Prolaris: prognostically accurate but of limited utility?


According to a media release from  Myriad Genetics, the company’s investigational genetic test (Prolaris™) can accurately predict prostate cancer-specific mortality at 10 years in men known to have clinically localized prostate cancer. This claim  is being made on the basis of data presented earlier this week at the European Society for Medical Oncology (ESMO) annual meeting in Milan, Italy.

Cuzick and his colleagues used the Prolaris genetic test to predict prostate cancer-specific mortality in a cohort of 337 patients who were diagnosed with clinically localized prostate cancer between 1990 and 1996. All diagnoses were made on the basis of pathological examination of “chips” of prostate tissue removed during transurethral resection of the prostate (TURP). These men were all managed “conservatively,” implying that they received no treatment until they required hormone therapy for symptomatic or evident metastatic prostate cancer. Since they were followed over time, their actual prostate cancer-specific mortality is known, allowing confirmation of the predictions of the Prolaris test. Unfortunately, the actual abstract of the presentation by Cuzick is only available to attendees of the ESMO meeting, so we have not been able to see this abstract for ourselves.

The Prolaris test is based on a panel of 46 genes, most of which are involved in cell cycle progression and cell growth. By measuring the capacity of prostate cancer tissue’s ability to divide and grow, the test can be used to assess whether a patient has a relatively indolent or a relatively aggressive form of prostate cancer using standard prostate tumor tissue of the type available to pathologists after TURP or after radical prostatectomy.

Data from the study by Cuzick et al. (“Prognostic value of a cell cycle expression profile score among men with conservatively treated localized prostate cancer”) showed that:

  • Prolaris was capable of accurate prediction of prostate cancer-specific mortality with very high statistical significance (P = 1 x 10-21).
  • 25 percent of patients in this cohort had a “favorable” Prolaris score and  a 98.5 percent probability of surviving their cancer over a 10-year period.
  • 25 percent of patients in this cohort had an “unfavorable” Prolaris score and a 42.4 percent probability of surviving their cancer over a 10-year period.
  • Prolaris was superior to the Gleason score, baseline PSA levels, and other prognostic factors in predicting prostate cancer-specific mortality.

In theory, these data suggest that the Prolaris test should be capable of discriminating between good and poor candidates for active surveillance. However, there may be a fly in the ointment. As yet we are not aware of any data showing that the Prolaris test can make accurate predictions based on tissue from an initial prostate biopsy, but such biopsy-derived tissue is what is available to the pathologist in the vast majority of cases prior to a decision whether to treat or to monitor individual patients. If the Prolaris test can only be used with accuracy on tissue available after a TURP or a radical prostatectomy, this will severely curtail the potential utility of this test.

The “New” Prostate Cancer InfoLink commented on the Prolaris test when it was first introduced, immediately prior to the annual Genitourinary Cancers symposium (in San Francisco earlier this year). At present we continue to take the position that validation of the utility of the Prolaris test will require a clear demonstration that the test can accurately predict prostate cancer mortality based on prostate biopsy tissue. If the test can be used to do that, then its value may well be considerable (although whether the insurance industry will be willing to pay the current asking price is still open to question).

4 Responses

  1. The abstract of the presentation by Dr. Cuzick at ESMO is available to the public as Abstract no. 1580 in the session on “Biomarkers” in the Annals of Oncology Vol. 21 (Supplement 8): viii63–viii77, 2010.

  2. Thanks for this additional information Mr. King.

  3. Have any of you examined claims made by the Aureon company for their Prostate Px test regarding their proprietary formula for predicting high- and low-risk prostatic adenocarcinoma based upon morphometric analysis of tumors using immunofluorescence, etc., and analyzing such factors as nuclear polarity, nuclear size, etc? … Similar claims to Polaris and possibly less transparency.

    It would be helpful to see your analysis of their claims. Their papers are written almost entirely by company employees — and very aggressively marketed in the New York area where they are based.

  4. Dear Dr. Rubin:

    We don’t have the capacity to conduct what I would describe as an “analysis” of these claims. Having said that, I can add that I have yet to see any data that I would consider to be scientifically or clinically compelling regarding the value of this test.

    Certainly there has been no prospective, randomized trial that would allow anyone to suggest that this test is any more specific or selective than a PSA test in the real world. And a number of specialists (on a “not-for-attribution” basis) have suggested to me that the Px test is not something they would recommend to any patient. Interestingly, no patient has asked us for an opinion on this test — although there have been a few who believe it was of benefit to them.

    Also … as far as I am aware … Aureon Laboratories has made no attempt to seek FDA approval to market this test, which may place them at risk from evolving FDA guidance about direct marketing of diagnostic tests to patients.

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 )

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: