Using AI-Driven Tools to Explore Osteosarcoma’s Genomic Chaos

Christina Curtis, PhD, MSc, a leader in cancer genomics at Stanford University, is applying her AI-driven research approaches that advanced breast cancer understanding to uncover osteosarcoma’s origins and new perspectives on therapy.

Osteosarcoma remains one of the most challenging cancers to treat effectively, but researchers like Christina Curtis, PhD, MSc, are shedding new light on how it begins and grows. Dr. Curtis is focused on understanding how these tumors form and evolve, a foundational step toward developing more effective therapies.

Christina Curtis, PhD, MSc

“I am a fundamental believer that if we understand the origins, we can intervene more optimally and earlier,” says Dr. Curtis, the RZ Cao Professor of Medicine, Genetics, and Biomedical Data Science at Stanford University, where she is also the Director of Artificial Intelligence and Cancer Genomics and of Breast Cancer Translational Research. “Knowing what is driving malignancy is really crucial to targeting it because these dependencies are established early, despite ongoing evolution.”

With grant funding from The Osteosarcoma Institute in 2023, Dr. Curtis is building on her renowned work in breast cancer genomics to explore osteosarcoma’s distinctive genetic chaos. Her goal: to identify the biological triggers and genomic patterns that could guide earlier, more precise interventions.

A Closer Look at Cancer’s Genomic Instability

Osteosarcoma is known for its chaotic genome. “This refers to the rearrangement of genes on chromosomes,” Dr. Curtis explains.

In healthy cells, each chromosome comes in a pair (one from mom and one from dad) and the genes are arranged in a neat, linear order that is straightforward to map. “But in cancer, and maybe the best example is osteosarcoma, those chromosomes can shatter and get stitched back together like a quilt, creating a mosaic of different parts that are interconnected and in places they should not have been,” Dr. Curtis says.

This kind of genetic chaos is also seen in certain breast cancers. Through her lab’s computational modeling and molecular profiling, Dr. Curtis helped define the molecular subtypes of breast cancer and uncover how those tumors develop, evolve, and spread.

“I thought, if we could apply a fraction of the knowledge and the tools and the methods we had developed (studying breast cancer) to osteosarcoma, maybe we could really learn something,” she says. “I am happy to say we have been able to learn a lot by applying the same kind of principles to osteosarcoma.”

Using AI to Understand Tumor Origins

Dr. Curtis’s lab uses advanced computational tools, including machine learning and mathematical modeling, to trace tumors back to their founding cell. “It is pattern discovery,” she explains. “The cells are a family. There is a mother cell to begin with that divides and there are the progeny. We use the patterns of alterations to trace back to the founding cell to better understand the originating genetic event that is most responsible for the cancer.”

By studying patient samples taken at diagnosis, throughout treatment, and when disease spreads, her team is uncovering how osteosarcoma evolves over time. This longitudinal approach has revealed two distinct evolutionary processes in osteosarcoma, findings that may help explain differences in patient outcomes and guide more tailored therapies in the future.

“A big part of the work that we want to do is to figure out: Now that we know more about this disease, how do we do better?” —Christina Curtis, PhD, MSc

“I believe that without the right diagnosis (the specific biological features that distinguish one osteosarcoma from another), you cannot tailor therapy,” Dr. Curtis says. “A big part of the work that we want to do is to figure out: Now that we know more about this disease, how do we do better?”

The Power of Collaboration

Dr. Curtis’s discoveries are driven by collaboration across institutions and disciplines. Her lab partners closely with other OSI-funded investigators, including Alejandro Sweet-Cordero, MD, of UCSF, to gather patient samples and develop laboratory models that make this kind of high-resolution analysis possible.

“Dr. Curtis has a unique set of skills as a biological scientist and a computer scientist to understand genomes in cancer evolving over time,” says OSI Strategic Advisory Board Chair Chand Khanna, DVM, PhD. “A strategic priority of OSI was to bring people from outside of the osteosarcoma world, and she is the best example of us doing that.”

Beyond OSI, Dr. Curtis contributes to Break Through Cancer, a collaborative focused on several of the world’s most challenging cancers, including osteosarcoma. “There is real magic that happens at the intersection of disciplines,” she says. “I am delighted to have been supported by OSI and to serve on its Strategic Advisory Board because I think there is so much more we can do and it’s beginning to move so much faster.”

Looking Ahead

Dr. Curtis and her team are now validating their osteosarcoma findings in larger patient cohorts and engineering new laboratory models to study the mechanisms behind chromosome shattering. Their goal is to identify biomarkers that can predict patient outcomes and ultimately inform better treatment strategies.

“We have more ideas than resources, which is a good problem to have,” she says. “We’re energized by what we’re discovering, and we’re just getting started.”