Cancer of the pancreas typically is not detected early because the organ is not easily accessible. Thus, once diagnosed, fewer than 5 percent of all pancreatic cancer patients survive for five years. Researchers from Northwestern University and from Evanston Northwestern Healthcare, both in Evanston, Ill., recently tested an optical technique that could change that outcome.

The method is based on the subtle physical changes that occur in the surrounding tissue because of the cancer. The researchers used an endoscope to take a biopsy from the uppermost section of the small intestine of study participants and examined the samples with a single optical instrument that performs 4-D elastic light scattering fingerprinting and low-coherence-enhanced backscattering spectroscopy. Both techniques were developed in the laboratory of Vadim Backman, the senior author of the study.

The optical instrument measures how cellular structures refract light from an intense white xenon lamp. The light refracts through the outermost layer of tissue and scatters into a spectrograph, and the detected light is analyzed by a computer to determine whether the cells are normal or cancerous.

In the study, biopsies were taken from 19 people previously diagnosed with pancreatic cancer and from a control group of 32 healthy people. The technique detected cancer with 95 percent sensitivity and 91 percent specificity, and it successfully diagnosed all 10 patients with early-stage cancer. The work is detailed in the Aug. 1 issue of Clinical Cancer Research.

While this study was done on biopsied tissue, the optical instrument can be turned into a portable fiber optic unit to be used during an endoscopy. In fact, the researchers have developed such an instrument for colon cancer screening, and it is undergoing clinical studies.

Using the same technique, the researchers would like to study patients with serious but noncancerous pancreatic diseases to discover the biological changes similar to those found in cancer of the pancreas. They also would like to perform studies on larger groups of people with pancreatic cancer using different types of control groups. If the technique proves successful, it could find clinical use in three to five years, providing doctors with a way to routinely screen patients at high risk for pancreatic cancer.