Osaka, Japan – One main goal of anti-cancer therapies is to kill tumor cells without affecting the surrounding normal cells. Therefore, many drugs are designed to target tumor-specific antigens, which are molecules only expressed by cancer cells. However, it has proven difficult to identify such specific antigens in certain cancer types, including acute myeloid leukemia (AML).

AML patients are often treated using allogeneic hematopoietic stem cell transplantation (allo-HCT), where they receive stem cells from a donor. Unfortunately, despite advancements with allo-HCT, many AML patients relapse.

In a recent article published in Nature Cancer, a multi-institutional research team led by The University of Osaka describes how a molecule called HLA-DRB1 can be used as a target for chimeric antigen receptor (CAR)-based therapy for AML. In CAR-based therapy, T cells are engineered to target and kill cells that express a specific molecule. CAR T cells have been highly successful in individuals with B cell leukemia/lymphoma and multiple myeloma (MM). However, most of the CAR T cell targets currently in clinical trials for AML are also expressed in normal cell types, leading to potential toxicity.

“In our previous work in MM, we screened monoclonal antibodies (mAbs) to identify any that could react with human MM samples but not with normal blood cells,” says Shunya Ikeda, lead author of the study. “We aimed to use that same strategy to find AML-specific antigens.”

The team began screening thousands of mAbs raised against AML cells, narrowing this list down to 32 that bind specifically to AML cells. One mAb, named KG2032, clearly bound to AML cells in over 50% of patient samples tested. Using a sequencing strategy, the researchers determined that KG2032 bound to HLA-DRB1.

“Interestingly, we found that KG2032 reacted with a specific HLA-DRB1 subset in which the protein has an amino acid other than aspartic acid in the 86^th position,” explains Naoki Hosen, senior author of the article. “KG2032 would therefore only be reactive to AML cells in individuals with mismatched HLA-DRB1, meaning the patient carries this amino acid residue but the allo-HCT donor does not.”

This finding indicates that HLA-DRB1 can be a potential target in treating certain patients with AML who have relapsed after allo-HCT.

The team then engineered KG2032 CAR T cells without the reactive HLA-DRB1 allele to test this finding. The KG2032 CAR T cells displayed strong and specific anti-AML effects in vitro with cell culture experiments, as well as in vivo with a mouse model. The treated mice did not display any overt signs of toxicity. Engineered cord blood-derived CAR natural killer (NK) cells showed similar results.

Overall, these very promising findings indicate that KG2032-derived CAR T or NK cells may be a lifesaving intervention for AML patients who have relapsed following allo-HCT. Clinical trials are currently being planned for both cell types.

###