Can the Decipher test accurately project risk for metastasis post-surgery?

A newly published article in the Journal of Clinical Oncology suggests that the Decipher test may be able to predict risk for metastatic prostate at 10 years post-surgery. However, there are serious questions about whether the use of the Decipher test in this way is really “ready for prime time” in actual clinical practice.

The abstract of the original article be Spratt et al. as well as a commentary on the MedPage Today web site are both available on line. So too is the full text of an editorial comment in the Journal of Clinical Oncology by Dr. Anthony D’Amico.

The bottom line seems to be this:

  • A meta-analysis of data from five different studies was carried out by Spratt et al.
  • The five studies included 975 patients and data from 855 of those patients were suitable for analysis.
  • The average (median) follow-up on the 855 patients was 8 years.
  • As classified by the Decipher test, the 855 patients could be categorized as follows:
    • Low risk (a Decipher score of < 0.45), 60.9 percent or about 521 patients
    • Intermediate risk (a Decipher score of 0.45 to 0.60), 22.6 percent or about 193 patients
    • High risk (a Decipher score of > 0.60), 16.5 percent or about 141 patients
  • The incidence of metastatic disease at 10 years of follow-up was
    • 5.5 percent among the Decipher low-risk patients
    • 15.0 percent among the Decipher intermediate-risk patients
    • 26.7 percent among the Decipher high-risk patients
  • The overall (“pooled”) hazard ratio (HR) for metastatic disease across all five studies, based on the original Decipher score was 1.52 per each 0.1 unit increase in the Decipher score (which is certainly statistically significant).

However, it should be noted (as Dr. D’Amico carefully points out) that the authors state, in the conclusion to their article, that:

Future study as how best to incorporate genomic testing in clinical decision making and subsequent treatment recommendations is warranted.

Dr. D’Amico is in complete agreement with that comment, and he raises numerous careful questions about detailed aspects of this study suggesting the need for great caution before anyone decides that all we need is a Decipher score to project long-term risk for metastasis after a radical prostatectomy, including the obvious one of how many of the patients had any type of adjuvant or salvage therapy after their surgery and before it was determined that they were metastatic.

And then, of course, it should be noted that 60.9 percent of these patients (about 520 out of 855) had Decipher low-risk disease — and so immediate radical prostatectomy would not be a recommended first-line treatment for many of these men today. By contrast, only 16.5 percent (about 141 out of 855) had high-risk disease, which is a relatively small percentage of the entire database.

In addition (and perhaps most importantly), among the men enrolled in the study, based on post-surgical data, 222 (26.0 percent) had a Gleason score of 8 to 10; 238 (27.8 percent) had seminal vesicle invasion, and 48 (5.6 percent) had lymph node invasion. In other words, at least 238/855 patients (and maybe more) were high-risk patients based on standard, post-surgical indicators of risk, which is far higher than the 141/855 patients classified by their pre-surgical Decipher score.

The “New” Prostate Cancer InfoLink agrees with Dr. D’Amico that, while there is certainly good reason to try to determine whether genomic tests like the Decipher test can be used to project risk for metastatic disease after first-line treatment (particularly in intermediate- and high-risk patients), the data from this study can’t be used to justify the clinical use of the Decipher test in this manner yet. We are going to need more and better data — preferably from prospective clinical trials.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: