Call it a “gut feeling.” For some stomach cancers, the first signs may appear years before a diagnosis.

At Sutter Health Research Institute, Dr. Su-Ying Liang is helping lead that work. Through Gastroshare, she and researchers at Stanford University are using artificial intelligence and electronic health record data to study early changes in the stomach and what they may reveal about a person’s cancer risk.

The project recently received a highly competitive National Institutes of Health research award, led by Stanford University principal investigator Dr. Robert Huang.

For Dr. Liang, the project is the latest chapter in a career dedicated to using data to improve patient care. Her work has spanned breast, lung, liver, ovarian and other cancers, with a focus on prevention, early detection and better outcomes.

What keeps her going is the possibility that research can change what happens for patients.

Dr. Su-Ying Liang

“It’s an exciting time to be at the intersection of healthcare and AI,” Dr. Liang says. “The answers we need may already exist in the data. Finding them could help us identify risk earlier and get ahead of cancer.”

An Earlier Shot at Prevention

Stomach cancer is one of medicine’s most frustrating paradoxes: serious, yet often preventable.

Many patients are diagnosed only after the disease has advanced.

One known risk factor is H. pylori, a common bacterial infection that lives in the stomach. Research has shown that treating the infection can reduce the risk of stomach cancer.

But some researchers believe the stomach may reveal clues long before cancer appears.

“An endoscopy gives us a window into changes in the stomach lining before cancer develops,” says Dr. Liang. “Gastroshare is helping us understand which of those signals matter most and how they can guide prevention.”

Finding Signals in a Sea of Data

Finding a needle in a haystack is hard. Finding thousands of needles is where AI comes in.

Dr. Liang and team will use AI to analyze pathology records, starting with an estimated 19,000 patients with early changes in the stomach lining linked to cancer risk. AI could help researchers identify even more patients whose conditions are documented elsewhere in their medical records.

The team will then connect those records with the California Cancer Registry to see which patients later developed stomach cancer.

That approach will give the team a chance to ask these important questions:

• Why do some people develop cancer while others don’t?
• Which patients may benefit from closer monitoring before cancer develops?

Sutter’s large patient population gives researchers real-world health data to study stomach cancer risk across different populations.

Gastroshare also builds on years of collaboration between Sutter and Stanford. Previous projects have used electronic health record data to study breast and lung cancer. Now, Dr. Liang and team are bringing that experience to a disease she says remains understudied in the United States.

The Trial That Never Happened

Gastroshare is also taking on a question that would be difficult and expensive to answer through a traditional clinical trial: Can screening people with these early stomach changes help prevent cancer?

A traditional study could take years and cost hundreds of millions of dollars.

Instead, the researchers plan to use existing electronic health records to recreate key parts of a clinical trial, allowing them to study whether screening the stomach with an endoscopy, a procedure that uses a small camera to look inside the digestive tract, may affect outcomes.

The findings could help the team better understand who may benefit from screening and help guide future studies.

For Dr. Liang, the new NIH support also comes with responsibility.

“This support gives us an opportunity to look for answers at a scale that wasn’t possible before,” she explains. “Our responsibility is to make sure those answers translate into better health for patients.”

And what researchers learn from Gastroshare could go well beyond this project.

Dr. Amanda Wheeler

“Some of our greatest opportunities in cancer care come before a diagnosis,” says Dr. Amanda Wheeler, breast surgeon with Sutter’s West Bay Medical Group and chair of Sutter’s advanced cancer service line. “If we can better identify who is at risk and intervene sooner, we have a chance to change the course of disease before cancer ever takes hold.”

Learn more about research and clinical trials at Sutter Health that are informing clinically distinct cancer care.