ORE than 240,000 new cases of breast cancer are diagnosed each year and one of every eight wom- en will suffer from breast cancer in her lifetime. Breast cancer ranks second among deaths from cancer in women, with an estimated 41,000 deaths each year. Brain metastases from breast cancer occur in approximately 10% of patients, 2 a rate second only to lung metastases. The pres- ence of intracranial lesions is considered an inauspicious prognostic indicator. Standard treatments for brain metastases include WBRT, chemotherapy, resection, and GKS. The untreated pa- tient with intracranial lesions will survive a median of 1 month after diagnosis, whereas a regimen of corticosteroids may improve the median duration of survival to 2 months. 2 Whole-brain radiation therapy may extend the median du- ration of survival to as many as 4 months, but response rates vary markedly. Resection, in addition to WBRT, improves the median duration of survival and local tumor control rates. Wronski and colleagues 10 reported a median survival time of 14 months after craniotomy with a local tumor con- trol rate of 82%. Unfortunately, neurosurgical candidates are limited to patients with surgically accessible lesions and a good neurological status; in addition, the incidences of morbidity and mortality associated with resection cannot be ignored. Using the Leksell Gamma Knife, GKS is an appealing al- ternative to resection because it can be applied to any site in the brain, especially deep-seated lesions, and it avoids the potential hazards of craniotomy. Brain metastases are an attractive target for GKS; they tend to have a well-defined border, allowing for planning a rapid fall-off in radiation dose, and they are usually spherical, which makes treatment J Neurosurg 103:218–223, 2005 218 Gamma knife surgery for the treatment of intracranial metastases from breast cancer SHARAD GOYAL, M.D., DHEERENDRA PRASAD, M.D., M.CH., FRANK HARRELL JR., PH.D., JULIE MATSUMOTO, M.D., TYVIN RICH, M.D., AND LADISLAU STEINER, M.D., PH.D. Department of Radiation Oncology, Howard University Hospital, Washington, DC; Departments of Radiation Oncology, Health Evaluation Sciences, Radiology, and Neurological Surgery, University of Virginia Health System, Charlottesville, Virginia; and Department of Biostatistics, Vanderbilt University, Nashville, Tennessee Object. The goal of this study was to evaluate the effectiveness and limitations of gamma knife surgery (GKS) in the treatment of intracranial breast carcinoma lesions. Methods. A retrospective analysis of the GKS database at the University of Virginia Health System identified 43 pa- tients with a total of 84 lesions who were treated between 1989 and 2000. All patients who received treatment were includ- ed in this study. Imaging studies were available in 35 patients with 67 treated lesions. The overall duration of median survival was 13 months (95% confidence interval [CI] 7–16 months) after radiosurgery. A univariable Cox regression analysis revealed that a single lesion (p = 0.035), a high Karnofsky Performance Scale (KPS) score (p = 0.019), and a high Score Index for Radiosurgery (SIR) in Brain Metastases (p = 0.036) were associated with a significantly lengthened time to local treatment failure. The median duration of survival for patients grouped according to the SIR as low, middle, and high was 3, 8, and 21 months, respectively (p = 0.00033). A multivariable analysis showed that a high KPS score (p = 0.006), a high SIR (p = 0.014), and advanced age (0.038) were predictive of survival. The 1-, 2-, 3-, and 5-year survival rates were 49, 23, 12, and 2%, respectively. The overall median time to local treatment failure was 10 months (95% CI 6–14 months) after GKS. A univariable anal- ysis demonstrated that a single lesion, higher KPS score, and a higher SIR were associated with a significantly longer time until local treatment failure. A multivariable analysis showed that a higher KPS score and SIR and patients who had re- ceived chemotherapy were associated with a significantly longer time to local treatment failure. Neuroimaging scores given for the enhancement pattern (ring-enhancing, heterogeneous, and homogeneous signal), amount of necrosis (none,  50%, and  50%), and mass effect (none, mild, moderate, and severe) of each treated lesion did not correlate with survival or local treatment failure. Conclusions. The SIR and the KPS score are prognostic factors in patients whose intracranial breast cancer metastases are treated with GKS. The SIR, which includes the KPS score, patient age, systemic disease status, largest lesion volume, and number of lesions, can be used to identify those patients with breast cancer metastasis who would benefit from GKS better than KPS score alone. The contribution of whole-brain radiation therapy to GKS with regard to local tumor control or survival could not be identified. KEY WORDS • gamma knife • breast cancer • brain metastasis M J. Neurosurg. / Volume 103 / August, 2005 Abbreviations used in this paper: CI = confidence interval; CT = computerized tomography; GKS = gamma knife surgery; KPS = Karnofsky Performance Scale; MR = magnetic resonance; SIR = Score Index for Radiosurgery; WBRT = whole-brain radia- tion treatment.