A phase III, randomized trial of docetaxel plus carboplatin (TP) versus epirubicin plus cyclophosphamide followed by docetaxel (EC-T) as adjuvant treatment for triple-negative, early-stage breast cancer in Chinese patients.
TPS1135 Background: Triple-negative [estrogen receptor (ER)-/progesterone receptor (PR)-/HER2-] breast cancer (TNBC) accounts for about 15% of all breast cancers and is associated with very poor prognosis. A combination of anthracycline and taxanes represents the most commonly used adjuvant chemotherapy for TNBC patients. BRCA1, a protein responsible for repair upon DNA damage, is often dysfunctional in TNBC. As a result, TNBC is often sensitive to DNA-damaging agents (e.g., cisplatin and carboplatin). A phase II study of docetaxel plus carboplatin as neoadjuvant treatment for TNBC achieved promising response rate (Chang, CANCER, 2010). Encouraged by this important finding, we are currently conducting a phase III trial to examine docetaxel plus carboplatin as adjuvant chemotherapy in patients with early-stage TNBC (registration number at www.ClinicalTrials.gov: NCT 01150513). Methods: All participants received radical mastectomy or breast-conserving surgery for pT1-3N0-2 breast cancer. All cancers were negative for ER, PR, HER2/Neu. All participants had adequate organ function and performance status. The age ranged from 18 to 70 years. Patients randomly received a TP regimen (docetaxel 100 mg/m2 plus carboplatin AUC=6 on day 1, 21 days a cycle for 6 cycles) or a EC-T regimen (epirubicin 90 mg/m2 plus cyclophosphamide 600 mg/m2 on d1, 21 days a cycle for 4 cycles; followed by docetaxel 100 mg/m2 on d1, 21 days a cycle for 4 cycles). Adjuvant radiotherapy was permitted in both arms. The primary endpoint is disease-free survival (DFS). Secondary endpoints included adverse event and quality of life (QoL). The study planned to recruit 500 subjects over a period of 5 years, with 5-year follow-up (once every 6 months). Target hazard ratio is 0.86 (5 year DFS of 74.0% vs. 71.0%). The estimated power is 0.80; 1-sided type 1 error was set at 0.05). The study was activated in June 2010 with enrollment of 110 subjects.