There is a lot of media coverage from late yesterday about a “new” urine test for diagnosis of prostate cancer that is being developed by the University of Michigan in combination with San Diego-based Gen-Probe. Whether the actual value of the test can live up to the media coverage is going to take a while to determine.
The media coverage is based on an article by Tomlins et al., published on line yesterday in Science Translational Medicine and on a media release on the web site of the University of Michigan Health System. Commentaries have appeared in everything from USA Today and the MSNBC web site to the Los Angeles Times Booster Shots blog and MedPage Today.
The new test is actually based on combining data from a urine sample which is tested for elevated levels of prostate cancer gene 3 (PCA3) and for the presence of fusions between two genes: the transmembrane protease, serine 2 (TMPRSS2) gene and the v-ets erythroblastosis virus E26 oncogene (ERG). This gene fusion is commonly known as TMPRSS2:ERG. More than half of all prostate cancers historically found by PSA testing are found to have such gene fusions.
So what did Tomlins and his colleagues actually do and show?
They prospectively collected whole urine from 1,312 men with elevated PSA levels who were scheduled for a biopsy at a number of different academic and community medical centers. (We assume that this urine had to be collected after vigorous prostate massage, since that is an essential component of the collection of urine for the PCA3 test.) They then tested this urine for its PCA level and for the presence of the TMPRSS2:ERG gene fusion, and they used these data, along with other data normally used in the Prostate Cancer Prevention Trial risk calculator, to project the risk that the patient would actually have prostate cancer. Finally, they compared their projected results to the actual results found for these patients on a prostate biopsy.
They claim that their results show the following:
- The presence of TMPRSS2:ERG in the urine sample, in combination with PCA3 data, improved the performance of the multivariate Prostate Cancer Prevention Trial risk calculator in predicting cancer on biopsy.
- Men in the highest and lowest of three TMPRSS2:ERG + PCA3 score groups had distinctly different rates of cancer, of clinically significant cancer by the Epstein criteria, and of high-grade cancer on biopsy.
- The presence of TMPRSS2:ERG in the urine sample, in combination with urine PCA3 levels, enhances the utility of serum PSA for predicting prostate cancer risk and clinically relevant cancer on biopsy.
The authors state that these data allowed them “to reliably predict whether the patient had prostate cancer and suggested which men were at higher risk of aggressive disease.”
However … there are a lot of cautions that come along with the data from this study.
In the first place, the authors themselves have pointed out that most of their test subjects were Caucasians, so whether this test is any more accurate in African-Americans or other racial groups is completely unknown at this point in time.
Secondly, this test is probably going to allow doctors not to give a patient a biopsy (yet, on the basis of the currently available data). Again, to quote the authors (see the commentary on HealthDay), “there is always a chance you are going to miss a cancer that doesn’t have either of these two markers.” In the longer term, it is possible that some could be followed with such a test under active surveillance without the need for a biopsy, but The “New” Prostate Cancer InfoLink believes that such a possibility will need a great deal more data before this strategy could become widely accepted.
Asked for his expert opinion on the potential value of this new test, Dr. Anthony D’Amico (from Harvard Medical School and Brigham and Women’s Hospital in Boston) is quoted as saying that this testis “a step forward, but we still have a ways to go.” He indicated that what this test tells us is that “On average the risk is higher in people with both markers and lowest in people who have neither, but that’s on average.”
The bottom line is that (currently) this test does provide some more information about which men with an elevated PSA level are at greater risk for both prostate cancer and clinically significant prostate cancer. However, the clinical utility and accuracy of that information in the assessment of risk for an individual patient may be relatively small. We are going to need a lot more data before the use of a test like this becomes compelling for the vast majority of men with a PSA level somewhere between 2.5 and 10 ng/ml. Is it, for example, any more useful than a much simpler %free PSA test? As yet, we just don’t know.
The research team currently plans to make this test available “soon” at the University of Michigan’s prostate cancer center.