September 25, 2013 — When scientists and doctors disagree over the finer points of our medical care, the best the rest of us can do is talk with our own doctors, read about the debate so we are as informed as possible, and perhaps do a bit of advocacy to fund the research needed to solve the debate. In May of last year, one such debate erupted when a group of physicians and scientists convened by the U.S. government to make health care recommendations for the general public changed course from prior years and called for an end to the routine use of a blood test that screens men for prostate cancer.
Currently there is sharp disagreement over that recommendation within the medical profession. For example, a survey of doctors conducted by Johns Hopkins Medicine found “serious pushback” from primary care doctors over ending these routine cancer screens. Formal evidence of pushback against the governmental task force’s recommendation to end routine screening came in May of this year, with the American Urological Association’s issuance of their own set of guidelines that called for selective routine screening.
The test in question is called PSA for prostate-specific antigen; it measures the blood level of the PSA protein that is produced by the prostate gland in all men. As PSA levels rise, so do the chances a man has prostate cancer; but benign conditions can also cause high levels of PSA in the blood. This means that the PSA protein is tissue specific in nature—only the prostate gland can make the protein—but not cancer specific.
Because higher levels of PSA in the blood—generally greater than 4.0 ng/ml—are considered suspect for a cancerous prostate gland, most doctors have routinely recommended a needle insertion biopsy for men with PSA levels of 4.1 or greater to test, more definitively, for the presence of cancer. While less than half of those biopsies find cancerous cells, the cancers found are identified early, at a stage when cure is likely. (Research has also shown that 15% of men with PSA levels of 0 to 4.0 ng/ml have prostate cancer and of those men, 15% have high-risk disease, thus PSA is far from perfect at capturing clinically relevant cancers.) And even in men whose PSA levels are high enough to trigger a biopsy, accurate results are not assured: biopsy tests may miss some cancers (on average less than 1% of the prostate gland is sampled at biopsy). And because up to 44% of those PSA-induced biopsies find cancerous cells that are non-lethal—meaning they’d not shorten the lifespan of the man—many men are overtreated for prostate cancer.
Thus, overall, the PSA test has serious drawbacks. In essence, it leads to finding too many prostate cancers and, ironically, not enough.
From its inception in the mid 1980s, to its widespread use as a screening tool for asymptomatic men in the early 1990s, the PSA test has been incredibly effective at reducing the U.S. death rate from prostate cancer in much the same way routine Pap smears reduced the death rate of cervical cancer in women after the test’s mainstream introduction in mid 20th century. (A study published last August in the journal Cancer calculated that without widespread PSA testing in use today, the number of men diagnosed with advanced, metastatic prostate cancer would triple.) Now the challenge is to reduce the treatment rate of what doctors call indolent prostate cancers—those that do contain abnormal prostate cells, but whose abnormalities would likely never morph into aggressive, lethal, symptomatic forms of the disease—while simultaneously assuring that high-risk disease is not undertreated. One way to accomplish that would be to introduce a test that is prostate-cancer specific as opposed to prostate-tissue specific, with the aim of better informing patients and their doctors of the need for biopsy after PSA screening.
Sampling urine too
A new urine test has been developed that is based on two extremely prostate-cancer specific biomarkers and is now available to patients through the University of Michigan MLabs. The test measures two molecular markers that are distinct to prostate cancer and only prostate cancer. One marker is a measurement of RNA made from the PCA3 gene; prostate cancer cells churn out extremely high levels of this PCA3 RNA which can be detected in urine and more than 95% of all prostate cancers overexpress PCA3 RNA. (A recent study by the National Cancer Institute showed that a positive PCA3 test predicted a positive biopsy 80% of the time at initial biopsy; and, for men undergoing repeat biopsy, a negative urine test predicted a negative biopsy 88% of the time.) The second marker is a measurement of RNA made from the fusion of two genes—TMPRSS2 and ERG. Normally the ERG gene, a potent oncogene, is turned off in prostate cells and its protein is not expressed or produced. But in 50% of prostate cancer foci, the ERG gene fuses abnormally with the TMPRSS2 gene, which is located close by on the same chromosome—number 21. When this happens, ERG is turned on and high levels of a slightly shortened form of the ERG protein is made. When this aberrant fusion RNA is detected in men’s urine at high levels it is ultra specific for the presence of prostate cancer.
PCF-funded Scott Tomlins, MD, PhD, an assistant professor of pathology and urology at the University of Michigan and a co-discoverer in 2005 of what is now commonly known as the TMPRSS2:ERG fusion, says that evidence shows that if TMPRSS2:ERG is detectable at high levels in urine, a man likely has prostate cancer, whether or not his biopsy is positive for cancer. “If you wanted to design a way to cause prostate cancer, this is what you’d do: fuse a gene that is normally turned on in the prostate (like TMPRSS2) with an oncogene that is normally turned off (ERG),” says Tomlins. (In 2007, the Safeway Foundation provided unrestricted funding to the Prostate Cancer Foundation for biomarker research. The Foundation also provided PCF-Young Investigator funding for Dr. Scott Tomlins, who at the time was just finishing his PhD training at the University of Michigan. Also, it was Tomlins and colleagues, including Dr. Daniel Rhodes also of the University of Michigan, who developed a novel bioinformatics algorithm called the Cancer Outlier Profile Analysis, or COPA, that led to the discovery of the overexpression of ERG and the TMPRSS2:ERG fusion.)
In February of 2012, the FDA granted the California-based biotech company Gen-Probe approval to offer the PCA3 urine test to men who are considering repeat biopsy after an initially negative result. Its trade name is PROGENSA PCA3.
While that was a welcome milestone for men, research has shown that testing urine for both PCA3 and TMPRSS2:ERG levels is an even better way to stratify men suspected of having prostate cancer. Incorporating both PCA3 and TMPRSS2:ERG will significantly improve upon PSA testing as a means to predict if a man has prostate cancer, says Tomlins. It will be an especially important tool for men with PSA levels below 10, to determine if they may be able to delay invasive biopsy testing, and opt for a program of monitoring their PSA and TMPRSS2:ERG and PCA3 levels over time for signs of progression.
The combination test may also offer men additional information on the size of their tumors which can be an indication of how aggressive that tumor would become if not surgically removed.
Gen-Probe has collaborated with the University of Michigan and the National Cancer Institute—under an initiative of that institution (Early Detection Research Network, or EDRN—to determine the best way to utilize TMPRSS2:ERG in combination with PCA3 and other markers for better patient management. “The ongoing NCI/EDRN study will complement the efforts at University of Michigan by providing independent validation of urine test clinical utility for men with elevated PSA,” says Dr. Jack Groskopf, director of Oncology Research and Development at Gen-Probe. In addition to PROGENSA PCA3, the company is also developing a commercial urine test that targets both PCA3 and TMPRSS2:ERG.
Data sets on the TMPRSS2:ERG and PCA3 urine tests
In August of 2011, Dr. Tomlins as first author and Dr. Arul M. Chinnaiyan (Chinnaiyan is also PCF-funded) as senior author published a paper in Science Translational Medicine detailing the ability of urine testing for the TMPRSS2:ERG fusion (both singularly or in combination with PCA3) to stratify prostate cancer risk in men with elevated PSA blood levels.
The researchers found that testing urine levels of both TMPRSS2:ERG and PCA3 was more effective at predicting prostate cancer than testing blood PSA levels alone. Looking at urine samples for 1,312 men with elevated PSA levels who had gone on to have either a biopsy or surgery to remove their prostate gland, the researchers found the highest rates of cancer in men with the highest levels of TMPRSS2:ERG and PCA3 in their urine. The men in the study were stratified into three groups based upon the levels of TMPRSS2:ERG and PCA3 in their urine: low, intermediate and high levels, or scores. Cancer was diagnosed in each of the groups respectively: 21%, 43%, and 69%. High-grade prostate cancer, defined in the study as a Gleason score greater than 6, also occurred at different frequencies in the three groups with 7%, 20%, and 40% diagnosed in each group respectively. (Gleason scores are based upon a microscopic examination of biopsied prostate tissue; the higher the score the more likely a cancer is or will become aggressive.)
In addition, adding the urine test to a standard method now in employ to calculate prostate cancer risk—the Prostate Cancer Risk Calculator, a tool that combines a patient’s clinical and family history with PSA levels to estimate a man’s risk of developing prostate cancer—significantly improved its predictive ability.
