The Handle That Grabs a T Cell Anti-CD3 Antibodies for Bispecific Cancer Therapy

This application covers monoclonal antibodies and their antigen binding fragments, variants, multimeric versions and bispecific forms that specifically bind CD3 epsilon, together with methods of making and using them across therapeutic, diagnostic and prophylactic indications. The applicant is NovImmune SA, the Swiss antibody company now trading as Light Chain Bioscience.

The Problem

CD3) is a cell surface complex on T cells, sitting alongside the T cell receptor and required for activating both CD8 positive and CD4 positive T lymphocytes. It is built from three related chain types: one gamma chain, one delta chain and two epsilon chains, pairing into epsilon-gamma and epsilon-delta heterodimers which combine with the T cell receptor and a signal transducing zeta chain homodimer to form the complete receptor complex.

The reason CD3 matters commercially is a class of drugs called T cell engagers. These are bispecific antibodies with two different arms: one binds CD3 on a T cell, the other binds an antigen found on tumour cells. The molecule physically tethers the two together, forming an immunological synapse, activating the T cell locally, and prompting it to destroy the tumour cell with perforin and granzyme. Crucially this works independently of MHC presentation, so tumours that hide by downregulating antigen presentation remain vulnerable.

Blinatumomab, marketed as Blincyto for B-cell acute lymphoblastic leukaemia, proved the mechanism clinically. It also exposed the difficulty. The CD3 binding arm is what makes these drugs work and what makes them dangerous: bind CD3 too strongly or too indiscriminately and T cells activate systemically, producing cytokine release syndrome, which can be severe. Every T cell engager programme is therefore constrained by the properties of its CD3 arm, and the field has needed a wider range of them.

What This Invention Does

The disclosure provides monoclonal antibodies that specifically bind CD3 epsilon, along with fragments, variants, multimeric versions and bispecific formats built from them, and the methods of making and using them.

Targeting the epsilon chain in particular is deliberate. Epsilon is present in both heterodimers of the complex, appearing twice where gamma and delta appear once, and it is the chain that most clinical CD3 binders engage. What differentiates one anti-CD3 antibody from another is the character of the binding: affinity, epitope, and how those translate into T cell activation strength. A weaker or differently positioned binder can retain tumour killing while producing less systemic cytokine release, which is why companies pursue their own CD3 arms rather than licensing a single one.

Covering fragments, variants and multimeric versions rather than just the parent antibody reflects how these molecules are actually deployed. A CD3 arm is a component, recombined into whatever bispecific architecture a given programme uses, and the format affects both potency and half life.

The sequence listing referenced in the application was originally created in March 2019 and resubmitted in the current standard format in March 2026, which locates the underlying invention firmly in the period when T cell engagers were expanding rapidly.

Key Features

• CD3 epsilon specificity. The antibodies bind the epsilon chain, present twice in the CD3 complex and the standard clinical target.
• Multiple formats covered. Claims extend to antigen binding fragments, variants, multimeric versions and bispecifics.
• Bispecific application. The antibodies are intended for use as the T cell recruiting arm of tumour targeting bispecific molecules.
• MHC independent killing. The mechanism activates T cell cytotoxicity without requiring MHC mediated antigen presentation.
• Manufacturing methods. Methods of making the antibodies are claimed alongside their use.
• Broad indication scope. Therapeutic, diagnostic and prophylactic uses are all within scope.

Who Is Behind It

NovImmune SA is a Geneva based antibody company, since renamed Light Chain Bioscience, known for its kappa lambda body platform for producing bispecific antibodies in a fully human format without the engineering artefacts that complicate manufacturing in other approaches. The named inventors are Nicolas Fischer, Ulla Ravin, Giovanni Magistrelli and Franck Gueneau. Nicolas Fischer has led research at the company for many years.

The application is a divisional of Australian application 2019235523.

Why It Matters

T cell engagers have moved from a single approved product to one of the more crowded areas in oncology drug development, with approvals across multiple myeloma, lymphoma, leukaemia and small cell lung cancer, and dozens of candidates in trials. Almost all of them need a CD3 binding arm, and the properties of that arm shape the therapeutic window of the whole molecule.

That has made anti-CD3 antibodies a strategically valuable asset class in themselves. A company with a good CD3 arm has a component that fits into many programmes, its own and potentially those of partners. Patent coverage over specific antibodies and their variants is what makes that position defensible.

The clinical problem being worked on is more specific than it might sound. Early T cell engagers required careful step-up dosing and hospital monitoring to manage cytokine release, which limits where they can be given and to whom. Newer CD3 arms designed for gentler activation aim at drugs that can be administered in an outpatient setting. That is not a small distinction for patients who would otherwise spend the first weeks of treatment under observation.

Related Concepts

• Bispecific monoclonal antibody – the drug format these antibodies are built into.
• CD3) – the T cell surface complex being targeted.
• T cell – the immune cell the therapy recruits against tumours.
• Blinatumomab – the first approved T cell engager, cited in the background.
• Cytokine release syndrome – the toxicity that CD3 arm design tries to manage.
• Cancer immunotherapy – the treatment field this technology belongs to.

---

AU 2026202078 was published in the Australian Official Journal of Patents on 9 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.