Perspective: Designing T-Cell Engagers With Better Therapeutic Windows

Abstract

This perspective highlights the history and challenges of developing CD3-based bispecific T-cell engagers (TCEs) as cancer therapeutics as well as considerations and potential strategies for designing the next generation TCE molecules. The goal of this article is to raise awareness of natural T-cell biology and how to best harness the tumor cell killing capacity of cytotoxic T-cells with TCEs. In light of 30 years of concerted efforts to advance TCEs in early clinical development, many of the first-generation bispecific antibodies have exhibited lackluster safety, efficacy, and manufacturability profiles. As of January 2020, blinatumomab remains the only approved TCE. Many of the current set-backs in early clinical trials implicate the high-affinity CD3 binding domains employed and the respective bispecific platforms as potential culprits. The underlying conviction of the authors is that by taking corrective measures, TCEs can transform cancer therapy. Through openness, transparency, and much needed feedback from ongoing clinical studies, the field can continuously improve the design and effectiveness of next generation T-cell redirecting therapeutics.

Keywords: CD3 redirection; T-cell engager; bispecific; cancer; cytokine release; therapeutic window.

Figure 1

Common structures of TCE proteins. This figure illustrates common molecular formats used to create TCE proteins. (A) knob-into-hole format for Fc and light-chain heterodimerization. (B) knob-into-hole format using a common light chain. (C) knob-into-hole triple-chain format, HC:LC Fab paired with scFv (Xencor) and (D) the 2+1 format including a second Fab (Xencor). (E) knob-into-hole triple-chain format, HC:LC Fab paired with heavy-chain only binding domain (Teneobio). (F) Fab arm exchange, DuoBody® (Genmab). (G) knob-into-hole Cross-MAb 1+1 format (Roche) and (H) knob into hole CrossMAb 2+1 format (Roche). (I) tetravalent scfv Fc fusion and (J) tetravalent HC:LC and scfv fusion (NV Cheung, MSKCC). (K) TandAb diabody (Affimed). (L) tandem scFv, first generation BiTE®format (Amgen).

Figure 2

Design considerations of a TCE molecule. The three binding domains of a typical TCE molecule are the CD3 targeting arm, the tumor targeting arm, and the Fc domain. When designing the CD3 targeting arm, important considerations are: the binding epitope, the binding affinity, and species cross-reactivity. For the tumor targeting arm, important considerations are: the epitope and affinity of tumor antigen binding, competition with soluble target or a ligand of the target, and on-target, off-tumor specificity. Considerations for the Fc domain are: maintaining FcRn binding for half-life and silencing variants to eliminate FcR binding, complement activity, and non-specific CD3 clustering and T-cell activation. For the holistic design of a TCE molecule, these design considerations must be made in the context of the interdependencies of the different domains.