The Anti-HER3 mAb Seribantumab Effectively Inhibits Growth of Patient-Derived and Isogenic Cell Line and Xenograft Models with Oncogenic NRG1 Fusions

Abstract

Purpose: Oncogenic fusions involving the neuregulin 1 (NRG1) gene are found in approximately 0.2% of cancers of diverse histologies. The resulting chimeric NRG1 proteins bind predominantly to HER3, leading to HER3-HER2 dimerization and activation of downstream growth and survival pathways. HER3 is, therefore, a rational target for therapy in NRG1 fusion–driven cancers. Experimental Design: We developed novel patient-derived and isogenic models of NRG1-rearranged cancers and examined the effect of the anti-HER3 antibody, seribantumab, on growth and activation of signaling networks in vitro and in vivo. Results: Seribantumab inhibited NRG1-stimulated growth of MCF-7 cells and growth of patient-derived breast (MDA-MB-175-VII, DOC4-NRG1 fusion) and lung (LUAD-0061AS3, SLC3A2-NRG1 fusion) cancer cells harboring NRG1 fusions or NRG1 amplification (HCC-95). In addition, seribantumab inhibited growth of isogenic HBEC cells expressing a CD74-NRG1 fusion (HBECp53-CD74-NRG1) and induced apoptosis in MDA-MB-175-VII and LUAD-0061AS3 cells. Induction of proapoptotic proteins and reduced expression of the cell-cycle regulator, cyclin D1, were observed in seribantumab-treated cells. Treatment of MDA-MB-175-VII, LUAD-0061AS3, and HBECp53-CD74-NRG1 cells with seribantumab reduced phosphorylation of EGFR, HER2, HER3, HER4, and known downstream signaling molecules, such as AKT and ERK1/2. Significantly, administration of seribantumab to mice bearing LUAD-0061AS3 patient-derived xenograft (PDX) and OV-10-0050 (ovarian cancer with CLU-NRG1 fusion) PDX tumors induced regression of tumors by 50%–100%. Afatinib was much less effective at blocking tumor growth. Conclusions: Seribantumab treatment blocked activation of the four ERBB family members and of downstream signaling, leading to inhibition of NRG1 fusion–dependent tumorigenesis in vitro and in vivo in breast, lung, and ovarian patient-derived cancer models.