Abstract Background: Risk factors and timing of BrM development among patients receiving NAC for BC are not well defined. Methods: A retrospective study of two cohorts of patients at the Sunnybrook Odette Cancer Centre was conducted. Cohort 1 included 469 patients treated with NAC for BC from 2008 - 2019. Cohort 2 included 801 patients treated for BC BrM with radiation and/or surgery from 1995 - 2023, among whom 117 received NAC. After unduplicating records, 586 patients were included in the final analysis. Descriptive statistics were used to summarize patient and treatment characteristics. Cox proportional hazards regression was used to estimate hazard ratios (HR) and identify predictors of BrM. Univariable analyses (UVA) were performed for all covariates. Bayesian Information Criteria determined best models from subsets of covariates for multivariable analyses (MVA). The proportional-hazards assumption was appraised by visually inspecting graphs of Schoenfeld residuals and testing their independence over time. Statistical significance was defined as p.05. Results: In the combined cohort of 586 patients who received NAC for BC, the median age at BC diagnosis was 49.0 (42 - 58) years. The most common BC subtype was HER2+ (n=225, 38.4%), followed by hormone receptor (HR)+/HER2- (n=213, 36.3%) and triple negative BC (n=132, 22.5%). In total, 77 patients (13.1%) had inflammatory disease. Following NAC, 134 patients (22.9%) achieved a pathologic complete response (pCR) and 409 (69.8%) had residual disease; 276 (47.1%) had residual nodal involvement. In total, 152 (25.9%) patients developed BrM with a median follow-up of 38.0 months (IQR 19.1 - 70.9). The median time from BC diagnosis to development of BrM was 29.6 (16 - 50.5) months in the overall cohort; it was shortest for patients with triple negative BC 18.0 (15.0 - 31.0) months and longest for those with HR+/HER2- disease 49.1 (19.8 - 76.3) months. In a Cox proportional hazard regression, factors associated with a shorter time to BrM development included lack of pCR HR 4.17 (95% CI 2.04 - 8.33), p0.0001, inflammatory BC HR 3.81 (95% CI 2.63 - 5.52), p0.0001, residual nodal involvement HR 2.76, (95% CI 1.79 - 4.27), p0.0001), presence of lymphovascular invasion HR 2.38 (95% CI 1.51 - 3.75, p=0.0002 and HER2+ or triple negative BC HR 1.75 (95% CI 1.21 - 2.51, p= 0.002. In a MVA, residual nodal involvement, inflammatory BC, as well as HER2+ or triple negative subtype were independently associated with a shorter time to BrM development (Table). Conclusions: The median time to BrM development was 24 months among patients with triple negative or HER2+ BC. Given that patients with residual nodal involvement and those with inflammatory BC have a particularly high risk of developing BrM, the possible utility of BrM screening in select populations of patients receiving NAC for BC warrants further study. Citation Format: I. Fernandes, J. W. Zhu, I. Konjundzic, W. Tran, V. Moravan, K. J. Jerzak. Clinical and pathological variables associated with brain metastases (BrM) development among patients receiving neoadjuvant chemotherapy (NAC) for breast cancer (BC) abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PD13-03.
Fernandes et al. (Tue,) studied this question.