Public Release: 

Protein patterns may aid prostate cancer detection

University of North Carolina Health Care

CHAPEL HILL - A test that detects a specific pattern of proteins in blood may distinguish benign prostate conditions from prostate cancer more effectively than the current biomarker for the disease: protein specific antigen, or PSA.

According to a report in Wednesday's (Oct. 16) issue of the Journal of the National Cancer Institute, the test detected 95 percent of prostate cancer cases from a single drop of blood from each patient. Moreover, it helped rule out prostate cancer for 71 percent of men with intermediate PSA scores (four-10), which would have allowed them to avoid an unnecessary, invasive biopsy procedure. Currently, most men with PSA scores between four and 10 are recommended for a biopsy, even though 75 percent to 80 percent of them do not have prostate cancer.

"This new technology has the potential to revolutionize how men are diagnosed with prostate cancer," said Dr. David K. Ornstein, a co-author of the study, assistant professor of surgery and a member of the Lineberger Comprehensive Cancer Center at the University of North Carolina at Chapel Hill.

"It's likely that it will be possible to use a simple blood test to accurately identify men who are affected with a harmful prostate cancer but spare healthy men from undergoing unnecessary biopsies."

In the new study, Ornstein and collaborators from the U.S. Food and Drug Administration- National Cancer Institute Clinical Proteomics Program and Correlogic Systems Inc. in Bethesda, Md., used similar artificial intelligence bioinformatics technology to that employed in a recently reported ovarian cancer test.

Here the method was used to analyze serum samples from men with prostate cancer and men with benign prostate conditions.

The researchers identified a proteomic pattern, or "computational disease model," that best discriminated between these conditions. The pattern was then tested to determine how well it predicted disease in 266 men known to have either prostate cancer or benign prostate disease.

The test correctly identified 36 of the 38 (95 percent) cases of prostate cancer. Among men with normal PSA levels, it correctly identified 70 of 75 (93 percent) asymptomatic men who had benign prostate conditions. Among men whose PSA levels were in the indeterminate range, the tool correctly classified 97 of 137 (71 percent) men as having benign conditions. The test also correctly classified 10 of 16 (63 percent) men with high PSA levels and a negative biopsy as having benign conditions. While the new findings provide further validation that the protein pattern approach can be effective in cancer detection, Ornstein and his co-authors added that it cannot replace biopsy as a definitive cancer detection tool. However, they said, "serum proteomic pattern analysis may be used in the future to aid clinicians so that fewer men are subjected to unnecessary biopsies."

In collaboration with Correlogic Systems Inc. and the FDA-NCI Clinical Proteomics Program, the UNC Lineberger Comprehensive Cancer Center is conducting a clinical trial to study if this new technology can be used determine the need for prostate biopsy among men with elevated PSA tests or abnormal digital rectal exams. "We're attempting to refine and further evaluate this new tool. The goal is to develop a clinically useful test for the benefit of patients," Ornstein said.


Note: Contact Ornstein at (919) 966-2572 or (919) 966-9251 or
Lineberger Center contact: Dianne Shaw, (919) 966-5905 or
School of Medicine contact: Les Lang, (919) 843-9687 or

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.