Pancreatic cancer is difficult to detect early because the pancreas is deep inside the abdomen, making potentially cancerous cells hard to reach and identify without surgery.
Researchers funded by the National Science Foundation (NSF) developed a new light-based technique that can identify precancerous and cancerous cysts—small, fluid-filled cavities in the body—by piggybacking on a standard diagnostic procedure.
“This approach can be called a virtual biopsy, as it does not collect any tissue,” said Lev Perelman, professor at Harvard University and director of the Center for Advanced Biomedical Imaging and Photonics at Beth Israel Deaconess Medical Center, whose team developed the tool with the support of the NSF Directorate for Engineering Biophotonics program.
One-fifth of pancreatic cancers come from cysts. While magnetic resonance imaging and other non-invasive imaging techniques in use today can scan for cysts, they have limited accuracy in determining which cysts are malignant and which are benign.
The best available diagnostic approach involves inserting a thin tube called an endoscopic ultrasound through a person’s mouth into the stomach and the first part of the small intestine. There, a small needle is inserted through stomach wall or intestine into the pancreas, where it can puncture a cyst. Medical personnel then collect and analyze cyst tissue. But lab results take time and are often inconclusive.
Perelman, who explores the ways in which light interacts with biological tissue, thought that a technique based on the physical principles of light scattering would be able to non-invasively determine the properties of subcellular structures (such as cell nuclei) in organs, providing physicians with diagnostic information.
Imagine a cell or a cellular nucleus, a small ball with a diameter of a fraction of the width of a hair, Perelman explains. If you shine white light on that ball, only part of that light will be reflected back at you and it will come back changed. Certain wavelengths will also be absorbed depending on the ball’s composition. The vast majority of cells in the human body only reflect visible light, though some, such as red blood cells, also absorb it.
His team applied light-scattering spectroscopy to several organs before discovering its potential to help overcome the unique challenges posed by the pancreas.
“Pancreatic cysts are a major clinical problem for gastroenterologists,” said Douglas Pleskow, clinical chief in the division of gastroenterology at Beth Israel Deaconess Medical Center and associate professor of medicine at Harvard Medical School, who collaborates with Perelman. “We need better tools, and that’s where the spectroscopy comes into play. The question is: Can this be used as a reliable tool to tell us whether a cyst is cancerous or not?”
Applying the basic premise about light scattering, the researchers inserted a tiny fiber optic probe connected to a broadband light source into the needle used to collect fluid samples in pancreatic cysts. They gathered the photons reflected from a cyst’s surface and then used an algorithm to process the data, providing an immediate result.
“The goal is to make this a real-time diagnosis, and eventually avoid having to puncture the cyst at all,” Pleskow said, emphasizing that the tool is still in the research stage. “We anticipate this could replace everything, but medicine takes a long time to change.”
In a study published in the journal Nature Biomedical Engineering, the team reported they accurately identified 95 percent of cysts in 25 patients.
“What makes NSF grants so valuable,” Perelman said, “is that they provide seed money to figure out basic principles about how stuff works before applying for larger funding needed to advance technology toward making it clinically available. The NSF Biophotonics program, directed by Leon Esterowitz, played an especially critical role in its long-term support for the light-scattering spectroscopy research.”
Lei Zhang et al. Light scattering spectroscopy identifies the malignant potential of pancreatic cysts during endoscopy, Nature Biomedical Engineering (2017). DOI: 10.1038/s41551-017-0040