Abstract S3-1: Lapatinib vs Trastuzumab in Combination with Neoadjuvant Anthracycline-Taxane-Based Chemotherapy: Primary Efficacy Endpoint Analysis of the GEPARQUINTO STUDY (GBG 44)

Abstract

Abstract Background: The tyrosine-kinase inhibitor lapatinib (L) has shown to improve efficacy of cytotoxic and endocrine treatment in HER2-positive metastatic breast cancer (BC). Improved pathological complete response (pCR) rates were demonstrated by adding trastuzumab (T) to neoadjuvant chemotherapy. However, so far no head-to-head comparison of the two anti-HER2-agents is available. One primary aim of the GeparQuinto study was to improve the pCR rate by adding L instead of T to anthracycline-taxane-based neoadjuvant chemotherapy. We previously reported interim safety data of this study showing more diarrhea, skin changes, and hot flushes, but no cardiac events with L compared to T (von Minckwitz G et al, Ann Oncol 2010 in press). Patients and Methods: Patients (P) with untreated HER2-positive BC were eligible if they had cT3/4a-d; or estrogen (ER) and progesterone (PgR) receptor-negative; or ER/PgR-positive tumors with clinically N+ (for cT2) or pNSLN+ (for cT1) disease, and no increased cardiac risks. P were randomized to receive 4 cycles epirubicin/cyclophosphamide (EC) (90/600 mg/m2) q3w followed by 4 cycles docetaxel (D) (100mg/m2) given in combination with either T 6 (loading dose 8) mg/kg every 3 weeks or L 1000-1250 mg/d throughout all cycles. pCR was defined as no invasive or non-invasive tumor residuals in breast and nodes. We assumed a pCR rate of 26% with ECT-DT (based on GeparQuattro) and expected a pCR of 37% for ECL-DL (odds ratio 1.67). A two-sided Pearson's Chi2 with α=0.05 and β=0.20 calculated a sample size of 613 P. Results: Between May ‘07 and June ‘10 597 P were randomized to ECT-DT (N=299) and ECL-DL (N=298). Median tumor size was 40/40 [T/L] mm (clinically) and 28/29 mm (sonographically); 4.7%/4.3% had T4a-c, 14.8%/14.2% T4d, 2.9%/1.8% bilateral, 17.0%/17.7% multifocal, and 9.0%/12.1% multicentric disease, 96.7%/97.9% had non-lobular, 45.6%/48.9% grade 3, 70.0%/67.7% node-positive, and 56.5%/56.0% ER and PgR-negative disease. Baseline characteristics were well balanced between the treatment arms. The last randomized P will have surgery early Dec'10. Final results on histological response and surgical outcome will be reported. Conclusion: The GeparQuinto trial will provide for the first time randomized phase III efficacy data on the comparison of L and H in combination to chemotherapy for patients with early breast cancer. As pCR has been confirmed as being a surrogate marker for long-term outcome after T treatment, this result will give insight on the overall efficacy of L in patients with early breast cancer. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr S3-1.