraction SRS dosimetry plans for patients with surgically resected BrM.

Methods: The authors queried their institutional database to retrospectively identify patients who underwent postoperative Gamma Knife SRS (GKSRS) after resection of BrM between January 2014 and January 2021. Exclusion criteria were prior radiation delivered to the lesion, age < 18 years, and prior diagnosis of LMD. Once identified, a simulated preoperative SRS plan was designed to treat the unresected BrM and compared with the standard postoperative treatment delivered to the resection cavity per Radiation Therapy Oncology Group (RTOG) 90-05 guidelines. Numerous comparisons between preoperative and postoperative GKSRS treatment parameters were then made using paired statistical analyses.

Results: The authors' cohort included 45 patients with a median age of 59 years who were treated with GKSRS after resection of a BrM. Primary cancer origins included colorectal cancer (27%), non-small cell lung cancer (22%), breast cancer (11%), melanoma (11%), and others (29%). The mean tumor and cavity volumes were 15.06 cm3 and 12.61 cm3, respectively. In a paired comparison, there was no significant difference in the planned treatment volumes between the two groups. When the authors compared the volume of surrounding brain that received 12 Gy or more (V12Gy), an important predictor of radiation necrosis, 64% of patient plans in the postoperative SRS group (29/45, p = 0.008) recorded greater V12 volumes. Preoperative plans were more conformal (p < 0.001) and exhibited sharper dose drop-off at the lesion margins (p = 0.0018) when compared with postoperative plans.

Conclusions: Comparison of simulated preoperative and delivered postoperative SRS plans administered to the BrM or resection cavity suggested that preoperative SRS allows for more highly conformal lesional coverage and sharper dose drop-off compared with postoperative plans. Furthermore, V12Gy was lower in the presurgical GKSRS plans, which may account for the decreased incidence of radiation necrosis seen in prior retrospective studies.