More data on accuracy of the PCA3 test in the real world

The prostate cancer gene 3 or PCA3 test has been approved exclusively as a means to assess the potential value of a repeat biopsy for a man who has already had a negative biopsy but whose PSA values (and/or other indicators) continue to suggest significant risk for prostate cancer.

We would emphasize that the PCA3 test has not been approved for indication of risk prior to a first biopsy, and we know of no good data to suggest that it has clinical value in that setting.

Since its original approval, there have been numerous indications of the relatively limited clinical value of the PCA3 test — even in its approved indication.

Italian clinical researchers (Barbera et al.) recently set out to evaluate the accuracy of the PCA3 test in its approved indication in a reasonably large cohort of men tested at their institution in Sciacca, Italy.

Between January 2010 and March 2012, they studied a sequence of 177 patients who had had a negative initial biopsy but any one of the following indications for re-biopsy:

  • A PSA value of > 10.0 ng/ml
  • A PSA value of between 4.1 and 10.0 ng/ml and a %free PSA value of < 20%
  • A PSA value of between 4.1 and 10.0 ng/ml and a %free PSA value of < 25%

Patients who met any one of these criteria were (a) given a PCA3 test and (b) given a repeat (saturation) biopsy. The investigators then analyzed the resulting data to assess the degree to which data from the PCA3 test results correlated with the results from the saturation biopsies. (The abstract of the paper does not tell us exactly how many biopsy cores were taken in the saturation biopsies, but it is fair to assume that the number was the same for all patients and must have been more than 12.)

Here are their basic findings:

  • The average (median) age of the patients was 64 years
  • The average (median) PSA value of the patients was 9.5 ng/ml (range, 3.7 to 28 ng/ml).
    • 74/177 men (41.8 percent) had a PSA value > 10 ng/ml.
    • 99/177 men (56 percent) had a PSA value between 4.1 and 10.0 ng/ml.
    • 4/177 men (2.2 percent) had a PSA value between 2.6 and 4.0 ng/ml (which suggests that they didn’t actually meet the study entry criteria).
  • The average (median) PCA3 score was 52 (range, 3 to 273).
    • 140/177 men (79 percent) had a PCA3 score > 20.
    • 100/177 men (56.5 percent) had a PCA3 score > 35.
  • 48/177 men (27.1 percent) were found to have localized prostate cancer on saturation biopsy.
    • Among the 48 men found to have cancer on saturation biopsy, the median PCA3 score was 60 (range, 7 to 208).
    • Among the 129 men who appeared to have no cancer on saturation biopsy, the median PCA3 score was 34 (range, 3 to 268).
    • This difference was statistically significant (p < 0.05).
  • When one considers the two different cut-off points for the PCA3 score of 20 and 35, one finds that
    • For a PCA3 cut-off point of 20
      • The diagnostic accuracy of the PCA3 test was 43.5 percent.
      • The sensitivity of the test was 91.7 percent.
      • The specificity of the test was 25.6 percent.
      • The positive predictive value (PPV) of the test was 31.5 percent.
      • The negative predictive value (NPV) of the test was 89.5 percent
    • For a PCA3 cut-off point of 35
      • The diagnostic accuracy of the PCA3 test was 50.2 percent.
      • The sensitivity of the test was 73.0 percent.
      • The specificity of the test was 41.8 percent.
      • The positive predictive value (PPV) of the test was 35.0 percent.
      • The negative predictive value (NPV) of the test was 80.6 percent.

