CMET-42. USE OF STEREOTACTIC RADIOSURGERY AND OUTCOMES FOLLOWING RADIOTHERAPY FOR BRAIN METASTASES IN PATIENTS WITH PRIMARY LUNG CANCER – AN AUSTRALIAN POPULATION-BASED STUDY

Abstract

Abstract INTRODUCTION We aim to evaluate the use of stereotactic radiosurgery (SRS) among lung cancer (LC) patients who received radiotherapy (RT) for brain metastases (BM), and the associated outcomes. METHODS This is a population-based cohort of LC patients who received RT for BM between 2013 and 2016, as captured in the Victorian Radiotherapy Minimum Dataset. Brain RT was classified as SRS (including multi-fraction stereotactic RT) and non-SRS. Mortality data was captured through data linkage with the Victorian Cancer Registry (VCR). The primary outcomes were: proportion of patients who had SRS for BM, and overall survival (OS) following brain RT. Multivariable logistic regression was used to evaluate factors associated with SRS use. Kaplan-Meier method was used to estimate OS. Multivariable Cox regression was used to evaluate factors associated with OS. RESULTS 1,002 LC patients were included in the study. 1,395 courses of RT for BM were delivered, of which one quarter (362/1,395) were SRS. Almost all SRS were delivered in metropolitan centres (347/362), and two-third in public institutions (235/362). In multivariate logistic regressions, increasing age (OR=0.91;95%CI=0.82–0.99;P=0.04) and treatment in regional centres (OR=0.13;95%CI=0.06–0.28;P< 0.001) were independently associated with lower likelihood of SRS use. Median follow-up was 3.3 months (IQR=1.3–7.8 months). 876 deaths were observed with 12-month OS of 16%. The 12-month OS for patients who had SRS were 38% vs. 12% in patient did not have SRS (P< 0.001). In multivariable Cox regressions, the use of SRS (HR=0.44,95%CI=0.37–0.54;P< 0.001) and female (HR=0.76;95%CI=0.67–0.87;P< 0.001) were associated with improved OS, while increasing age was associated with worse OS (HR=1.02;95%CI=1.01–1.03;P< 0.001). CONCLUSION This is the largest Australian study reporting on SRS use for BM in LC patients. We observed geographical variations in SRS use. We believe the impact of SRS use on OS is most likely due to patient selection for SRS.