114-candidate gene panel can detect nearby prostate cancer in stromal prostate tissue

There has been a lot of media “chatter” over the past 3 days about an article published in the journal Cancer Research. The article suggests that an investigational genetic test can project the presence of prostate cancer in biopsy-based specimens with an average accuracy of 97 percent.

Now this article by Jia et al. is certainly interesting from a biologic point view, because what is really does is show that genetic tests in (stromal) tissues near to actual prostate cancer tissues can indicate the presence of prostate cancer. This finding would theoretically allow biopsy tissues to be genetically screened using the 114-candidate gene panel described by the authors. Such genetic screening could then be used to diagnose prostate cancer even in men who were suspicious for prostate cancer but who had had one or more negative biopsies.

On the other hand, of course, this implies that we may be able to find another 50,000 men a year (or perhaps more) with very early stage prostate cancer, many of whom will have low-risk disease and may not benefit in any way from invasive treatment. All of these men would need a biopsy to obtain sufficient tissue for genetic testing, and the authors offer no suggestion that their gene panel can help to differentiate between aggressive and indolent forms of prostate cancer.

We are increasingly going to be faced with situations like this, in which technical capabilities to carry out certain types of sophisticated test carry the possibility of clinical benefit, but where the actual clinical benefit (and the actual clinical risk) of the test may be completely unknown. This type of genetic profiling is by no means the first example of such a situation.

The question we are going to have to ask ourselves as a society — and it is a very serious question — is how one manages the use of such a test to maximize its value.

It would appear clear that this is a pointless test in anyone with a positive biopsy finding for prostate cancer. We already know that such a patient has prostate cancer, so there would be no point in running this test (or any test like it) in such a patient. However, what should we do with a 62-year-old African American male who has had two negative biopsies despite (say) a prior PSA of 3.9 ng/ml and a current PSA of 4.7 ng/ml, a negative physical exam, and a family history of prostate cancer?

According to the Prostate Cancer Calculator, this man has a 33.2 percent chance of biopsy-detectable prostate cancer — but we aren’t finding anything. Should every man like this be given this type of genetic test? Again according to the Prostate Cancer Calculator, this man’s risk for high-grade prostate cancer is only 5.9 percent.

It is only fair to the authors of this paper to note that they have published a remarkable piece of work. When they used their “gene classifier” to study tissues from a total of 364 prostate specimens –243 of which were known to be tumor-bearing samples and 121 of which were known to be normal specimens, the 114-gene panel was able to correctly identify 241/243  tumor-carrying specimens, which really is a very high degree of accuracy. However, scientific capability does not always translate well into clinical value, and The “New” Prostate Cancer InfoLink would like to see just how accurate this gene panel is when used prospectively to help make diagnostic decisions about (say) potentially high-risk men with a history of negative bi0psy findings before we rush into over-use of this test.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: