Validity of Brain Metastasis Risk Predictive Model in Hormone Positive Breast Cancer Patients

• Data collection:

Age at BC diagnosis, tumor size, axillary adenopathy, clinical stage, HER2 status, Ki67 proliferation index and the modified Scarff-Bloom-Richardson grade of differentiation.

• Analysis Plan:

Continuous variables as means and standard deviation (SD) if they follow a normal distribution; if not, as medians and IQR (interquartile range). Numbers and percentages display categorical variables.

• Model validation:

The association between covariates and the diagnosis of brain metastases by performing logistic regression analyses; odds ratio (OR), 95 % confidence intervals (CI), and p-value will be estimated. Variables with a p-value < 0.05 in the univariate logistic regression analysis will be further assessed in a multivariate logistic regression analysis.

For the clinical use of the model, we will use a score to each variable as that designed by Cacho-Díaz B. et. al. then, we will categorize patients into three risk groups according to the sum of the scores. We will employe a Cox regression analysis to assess each group's risk of developing brain metastases. Adjusted hazard ratios (aHRs), 95 % CIs, and p values will be evaluated.

Inclusion criteria

• Confirmed diagnosis of HR-positive primary invasive breast cancer.

Exclusion criteria

• Patients if they were diagnosed with brain metastases at the time of initial breast cancer diagnosis or within the subsequent three months.
• Bilateral breast cancer.
• Synchronic cancer (i.e., lung cancer, melanoma).
• Paitents presenting neurologic symptoms without follow-up or lacking a contrast-enhanced magnetic resonance image (cMRI).
• Incomplete clinicopathological information at the time of diagnosis.