Phase III Randomized Trial Comparing Preoperative Hypofractionated Radiosurgery (HSRS) to Postoperative Hypofractionated Radiosurgery (HSRS) for Patients With Large Brain Metastases (= 2.1cm) Suitable for Surgical Resection

The occurrence of BMs is a huge and challenging issue affecting about 20-40% of patients with solid primary tumors. Among these, about 25% of patients harbored large BMs, defined as ≥ 2.1 cm. Single dose SRS, using the dose guidelines suggested by the Radiation Therapy Oncology Group (RTOG) 90-05 study, obtains an unsatisfactory local control (LC) rate ranging from 45-49%. In this subset of patients other treatment pathways have been investigated. In the 1990s, Patchell and colleagues determined that patients with good functional status, and solitary intracranial metastases should undergo surgical resection. Unfortunately, surgery alone is able to control tumor in only 50% of patients, and an adjuvant radiation therapy (RT) is required. For several years, adjuvant whole brain radiation therapy (WBRT) has been considered the standard of cure, but a high risk of impairment in neurological functions was recorded, without an actual benefit on survival. Different RT approaches have been inquired with the aim to reduce neurological toxicity preserving the same brain tumor control. Recent randomized trials showed that single dose SRS on the tumor bed might be a valid, and less toxic alternative to WBRT, although an increased risk of radio necrosis (RN) was noticed when large surgical cavities are treated. In the last years hypofractionated stereotactic radiosurgery (HSRS) has gained interest. Its goal is to reduce the risk of RN compared to single dose SRS, while providing similar, or perhaps, improved LC, probably in relation to the need of reducing the dose prescribed in cases of larger lesions using SRS.

However, there has been increasing evidence that patients treated with postoperative SRS have an increased rates of leptomeningeal disease (LMD) occurrence than what was observed when postoperative WBRT was used as the standard. Several retrospective studies have demonstrated a LMD rates up to 31% in the postoperative SRS setting.

The proposed mechanism of this increased risk is iatrogenic tumor dissemination into the cerebrospinal fluid (CSF) at the time of surgical resection, which was not as apparent when the entire intracranial CSF space was treated with routine postoperative WBRT, but has become more apparent with increasing use of postoperative SRS only. It is important to note that a standardized definition of radiographic LMD does not exist and ascertainment bias as to what constitutes radiographic LMD (vs local or distant meningeal failure as an example) is an unresolved issue.

Due to the perceived drawbacks of postoperative SRS, namely the need for cavity margin expansion due to target delineation uncertainty, the variable postoperative clinical course and potential delay in administering postoperative SRS, and the theoretical risk of tumor spillage into CSF at the time of surgery, investigators began to study the use of preoperative SRS as an alternative paradigm to maximize local control of the resection cavity and minimize neurocognitive detriment associated with WBRT. Preoperative SRS has several potential advantages compared to postoperative SRS consisting in :

• a better target delineation to an intact lesion
• the reduction of normal brain irradiated considering the useless of additional margins
• the potential prevention of any cells spilled during resection
• a greater oxygenation ratio of the intact region
• a sterilization effect
• the resection of the majority of irradiated tissues Based on this background we designed this phase III randomized trial comparing preoperative HSRS to postoperative HSRS in patients with large at least one BMs from solid tumors suitable for surgical resection.