S5-7: A Phase 2, Randomized, Open-Label, Study of Neratinib (HKI-272) vs Lapatinib Plus Capecitabine for 2nd/3rd-Line Treatment of HER2+ Locally Advanced or Metastatic Breast Cancer.

Abstract

Abstract Background: Neratinib (N), an irreversible pan-tyrosine kinase inhibitor (TKI), with activity against HER1, -2 and -4, has shown antitumor activity in patients (pts) with HER2+ breast cancer (BC). Lapatinib (L), a reversible HER1 and -2 TKI is approved in combination with capecitabine (C) for treatment of pts with HER2+ advanced or metastatic BC who had prior therapy. Materials and Methods: This phase 2, randomized, open-label study evaluated safety and efficacy of N 240 mg/day vs L 1,250 mg/day plus C 2,000 mg/m2/day (14 days/21 day cycle) in pts with HER2+ locally advanced or metastatic BC. Eligible pts had: ≤2 prior trastuzumab regimens, prior taxane treatment, and no prior anthracycline treatment with cumulative dose >400 mg/m2 doxorubicin, >800 mg/m2 epirubicin, or >equivalent dose of other anthracycline. Primary endpoint was progression-free survival (PFS; investigator-assessed); secondary endpoints included safety, overall survival (OS), objective response rate (ORR), and clinical benefit rate (CBR;% pts with complete response, partial response, or stable disease ≥24 wks). Tumor assessments were every 6 wks for the first 48 wks, then every 12 wks until progressive disease (PD; RECIST 1.0) or initiation of new anticancer therapy. Results: Overall, 117 pts were randomized to N and 116 to L plus C (LC). Mean age (SD; range) was 53.9 y (10.3; 28–79); 60% were White, 34% Asian, and 6% other. Prior treatments included: trastuzumab (229 pts: 168 metastatic or locally advanced, 51 adjuvant, 10 neoadjuvant), taxanes (230 pts), and anthracycline (156 pts). Median treatment duration (range) was 126.5 days (1-636) for N and 201 days (13-622) for LC. Median relative dose intensity (actual/expected exposure) for N was 100%. As of data cutoff, 84% had discontinued treatment; 65% from PD (N 63%, LC 67%), 9% for adverse events (AEs; N 7%, LC 11%). In the ITT cohort, for N and LC, respectively, median PFS (95% CI) was 4.5 mo (3.1−5.7) and 6.8 mo (5.9−8.2; P = 0.091; hazard ratio = 1.3 [95% CI, 1.0−1.8]); median OS (95% CI) was 19.4 mo (19.4−22.2; 41 deaths) and 19.0 mo (16.9-NA; 35 deaths; P = 0.180); ORR (95% CI) was 29% (21-38) and 41% (32-50; P = 0.067); CBR (95% CI) was 44% (35-54) and 64% (54-73; P = 0.003). Most common drug-related treatment-emergent AEs (TEAEs; any grade) were diarrhea (N 84%, LC 67%), nausea (34%, 38%), palmar-plantar erythrodysesthesia (PPE; 5%, 63%), and rash (18%, 34%); for grade ≥3, diarrhea (28%, 10%) and PPE (0, 14%). Dose reductions/discontinuations from diarrhea occurred in 13/3 pts on N, and 15/7 pts on LC. Study deaths for N and LC, respectively, included: 36 pts (31%) and 32 pts (28%) from PD; 5 pts (4%) and 3 pts (3%) from AEs unrelated to study drug. Discussion: In this setting of pts less heavily pre-treated than in the pivotal LC trial, single agent N demonstrated high anti-tumor activity (ORR 29%), confirming results from prior N phase 2 trials. N alone did not appear to be as effective as LC. No unexpected TEAEs were observed; N was well tolerated in pts with HER2+ locally advanced or metastatic BC; while diarrhea was more frequent on N than LC, it was manageable with antidiarrheals and did not lead to more treatment discontinuations. Citation Information: Cancer Res 2011;71(24 Suppl):Abstract nr S5-7.