Abstract Background: NRG1 fusions are oncogenic drivers of various cancers including pancreatic and lung adenocarcinomas. NRG1 fusion proteins bind to HER3, leading to HER2/HER3 heterodimerization, increased downstream signaling, and tumor growth. MCLA-128 is a bispecific antibody directed against HER2 and HER3 that docks on HER2 and blocks ligand binding to HER3, thereby preventing downstream signaling. In contrast to tyrosine kinase inhibitors or anti-HER3 monoclonal antibodies, MCLA-128 was shown in vitro and in vivo to potently inhibit ligand-driven tumor growth at high NRG1 levels present in NRG1 fusion-positive cancers. MCLA-128 offers a novel therapeutic paradigm for NRG1 fusion-positive cancers. MCLA-128 has a very well tolerated safety profile with <5% of patients reporting grade 3-4 suspected related AEs, and a notable lack of cardiotoxicity and severe gastrointestinal or skin toxicity. Methods: Cell line/xenograft models with NRG1 fusions (MDA-MB-175, OV5383, OV10-0050) were treated with MCLA-128. Patients with cancers harboring NRG1 fusions were identified using prospective molecular profiling by DNA/RNA-based next-generation sequencing (NGS). Patients with NRG1 fusion-positive tumors were treated with MCLA-128 (750 mg intravenously, every 2 weeks) on FDA-approved single-patient protocols. Results: Treatment with MCLA-128 inhibited proliferation of NRG1-fusion positive cell lines in vitro and resulted in rapid tumor shrinkage in NRG1 fusion-positive xenograft models in vivo. NGS identified 29 patients with NRG1 fusions across 8 tumor types (pancreas, lung, breast, sarcoma, prostate, gallbladder, unknown primary, and DLBCL). Of these 29 patients, 3 with chemotherapy-resistant metastatic cancer were treated with MCLA-128 and experienced dramatic clinical and radiographic responses. A 52-year-old man with ATP1B1-NRG1 fusion-positive pancreatic ductal adenocarcinoma (PDAC) with liver metastases, worsening fatigue, and weight loss, achieved rapid clinical and pharmacodynamic responses (CA19-9 decrease from 262 to 56). Imaging at 8 weeks demonstrated a partial response (-44%) by RECIST v1.1 and a complete response by PERCIST. A 34-year-old man with ATP1B1-NRG1 fusion-positive PDAC and longstanding tumor-associated abdominal pain also achieved rapid resolution of his pain, and normalization of CA 19-9 (418 to 11) upon treatment with MCLA-128. Imaging at 6 weeks showed tumor reduction (-22%) and that the liver metastases were non-FDG avid. A third patient with CD74-NRG1 fusion-positive non-small cell lung cancer (NSCLC) metastatic to the brain was started on MCLA-128. Despite progression on 6 prior lines of systemic therapy including afatinib, he rapidly responded to MCLA-128 with scans showing a partial response (-33%) by RECIST v1.1 at 8 weeks and tumor shrinkage in the brain. All patients remain on therapy (6+ months into treatment for the PDAC patients, 2+ months for the NSCLC patient) with no substantial toxicity. Conclusions: MCLA-128 leads to clinical responses in patients with NRG1 fusion-positive cancers through inhibition of ligand-driven activation of the HER3 pathway. A global, multicenter phase 2 basket trial for NRG1 fusion-positive cancers is now accruing patients. Citation Format: Alison M Schram, Eileen M O’Reilly, Romel Somwar, Ryma Benayed, Sara Shameem, Thrusha Chauhan, Jean Torrisi, Jim Ford, David Maussang, Ernesto Wasserman, Marc Ladanyi, David M Hyman, L. Andres Sirulnik, Alexander Drilon. Clinical proof of concept for MCLA-128, a bispecific HER2/3 antibody therapy, in NRG1 fusion-positive cancers [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr PR02. doi:10.1158/1535-7163.TARG-19-PR02