Re-TREAT: A Phase 2 Study of Stereotactic Re-irradiation of Relapsed Brain Metastases

Background It is estimated that 10-20 % of patients with cancer develop brain metastases (BM) and the number is increasing due to prolonged patient survival as a result of improvements in surgery and systemic treatment. With few exceptions however, chemotherapy is not effective in treating disease within the central nervous system and therefore BM are often treated as a separate compartment independently from the rest of the disease throughout the body. Treatment modalities include surgical resection, stereotactic radiosurgery (SRS) and whole brain irradiation (WBI) or a combination.

SRS is a technique where a high dose of radiation is delivered through multiple fields, often as rotational intensity modulated radiation therapy (IMRT), where the beam is aimed at the tumor from everchanging angles, in order to minimize the dose to the surrounding tissue. In contrast to other radiation modalities, where a margin is added to the target in order to eradicate tumor cells that are not visible on imaging, SRS employs very small margins because brain metastases tend to be localized with sharp boundaries to surrounding tissue. The relative low dose absorbed by the surrounding tissue allows for delivery of the whole radiation dose to the tumor in one to three treatment sessions ('fractions'), as opposed to conventionally fractionated radiotherapy which is delivered in multiple fractions for the protection of surrounding healthy tissue.

As a whole, SRS results in high rates of local tumor control (65-90%) and low rates of toxicity (10-15%), which is usually mild. The most significant adverse event following SRS is radionecrosis, which is estimated to occur in 5-25% of patients. This is a condition which is poorly defined but is commonly used to describe a situation where the contrast enhancing area in the irradiated volume increases, often accompanied by increasing edema in the surrounding tissue. The condition may or may not cause symptoms. In cases of symptomatic, progressive radionecrosis, oral corticosteroids such as prednisolone may be needed for longer periods of time. Differentiation between relapsed metastasis and radionecrosis is notoriously difficult and represents one of the major diagnostic challenges in all of neuro-oncology. Advanced imaging techniques such as dynamic MRI (e.g. perfusion weighted images) or PET (positron emission tomography) have been used to aid in differentiation in experimental studies, but this currently remains an area of research. The development of the brain lesion over time and use of close follow-up using MRI will most often aid in interpretation of the images by the multidisciplinary team conference consisting of a radiologist, a radiation oncologist and a neurosurgeon.

There is no standard treatment in case of local tumor relapse following SRS. Patients who have not previously received WBI may be offered this, but the possible detrimental effects on neurocognition and quality of life are well documented4 and many clinicians as well as patients prefer to avoid this for as long as possible. Few patients in this category are recommended surgical resection.

Salvage SRS (repeated SRS to the same area at time of relapse) is commonly practiced at some institutions, Rigshospitalet being one of them, but the evidence to support routine use is sparse. The treatment has been evaluated only in retrospective studies. These have found high rates of local control (70-90%), but as adverse events and side effects cannot be evaluated from retrospective studies without risk of critical bias, the safety of this treatment remains to be established.