Objective: This preoperative study aimed to evaluate the efficacy of 4 cycles of TX compared with 4 cycles of T, following 4 cycles of FEC in operable breast cancer patients. Methods: Women with operable breast cancer (T1C-3N0M0/T1-3N1M0) were randomly assigned to receive either T (75 mg/m 2 , q3) plus X (825 mg/m 2 bid, days 1 to 14) or T alone (75 mg/m 2 , q3) after completion of FEC (5-fluorouracil 500 mg/m 2 , epirubicin 100 mg/m 2 , cyclophosphamide 500 mg/m 2 , q3). Patients who had disease progression on FEC were excluded from randomization. The primary endpoint was pathological complete response (pCR). Predictive factor analysis was conducted using clinicopathological markers such as hormone receptors, Ki67 labeling index (Ki67LI) and thymidine phosphorylase (TP). Results: From November 2005 to October 2009, 504 patients were enrolled and 239 and 238 patients were assigned to TX and T group, respectively (median age 49 years, median tumor size 3.5 cm and node positive rate 56%). Treatment discontinuation was observed in 22% of TX group and 5% in T group (p According to the analysis for evaluating predictive values of biomarkers, a multivariate logistic regression analysis showed that HER2 (+/−: odds ratio 4.107, p Conclusions: Addition of X to T showed no superiority to T alone following FEC in neoadjuvant treatment of these patients with operable breast cancer. Treatment discontinuation rate was significantly higher in TX than T group, however the pCR rate in patients in TX group who required treatment discontinuation or dose-reduction was similar to that in patients who completed as scheduled, which was different from T group. Determination of pre-/ post-treatment Ki67LI looks useful for predicting pCR and DFS. Citation Information: Cancer Res 2012;72(24 Suppl):Abstract nr P1-14-02.