Interpreting the improved outcome of patients with central nervous system metastases managed in clinical trials compared with standard hospital practice

The aims were to determine the median survival and prognostic factors of patients with central nervous system (CNS) metastases managed with whole-brain radiation therapy (WBRT), and to explore selection criteria in recently published clinical trials using aggressive interventions in CNS metastases. A retrospective audit was performed on patients managed with WBRT for CNS metastases. Potential prognostic factors were recorded and analysed for their association with survival duration. The proportion of patients with these factors was also compared with those of patients managed under three recently reported studies investigating aggressive interventions, such as radiosurgery and chemotherapy for CNS metastases. Seventy-three patients were treated with WBRT for cerebral metastases over a 12-month period. The median survival of the population was 3.4 months (95% confidence interval: 2.7-4.1), with 6- and 12-month survival rates of 30 and 18%, respectively. Significant prognostic factors for prolonged median survival were Eastern Cooperative Oncology Group status 0-2 (P = 0.015), Medical Research Council neurological functional status 0-1 (P = 0.006), and Recursive Partitioning Analysis Class 2 versus Class 3 (P = 0.020). On multivariate analysis, younger patient age (P = 0.02) and better performance status (P<0.01) were associated with improved outcome. When comparing these characteristics with selected published studies, our study cohort demonstrated a higher proportion of patients with poor performance status, a greater number of metastases per patient and a higher incidence of extracranial disease. This reflects the selected nature of patients in these published studies. Central nervous system metastases confer a poor prognosis and, for the majority of patients, aggressive interventions are unlikely to improve survival. The use of potentially toxic and expensive treatments should be reserved for those few in whom these studies have shown a potential benefit. Copyright (2005) Blackwell Science Pty Ltd