Brain metastases (BM) are the most common form of intracranial cancer. The incidence of BM seems to have increased over the past decade. Recursive partitioning analysis (RPA) of data from three Radiation Therapy Oncology Group (RTOG) trials (1200 patients) has allowed three prognostic groups to be identified. More recently a simplified stratification system that uses the evaluation of three main prognostics factors for radiosurgery in BM was developed.
To analyze the overall survival rate (OS), prognostic factors affecting outcomes and to estimate the potential improvement in OS for patients with BM from breast cancer, stratified by RPA class and brain metastases score (BS-BM). From January 1996 to December 2004, 174 medical records of patients with diagnosis of BM from breast cancer, who received WBRT were analyzed. The surgery followed by WBRT was used in 15.5% of patients and 84.5% of others patients were submitted at WBRT alone; 108 patients (62.1%) received the fractionation schedule of 30 Gy in 10 fractions. Solitary BM was present in 37.9 % of patients. The prognostic factors evaluated for OS were: age, Karnofsky Performance Status (KPS), number of lesions, localization of lesions, neurosurgery, chemotherapy, absence extracranial disease, RPA class, BS-BM and radiation doses and fractionation.
The OS in 1, 2 and 3 years was 33.4 %, 16.7%, and 8.8 %, respectively. The RPA class analysis showed strong relation with OS (p < 0.0001). The median survival time by RPA class in months was: class I 11.7, class II 6.2 and class III 3.0. The significant prognostic factors associated with better OS were: higher KPS (p < 0.0001), neurosurgery (P < 0.0001), single metastases (p = 0.003), BS-BM (p < 0.0001), control primary tumor (p = 0.002) and absence of extracranial metastases (p = 0.001). In multivariate analysis, the factors associated positively with OS were: neurosurgery (p < 0.0001), absence of extracranial metastases (p <0.0001) and RPA class I (p < 0.0001).
Our data suggests that patients with BM from breast cancer classified as RPA class I may be effectively treated with local resection followed by WBRT, mainly in those patients with single BM, higher KPS and cranial extra disease controlled. RPA class was shown to be the most reliable indicators of survival.