Marianne Guenot Ph.D. Contributor GEN

Plotting Molecular Pincer Movements, Denying Cancer Room to Maneuver

In December 2014, for the first time, a bispecific antibody, blinatumomab (Blincyto©), was approved for therapeutic use. Since then, over 120 bispecific molecules have entered the clinical pipeline. More bispecific drugs are expected to hit the market within five to six years, and the global market for 2024 is estimated to hit a staggering $5.8 billion annually, according to a Research and Markets report (“Bispecific Antibody Therapeutics Market, 2014–2024”).

With such a market opening up, biotech companies are investing heavily in bispecific approaches. This is what encouraged heavy hitters and up-and-comers of the field to convene at two recent events, the 16th annual PEPtalk, in San Diego, and the 7th annual World Bispecific Antibody Summit, in Boston.

Companies such as Roche, Merus, Ablynx, F-star Biotechnology, and Biomunex Pharmaceuticals discussed the variety of creative ways they are endowing existing antibody structures with extra specificity. By snipping offshoots from and pasting bits and pieces to their molecular creations, these companies hope to open exciting new opportunities for cancer treatment.

Bispecific antibodies are products of genetic engineering that allow one antibody-like molecule to bind several different antigens at once. Amgen’s blinatumomab, for instance, is a molecule composed of the single-chain variable region (scFv) of one antibody targeting the T-cell activating molecule CD3, linked to another scFv, which binds to the B-cell antigen CD19. In the context of acute lymphoblastic leukemia (ALL), this bispecific antibody brings the T cell’s potent cytotoxicity close to the malignant B cells, allowing them to be specifically targeted. This approach has shown significant results as a second-line treatment to ALL.

Novelty in the Constant Region

“There’s now a large heterogeneity of products out there in the market” said Neil Brewis, Ph.D., the chief scientific officer of F-star Biotechnology. “Each of those have their various strengths and weaknesses, but some of the rules you need to be successful are becoming clearer.”

For Dr. Brewis, bispecific antibodies should bring clinical benefits beyond those brought by the injection of two monospecific antibodies. “They potentially bring demonstrable innovation into the marketplace, and they should address a lot of clinical unmet needs,” he added

The industry is now also looking for their products to be highly developable. “We want them to have good behavior when we dose them in humans, but they should also be developable in terms of manufacture,” explained Dr. Brewis. “I think it’s important that bispecific antibodies maintain that direction of travel to keep reducing costs of goods to make medicines more affordable in general.”

In F-star’s product, mAb2 ™, the bispecificity is generated by introducing a new binding site to the Fc region of the antibody.

“From a gross perspective, our bispecific antibodies look exactly like a monoclonal antibody, but we can bind bivalently at both ends at the same time,” asserted Dr. Brewis. “We understand where to introduce diversity into the loops that are exposed so we don’t interfere with Fc receptor binding, which is important for pharmacokinetics and effector functions.”

This new Fc-like domain has been termed Fcab™. An Fcab domain, such as the one directed against HER2 (and which is currently in a Phase I trial), can be added to Fabs binding a variety of cancer targets. F-star also has a bispecific antibody that targets LAG-3 and PD-L1. This bispecific antibody is currently in preclinical testing, and it will be the subject, F-star expects, of an IND filing by the end of the year.

Like other approaches using IgG-like structure, “in terms of production, it’s exactly the same as a monoclonal antibody,” says Dr. Brewis, “and whatever technical advances are being made for monoclonal antibody manufacture, we would just piggyback onto them. … Not least, at the end of it, you’re getting twice the drugs for the same ‘cost.’”

Knobs, Holes and a Light Touch

Larger companies such as Roche can follow multiple approaches. Roche has chosen to apply a “toolbox” approach, integrating and developing a variety of technologies to provide the best solution for their bispecific antibodies.

