Computer Science Reveals Possible Drug Target for Deadly Childhood Leukemia
When Hamid Bolouri, PhD, was a computer science professor in England, one project changed the trajectory of his career. It was the late 1990s and his team was examining exactly how the brain forms new networks, aiming to inform artificial intelligence (AI) technology.
“I just got absorbed into biology and never looked back,” he says. “I started using AI methods to analyze biology rather than using biology to inspire computer science.”
Now at BRI, Dr. Bolouri is an expert in using computer science to advance biomedical research. Working with Soheil Meshinchi, MD, PhD, and his team at Fred Hutch Cancer Center, they made a recent discovery that may lead to new treatments for certain forms of pediatric acute myeloid leukemia (AML).
“We have effective treatments for most pediatric AML, but some types of AML still have a very poor prognosis,” Dr. Bolouri says. “We discovered that inflammation plays a role in several of these deadly subtypes of AML. And we may have found a way to target that inflammation to get better clinical outcomes.”
Genetics Research Leads to Pediatric Cancer Breakthroughs
Scientists have long understood that genetic changes (mutations) play a role in leukemia and other cancers. Initially, scientists assumed that the same genetic changes caused both adult leukemias and pediatric leukemias. But treatments that worked for adult cancers weren’t as successful in children.
“We and others have shown that mutations connected to leukemia in children are very different from those in adult leukemia,” Dr. Bolouri says. “It makes sense that adult treatments didn’t work well for children because the underlying genetic cause was very different.”
Pinpointing underlying genetic changes has helped lead to better treatments for most leukemias and all childhood cancers. But about 30 percent of children who develop AML have what’s considered “high-risk” AML — cancers that are difficult to treat and likely to spread or come back after treatment. Dr. Bolouri’s team wanted to focus on those types of AML, hoping to make insights that could lead to better treatments.
Too Many Types of High-Risk AML
There are many different types of high-risk AML, each linked to different genetic mutations. Creating a specific treatment for each different type would be very difficult and require huge investments of time and money. So Dr. Bolouri’s team set out to look for common targets across different forms of high-risk AML.
“We know that inflammation contributes to many diseases and it’s not always clear how, so we decided to examine the role inflammation plays in these AML subtypes,” Dr. Bolouri says.
Dr. Bolouri’s research team was able to access the largest database of AML bone marrow data ever collected, thanks to a partnership with the Meshinchi Laboratory and the Children’s Oncology Group, an international pediatric cancer research consortium. Dr. Bolouri’s team used every tool in the BRI arsenal to look for meaningful patterns in the pediatric AML samples. Their findings were both heartbreaking and hopeful.
“Children with inflammation almost always had worse outcomes. They could go through rigorous chemotherapy treatment with terrible side effects, but their cancers generally came back within two years,” Dr. Bolouri says. “But that pattern also gave us a target — something to investigate further in hopes of changing those odds.”
A Roadblock for Inflammation
Dr. Bolouri and his collaborators dug deeper into how and why this inflammation happens, and made another promising discovery: A wide range of genetic mutations caused this inflammation. But they all use the same cellular pathways and processes.
“Think of the pathway we discovered as a highway, say the I-5. The inflammation starts from different sources. But regardless of where it starts, it needs to get on the I-5 at some point to reach the destination. So if you block the I-5, you can block all types of inflammation, regardless of which mutation causes it,” Dr. Bolouri says.
The implications could be huge. The research team is now applying for grants to test various therapies that could “block the I-5” in hopes of slowing down this inflammation and making treatments more effective.
“All immune cells start in the bone marrow before they mature and travel to various places in the body. And so learning more about this very fundamental process can not only teach us how it becomes dysregulated in leukemias, but in other immune-related diseases too,” Dr. Bolouri says.
How does cancer research fit into BRI's mission?
BRI’s mission is to predict, prevent, reverse and cure immune system diseases — including autoimmune diseases, allergies, infectious diseases and cancer. Cancer and autoimmune disease have an interesting inverse relationship: Autoimmune diseases happen when the immune system mistakenly attacks healthy tissue. Cancer happens when the immune system fails to fight off cancer cells. Investigating these relationships and the processes that underlie all immune system disease is a key part of our work. This approach enables us to use progress against one immune system disease to make progress against them all.
Immuno-what? Hear the latest from BRI
Keep up to date on our latest research, new clinical trials and exciting publications.