Under the Microscope: Helping immunotherapy work for more children with leukemia

B-cell acute lymphoblastic leukemia, one of the most common childhood cancers, continues to confound researchers, even after years of efforts to develop new treatments.

Although a breakthrough immunotherapy treatment known as chimeric antigen receptor (CAR) T-cell therapy was recently developed for particularly high-risk cases, it ultimately fails more than a third of such patients. Also, because the treatment kills all B cells, children who do respond are left partly immunocompromised, perhaps for the rest of their lives, making them potentially susceptible to COVID-19 and other infections.

When CAR T-cell therapy was first developed, it was thought to be a miracle cure.

Andrei Thomas-Tikhonenko, Ph.D.

“When CAR T-cell therapy was first developed, it was thought to be a miracle cure,” said Andrei Thomas-Tikhonenko, Ph.D., from Children’s Hospital of Philadelphia. “However, while this immunotherapy works in many patients, some of them eventually relapse after initial successful rounds of treatment. This motivates our research, which is aimed at designing therapies that could help those who don’t respond to conventional CAR T-cell therapy and that would spare healthy B cells.”

Saving healthy B cells

B cells help our body fight infections, but in children with cell acute lymphoblastic leukemia, these white blood cells turn into cancerous leukemia cells. Cancerous B cells don’t function like normal B cells, yet they reproduce quickly and outlive normal cells. After building up in the bone marrow, they often move into the bloodstream and affect other organs in the body. The most effective conventional CAR T-cell therapy targets a protein called CD19 that marks all B cells, meaning the therapy destroys cancerous and healthy B cells.

With V Foundation funding, Thomas-Tikhonenko and his research team are hunting for a marker that is found on cancer cells but not normal B cells. They are some of the growing number of researchers looking at RNA splicing, a cellular process that typically doesn’t work properly in cancerous cells, to see if it offers a way to distinguish between leukemia cells and healthy B cells.

RNA splicing is a key part of the process cells use to make proteins. When “copy editing” the genome to produce instructions for making proteins, messenger RNA removes non-coding sequences of genes and joins protein-coding sequences together. In cancerous cells, problems with this splicing process can produce abnormal proteins that don’t exist in a normal cell.

“For B-cell acute lymphoblastic leukemia, a protein known as CD22 is mis-spliced in a way that creates junctions between parts of the protein that are not normally together,” said Thomas-Tikhonenko. “Our work is focused on making antibodies that recognize this junction as foreign. These antibodies could be used to create a therapy that targets the cancer cells without affecting healthy B cells.”

So far, Sisi Zheng, M.D., in the Thomas-Tikhonenko laboratory has developed one first-in-class animal antibody that recognizes abnormally spliced CD22, and she has shown it can readily distinguish between tumor cells and normal cells. She and her colleague Zhiwei Ang, Ph.D., are now looking for other targets and are also working through the difficult process of figuring out how to turn novel antibodies into drugs, which requires converting the animal antibody to a humanized antibody.

New tools, new collaborations

Recently, Thomas-Tikhonenko’s research team began collaborating with Jamie Spangler, Ph.D., at Johns Hopkins University, who uses a new technology to express all known human antibodies in yeast cells. This approach could prove helpful for screening billions of existing human antibodies to identify ones that recognize CD22 or other proteins that mark cancerous B cells.

“Not only does this high-throughput approach keep us from having to go through antibodies one by one using animal models,” said Thomas-Tikhonenko. “It also means that we don’t have to generate a humanized version of a mouse antibody, which can take a long time. Thus, this approach could shorten the time it takes to translate findings from the lab into clinical trials.”

The researchers are also studying data from patients who undergo CAR T-cell therapy to find out why the therapy doesn’t work in everyone. They’re examining differences in RNA splicing between responders and non-responders to see if mis-splicing might help some cancers resist therapy.

The V Foundation does a great job of funding high-risk, high-reward projects like ours.

Andrei Thomas-Tikhonenko, Ph.D.

“The V Foundation does a great job of funding high-risk, high-reward projects like ours,” said Thomas-Tikhonenko. “We were trying to develop a new type of drug based on a hypothesis that hasn’t been validated.”

This work, made possible through the current and previous V Foundation support enabled the research team to join forces with Yoseph Barash, Ph.D., and his computational biology lab at the University of Pennsylvania and successfully compete for new NIH Beau Biden Cancer Moonshot Initiative funding. Through that award mechanism, Thomas-Tikhonenko and his collaborators are actively participating in the Pediatric Immunotherapy Discovery and Development Network focused on developing new immunotherapies.