Is the Decipher test really as useful as it’s manufacturer seems to claim?

A newly published article in the journal PLoS One questions whether the Decipher® test (a 22-gene expression assay test) has actual clinical utility in helping physicians to make good decisions about the follow-up treatment of men with prostate cancer after an initial radical prostatectomy.

According to the developers of the Decipher test (in statements on the GenomeDx Biosciences corporate web site):

Decipher is a transformative genomic test that informs decisions about the course of treatment for men after prostate surgery.

The Decipher test predicts the probability of metastasis after surgery and provides an independent assessment of tumor aggressiveness; information distinct from that provided by Gleason score or PSA.

The new paper by Marrone et al. (available on line as a full-text article) is very clear that the Decipher test does indeed have

good discriminatory ability (e.g., ability to predict clinical metastasis) in both the discovery and validation studies

based on the data available to date. In other words, yes, the Decipher test does appear to be able to accurately predict risk for metastasis among radical prostatectomy patients at 5 years post-surgery. However, this risk can also be projected with relatively high accuracy through the use of nomograms based on the patients’ pre-surgical PSA levels and post-surgical data such as pathological Gleason scores and extent of disease.

Marrone et al. are equally clear that:

No direct evidence for clinical utility for the genomic classifier is currently available showing changes in selection of primary or adjuvant treatment based on the genomic classifier or showing that the genomic classifier can successfully predict treatment outcomes.

In other words, there is no good evidence that the results of the Decipher test offer any meaningful value as to the implementation and outcomes of adjuvant or other clinical treatment of patients following surgery.

If new tests like the Decipher test — which are expensive — are unable to provide better information about how to treat men after surgery than the information already available, and there are no reliable data demonstrating that decisions made through the use of Decipher-like tests can actually change the outcomes of such patients as a consequence of changes in their treatment, then this really does raise questions about the value of such tests. And the only way to truly establish whether tests like the Decipher test do have real value is through the use of randomized clinical trials.

It is worth remembering, in this context, that laboratory-developed tests like the Decipher test, which can be carried out only by specialized laboratories licensed under the Clinical Laboratory Improvement Act (CLIA), do not have to be approved for clinical use by the Food and Drug Administration (at least here in the USA). The quality of the data used to promote and market such tests is therefore not based on the same levels of evidence as the data used to promote other, more widely used tests that can be carried out at a wide variety of clinical laboratories based on materials provided by the tests’ manufacturers.

As Marrone et al. also point out, what we need is a test that really can tell us, immediately after surgery, which men will have better outcomes if they have immediate adjuvant therapy and which men will do just as well (or better) if they are simply monitored.

Currently, for each 1,000 men undergoing a radical prostatectomy in America each year

  • About 190 to 300 will have biochemical recurrence (i.e., a PSA level that rises to > 0.2 ng/ml) within 5 to 10 years after surgery.
  • Of these 190 to 300 men,
    • About 70 to 110 will develop distant metastasis within 5 years if not treated.
    • About 30 to 50 will die from prostate cancer within 6 years of biochemical recurrence.

If they are to have high clinical value, tests like the Decipher test need to be able to tell us three things with a high level of accuracy:

  • Which of those 1,000 men are at high risk for metastatic disease
  • Whether early treatment will actually lower their risk for distant metastasis
  • Whether early treatment will extend their long-term survival

We can already make good assessments of the first of these three risks, without the need for genetic/genomic testing. It is the second and the third of these risks that we need help with.

Leave a Reply

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

You are commenting using your 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: