When a prostate-specific antigen (PSA) blood test produces an abnormal result, the next step is usually a prostate biopsy. A biopsy can confirm or rule out a cancer diagnosis, but it also has certain drawbacks. Prostate biopsies are invasive procedures with potential side effects, and they often detect low-grade, slow-growing tumors that may not need immediate treatment — or any treatment at all.
Researchers are exploring various strategies for avoiding unnecessary biopsies. Specialized magnetic resonance imaging (MRI) scans, for instance, can be useful for predicting if a man’s tumor is likely to spread. A blood test called the Prostate Health Index (PHI) measures various forms of PSA, and can help doctors determine if a biopsy is needed.
In April, researchers at the University of Michigan published results with a test that screens for prostate cancer in urine samples. Called the MyProstateScore 2.0 (MPS2) test, it looks for 18 different genes associated with high-grade tumors. “If you’re negative on this test, it’s almost certain that you don’t have aggressive prostate cancer,” said Dr. Arul Chinnaiyan, a professor of pathology and urology at the University, in a press release.
Gathering data and further testing
To create the test, Dr. Chinnaiyan and his colleagues first turned to publicly-available databases containing over 58,000 prostate cancer-associated genes. From that initial pool, they narrowed down to 54 genes that are uniquely overexpressed in cancers classified as Grade Group 2 (GG2) or higher. The Grade Group system ranks prostate cancers from GG1 (the least dangerous) to GG5 (the most dangerous).
The team tested those 54 genes against archived urine samples from 761 men with elevated PSA who were scheduled for biopsy. This effort yielded 18 genes that consistently correlated with high-grade cancer in the biopsy specimens. These genes now make up MPS2.
Then the team validated the test by performing MPS2 testing on over 800 archived urine samples collected by a national prostate cancer research consortium. Other researchers affiliated with that consortium assessed the new urine test’s results against patient records.
Interpreting the results
Study findings showed that MPS2 correctly identified 95% of the GG2 prostate cancers and 99% of cancers that were GG3 or higher. Test accuracy was further improved by incorporating estimates of the prostate’s size (or volume, as it’s also called).
According to the team’s calculations, use of the MPS2 would have reduced unnecessary biopsies by 37%. If volume was included in the measure, then 41% of biopsies would have been avoided. By comparison, just 26% of biopsies would have been avoided with the PHI.
Dr. Chinnaiyan and his co-authors emphasize that ruling out high-grade cancer with a urine test offers some advantages over MRI. The specialized multi-parametric MRI scans needed to assess high-grade cancer in men with elevated PSA aren’t always available in community settings, for instance. Moreover, the interpretation of mpMRI results can vary from one radiologist to another. Importantly, the MPS2 can be updated over time as new prostate-cancer genes are identified.
Commentary
Dr. Boris Gershman, a urologist at Harvard-affiliated Beth Israel Deaconess Medical Center in Boston, and a member of the advisory and editorial board for the Harvard Medical School Guide to Prostate Diseases, described the new study results as promising. “It does appear that the performance of the 18-gene urine test is better than PSA alone,” he says.
But Dr. Gershman adds that it will be important to consider how such a test will fit into the current two-stage approach for PSA screening, which entails prostate MRI when the PSA is abnormal. Where MRI delivers a yes/no result (meaning lesions that look suspicious for cancer are either present or not), the MPS2 provides numerical risk estimates ranging between 0% and 100%. “The challenge with clinical implementation of a continuous risk score is where to draw the line for biopsy,” Dr. Gershman says.
“This research is very encouraging, since many men in rural areas may not have access to prostate MRI machines or the added sophistication that is needed in interpreting these MRI scans,” says Dr. Marc Garnick, the Gorman Brothers Professor of Medicine at Harvard Medical School and Beth Israel Deaconess Medical Center. “A widely available urine test may eventually help provide more precision in determining who should undergo a prostate biopsy, and may also help to assess the probability that a cancer is clinically significant and in need of treatment.”