The Impact of Dose on Treatment Outcomes in Patients Treated for Brain Metastases With a Hypofractionated Technique on Gamma Knife

Abstract

Purpose: Brain metastases are the most common type of diagnosed central nervous system tumours, occurring in approximately 40% of all cancer patients. Historically, treatment consisted of whole brain radiotherapy (WBRT), however current practice for patients with limited intracranial disease, is to offer stereotactic radiosurgery (SRS). In 2016, our centre acquired a Gamma Knife (GK) unit, designed to treat brain metastases with a stereotactic approach, using an invasive frame based immobilization system. In 2017, a mask based frameless system was introduced, expanding indications for patients and allowing for the treatment of hypofractionated courses on this unit. Given the unique dosimetric characteristics of the GK, our research focused on patients treated for intact brain metastases, with hypofractionated doses, and aimed to report adverse radiation effects and local control. Methods: This was a retrospective study, including all patients with intact brain metastases, treated on the GK unit, with hypofractionated (5 fraction) treatments, between 2017-2020. Hypofractionated treatments are generally given to larger lesions (≥2cm) or those lesions situated in eloquent areas of the brain, such as the motor cortex or brainstem. Patients were excluded if the specific metastases received prior SRS or hypofractionated treatment, however those who received previous whole brain radiotherapy were included. Those without post -treatment imaging follow up were also excluded from the analyses. Competing risk methods were used to determine local control and symptomatic adverse radiation effect (ARE) rates per lesion and Kaplan Meier methods were used to determine survival estimates. The Common Terminology Criteria for Adverse events (CTCAE) v 5.0 was used to determine symptomatic ARE. Results: A total of 146 patients and 299 metastases were identified. The most frequent histologies were lung (38.5%), breast (31.4%) and melanoma (13.7%). 53.4% of patients had a KPS > 70% and the majority had multiple lesions treated at the time of their hypofractionated radiosurgery (62.3%). Only 18% had received prior WBRT. Prescribed doses ranged from 20-27.5Gy with the median being 27.5Gy in 5 fractions prescribed in 54.2% of cases. The median volume of the treated lesions was 2.54cc (range 0.004-24.92cc). Clinical follow-up ranged from 0.5 to 49.9 months, with the median being 13.8 months. The 1 year local control was 84.8%, and on multivariable analyses, 1 year local failure rates increased (8.3% (95% CI, 4.6-13.2) vs 23.5% (95%CI,16.6-31)) with decreasing prescribed doses (27.5Gy vs 20-25Gy) respectively. In terms of adverse radiation effects, there were 32 (10.7%) incidents identified, of which 14 (44%) were symptomatic. For symptomatic AREs, the 1 and 2 year rates were 1.7% (95% CI, 0.7-3.8) and 4.5% (95% CI, 2.4-7.7), respectively. Overall survival for this patient population was 12.7 months, with the 1 and 2 year rates being 51.3% (95% CI 42.6-59.2) and 36.8% (95%, CI 28.5-45.1), respectively. Conclusion: For patients with intact brain metastases, GK hypofractionated treatments are a safe and effective treatment modality. A 27.5Gy prescription dose in 5 fractions is associated with improved local control and a low risk of symptomatic ARE at one year.