Sophia Analysis By Chemotherapy (Ctx) Choice: A Phase Iii (P3) Study Of Margetuximab (M) Plus Ctx Versus Trastuzumab (T) Plus Ctx In Patients (Pts) With Pretreated Her2+Metastatic (Met) Breast Cancer (Mbc).

1040 Background: Despite advances, pretreated HER2+ MBC remains incurable with ongoing need for new therapies. Investigational M has similar HER2 binding and antiproliferative effects as T. Relative to T, M Fc engineering increases binding affinity for both variants of activating Fc receptor (FcR) CD16A and decreases affinity for inhibitory FcR CD32B, coordinately activating innate and adaptive immunity. In a Phase 3 (P3) trial, M prolonged PFS over T (Table). Second interim OS results from Sept 2019 also favor M (hazard ratio [HR], 0.89; 95% CI 0.69–1.13; nominal P=0.326). Methods: SOPHIA (NCT02492711), an open-label P3 trial, enrolled pts with HER2+ MBC after pertuzumab and 1–3 lines of prior treatment (Tx) for MBC. Randomization was 1:1 to M (15 mg/kg IV q3w + Ctx) or T (6 [8 for loading dose] mg/kg IV q3w + Ctx), stratified by met sites (≤2, >2), lines of Tx for met disease (≤2, >2), and Ctx choice, including capecitabine (Cap), eribulin (Eri), gemcitabine (Gem), or vinorelbine (Vin). Primary endpoints were central blinded PFS and OS, assessed sequentially using the stratified log-rank test. Results: Investigator chemotherapy choices and results by chemotherapy are shown in the table. Subjects receiving Eri and Gem had the lowest PFS hazards ratios (HRs), favoring M over T, although no statistical significance of individual chemotherapy subgroups was seen. There was variable toxicity among Ctx subgroups, and fewer subjects receiving Cap had Ctx related Grade 3 or higher (>=Gr 3) AEs. In this unblinded study, more subjects on M than T in all subgroups discontinued Ctx while continuing study antibody. Conclusions: In combination with chemotherapy in pretreated HER2+ MBC, M improved PFS over T. Safety was manageable in all Ctx subgroups. Differences among HRs for chemotherapy subgroups may be driven by selection bias and/or sensitivity differences. Clinical trial information: NCT02492711 . [Table: see text]