Nomogram for predicting development of brain metastasis in patients with inflammatory breast cancer.

Abstract

243 Background: Inflammatory breast cancer (IBC) is associated with poor prognosis and high rates of distant metastasis. Brain metastasis is particularly important due to its morbidity and mortality. The purpose of this study was to construct a nomogram from clinicopathological factors that could predict the risk of occurrence of brain metastasis in patients with IBC. Methods: We performed a retrospective analysis of 406 patients with IBC, who were seen at the University of Texas M. D. Anderson Cancer Center between 1991 and 2009. Data was collected from IBC registry. Patients were divided in a randomized manner into a training cohort (n = 206) and a validation (n = 200) cohort. A multivariate cox proportional hazards regression analysis of selected prognostic features was performed on the training cohort. The model estimated in the training cohort was cross-validated with the validation cohort. Based on this model a nomogram was constructed to assess 5- and 10-year risk of developing brain metastasis. Results: A total of 115 (28%) patients developed brain metastasis. Univariate analysis identified menopausal status, progesterone receptor expression, number of distant metastasis, adjuvant taxane-based therapy and age at diagnosis as significant factors associated with brain metastasis (p < 0.05). Subsequent multivariate analysis showed menopausal status and number of metastasis were independently associated with development of brain metastasis (p < 0.05). Nomogram constructed using above model was evaluated for its ability to discriminate among likelihood of occurrence of brain metastasis in patients, using the validation set. The validation set showed good discrimination with the area under the receiver operating characteristic curve of 0.68. Conclusions: We have formulated a distinct predictive tool to evaluate 5 and 10-year risk for developing brain metastasis in patients with IBC. This nomogram can be potentially helpful for designing risk adapted therapeutic trials. Identifying patients at high risk for brain metastasis in IBC can help personalize therapy, surveillance, and prognostication. [Table: see text]