Research Seeks to Speed Up Rare Disease Diagnosis

Print

Published on October 17 2024 9:19 am
Last Updated on October 17 2024 9:20 am

The Verdeyen family may cheer for different sports teams but they are all on team Todd as the father of four fights Pompe Disease symptoms. They are excited about new research led by Dr. Adam Cross to create a tool to help patients get quicker diagnosis of rare diseases. From left Anneke on the shoulders of Pierce, Tina, Todd, and Makenna on the shoulders of Quinn Verdeyen. (Photo courtesy of the Verdeyen family)

By PHYLLIS COULTER FarmWeek

Todd Verdeyen, an Effingham County pharmacist, no longer had the stamina to travel with his wife or keep up with their four teens.

“At the rate I was going, I was convinced that I would be on oxygen or in a wheelchair soon,” Verdeyen told FarmWeek.

Genetic testing changed everything for him. He finally received a firm diagnosis after decades of searching and started treatment that has slowed the progression of a rare disease.

Like Verdeyen, one in 10 people have a rare disease, according to the World Health Organization. Getting diagnosed can be challenging, especially in rural or medically under-resourced areas.

But all that could change with the work of a University of Illinois Chicago and OSF HealthCare in Peoria team, led by Dr. Adam Cross, a pediatric hospitalist and clinical researcher for the University of Illinois College of Medicine.

His research uses artificial intelligence (AI) and knowledge graphs to create a model that uses a doctor’s notes about a particular patient’s symptoms and securely lifts the relevant information. That information can be paired with databases of 10,000 rare diseases to help the physician determine what further tests need to be done for a diagnosis.

The goal is to create a tool for primary care providers to help patients get an early diagnosis of a rare disease.

“Rare diseases must be caught early to be treated. A lot of people go undiagnosed for a long time,” said Cross, also the director of the Children’s Innovation Lab housed within the OSF Healthcare Jump Trading Simulation and Education Center.

Genetic testing is typically the only way to confirm a rare disease. That testing is expensive and not readily available in primary care settings, especially in rural communities. Patients have long waits and often must travel in hopes of getting a diagnosis, Cross said.

Machine learning with AI can provide physicians with the information they need, he said.

Dr. Adam Cross, a pediatric hospitalist and a clinical researcher for the University of Illinois College of Medicine Peoria, displays “knowledge graphs” he’s creating to show relationships between symptoms and rare diseases. (Photo courtesy of Colleen Reynolds, OSF HealthCare)

Cross is motivated by knowing what a difference this tool could make in people’s lives, especially in rural areas such as Mason County, where his grandpa was a farmer.

“I always want to make sure that my education and expertise can help people like my grandparents,” said Cross. Members of his family are McLean County Farm Bureau members living near Heyworth, so his rural connections remain strong.

Now with funding from Jump Applied Research for Community Health through Engineering and Simulation grant money, researchers aim to further use AI to tweak the system to make it most effective, including lessening the chances of false positives and avoiding over-testing.

“We can adjust the model to be precisely tuned for its intended purpose,” Cross said.

Ideally, the tool will be ready in three to four years. “That’s a pretty fast timetable,” Cross acknowledged, but he is confident it can be attained.

“We’re experienced in creating tools to support physicians in decision-making,” he said of the team.

The life-changing potential of this tool is clear to Verdeyen.

His symptoms started when he was in high school and varied over decades. By 2019, when he was in his 40s, things got much worse.

“If I walked 500 feet I was completely out of breath and gasping for air,” he said.

He was sent to a pulmonologist and a cardiologist. It wasn’t cardiac problems. Pneumonia was considered for his respiratory issues. He got an MRI to address other symptoms. Muscular Dystrophy was a possible diagnosis. When the soonest he could see a specialist at Barnes Jewish Hospital in St. Louis was in February 2025, Verdeyen was discouraged.

However, when he was referred for genetic testing, this changed.

With results in hand, Jennifer Burton, a clinical genetic counselor with OSF Healthcare Children’s Hospital, diagnosed him with Pompe Disease.

Coincidently, Verdeyen had tried 23andMe “for fun.” It provides DNA testing for users to learn more about their ancestry, possible health predispositions and genetic carrier status. The day before his appointment with Burton, he reread the results, which showed that he could be a carrier for Pompe Disease.

During the appointment, Burton explained that based on his symptoms, Pompe Disease was a strong possibility. Genetic testing identified two genetic mutations for Pompe Disease, confirming the diagnosis. Burton cautions people not to put too much faith in recreational DNA testing services, which identified one genetic marker for Pompe Disease in this case; they are not diagnostic tests.

Today, at age 53, Verdeyen receives infusion treatments five or six hours every other week. At first, the southern Illinois pharmacist had to travel to Peoria for treatment. Now he can get it close to home in Altamont.

While his improved health allowed him to go on an Alaskan cruise, Verdeyen will be taking the regular infusions for the rest of his life unless new treatments are found.

“Pompe is a progressive disease (which causes muscle weakness), and the goal of treatment is to stop progression,” Burton said. “We do not always see improvements, so we are thrilled to see any improvements in Todd’s health, daily activity and quality of life.”

Verdeyen wonders how his life would be different if they had AI and Cross’s new tool while he was seeking a diagnosis, but he is hopeful for others.

“I do think AI could help people get a diagnosis even if there is no treatment,” he said.