Pattern of Metastasis as a Risk Factor for Brain Metastasis in Patients With Breast Cancer: A Time-Dependent Cox Regression.

[AIM] Brain metastasis is associated with poor prognosis in patients with breast cancer. Conventional predictive models rely on fixed baseline characteristics and provide limited clinical utility. We aimed to analyze the factors associated with brain metastasis using time-dependent Cox regression.

[METHODS] We retrospectively reviewed the records of consecutive patients who underwent breast cancer surgery between 2002 and 2020 at a single institution. The number of metastatic organs was defined as a time-dependent covariate. Multivariable Cox regression analysis was performed to identify factors associated with brain metastasis.

[RESULTS] Of the 2459 patients included in the study, 58 (2.4%) patients developed brain metastasis during a median follow-up of 62 months. On time-dependent multivariable Cox regression, advanced T stage, hormone receptor negativity, and the number of metastatic organs were independent predictive factors of brain metastasis. Hazard ratios (HRs) increased with metastatic burden: 30.06 for single-organ metastasis, 73.39 for double-organ metastases, 265.10 for triple-organ metastases, and 278.60 for quadruple or more organ metastases. Compared with patients whose first distant metastasis occurred at other sites, those with bone metastasis were less likely to develop brain metastasis (HR, 0.280; 95% confidence interval [CI], 0.099-0.792; p = 0.016), whereas those with soft tissue metastasis were more likely to develop brain metastasis (HR, 4.048; 95% CI, 1.391-11.782; p = 0.010).

[CONCLUSION] The number and site of metastatic organs are key risk factors of brain metastasis. Personalized brain screening may facilitate early detection of brain metastasis in high-risk patients.