Purpose: We conducted a meta-analysis of randomised trials evaluating pathological complete response (pCR) and surgical outcomes after neoadjuvant systemic therapy (NST) in patients with early breast cancer (EBC).
Patients and methods: The primary outcome was breast-conserving surgery (BCT) rate. Secondary outcomes were pCR rate and association to BCT. Meta-analyses were performed using random effects models that use inverse-variance weighting for each treatment arm based on evaluable patients. Point estimates are reported with 95% confidence interval (CI), and p < 0.05 was considered statistically significant.
Results: Thirty-six studies were identified (N = 12,311 patients). We selected for the analysis 16 of 36 studies reporting both pCR and BCT for at least one treatment arm. Arms per study ranged from one to six; 42 independent units were available to evaluate the association between pCR and BCT. BCT rate ranged 5-76% across arms with an average BCT of 57% (95% CI 52-62%). Significant heterogeneity was observed among the trials (Cochrane Q = 787, p < 0.001, I2 = 97%). In the meta-regression model, BCT rates were not significantly associated with year of first patient-in (p = 0.89), grade (p = 0.93) and hormone-receptor status (p = 0.39). Clinical N-stage (p = 0.01) and human epidermal growth factor receptor (HER2) status (p = 0.03) were significantly associated with BCT. pCR rate ranged 3-60% across studies. The average pCR across all study arms was 24% (95% CI 19-29%). No association was observed between pCR rate in a study arm and the resulting BCT rate in a univariate model (p = 0.34) nor after adjusting for HER2 and clinical nodal status (p = 0.82). In the subset of 14 multi-arm studies, no significant association was seen between the differences in pCR and BCT between treatment arms (p = 0.27).
Conclusions: pCR does not increase BCT in patients receiving NST for EBC.
Keywords: Breast cancer; Mastectomy; Neoadjuvant therapy; Quadrantectomy.
Copyright © 2018 Elsevier Ltd. All rights reserved.