Purpose: A retrospective study of patients treated at the Peter MacCallum Cancer Institute for brain metastases from primary carcinoma of the lung is presented.
Methods and materials: The medical records of 416 patients with the diagnosis of primary carcinoma of the lung who presented with, or subsequently developed, brain metastases during the period January 1984 to December 1987 were reviewed. Information on a number of factors of potential prognostic significance (sex, age, histology, performance status and interval between diagnosis of the primary and brain metastases) was collected. Details of surgery, radiation and steroid usage were recorded, and any steroid side effects documented. Survival was calculated from the date of diagnosis of brain metastases. Stepwise regression based on Cox's proportional hazards model was used to determine significant prognostic factors affecting survival. Patients with and without steroid side effects were compared using Yates's corrected chi-square test.
Results: The overall estimated median survival was only 3.3 months (95% confidence interval 2.9-3.7 months). Only two factors were found to be associated with a significantly improved survival--surgical intervention and good performance status. After taking these two factors into account, the dose of radiation used (< 30 Gy or > or = 30 Gy) did not influence survival. There was a 3% incidence of gastric bleeding or perforation in patients taking steroids, with a 40% fatality rate. Predisposing factors to gastric side effects were a prior history of peptic ulcer and/or aspirin or nonsteroidal anti-inflammatory drug consumption.
Conclusion: Radiation of brain metastases from primary lung cancer results in modest survival benefit. Radiation dose (< 30 Gy or > or = 30 Gy) is not a significant determinant of survival. Other treatment modifications, such as concurrent radiation and chemotherapy, should be explored. Steroids should be used with caution as fatal side effects can occur.