“We think it’s essential to have a toolbox which is not just one but multiple solutions which we can then draw from accordingly,” insisted Hubert Kettenberger, Ph.D., senior principal scientist at Pharma Research and Early Development (pRED) Roche. “There’s no one-size-fits-all solution, but given the nature of the biology we want to modify, we need the different formats.”

Dr. Kettenberger argued that because Roche pRED’s technology, Crossmab©, doesn’t use any chemical linkers or connectors, “you can use existing antibodies and upcycle them into a Crossmab with a very small bit of engineering.”

The Crossmab technology relies on protein asymmetry, which ensures that the correct pairing of the light and heavy chains happens spontaneously, without need for further crosslinking. To do so, Roche built on established knob-into-hole technology, which allows efficient heterodimerization of the antibody heavy chains. One heavy chain carries a “knob” and the other a “hole,” an asymmetry which ensures correct pairing of the two different heavy chains.

“The knob-into-hole principle is fairly old,” Dr. Kettenberger pointed out. “It was invented in the 1990s by a colleague at Genentech, which used it as a first step in making bispecifics, but the company hadn’t solved the light-chain miscarrying issue yet.”

“Roche came in about 10 years ago, with the idea of creating this extra bit of asymmetry in the light chain–heavy chain interface,” Dr. Kettenberger continued. “This was the Crossmab, essentially.”

The asymmetry is mediated by swapping either the variable or constant domain of the heavy and light chains.

Because these resemble traditional antibodies, Roche can use its extensive infrastructure for production. “We have a clinical manufacturing progress in place that is large enough to produce several kilos of Crossmabs to supply the upcoming clinical studies, which not many others can do,” asserted Dr. Kettenberger.

In 2014, Roche expanded their toolbox by acquiring Dutalys and that company’s proprietary DutaMabs™ technology. “With traditional antibodies and Crossmabs, each arm can bind exactly one target,” explained Dr. Kettenberger. “Dutamabs can be made so that each of the arms bind two. It still looks like a traditional antibody, but molecules can bind two times two copies of the ligand.”

Roche currently has several bispecific antibodies in Phase I trials directed against cancer targets, and are exploring exciting new areas for development. “We launched programs that bind to three or four different targets,” Dr. Kettenberger told GEN. “That’s something we are eagerly verging on at the moment.”

Expecting Unexpected Biology

For Merus, the priority is that the bispecific antibody be considered as a drug modality in its own right. “We’re firm believers that if you want novel activities and mechanisms that can only be achieved by bispecificity, you have to do it through empirical screening of bispecific antibodies,” said Mark Throsby, Ph.D., chief science officer and executive vice president of Merus.

Merus’ Biclonics© platform allows for a rapid generation of bispecific antibodies to efficiently generate panels of human antibodies with heavy-chain diversity but sharing a common light chain.

“Other approaches often depend on repurposing existing monoclonal antibodies that get pasted together to form the bispecific antibody candidate,” noted Dr. Throsby. “We make large panels of bispecifics that combine different binding arms in a high-throughput way, meaning we can rapidly perform the first unbiased screen of the functionality of these bispecific antibodies in the final therapeutic format.

“We prefer this intensive approach to identifying new candidates at the discovery phase because it allows us to be surprised by novel biology. Because our format involves a lot of constant elements, we can leverage all the developability knowledge that has accumulated around antibody production to secure aggressive timelines from the moment we have a lead candidate to the point where we’re going to the clinic.”

Merus currently has two antibodies in Phase I/II development that illustrate two of the company’s approaches to cancer treatment. One bispecific antibody, MCLA-117, bridges T cells to leukemic stem cells in acute myeloid leukemia (AML) by binding CD3 and CLEC12A. The other bispecific antibody, MCLA-128, reflects the company’s dock-and-lock approach.

“Here, one of the arms is binding to an overexpressed target on a tumor cell (‘dock’), and it is associated with a functional target in close proximity,” explained Dr. Throsby. “This can lead to much more potent inhibition (‘lock’) than with a monoclonal antibody against the same functional target.”

