Abstract P2-16-05: BEVERLY2, a Phase II Study Evaluating Bevacizumab (BEV) Combined with Chemotherapy (CT) and Trastuzumab (H) as Neoadjuvant Therapy for HER2-Positive Inflammatory Breast Cancer (IBC): First Efficacy Results

Abstract

Abstract Background: The efficacy of first-line CT for HER2-negative metastatic breast cancer is significantly improved by combining CT with BEV, as shown in three randomized phase III trials. In HER2-positive breast cancer, the role of H is established but there is less information on the impact of BEV in this setting, particularly in IBC. Patients and methods: The primary objective of this single-arm, open-label, phase II study is to determine the rate of pathologic complete response (pCR; Sataloff classification based on central pathologic review) in patients (pts) receiving neoadjuvant CT, BEV, and H for IBC. Secondary endpoints include disease-free survival, recurrence-free interval, overall survival, safety (CTCAE v3), cardiac safety, and potential predictive markers (biomarkers, genomics, imaging). Pts with HER2-positive IBC (IHC 3+ or FISH/CISH+, T4d, any N) received 8 cycles of neoadjuvant therapy (FEC100 + BEV 15 mg/kg, d1 q21d x 4, followed by docetaxel [DOC] 100 mg/m2 + BEV 15 mg/kg + H 8≥6 mg/kg q21d x 4). Pts underwent surgery (mastectomy and axillary node dissection) 4 weeks after the last dose of neoadjuvant therapy, followed by radiotherapy. H was continued during surgery and radiotherapy and as adjuvant therapy. BEV was re-introduced as adjuvant therapy for 10 cycles, giving a total duration of BEV and H of 54 weeks. The study had a two-stage Simon design, with 24 pts planned in the first stage. Results: Among 24 pts reviewed centrally for pCR in the first stage, median age was 51 years (range 38-68), Scarff-Bloom-Richardson grade was II in 42% of pts and III in 54% (unknown in 4%), and 50% had hormone receptor-positive disease. All 8 cycles of neoadjuvant therapy were completed by 19 pts (79%). Of the remaining 5 pts, 1 discontinued BEV and DOC (cutaneous toxicity, cycle 8), 1 omitted cycle 1 of BEV (dental extraction), 1 omitted cycle 2 of BEV (hypertension), and 2 discontinued DOC but not BEV (nail toxicity, cycle 5; hand-foot syndrome, cycle 5). Grade ≥3 adverse events occurred in 15 pts (63%) and were considered related to BEV in 3 pts (13%). The safety profile was consistent with the known profiles of the component agents. Clinical responses were seen in 18/21 pts (86%) after cycle 4 and 19/20 (95%) after cycle 8. The pCR rate was 63% (95% CI 42-83) according to central review, as shown below (2 pts without evaluable histologic samples counted as non-responders), and 71% by local review. As the pCR rate in the first 24 pts exceeded the prespecified 40% pCR rate considered to prove efficacy of the regimen, recruitment continued to a total of 52 pts. Local pathologic review indicated a pCR in 33/52 pts (63%; 95% CI 49-78; 2 pts without histologic samples and 4 who had no surgery counted as non-responders). Conclusions: This neoadjuvant regimen of BEV, H, and CT demonstrated high activity in IBC. The safety profile of treatment was acceptable and predictable. Results from central review of all 52 pts will be presented. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P2-16-05.