Prognosis prediction and risk factors for triple-negative breast cancer patients with brain metastasis: A population-based study.

BACKGROUND

Brain metastasis (BM) in triple-negative breast cancer (TNBC) patients is associated with significant morbidity and mortality. In this research we aimed to develop a nomogram to predict the prognosis of TNBC patients with BMs (TNBC-BM) and explore the potential risk factors.

METHODS

We used data from the Surveillance, Epidemiology, and End Results (SEER) database. A prognostic nomogram was built and validated based on patients with BM at newly diagnosed TNBC (nTNBC-BM). Its effect on TNBC patients with BM was also validated in an extended group. The prognostic effect of treatment and risk factors for nTNBC-BM were further tested.

RESULTS

A nomogram was constructed and validated to predict overall survival (OS) in TNBC-BM patients. For patients with BM diagnosed at the initial treatment or later course, the C-index (0.707, 0.801, and 0.685 in the training, validation, and extended groups, respectively) and calibration plots showed the acceptable prognostic accuracy and clinical applicability of the model. Surgery on the primary tumor and chemotherapy were found to confer significantly better OS (11 months vs. 4 months; 5 months vs. 3 months, respectively). In addition, advanced tumor/nodal stage and bilateral cancer were associated with a higher risk of nTNBC-BM.

CONCLUSION

We developed a sensitive and discriminative nomogram to predict OS in TNBC-BM patients, both at initial diagnosis and the latter course. nTNBC-BM patients may benefit more from surgery and chemotherapy than from radiotherapy. In addition, in the predictive model, TNBC patients harboring advanced tumor/nodal stages and bilateral tumors were more likely to have BM at initial diagnosis.