Preoperative or Postoperative Radiosurgery for Brain Metastases?

Aims: Preoperative radiosurgery (RS) facilitates target definition enabling a reduced planning margin. Our aims were (1) to review the outcome of patients treated with postoperative hypofractionated RS, (2) to compare matched preoperative GTV and PTV with postoperative CTV and PTV, (3) to explore dosimetry of preoperative RS, (4) to evaluate the patients in the context of the two current studies of preoperative RS.

Methods: We reviewed MRI imaging of 20 patients treated with postoperative RS (5 × 6 Gy to the 80% isodose). Brain metastases (BMs) were contoured retrospectively on the preoperative MRI to generate a theoretical preoperative GTV. A planning margin of 1 mm was added to the preoperative GTV and 2mm to the postoperative CTV. 12 Gy and 15 Gy plans were created (iPlan, Brainlab) for brain metastases > 3cm diameter. Patients were evaluated against the eligibility criteria for current Phase I and Phase II trials of preoperative RS (RAD001, NCT02514915).

Results: 5 patients declined MRI follow-up. MRIs showed 80% local control of the tumor bed [12 of 15 evaluable patients, median FU 15 mths, (range 3–55 mths)]. 3 patients (20%) had a local recurrence at the time of distant intracranial brain failure (median 8 mths), of whom one patient had leptomeningeal failure (6.6%). These outcomes are comparable to reports of preoperative RS (local control 85.6%, leptomeningeal relapse 4.5%). The postoperative CTV was larger than the preoperative GTV in 13/20 of our cases. Correspondingly, the postoperative PTV was larger than the preoperative PTV in 14/20 patients. The mean increase in PTV was 160% (18–530%), with increases >200% in the case of metastases <2 cm3. Where a decrease in volume was seen postoperatively, the mean preoperative GTV (often cystic) was 12.9 cm3 (11.3–76.2 cm3) and the average volume decrease was 37.6% (18–60.5%). 5 patients had synchronous BMs that were treated with radiosurgery. 13/20 patients would be eligible for current studies.

Conclusion: Preoperative radiosurgery is compelling due to more precise target definition, the potential for less normal brain toxicity due to a smaller irradiated volume and a dose reduction, a low incidence of leptomeningeal disease, enhanced patient convenience, reduced disruption of systemic therapy and radiosurgery of synchronous BMs. We propose a multidisciplinary registry study of preoperative radiosurgery for BMs indicated for resection with < 5mm midline shift.