Perhaps no other cancer screening test has been so well-known—and more recently, criticized—as prostate-specific antigen (PSA). The limitations of PSA as a screening tool are now widely acknowledged, despite the fact that it almost certainly has decreased prostate cancer mortality and that urologists and oncologists rely heavily on it for managing care. Prostate cancer incidence peaked in the early 1990s in the U.S., yet prostate cancer mortality decreased over the same decade. Most experts attribute those changes to the introduction of large-scale PSA screening that detected tumors earlier.
“By 1989, prostate cancer became the most common cancer diagnosed in men in the United States,” noted E. David Crawford, MD, head of Urologic Oncology at the University of Colorado Medical Center. “We were finding a lot of prostate cancers, but we were also finding a lot that we didn’t need to find.”
Now, 30 years after the Food and Drug Administration (FDA) approved the test, several new tests for prostate cancer detection and management have won regulatory sanction and many more are in the pipeline. The fruit of decades of prostate cancer research are coming at a time when both patients and providers appear ready for a change in strategy: data published in December 2015 showed that PSA screening dropped 18% from 2010–2013, coinciding with a strong recommendation against PSA screening from the United States Preventive Services Task Force (USPSTF) (JAMA 2015;314:2054–61).
From Prize to Pariah
The problem with PSA as a screening test is that it is specific for prostate-related issues, not prostate cancer. This means raised PSA levels could indicate benign prostate conditions, such as infections. FDA initially approved the use of the PSA test in 1994 in conjunction with a digital rectal exam (DRE) to test asymptomatic men for prostate cancer. But by 2008, experts started calling for limiting the use of PSA testing out of concern it was doing as much harm as good. That year, USPSTF recommended against PSA screening for men older than age 75. In 2012, USPSTF issued a more unfavorable recommendation: regardless of age, men without symptoms should not be screened routinely.
“We get a lot of false positives with PSA,” said Marc Garnick, MD, Gorman Brothers clinical professor of medicine at Harvard Medical School and an internationally renowned expert in medical oncology and urologic cancer. “But PSA is very useful in men who have a known diagnosis of prostate cancer. It can help guide treatment if there is a marked elevation.”
After an abnormal PSA screening result, biopsy is usually the next step, which comes with risks of pain, infection, and bleeding. Many biopsies are unnecessary, but with traditional testing methods, it has been almost impossible to determine who should or should not get a biopsy, according to Ashley Ross, MD, PhD, assistant professor in the Departments of Urology, Oncology, and Pathology at Johns Hopkins School of Medicine in Baltimore. “With a PSA threshold of around 4 ng/mL, positive biopsies could range from about 17 percent to 40 percent positive.”
Neither does PSA distinguish between clinically important and unimportant cancers. “There has been a problem of overdiagnosis with prostate cancer. Many patients with clinically unimportant cancers are getting treatment,” said Andrew Stephenson, MD, director of the Center for Urologic Oncology at Cleveland Clinic. “But now there are better biomarkers to help identify patients who should undergo prostate biopsy in the setting of an elevated PSA and those who should probably forego the biopsy.”
In 1998, FDA approved the free PSA test for use in men with a PSA between 4 and 10 ng/mL. The biggest difference between this test and its precursor is that it measures the percentage of unbound PSA; the traditional test measures the total of both free and bound PSA. The percentage of free PSA can be used to estimate how likely it is that a biopsy will show cancer.
For more than a decade, no additional screening tests existed. But FDA approval of Beckman Coulter’s Prostate Health Index (phi) in 2012 was a turning point for new prostate cancer tests. The phi test is 2.5 times more specific in detecting prostate cancer than the traditional PSA test in patients with PSA values in the 4–10 ng/mL range, according to the company. It combines in one score total PSA, free PSA, and the p2PSA isoform.
Now, a range of improved tests and biomarkers are available or in development, including Exo106, ConfirmMDx, ProMark, PCA3, Oncotype DX, Prolaris, Decipher, and 4Kscore. Developed by OPKO Lab, 4Kscore is the only test on the market that assesses a patient’s risk for aggressive prostate cancer prior to a prostate biopsy. The test measures blood plasma levels of four different prostate-derived proteins: total PSA, free PSA, intact PSA, and Human Kallikrein 2 (hK2). These are combined with a patient’s age, DRE status, and prior biopsy status using an algorithm to calculate the probability of finding a Gleason Score 7 or higher, indicating cancer.
In September 2015, results from a multi-institutional prospective trial showed the test could possibly reduce biopsies by 30–58%, with a 0.82 area under the curve for predicting aggressive cancer (Eur Urol 2015;68:464–70).
“When new biomarkers like 4KScore and Prostate Health Index are done as adjuncts with the PSA, we can significantly reduce the number of biopsies that are performed while not missing any clinically important cancers,” Stephenson said.
Meanwhile, Exosome Diagnostics is working on the first exosomal RNA test for prostate cancer, called EXO106. It is not FDA-approved, but clinical trial results are promising. In June 2015, researchers presented data at the annual meeting of the American Urological Association showing that, in a study led by New York-Presbyterian Hospital and Columbia University Medical Center, this urine-based test had a 98.6% negative predictive value for high-grade prostate cancer. The study looked at 1,600 men older than age 50 without a history of prostate cancer at 25 urology practices across the country (Rev Urol 2015;173:179–89).
Researchers are also looking into combinations of protein biomarkers and genetic markers. The large Stockholm 3 study found that combining six protein biomarkers—including PSA, free PSA, intact PSA, hK2, MSMB, and MIC1—with more than 200 genetic polymorphisms could reduce the total number of biopsies by 32% and could avoid 44% of benign biopsies (Lancet Oncol 2015;16:1667–76). The study, conducted between 2012–2014, included 58,818 men from Stockholm ages 50–69.
The Stockholm 3 investigators also reported that this so-called STHLM3 test also found aggressive cancers in men with low PSA values of 1–3 ng/mL, cancers that currently may go undetected. “These are indeed promising results. If we can introduce a more accurate way of testing for prostate cancer, we’ll spare patients unnecessary suffering and save resources for society,” said principal investigator Henrik Grönberg, MD, PhD, professor of cancer epidemiology at Karolinska Institutet in a statement. “The STHLM3 test will be available in Sweden in March 2016 and we will now start validating it in other countries and ethnic groups.”
What role genomic testing—still in its infancy—will have on patient outcomes is unclear. “There are some genetic tests that lead to actionable mutations,” Garnick said. “For example, genetic tests that detect genes that switch on mutations that lead to cancer. That’s where the home run is.” But so far, this type of advance is happening with colon cancer, lung cancer, and melanoma, not yet in the prostate cancer field.
Genomic tests eventually could help determine the best courses of treatment for patients with certain genetic mutations for which drug treatments are available. But that’s the other current limitation of genomic testing: drug discovery and development of targeted cancer therapies is playing catch-up. And new drugs require new drug targets.
“Unfortunately, in prostate cancer we have a lot of genomic tests, but we don’t really have a target for treatment, with the exception of the Androgen Receptor Splice Variant 7,” Garnick said. “This receptor predicts who will or will not respond to treatment. If you have a mutation, you are less likely to respond to treatment.”
Some therapies for prostate cancer can be expensive, so genomic testing can help select therapeutic choices for those most likely to benefit from them.
As more tests are developed and gain approval, and as genomic testing comes of age, experts are confident that finding and diagnosing medically important prostate cancers will become more accurate. Crawford is optimistic about the current flurry of innovation. “This is an exciting time,” he said. “We have a lot of new promising markers."
*Correction: A previous version of this article stated that the 4Kscore was FDA approved. The test is not currently FDA approved.