RESUMEN
PURPOSE: Risk factors for local recurrence (LR) following breast-conserving therapy (BCT) inform the need for local therapy. A Danish population-based cohort study identified residual disease on reexcision as an independent risk factor for LR but was limited by incomplete data on biologic subtype (Bodilsen et al. 2015 in Ann Surg Oncol 22: S476-S485). We sought to elaborate this risk in an independent cohort with clearly defined biologic subtypes. METHODS: The study population included patients with localized invasive breast cancer with complete biologic subtype data treated with BCT with one or zero reexcisions at one institution from 1998 to 2008. Cumulative incidence of LR was calculated using competing risk analysis, and associated risk factors were evaluated using Cox proportional hazards regression. RESULTS: A total 1073 consecutive patients were included with a median follow-up of 10 years. The 10-year LR rates were 2.4% [95% confidence interval (CI) 1.4-3.9%] without reexcision, 6.0% (95% CI 3.8-8.9%) with reexcision, and 8.2% (95% CI 4.1-14.0%) with any reexcised residual disease. On univariate regression, residual disease [hazard ratio (HR) = 1.50, p = 0.31] was not significantly associated with LR. Subtype other than luminal A/luminal-HER2 (luminal B HR = 2.29, p = 0.033; HER2/triple-negative HR = 2.85, p = 0.004), age (HR = 0.95 per year, p < 0.001), and nodal involvement (HR = 1.12 per involved node, p = 0.001) remained significant on multivariate regression. The impact of residual disease was confounded by its association (p < 0.001) with nodal involvement. CONCLUSIONS: Our findings suggest that LR is associated with younger age, nodal involvement, and biologic subtype but not with residual disease at reexcision. The study's power is limited by the low incidence of LR despite detailed clinical data and long-term follow-up. Further study is required.