More than 60 years ago, scientists started making bispecific antibodies (bsAbs). As the name suggests, these antibodies can bind two specific targets. Moreover, bsAbs can bind two targets on the same cell or targets on different cells to bring them together, such as developing a cancer treatment by attaching an immune cell to a tumor cell. Despite the years of advances on bsAbs, including their introduction as therapeutics in 2009 with the European Medical Agency’s approval of catumaxomab, it remains challenging to make one—at least one that performs a desired task.

Most clinical studies using bsAbs involve a cancer treatment. In fact, a search of ClinicalTrials.gov on November 25, 2022, returned 200 studies of bsAbs in the recruiting stage, and almost all of those trials—96%—involved treatments for cancer. The remaining 4% included trials on treatments for chronic kidney disease, hemophilia, HIV, lupus, and psoriasis. Treating different conditions depends on engineering bsAbs for a specific purpose.

In a recent webcast, Michael Fiebig, PhD, CSO at Absolute Antibody, which is headquartered in the U.K., said, “When we talk about antibody engineering, we mean that you take the variable domains of an antibody that bind to its target, and you can engineer them onto different backbones.” He added, “There are lots of architectures or geometries that you can use to create a bispecific binder.” In doing so, a bsAb needs to meet various key attributes, including affinity and half-life.

Attributes can be counterintuitive

Sometimes, the desired attributes can be counterintuitive. Taking affinity as an example, Fiebig explained that higher binding is not always better. With tumor-associated antigens, for instance, a weaker-binding bsAb is more likely to target tumor cells than healthy ones, because the former contain higher levels of the target.

In addition, Fiebig pointed out that bispecifics extend beyond antibodies. “Not every binder that you include in a construct needs to be an antibody,” he said. “Nature’s already developed a lot of good binders for us, and it’s always worth considering if one might be able to use something like this in a construct.” That “something” could be, for instance, a protein that’s not an antibody.

Just making a bsAb that meets the therapeutic objectives is not enough, because one that cannot be made economically and efficiently is not likely to become a useful therapeutic. As an example, it can take several steps of purification to go from bioprocessing to a product. When discussing purification, Fiebig said, “that’s when it can get extra tricky.”

Fiebig emphasized the steps in commercial processing of a bsAb. “Do not underestimate purification and manufacturability,” he noted. “This is very important to bear in mind from the beginning, so that you don’t end up with something that looks good and maybe works well, but you just can’t make it at big enough quantities and at sufficient purity.”

So even after 60 years of work, many factors must be considered, and various obstacles overcome to create a bsAb that turns into a treatment. From affinity to purification, each step must work just right.