Abstract Purpose Pathological complete response (pCR) after neoadjuvant chemotherapy (NAC) improves outcomes in breast cancer (BC); however it may not prevent brain metastases. We evaluated central nervous system (CNS) recurrence patterns in early-stage BC following NAC according to pCR. Methods All consecutive stage I–III BC treated with NAC and surgery at a single center between 2007 and 2018 were analyzed. Endpoints included the impact of pCR on CNS recurrence across BC subtypes—hormone receptor-positive(HR)/HER2-negative, HER2-positive and triple-negative (TNBC), CNS recurrence patterns and overall survival (OS) after CNS relapse. Statistical comparisons included Fisher’s Exact test, Chi-square, Kaplan–Meier, and regression analyses. Results Among 1147 patients, 537 had HR-positive/HER2-negative, 301 HER2-positive, 309 TNBC, mostly stage III, treated with anthracycline + taxane NAC, and trastuzumab if HER2-positive. Three hundred sixty-five achieved pCR (59/537 HR-positive/HER2-negative, 158/301 HER2-positive, 148/309 TNBC). CNS recurrence occurred in 72 (6.2%) patients, with no difference between pCR and non-pCR (4.7 vs. 7.0%, p = 0.15). Across subtypes, there was no difference for HR-positive/HER2-negative (3.4 vs. 4%, p = 1.0), TNBC (5.4 vs. 9.3%, p = 0.2), however there was a reduction in HER2-positive (4.4 vs. 14.7%, p = 0.003) after pCR. Isolated CNS relapse was the predominant pattern of CNS metastasis (82.4%) in pCR, particularly in HER2-positive. Median OS after CNS relapse was 12 months. Multivariate analysis identified HER2-positive, TNBC, and cN2–3 status as independent predictors of CNS recurrence. Conclusion Although pCR was not associated with a lower overall risk of CNS recurrence, it predicted a reduced risk in HER2-positive disease. Isolated CNS relapse predominated, suggesting a sanctuary effect.
Leite et al. (2026) studied this question.