AI-Powered Digital Twins Offer Novel Hope for Brain Tumor Treatment

A groundbreaking study from the University of Michigan has harnessed the power of artificial intelligence to create digital twin replicas of gliomas, offering a new approach to understanding and treating these aggressive brain tumors. Published January 6, the research sheds light on the metabolic pathways driving tumor growth and paves the way for personalized treatment strategies.

Understanding Digital Twins in Cancer Research

Digital twins – virtual representations of physical systems – are revolutionizing medical research. According to Dr. Daniel Wahl, associate professor of radiation oncology and study co-author, “A digital twin is a virtual representation of some physical property that exists in the real world. Having a virtual representation of the system allows you to study and perturb it virtually, so you can determine what will happen to your real-life system if you make a modification.” This technology allows researchers to medically manipulate a digital copy of a patient’s brain tumor, testing different therapies and observing their effects without directly impacting the patient.

Metabolic Dependencies: A Key to Targeted Therapy

The study focuses on metabolic dependency, examining how cancer cells rewire their energy production and biomass creation. Researchers specifically investigated two crucial pathways: nucleotide synthesis and serine consumption. Nucleotide synthesis involves the production of DNA and RNA building blocks for cell proliferation, while serine is an amino acid vital for tumor growth and development.

Baharan Meghdadi, a doctoral student in chemical engineering and study co-author, explained that the AI-twin model can differentiate between these pathways and identify external factors influencing them. “The machine learning model…could distinguish between these pathways and find the contribution for each of these pathways in each patient,” Meghdadi said.

Mouse Trials Demonstrate Dietary Intervention

Initial trials using the AI-twin technology were conducted on mice. Researchers discovered that some mice were more reliant on serine production for tumor growth. By adjusting the mice’s diets to limit serine intake, they successfully reduced tumor size. This finding validates the predictive power of the AI model and suggests a potential therapeutic avenue.

Building a Personalized Treatment Plan

The researchers developed a first principles model grounded in basic biochemical laws to build each patient’s treatment plan. This model uses stoichiometric and isotopic simulations to understand how a tumor processes nutrients. This core framework is then combined with single-cell level metabolic activity data from patients, a process called “flux” analysis, to create a detailed metabolic profile of each tumor.

Dr. Wahl emphasizes the importance of this technology in addressing the gaps in understanding differing metabolic pathways. “We have no way to ask, really, ‘Is a metabolic pathway active in cancer?’ And so if we’re trying to develop metabolic therapies, we don’t know who to give (each) one to. And so that’s why we’ve developed this: now we can say, ‘Hey, patient one has this pathway active, we can block this pathway. Patient two doesn’t have that active? Well, we shouldn’t apply that drug — we should do something else.’”

Expanding the Patient Pool for Clinical Application

While the initial study involved data from only eight patients, researchers recognize the need for a larger patient pool to refine the technology and prepare it for widespread clinical use. Deepak Nagrath, professor of biomedical engineering and study co-author, stated, “The data was only available to us for eight patients, and ideally we would aim for hundreds of patients.” Expanding the dataset will allow oncologists to more accurately identify factors driving tumor growth and determine the most effective therapies.

Future Directions

The AI-powered digital twin technology holds immense promise for revolutionizing brain tumor treatment. As the patient database grows and the models become more sophisticated, this approach could lead to truly personalized therapies, maximizing treatment efficacy and improving outcomes for patients battling these challenging cancers.

Dr. Daniel Wahl is a physician scientist at the University of Michigan specializing in cancers of the central nervous system. His research focuses on developing new treatment strategies for brain tumors, particularly the interactions between radiation and abnormal metabolism in glioblastoma [University of Michigan Medical School].