In their conclusion, Barbera et al. note that

  • The PCA3 data do, in fact, correlate to a reduction in the number of unnecessary, repeat (saturation) biopsies
  • Using a PCA3 cut-off point of 20 in this patient cohort would have
    • Avoided 21.0 percent of the biopsies
    • Missed 4 /48 cases of prostate cancer (8.4 percent)
  • Using a PCA3 cut-off point of 35 in this patient cohort would have
    • Avoided 37.8 percent of the biopsies
    • Missed 13/48 cases of prostate cancer (27.0 percent)

When one considers these data with an unbiased eye, however, it certainly appears to The “New” Prostate Cancer InfoLink that the clinical value of the PCA3 test is rather less than stellar! At a cut-off point of 20, the test fails to identify 8.4 percent of men with prostate cancer and it only cuts the number of biopsies by 21 percent! And this cut-0ff point is way below the recommended cut-off point of 35 (which in this study appears to have been near to useless).

We would like to think that some of the newer tests that are coming to market soon would make the limited value of the PCA3 test redundant. We would also note that if this is the “real world” level of accuracy of the PCA3 test in its approved indication, we cannot imagine how anyone can justify the use of this test in men who have never been biopsied (except as a method of making money).

13 Responses

  1. It would be interesting to see what was the tumor grade of the cancers that were true positive and those that were false negative. Missing Gleason 7-10 and finding low-risk Gleason 6 in the long run may make the test even less worthwhile

  2. First biopsy in 2001: negative. PCA3 test in 2006: negative. PCA3 test in 2010: negative. Biopsy in December 2011: 12 cores, 6 positive (Gleason 4 + 3 and 3 + 4). Radical prostatectomy in March 2012: T3a; Gleason 4 + 3 in 15% of left lobe; negative surgical margins; negative lymph nodes.

    March 2013: PSA 0.6 ng/ml (detectable).

    Probably start salvage radiation soon even though 70% chance of micrometastasis. Still deciding. If not, next step hormones, though not sure when.

    Not a proponent of PCA3 test! Should have had another biopsy sooner.

  3. The PCA3 test certainly help me make a decision for a biopsy when my PSA test was measuring just a little over 4. With all the confusion about the PSA test it helped me overcome my denial and push forward to determine if I had prostate cancer. I did, with one core positive and a Gleason score of 3 + 4. So the PCA3 certainly added value to me and hopefully added years also.

  4. Indeed. … Unfortunately the original paper is in Italian so someone else would heave to help us out even if those data do exist in the full text! (Chodak? Hmmm … Nope, probably not Italian!)

  5. For me, a PCA3 score of 61 lead to a repeat biopsy with three cores positive for prostate cancer, Gleason 6, PSA 4.


    Based on the impressive success being recorded in active surveillance programs, Dr. Chodak’s point in the first response looks critical to determining whether the PCA3 test is worthwhile.

    I’m also wondering what results would have looked like under Dr. Chodak’s envisioned approach if the cut-off had been set substantially higher.

    There are several versions of the PCA3 test, and I’m wondering which was used and whether there is research on how they compare.

  7. What struck me the most was the incredibly high score of 268 for one of the patients who had a negative biopsy.

  8. Dear Jim:

    As far as I am aware, in Europe most people have been using the Progensa PCA3 test. It was approved in Europe at least a couple of years before it was approved in the USA (where commercial labs lik Bostwick Laboratories therefore developed their own versions of the test).

  9. For me this is a strong indicator of the false postive risks associated with this test.

  10. Based on my situation, I think the greater risk is the false negatives, but statistically it may still be a worthwhile test. Just because it didn’t work on me may be due to the unique circumstances of my particular prostate cancer. I’m not trying to dissuade avoiding unnecessary biopsies but, based on my results, I would hesitate to advise someone with a negative PCA3 result to breathe too easily.

  11. First biopsy negative. … Second biopsy negative. … Everything seems to be normal except PSA 9.3 and DRE also OK. … Now PCA3 a couple of weeks ago 37 (positive). … Making agenda for a third biopsy. … Hope it won`t be a third strike. …

  12. What percent of men with a PCA3 level below 20 had high volume disease or Gleason 7 or higher?

  13. Dear Sam:

    I have no idea. You have access to the same data that I do.

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 )

Connecting to %s

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

%d bloggers like this: