Abstract A040: Characterization of a humanized monoclonal antibody targeting cancer-expressed EGFR

Abstract

The epidermal growth factor receptor (EGFR), an established oncology target, is a transmembrane protein highly expressed on the surface of malignant tumors such as glioblastoma, lung, liver, bladder and breast cancers. In addition to overexpression, EGFR mutants are frequently associated with malignancy.  For example, EGFR variant III (EGFRvIII), deletion of exons 2-7, is only expressed on cancer cells and is associated with poor prognosis and chemotherapy resistance. To target cancer-expressed EGFR, including EGFRvIII, we generated the 40H3 monoclonal antibody which is reactive with the 287-302 loop that is fully exposed on the extracellular domain of EGFRvIII. The 40H3 antibody also showed strong reactivity for cells expressing gene-amplified EGFR but little or no reactivity for cells expressing normal levels of wild type EGFR. When conjugated with toxic payloads (either tesirine or deruxtecan), the resulting antibody-drug conjugates (ADCs) exhibited antitumor activity in several xenograft models. To improve its utility as a clinical candidate, we humanized the variable portions of the heavy and light chains of 40H3, without changing the original CDRs. The variable regions of 40H3, heavy and light chains, were compared to the closest human immunoglobulin families and then framework residues were altered to produce candidate humanized antibodies. The humanized variable chains were then fused with a human IgG1 heavy chain and a kappa light chain to generate full-sized humanized antibodies. A total of 9 humanized antibodies were generated by pairing 3 variable humanized heavy chains (VH1, VH2, VH3) with 3 variable humanized light chains (VL1, VL2, VL3). To evaluate the binding activity of humanized antibodies, we used either immobilized EGFRvIII (by ELISA or Octet) or various cancer cells expressing either EGFR or EGFRvIII (by flow cytometry).   All humanized antibodies retained binding activity to immobilized EGFRvIII but bound with different apparent affinities to cells overexpressing EGFR or cells transfected with EGFRvIII. The ‘A10’ antibody was the best candidate. Mutations introduced in A10 preserved the binding affinity of the 40H3 antibody but exhibited an increased affinity compared to the corresponding chimeric antibody constructed with the original mouse VH and VL. The other eight antibodies retained EGFRvIII binding activity, albeit with a lower cell binding affinity. Specifically, A10 showed improved binding to cells with EGFR gene amplification including two TNBC lines, MDA-MB-468 and BT-20, epithelial cancer cells, A431, head and neck squamous carcinoma, UMSCC17B, as well to cells expressing EGFRvIII, F98npEGFRvIII. No binding of A10 was detected to cell lines expressing wild type EGFR, WI-38 or U87MG. In summary, humanized candidates of the 40H3 monoclonal antibody retained EGFR binding. Of 9 antibodies, A10 exhibited the best cell binding activity elevating it to our top clinical candidate for the production of ADCs or similar antibody-based therapeutics.

Citation Format: Tamara G Fernandes Costa, Robert Sarnovsky, Maria Teresa Bilotta, Antonella Antignani, David FitzGerald. Characterization of a humanized monoclonal antibody targeting cancer-expressed EGFR [abstract]. In: Proceedings of the AACR-NCI-EORTC Virtual International Conference on Molecular Targets and Cancer Therapeutics; 2023 Oct 11-15; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2023;22(12 Suppl):Abstract nr A040.