Radiation necrosis and therapeutic outcomes in patients treated with linear accelerator-based hypofractionated stereotactic radiosurgery for intact intracranial metastases

Qichen Zhang, Daniel Hamilton, Paul Conway, Sophia Jing Xie, Neda Haghighi, Arian Lasocki

Research output: Contribution to journalArticleResearchpeer-review

1 Citation (Scopus)


Introduction: Balancing disease control and treatment-related toxicities can be challenging when treating higher-risk brain metastases (BMs) that are larger in size or eloquent anatomical locations. Hypofractionated stereotactic radiosurgery (hfSRS) is expected to offer superior or equal efficacy with lower toxicity profile compared with single-fraction SRS (sfSRS). We report the efficacy and toxicity profiles of hfSRS in a consecutive cohort of patients to support this predicted benefit from hfSRS for high-risk BMs. Methods: We retrospectively analysed 185 consecutive individual lesions from 152 patients with intact BMs treated with hfSRS between 1 July 2016 and 31 October 2019 and followed up to 30 April 2022 with serial brain magnetic resonance imaging (MRI). The primary endpoint was the event of radiation necrosis (RN). Local control (LC) rate and distant brain failure (DBF) were reported as secondary outcomes. Kaplan–Meier method was used to report the cumulative incidence of RN and overall survival and the incidence of DBF. Potential risk factors for RN were assessed using univariable Cox regression analysis. Results: The median follow-up was 38.0 months, and the median survival post-SRS was 9.5 months. The cumulative incidence rate of RN was 13.2% (95% CI: 7.0–24.7%), and 18.1% of patients with confirmed RN were symptomatic. Higher mean dose delivered to planning target volume (PTV) (HR 1.22, 95% CI: 1.05–1.42, P = 0.01), higher mean BED10 (biological equivalent dose assuming a tissue (Formula presented.) ratio of 10) (HR 1.12, 95% CI: 1.04–1.2, P < 0.001), and higher mean BED2 (HR 1.02, 95% CI: 1–1.04, P = 0.04) delivered to the lesion was associated with increased risk of RN. LC rate was 86% and the cumulative incidence of DBF was 36% with a median onset of 28.4 months. Conclusions: Our results support the predicted radiobiological benefit of the use of hfSRS in high-risk BMs to limit treatment-related toxicity with low risk for symptomatic RN comparable with lower risk population receiving sfSRS while achieving satisfactory local disease control.

Original languageEnglish
Pages (from-to)308-319
Number of pages12
JournalJournal of Medical Imaging and Radiation Oncology
Issue number3
Publication statusPublished - Apr 2023
Externally publishedYes


  • magnetic resonance imaging
  • neuroradiology
  • oncologic imaging
  • radiation oncology
  • radiation oncology imaging

Cite this