MCLA-128, which binds HER2 and HER3, is being tested for therapeutic activity in breast, gastric, ovarian. and endometrial cancers.

A Plug-and-Play Platform

Biomunex Pharmaceuticals, a company founded in 2014, proposes a different paradigm for therapeutic candidate identification. “We do not need to spend time trying to optimize our construct or our engineering,” said Eugene Zhukovsky, Ph.D., the company’s chief scientific officer. “We spend time on doing biological evaluation to select the most efficacious bispecific antibodies.”

The company’s proprietary technology, the BiXAb® platform, relies on the coupling of two native Fabs to the first two Fabs through two sets of linkers to create a tetra-Fab structure. “It is a turnkey approach for the development of bispecific antibodies,” explained Dr. Zhukovsky. “It enables us, in three months, to essentially generate a molecule that could be put into various types of evaluation for its activity.”

There is also an advantage to conserving an intact Fc portion in a bispecific antibody structure. The Fc domain contributes to enhance stability and productability of antibodies. Because of this, like other molecules that closely resemble antibodies, the Biomunex molecules “are very stable, don’t aggregate, and have excellent binding and half-life properties,” asserted Dr. Zhukovsky.

Biomunex is focusing on clinical-stage programs that are evaluating tetravalent, bispecific BiXAbs for dual targeting. One of the current programs targets the EGFR family of oncogenes to treat different solid tumors. The company is also exploring possibilities for tri- and tetra-specific antibodies.

A Nanobody Format

Ablynx, for its part, is exploring multivalency by utilizing an approach different from the traditional antibody format. “Traditional technologies are still, in essence, constrained within a traditional antibody framework,” said Antonin de Fougerolles, Ph.D., the company’s chief scientific officer.

Ablynx’ nanobody technology utilizes a specific immune response that is mounted in camelids. “Camelids, lamas, alpacas, camels can make a heavy chain–only immune response,” noted Dr. de Fougerolles.

“We’ve cloned out the heavy-chain domain, and we can link [several domains] together using genetic fusion to bind either the same target or different targets. In that way, we make multivalent or multispecific constructs,” explained Dr. de Fougerolles. “Because each nanobody is about 12–15 kDa, three or four nanobodies together make up a molecule that is still significantly smaller than a traditional antibody. Some molecules have up to seven nanobodies in a single construct.”

The small size of these molecules opens the path toward alternative delivery methods. Nanobodies can, for instance, be delivered via oral or inhalation routes, as they resist nebulization. However, as this molecule does not comprise an Fc domain naturally, Ablynx has had to engineer the structure’s stability.

“The nanobodies themselves have a very short half-life, 8 to 10 hours in man,” detailed Dr. de Fougerolles. “If we want to extend the half-life, we can graft our nanobodies directly onto an Fc domain, or link our nanobody building block to a building block that binds serum albumin. The nanobody will then circulate alongside the salbumin and, in man, raise the half-life up 10 to 20 days.”

Ablynx currently has several nanobodies that have moved into trials in that space, the most advanced of which is a VEGF/Ang2 bispecific nanobody, which is in Phase I.

It will be fascinating to follow the evolution of these bispecific antibody formats as they reach the final phase of clinical trials.

“We’re going to see some very interesting results in the field. [Bispecific antibodies] are emerging, and I think they’re going to keep on emerging in the next couple of years,” predicted Dr. Brewis. “There are going to be successes, and there are going to be failures, but I am sure there will be a number of successful formats that make it to the market.”


Ablynx says its proprietary Nanobody platform allows for the rapid generation and large-scale production of novel biological therapeutics that have potential in a wide range of human diseases.

Additional Resources
Scaling Bispecific Antibody Production
Quantifying Stability for Biologics

Previous articleThrough Development, Some Brain Cells May Favor Mom’s Genes, or Dad’s
Next articlePositive Phase III Data for Merck Candidates against CMV and Shingles