RESUMO
AIM: This pooled analysis aimed to evaluate locoregional recurrence (LRR) rates of breast cancer (BC) after neoadjuvant chemotherapy (NACT) and to identify independent LRR predictors. METHODS: 10,075 women with primary BC from nine neoadjuvant trials were included. The primary outcome was the cumulative incidence rate of LRR as the first event after NACT. Distant recurrence, secondary malignancy or death were defined as competing events. For identifying LRR predictors, surgery type, pathological complete response (pCR), BC subtypes and other potential risk factors were evaluated. RESULTS: Median followup was 67 months (range 0-215), overall LRR rate was 9.5%, 4.1% in pCR versus 9.5% in non-pCR patients. Younger age, clinically positive lymph nodes, G3 tumours, non-pCR and TNBC but not surgery type were independent LRR predictors in multivariate analysis. Among BC subtypes, 5-year cumulative LRR rates were associated with higher risk in non-pCR versus pCR patients, which was significant for HR+/HER2- (5.9% vs 3.9%; HR = 2.32 [95%CI 1.22-4.43]; p = 0.011); HR-/HER2+ (14.8% vs 3.1%; HR = 4.26 [94%CI 2.35-7.71]; p < 0.001) and TNBC (18.5% vs 4.2%; HR = 4.10 [95%CI 2.88-5.82]; p < 0.001) but not for HR+/HER2+ (8.1% vs 4.8%; HR = 1.56 [95%CI 0.85-2.85]; p = 0.150). Within non-pCR subgroup, LRR risk was higher for HR-/HER2+ and TNBC vs HR+/HER2- (HR = 2.05 [95%CI 1.54-2.73]; p < 0.001 and HR = 2.77 [95%CI 2.27-3.39]; p < 0.001, respectively). CONCLUSIONS: This pooled analysis demonstrated that young age, node-positive and G3 tumours, as well as TNBC, and non-pCR significantly increased the risk of LRR after NACT. Hence, there is a clear need to investigate better multimodality therapies in the post-neoadjuvant setting for high-risk patients.