When the PSA Test Can't Do It All
New Biomarkers To Fill in the Gaps
Carter: PSA doesn't always tell the whole story- so finding one PSA level that's "normal" for every man is impossible.
Low PSA-anything under 4.0-is good, right? The answer is, most of the time, but not always. And this prompts the next question: Would a lower PSA cutoff find more cancer? Yes, undoubtedly. But not all of these cancers would need to be treated, and some men with insignificant, slowgrowing disease would wind up getting prostate biopsies they don't need.
However, Carter doesn't believe a lower PSA cutoff will solve the problem. One reason, he says, is that PSA doesn't always tell the whole story-so finding one PSA level that's "normal" for every man is impossible. "High-grade cancers actually produce less PSA per gram of tissue than low-grade cancers." For this reason, even though higher- grade cancers are often larger in volume than low-grade cancers, the PSA level does not accurately indicate the volume of the cancer.
Also, as many as 75 percent of men with a PSA higher than 4 who get a biopsy turn out to have BPH, notes Daniel W. Chan, Ph.D., professor of pathology, oncology, urology, and radiology. "PSA is by no means perfect," he concurs. Even with more sophisticated PSA tests-of free or complex PSA- "it gives us some help in deciding who needs a repeat biopsy. However, with free PSA, less than 10 percent is considered high-risk, and greater than 25 percent is considered lowrisk. What happens if your free PSA is between 10 and 25 percent? This is another diagnostic gray zone."
Biomarkers to the Rescue
Instead, say Carter and Chan, what's needed are new biomarkers-new ways to detect cancer earlier, and determine the need for biopsy. "A biomarker," explains Robert Veltri, Ph.D., associate professor of urology, "is a cancer property that can be objectively measured, and used to monitor someone's disease. It can also tell us how well a particular therapy is working." Veltri, Chan, and several Brady scientists are exploring many promising new biomarkers, not only as means of improving diagnosis, but of predicting a man's stage of cancer and his risk of recurrence, and of sounding the earliest possible alarm that cancer has spread, so that it can be treated while the disease is most vulnerable.
Chan and colleagues are using a sophisticated computerized proteomic technology- minuscule chips of protein, added to just one drop of a man's blood. The protein chip acts like a tiny magnet, attracting all the proteins in that blood drop to stick to its surface. This process is a bit like dangling a hook into the water, and seeing what bites. In this case, Chan is fishing for proteins. A sophisticated computer technology called Mass Spectrometry allows Chan to see what he caught; each protein has its own characteristic "signature" of peaks and valleys.
|The protein chip acts like a tiny magnet, attracting all the proteins in the blood drop to stick to its surface. It's a bit like dangling a hook into the water, and seeing what bites.|
It also shows Chan what he didn't catch-any proteins that are missing, or mangled, or changed, are of interest, as well.
Chan's next job is to filter out the background noise, to find the real molecular signature of cancer-which may be different in each man, because prostate cancer itself is so complex, and made up of so many different types of cells and genetic configurations. "We want to decide which of these peaks are real proteins, and which have clinical value in terms of early detection of prostate cancer." He and colleagues at Hopkins have developed a set of "bio-informatic" tools- new ways to analyze all of this information.
|"High-grade cancers actually produce less PSA per gram of tissue than low-grade cancers."|
When he finds likely-looking proteins, he purifies them, sequences them, and determines their molecular fingerprint. Because prostate cancer is so variable from man to man, Chan isn't looking for one key suspect, but a whole gang of them-a panel, or a multiplex, of biomarkers. He then wants to test these biomarkers in as many patients as possible, on patients at Hopkins and centers worldwide. He also believes this technology will produce new tests that can help diagnose early recurrence of cancer.
Some biomarkers under study include:
- Prostate-breast overexpressed gene-1 (PBOV1): Veltri and colleagues discovered that this gene is, as its name suggests, overproduced in prostate and breast cancers. He is working to develop a new blood and biopsy test that can help determine if cancer is present, and whether it is likely to need treatment.
- GSTP1: Veltri and Partin are evaluating a test to measure GSTP1, which helps the body fight oxidative damage to DNA in cells. This test may be useful for men who have what appears to be prostate cancer, based on a rectal exam, but cancer-free biopsies.
- NMP44: Partin and Chan are working on a test to detect NMP44, a protein that binds with Vitamin D.