AI is helping us better identify pancreatic cysts, and an organ-saving procedure gives patients new options

One of the challenges facing pancreatic cancer care teams and their patients is that pancreatic cancer is often diagnosed in late stages, because most pancreatic cancers don’t have obvious symptoms and there’s no universal screening tool for pancreatic cancer.

That limits the effectiveness of treatment and the chances of long-term survival.

About a fourth of pancreatic cancers, though, start as cysts that may or may not develop into cancer. Identifying which cysts are mostly likely to turn into pancreatic cancer is one key to providing better treatment.

That’s where AI comes in. Researchers at The Ohio State University Comprehensive Cancer Center – Arthur G. James Cancer Hospital and Richard J. Solove Research Institute (OSUCCC – James) are now using artificial intelligence to help detect pancreatic cancer in earlier stages or before cysts even turn into cancer.

Why not remove all pancreatic cysts?

Not all precancerous cysts actually need to be removed, says Somashekar Krishna, MD, a gastroenterologist and researcher with the OSUCCC – James Molecular Carcinogenesis and Chemoprevention Research Program.

“About 50% likely did not need to be removed and, more than likely, they wouldn’t have progressed to cancer during a patient’s lifetime,” Dr. Krishna says.

Most patients affected by pancreatic cancer are over 60, he says, and many of them also have other medical conditions that can make surgery risky.

“If we can find a way to identify and risk-stratify lesions upfront, many patients can be treated with minimally invasive organ-saving ablation procedures to eliminate the cyst.”

How AI helps doctors accurately and quickly identify the most dangerous pancreatic cysts

Dr. Krishna’s team evaluates the pancreas through endoscopic ultrasound, in which laser optical cameras are sent into the pancreas with a tiny probe, giving them a microscopic view of the cyst wall.

Identifying all the cancer risk factors from that video from inside the pancreas can be difficult for a human, though.

“It’s like asking someone to count all the black, white and blue cars in a video where the cars are moving at 100 mph. It’s simply impossible without slowing down the video and analyzing hundreds, if not thousands, of individual frames,” Dr. Krishna says.

So, Dr. Krishna is working with an Ohio State computer science engineer, Harry Chao, PhD, to train an AI program to review these images and improve the accuracy of imaging results.

An AI algorithm could flag markers of pancreatic cancer in seconds instead of hours, Dr. Krishna says.

“It can not only save lives but also reduce the need for unnecessarily invasive surgeries and their associated costs,” he says. “Additionally, it lowers the risk of medical side effects from organ removal, such as the onset of diabetes and digestive complications.”

How pancreatic cancer is removed once it’s identified

Dr. Krishna has also developed a new endoscopic procedure to treat those precancerous cysts using targeted heat, which uses radiofrequency ablation to eliminate that abnormal tissue.

This much less invasive than the more standard Whipple surgical procedure, in which pancreatic cancer is removed by taking out the head of the pancreas, the part of the small intestine called the duodenum, the gallbladder and the bile duct. For older adults and people who have other serious health conditions, recovery from this large surgery can be difficult.

The minimally invasive endoscopic procedure can help people with early-stage pancreatic cancer or precancerous conditions avoid that major surgery, Dr. Krishna says. While it might not be the first-line option for every patient, it can be a good alternative for people who face higher risks with Whipple surgery.