Prof. Szeto predicts patient responses to immunotherapy

As a Washington Post article highlighted, some doctors are now using immunotherapy—stimulating the body’s own immune response to tackle disease in targeted ways—to treat cancer patients who are not responding to traditional chemotherapies, and increasingly in place of chemotherapies or in combination with them. The Food and Drug Administration recently received unanimous recommendation from an advisory committee that a new class of immunotherapy, a “living drug,” be approved for use in children and young adults with leukemia. “The treatment takes cells from a patient’s body, modifies the genes, and then infuses those modified cells back into the person who has cancer,” explains NPR.

The Multiple Myeloma Research Foundation awarded UMBC’s Greg Szeto, assistant professor of chemical, biochemical and environmental engineering, a $75,000 research fellow grant to continue his work on immunotherapy. Specifically, Szeto is developing experimental tests and computational models that could help physicians more accurately predict how individual patients with multiple myeloma will respond to a similar type of immunotherapy. Szeto’s preliminary work with collaborator Ivan Borrello, associate professor of oncology at Johns Hopkins Sidney Kimmel Comprehensive Cancer Center, correctly predicted clinical outcome in more than 80% of multiple myeloma patients from a prior immunotherapy clinical trial using “living drugs.”

“A major hurdle for the field is determining who will benefit from which immunotherapies, and making that determination as early as possible,” explains Szeto. “Though a long way off, the rewards of pursuing such research are key insights for better understanding how existing and emerging therapies work, enhancing the connections between STEM disciplines, and in the long run increasing patient quality of life and the ability of doctors and patients to make informed care decisions.”

In previous research, Szeto found that combining multiple immunotherapy agents can improve responses in mice treated for melanoma, and computational models could help predict and dissect these responses. A paper he coauthored with collaborators including Michael Zhang Ph.D. ’23, chemical engineering, who worked with Szeto as a research tech at MIT, appeared in Nature Medicine last year and indicated that combinations of immunotherapy agents can enable immune cells to more effectively infiltrate and reduce the size of the tumor. This size reduction was predicted by fusing experimental tests with computational models. Currently, Szeto is working to define when and how to test different samples to get the best predictive models.

Because treatments are tailored to individual patients based on a broad range of factors, understanding and assessing their function can be incredibly complex, and it can be challenging to replicate successful treatments across patients. Szeto hopes to help researchers develop more predictable treatment regimens that can use knowledge about individual patients, but with a firmer sense of when particular schedules of medication and dosage levels should be used, and what impact they are likely to have.

“There really hasn’t been any approval for this ‘living drug’ approach before, so what recently happened is a major milestone,” he says, adding that “it paves the way for approval of other drugs within this class, similar to those we are studying with Dr. Borrello.”