Defining treatment for brain metastases patients: nihilism versus optimism

Treatment of brain metastases patients has included whole brain radiotherapy (WBRT) for over 50 years, and there is much data showing this to be associated with short-term gains. The integration of resection and radiosurgery to these patients allows some better prognostic groups to experience long-term local control and improvement in quality of life. The recursive partitioning analysis of the Radiation Therapy Oncology Group (RTOG) has been used as a predictive model for over a decade to identify three classes of patients. Number of lesions has been used to define treatment for a good prognostic subgroup that is eligible for surgery or radiosurgery, but there are few prospective studies of poorer prognosis brain metastases patients to evaluate the influence of number of lesions on the prediction of outcome. We examined patient, treatment and outcome parameters of all brain metastases patients in a 5-year period so that we could measure outcome and evaluate various factors on survival.

This was a population-based study of all brain metastases patients in Southern Alberta between 2000 and 2005. It used an Excel spreadsheet database and STATA 8 software to analyze outcomes. The study included 568 patients representing 4.4% of our radiotherapy population. Median age, performance status and distribution of primary disease sites were comparable with other large series. Overall survival for the whole group was 3.05 months. Independent factors predicting for improved overall survival included younger age, KPS <70, less than four lesions and the use of stereotactic radiosurgery. Presence of extracranial disease or persistence of primary disease did not adversely impact survival outcome.

This series shows that the number of lesions is a strong predictor of outcome. Integration of this factor into a decision-making model allows for identification of not only good prognosis patients who will benefit from aggressive treatment but it also facilitates decision making for poorer prognosis patients who are less likely to benefit from WBRT. Recursive partitioning RTOG class 2 and 3 patients with more than three lesions did particularly poor and had an overall survival of 3 months with WBRT. We question the value of WBRT in this subgroup and wonder if best supportive care would be more justifiable given the low survival figures achieved.