Abstract PD5-5: Phase 2 study of single agent MLN8237 (alisertib), an investigational aurora A kinase (AAK) inhibitor, in patients (pts) with relapsed/refractory breast cancer (BrC)

Abstract

Abstract Background: AAK, a key mitotic regulator, is frequently amplified/overexpressed across a spectrum of tumors, including BrC. AAK overexpression is associated with poor prognosis. MLN8237 is an oral selective AAK inhibitor under evaluation in pts with advanced cancer as a single agent and in combination therapy. A phase 1/2 study (NCT01045421) evaluated MLN8237 in pts with different solid tumors; phase 2 data for the Brc cohort are presented. Methods: Females aged ≥18 y with relapsed/refractory BrC including HR+, HER2+ and triple-negative histological subtype, ECOG PS 0–1, measurable disease by RECIST v1.1 and ≤4 prior cytotoxic chemotherapy regimens (not including adjuvant, neo-adjuvant; no limitation on prior hormonal or HER2 targeted, immunological or biological agents) were enrolled. Symptomatic brain metastases were excluded (treated stable metastasis allowed). A Simon's optimal 2-stage design was used; 20 pts were initially enrolled, expansion proceeded if ≥2 objective responses were observed in these response-evaluable pts. Pts received MLN8237 50 mg BID for 7 days in 21-day cycles. Primary objective: overall response rate (ORR). Secondary objectives: safety, duration of response (DOR) and progression-free survival (PFS). An exploratory study was performed to assess clinical responses in relation to candidate biomarker dysregulation (mutation, amplification, and deletion) in banked tumor specimens. Results: As of April, 2013, 53 pts were enrolled: median age was 60 y (range 33–81), median of 4 cycles (range 1–21). 49 pts (92%) were response-evaluable (HR+, n = 26; HER2+, n = 9; triple negative, n = 14). 33% of pts received treatment for ≥6 months (HR+, n = 11 [69%]; HER2+, n = 3 [19%]; triple negative, n = 2 [12%]). ORR (all pts) was 18% and median PFS was 5.42 months. Efficacy data per subgroup are shown in the table. All 53 pts were included in the safety population; drug-related adverse events (AEs) were reported in 51 pts (96%), most frequent were neutropenia (55%), alopecia (49%) and diarrhea (45%). 38 pts (72%) had grade ≥3 drug-related AEs, including neutropenia (49%), leukopenia (21%) and febrile neutropenia (4%). G-CSF use was 32%. 2 pts (4%) discontinued due to AEs (sepsis [grade 4], n = 1; neutropenia [grade 3], n = 1); no on-study deaths were reported. Treatment is ongoing in 3 HR+ pts. Whole-exome sequencing of selected tumor samples was completed. Correlative analysis is ongoing to identify potential genetic markers/mutated pathways associated with clinical response. Preliminary results will be presented. Conclusions: MLN8237 appears to have a generally manageable toxicity profile and shows signs of single agent antitumor activity in pts with heavily pretreated (different molecular subgroups) of relapsed/refractory BrC supporting further evaluation of MLN8237 in this tumor type in different combination strategies. Efficacy HR+ (n = 26)HER2+ (n = 9)Triple-negative (n = 14)Best response,%ORR (PR)23227Stable disease653336Progressive disease124457Median DOR, months4.2-*-**Median PFS, months7.94.11.5*2 events, **1 event Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr PD5-5.