Skip to main content
Premium Trial:

Request an Annual Quote

T-Cell Immunotherapy Targeting AML With CBFB-MYH11 Fusions Headed for Human Studies


NEW YORK – Investigators at the Fred Hutchinson Cancer Research Center have engineered T cells that in preclinical studies recognized and killed acute myeloid leukemia cells with a particular gene fusion. The promising preliminary activity, they said, is a step toward developing an immunotherapy for the disease.

A subset of AML that particularly affects young people arises through the fusion of the CBFB and MYH11 genes. The fusion event occurs in about 1 in 10 patients and is an essential leukemia-initiating event. Most of these patients — about 90 percent — have identical fusions, indicating that a T-cell-based immunotherapy could be developed to treat this subset of AML patients.

"We've seen a huge revolution in T-cell immunotherapy for other kinds of leukemia and we're not quite there with AML yet," Melinda Biernacki, an associate in clinical research at the Fred Hutchinson Cancer Research Center, said in an interview. "There's an unmet need for effective T-cell immunotherapies for this disease."

She and her colleagues searched for T cells that naturally recognize proteins created by this fusion gene. As they described in a recent paper in the Journal of Clinical Investigations, they then engineered other T cells to express receptors that recognize the CBFB-MYH11 fusion neoantigen and kill AML cells. Following additional safety studies and further optimization, they plan to study their approach in humans though don't yet have a clinical trial planned.

"The survival these days for AML is only about 26 percent. It still hasn't improved greatly," senior author Marie Bleakley, a pediatric hematologist-oncologist at the Fred Hutch and Seattle Children's Hospital whose work in this area has been funded by Stand Up To Cancer, added. "There's a huge need for new therapeutics."

When searching for new immunotherapies, Biernacki said there are three factors that researchers consider. First, they want targets that are specific to cancer cells, so that the treatment's effect on healthy cells is minimized. But they also want targets that are present on nearly all cancer cells and that are essential to the cancer, so that all of the cancer is targeted and treatment escape is more difficult. 

Neoantigens produced by the CBFB-MYH11 gene fusion could fulfill these requirements as the fusion is a leukemia-initiating event. "When we think about fusion genes and CBFB-MYH11 in particular, because of the role that the fusion has in leukemia development, it means that the fusion is essentially present in 100 percent of leukemia cells and zero percent of normal cells," Biernacki said. "And it's hard for the cancer to lose the fusion because it depends on it to be a leukemia."

The researchers first developed a repertoire of synthetic proteins derived from the amino acid sequence encoded by the fusion gene. Using blood samples from healthy donors, they sought to find T cells that responded to the synthetic peptide.

They previously examined blood from AML patients to look for such T cells, but the immune system of AML patients is ravaged not only by the disease, but also by the chemotherapy used to treat it, and they were unable to find any T cells targeting the fusion protein. This, Biernacki noted, makes sense because if patients had functioning T cells targeting leukemia, then their disease should not have developed.

Instead, from within samples from healthy people, they uncovered naïve T cells that, likely by chance, recognized one of the fusion proteins.

The researchers tested whether the T cells they identified could kill AML cell lines that expressed the CBFB-MYH11 gene fusion. That way, they confirmed the T cells were not reacting to a synthetic protein that isn't generated by the disease. Additionally, the T cells killed primary AML cells obtained from patients, as well as in an in vivo mouse model of the disease.

Following this, they engineered T cells to express T-cell receptors (TCRs) specific for the CBFB-MYH11 neoantigen. 

T cells, Biernacki noted, can be engineered to express a range of receptors to enable them to recognize leukemias or other cancers. While engineering chimeric antigen receptors (CARs) may be more common and form the basis of CAR T-cell therapy, the researchers instead opted to engineer T cells to express T-cell receptors, which can also be used in immunotherapy.

While AML-targeting T cells are hard to find among leukemia patients, other T cells that are virus- or bacteria-specific are still present. These T cells unaffected by cancer can then be isolated, engineered to express the neoantigen-specific T-cell receptor, and reinfused in the patient as a treatment.

In this analysis, the researchers followed a series of steps to convince themselves that their engineered T cells only recognized the specific neoantigen of interest and not other, similar natural proteins. For instance, they sifted through protein databases to search for other proteins with similar sequences and checked whether the engineered cells would bind them, which they did not. In addition, they again generated artificial proteins based on the fusion protein and swapped out one amino acid after another to gauge whether their cells would bind the slightly altered proteins, finding that they did not.

They also tested the engineered cells on other AML cells without the fusion to make sure they were indeed reacting to AML driven by the fusion and not AML cells in general. 

There's more work to do to fine-tune, optimize, and test this approach to get it ready for the clinic. For instance, Biernacki said they are working to build a safety switch into their construct so that if anything goes wrong, for example, if the patient experiences an adverse event to treatment, they can shut down the TCRs, as well as other receptors or molecules that might improve its functionality.

At the same time, the researchers are searching for other targets. They aim to develop a suite of neoantigens so they can treat more than one subset of patients with AML.

"The one potential downside with neoantigens is that they do tend to be semi-personalized or personalized," Biernacki said. "Our goal is to build a toolbox of multiple neoantigens so we can really take care of as many people as possible."