Prognostic predictors for breast cancer patients with brain metastasis after radiotherapy

黄晓波,蒋国梁,陈佳艺

DOI: https://doi.org/10.3969/j.issn.1007-3639.2004.01.013

2004-01-01

Abstract:Purpose:To identify the prognostic predictors a nd to establish a prognostic index model for breast cancer patients with brain m etastasis(BM) after irradiation.Methods:From Jan. 1994 to Dec. 2000, 61 patients irradiated for BM from breast cancer, and who had both complet e medical charts and follow-up data available, were eligible for this retrospec tive analysis. A number of potential factors which might affect prognosis after irradiation were evaluated, including age,KPS, primary disease stage, ER status, menses status, first sign of BM, extension and numbers of BM lesions, control o f primary tumor, presence of extracranial metastasis and sites of metastasis, fr actionation of WBRT, total dose of radiotherapy, chemotherapy, etc. The signific ance of prognostic variables in the survival resulted from both univariate analy sis by Kaplan-Mierer combining with log-rank test and multivariate Cox regress ion model. The prognostic index(PI) was established based on Cox regression anal ysis and subgrouping values.Results:For the entire cohort, the median survival from the sta rt of radiation for BM was 8 months, and the actuarial overall survival rate was 30% and 10% at 1, 2 years respectively. Univariate analysis showed KPS, age, nu mbers of BM, extracranial systemic metastasis, total radiation dose and chemothe rapy were predictors of prognosis, and patients with higher total dose(50Gy) in solitary or limited BM subgroup seemed to survive longer. However in the Cox mu ltivariate analysis, only KPS, numbers of BM, extracranial systemic metastasis a nd chemotherapy were significant prognostic factors. Just as those of RPA subgro ups, the median survival from the start of WBRT was 3, 9, 16 months in 3 PI subg roups respectively.Conclusions:Overall survival in BM from breast cancer remains p oor. KPS, numbers of BM, extracranial systemic metastasis and chemotherapy were independent prognostic factors. RPA classification can predict the prognosis of patients with BM from breast cancer, so can the PI model.