Working Smarter: Improving Personalized Stem Cell Treatments for Kids

Stem cell therapies are improving recovery and survival rates for pediatric cancer patients. But the treatments can be risky. They can weaken the immune system, making children highly vulnerable to infections. And there are other potential long-term complications, including damage to tissues and organs.

A team of researchers at Georgia Tech has addressed this challenge, creating a new way to predict how these cutting-edge treatments might work in a particular patient. And it could revolutionize treatments for kids with complex immune system challenges.

Currently doctors inject stem cells almost blindly, taking kilograms of cells from bone marrow or umbilical cord blood without knowing exactly which will work best, or how they’ll react to immune responses. The new research provides a more precise way to understand these complex interactions.

“We essentially created a miniature laboratory that can show how stem cells behave inside a patient’s body,” said Ahmet Coskun, Bernie Marcus Early Career Professor in the Wallace H. Coulter Department of Biomedical Engineering at Georgia Tech and Emory University, who led the research team.

“It’s a predictive approach to guide clinicians in choosing the right treatment and understand exactly how these cells interact with a patient’s immune system,” added Coskun, whose team explained its innovative research recently in the journal BME Frontiers.

Mimicking the Body

Coskun’s team developed a specialized hydrogel that mimics the body’s internal environment, allowing researchers to examine stem cells and immune cells together. Using advanced 3Dspatial transcriptomics analysis techniques, the team can study 30 different spatially resolved genetic markers instead of the traditional two or three.

“This allows us to create a much more detailed graphically wired maps of how these cells interact,” said Coskun. “It’s like understanding a conversation by listening to, and understanding, 30 different speakers instead of just two. We can see how stem cells from different sources might behave differently, potentially helping doctors choose the most effective stem cells for a specific patient.”

The research could have significant implications for treating graft versus host disease (GVHD), a serious complication that can occur when stem cells transplanted from a donor recognize the patient’s body as foreign and start attacking the child’s tissues and organs.

“By better understanding how stem cells interact with immune cells before injecting them to patients, doctors might be able to reduce the risk of rejection and improve treatment outcomes,” said Coskun, whose collaborators included Children's Healthcare of Atlanta pediatric oncologist Edwin Horwitz and Georgia Tech biomedical engineer Ankur Singh, and lead author, Zhou Fang, a grad student in the Coskun lab. The study also involved six undergraduate researchers, highlighting the team's commitment to training the next generation of scientific talent.

“Our approach goes beyond traditional cell analysis by examining how cells communicate – not just whether they interact, but how they ‘speak’ to each other at a molecular level,” said Coskun. “We used advanced data science techniques to create complex networks that map cellular communication, similar to how social networks connect people.”

Coskun believes the research offers a promising new direction for personalized stem cell therapies. 

Citation: Zhou Fang, Kelsey Krusen, Hannah Priest, Mingshuang Wang, Sunwoong Kim, Anirudh Sriram, Ashritha Yellanki, Ankur Singh, Edwin Horwitz, Ahmet Coskun. “Graph-Based 3-Dimensional Spatial Gene Neighborhood Networks of Single Cells in Gels and Tissues,” BME Frontiers. DOI: 10.34133/bmef.0110

Funding: This research was funded by the 2022–2023 Regenerative Engineering and Medicine Seed grant, National Institutes of Health (NIH) grants R35GM151028, 5R01CA238745-03, 1R01CA266052-01A1, 1R01AI186314-01, 1R01AI181282-01A1, National Science Foundation grant (NSF) EEC-1648035 and an NSF CAREER award number (338935).