And in a study published in the American Journal of Clinical Pathology, with Tomlins as senior author, the researchers demonstrated that the amount of TMPRSS2:ERG in urine samples from men suspected of having prostate cancer correlated highly with the size of their tumor (known as tumor burden) when removed at prostatectomy. (This applies only to tumors positive for the fusion.) “So if a man has lots of TMPRSS2:ERG in his urine, these men turn out to have a large fusion-positive tumor focus or high tumor burden, and in general, these larger tumors are the ones that are more aggressive,” says Tomlins. Their data also suggest that in men who have a high urine TMPRSS2:ERG score, but are found to have low tumor burden on biopsy may have been under-sampled during biopsy and might be well served by repeat biopsy.
Not the whole solution, but strong headway
Tomlins admits their combination urine test is far from the be-all-end-all in determining how aggressive a prostate tumor is or will become. (A study in Cancer Epidemiology, Biomarkers & Prevention reported finding no association with having a tumor positive for TMPRSS2:ERG and recurrence of the cancer after removal of the prostate, although the study only looked for the presence of the genetic marker, not how much the tumor had produced or whether it could be detected in urine. It also did not report on long-term outcomes in men with TMPRSS2:ERG positive tumors who deferred surgery.) Prostate Cancer Foundation President and CEO, Dr. Jonathan W. Simons, says that being positive for the gene fusion seems to act as a tipping point to prostate cells becoming cancerous but it’s not the later instigator of transition to aggressive, deadly disease. This is seemingly borne out by research that shows metastatic prostate cancer sites tend to be either uniformly TMPRSS2:ERG positive or TMPRSS2:ERG negative, thus the mutation doesn’t tend to occur after the initial development of the cancer and is unlikely to be necessary to seed metastatic cancer sites. In the urine test, TMPRSS2:ERG functions as a highly-specific marker of cancer cells shed after physical manipulation of the prostate gland via a digital rectal examination. Larger, more invasive tumors may shed more cells, and this could explain the association between higher urine TMPRSS2:ERG scores and high-volume, high-grade disease.
“This combination test is not designed to say definitively whether you have aggressive prostate cancer at diagnosis,” says Tomlins. Yet, the test, he says, can provide men with a more accurate estimate of the likelihood that they do in fact have cancer, and the likelihood that they have an aggressive cancer.”
Dr. Howard Soule, the chief science officer at PCF says, “The hope with this new urine test is that it will lead to a more informed decision tree for men and their doctors without upticks in detection of incidental prostate cancers.”
The department of Pathology at the University of Michigan, under the Michigan Center for Translational Pathology and MLabs, have just began offering the a combined test, termed Mi-Prostate Score (MiPS), that incorporates serum PSA, urine PCA3 score and urine TMPRSS2:ERG score to provide a patient with their individualized risk of having prostate cancer on biopsy. The formulas used to derive MiPS were developed and validated on over an additional 1,900 pre-biopsy urine specimens. The MiPS varies from 0 to 100, with the score equivalent to the risk of having cancer (i.e., a MiPS of 30 indicates a 30% risk of having cancer). A separate score is provided that indicates the risk of having high Gleason score prostate cancer (Gleason score > 6).
For example, a man may receive a report that gives him a 15% risk of having prostate cancer, and a 7% risk of that cancer being high grade. That man might then feel more comfortable forgoing an immediate biopsy, instead opting for a less invasive program of blood and urine monitoring over time for rising levels of PSA, TMPRSS2:ERG and PCA3. Conversely, another man whose report shows an 85% risk of having cancer and a 45% risk of having high grade cancer, may feel more confident that immediate biopsy is the right choice for him. It is entirely feasible that both men might have started out with the same PSA level.
In an editorial that accompanied their study in Science Translational Medicine, the author wrote: “Here, Tomlins et al. improve on the PSA test by taking a new twist on a known gene fusion….demonstrating more accurate, individualized stratification of men at high risk for developing clinically significant prostate cancer.” The editorial concludes that the combination TMPRSS2:ERG + PCA3 urine test may help men and their doctors better estimate how urgently a follow-up biopsy is needed after a suspect PSA test.
Dr. Simons of the Prostate Cancer Foundation says, “To our knowledge and from our perspective, this could be the best new use of urine for diagnostics since the pregnancy test.”
The combined MiPS test is now available from the University of Michigan MLabs to provide patient-specific prostate cancer risk assessment. For questions on How to Send a Specimen, please call MLabs at 800-862-7284. You may also visit us at www.mlabs.umich.edu
PDF of PCFs comments in response to the U.S. Preventive Services Task Force’s (USPSTF) request to the Prostate Cancer Foundation to review the draft Recommendation Statement against prostate-specific antigen (PSA)-based screening for prostate cancer.