Engineering Antibody Immune Agonists

Monoclonal antibodies (mAbs) have been the most successful category of biologics in terms of revenue, volume produced, and patient benefit. As of mid-2022 the Umabs Antibody Therapies Database listed 162 mAbs that were approved in at least one regulatory jurisdiction, including 122 in the U.S., 114 in Europe, 82 in Japan, and 73 in China. This unprecedented success propelled global sales of mAbs to $186 billion in 2021, with that number projected to more than double, to $445 billion, by 2028.

Among the many variations on the mAb theme are whole (canonical) antibodies, antibody fragments (e.g., the antigen-binding Fab, F(ab’)2, Fab’, scFv, di-scFv dimeric), plus sdAb single-domain, BsAb bispecific, 3funct trifunctional, and BiTE bi-specific T-cell engager.

Despite attempts to commercialize nonstandard mAbs and fragments, canonical antibodies still make up the majority of approved mAbs, 115, followed by 14 antibody-drug conjugates, 7 bispecific antibodies, 8 antibody fragments, 3 radiolabeled antibodies, 1 antibody-conjugate immunotoxin, 2 immunoconjugates, and 12 Fc-Fusion proteins. These mAb variants operate on more than 90, mostly cancer-related, biological targets.

Slicing & dicing

There are at least two significant ways to slice and dice mAbs for drug development purposes: by chemical class, as in the preceding paragraphs, or by molecular target (or target type). The relatively low approval numbers for noncanonical antibodies suggests that mechanism and choice of disease modifier is a more fruitful approach than searching for molecular variations on a theme that may be (but are often not) active.

Most therapeutic mAbs are antagonists—that is, they block specific disease-enabling pathways. But we know that immunity involves not just inhibitory processes but stimulatory (agonistic) interactions as well. Investigators are now looking into this second immune-modulating approach through a mechanism-based mAb category known as agonist antibodies. Where conventional antibodies block undesirable pathways, agonist mAbs promote or enable disease-fighting immunity.

The search for new targets, pathways, and mechanisms based on agonists is an issue of necessity. Drug discovery has been overly engaged with tried-and-true, well-characterized targets, the “low-hanging fruit” on which 99% of our current pharmacopoeia is based. It is time, according to at least one author, to take a closer look at the "high hanging fruit”, by “developing antibodies with novel mechanisms of action.” To navigate this difficult discovery path will require, moreover, “overcoming existing obstacles to delivery” to relatively accessible tissues (e.g., the brain) and by searching for “novel mechanistic approaches to agonism and transport that are currently underserved by mAbs.” Prominent in the agonist-mAb approach will be the engineering of antibody–drug conjugates and bispecific antibodies.

Clinicaltrials.gov lists 158 agonist mAbs currently in clinical development, the vast majority for cancer indications.

Relatively small but growing number

Agonist mAbs are not a new idea. The first papers on this topic in a drug discovery setting began appearing in the 1980s and their number has steadily risen. Close to 6500 papers have appeared on agonist antibodies during the past decade.

Their popularity, however, has been tempered by low efficacy, off-target effects (leading to toxicity), and delivery to appropriate tissues. This has led to efforts at engineering these molecules to accentuate their positive qualities and mitigate their shortcomings. Unfortunately, molecular design advances that have led to successful antagonist-mAb products have not directly translated to the discovery of agonists.

The first two agonist mAbs, BMS-663513, a fully human anti-CD137 agonist mAb, and Utomilumab, a 4-1BB/CD137 mAb agonist, failed due to liver toxicity and lack of efficacy, respectively.

Incorporating additional, co-stimulatory 4-1BB domains in second-generation chimeric antigen receptor T-cell (CAR-T) therapies led to U.S. Food and Drug Administration (FDA) approval of two agonist mAbs: tisgenlecleucel, for acute lymphoblastic leukemia, in 2020, and ciltacabtagene autoleucel, for multiple myeloma, in 2022. 4-1BB, more commonly known as CD137, a transmembrane protein expressed on the surfaces of leukocytes and non-immune cells, belongs to the tumor necrosis factor (TNF) receptor family and is the object of considerable investigation as a player in co-stimulatory immune checkpoint events.

Discovered in 1989, 4–1BB is now recognized as a significant costimulatory receptor on several immune cells, including T cells. As early as 1997 the receptor was shown to induce anti-tumor T cell activation leading to the eradication of established tumors in mice. It was subsequently shown to improve the activity of CD8 T cells through cytokine secretion, cytotoxicity, long-lived memory formation, and other cancer-fighting pathways.

Variations on a theme

Current thinking in the design of agonist mAbs recapitulates the notion in chemotherapy generally, that two (or three) hits are better than one. This has led to interest in multispecific antibodies to effect functions of both agonists and antagonists. The idea can be extended, moreover, to tri- or even higher-order specific agonist mAbs that inhibit cancer-promoting pathways while simultaneously encouraging cancer-fighting immune responses.

Replicating the activity of two or more antibodies provides an orthogonal approach to cancer treatment by attacking the condition from several directions and thus, at least in theory, creating barriers to tumor escape or T cell exhaustion. It has also been argued that such treatments could lower healthcare costs by treating with one molecule instead of multiple mAbs.

The building blocks of trispecific mAbs may be any of the formats mentioned earlier, including fragments connected by short polypeptide linkers. These components may further be of the IgG or non-IgG type, which lack an Fc region. One objective of multi-specific mAbs is the ability to bind simultaneously to cancer cells and effector cells lacking affinity for those tumors, and perhaps to stimulate other tumor-fighting pathways through a third specificity.

The principal indication for trispecific agonist approaches is cancer, with its well-known cellular, immune system, and antigenic players. In bispecific format, one component binds to T cells or natural killer cells, while the other engages tumor-associated antigens. These molecules are known as immune cell engagers. Trispecifics may bind to two or more antigens on the tumor cell or, alternately, one receptor on the target and two costimulatory (and/or checkpoint) receptors on T cells.

Engineering also opens the way to improved safety or pharmacokinetics. Many proteins, particularly smaller antibody fragments, have extremely short circulating half-lives, as developers of bispecific agonist mAbs have discovered.

Fortunately, approaches to extending the circulating half-life of ordinary mAbs also work with multi-specific agonist mAbs. One such method, incorporating an albumin-binding domain to a multispecific agonist antibody, is widely used to extend the half-life of small biotherapeutics as well as proteins. Because of its size, albumin has a serum half-life of three weeks. The main advantages are more even drug concentrations and less frequent dosing. Many trispecific agonist mAbs currently under clinical investigation employ this strategy.