The role of the phi test prior to decisions about prostate biopsies

Yesterday morning, a newly published paper was brought to my attention. It reports that use of the Prostate Health Index or phi test in a partially prospective study can and did change urologists’ opinions about whether selected patients really do or don’t need to have a prostate  biopsy.

The phi test was designed to help urologists and their patients make better decisions about who really did and who really did not need a biopsy if the patient had an elevated PSA level of 4 to 10 ng/ml but no other significant indicator of risk for prostate cancer.

While this study by White et al. is not a perfect study (because it was not randomized), it may still be a very important one. (And for interested readers the link will take you to a full-text copy of the study report.)

What the study did was to enroll 506 men who appeared to have some degree of risk for prostate cancer and give them a phi test. These men were individually selected by urologists practicing at one of four large, US-based, urology practice groups, if they met the following citeria:

  • They had to
    • Be ≥ 50 years of age
    • Have a non-suspicious (i.e., negative) rectal exam (DRE)
    • Have a total serum PSA level of 4 to 10 ng/ml within 6 months of the DRE
  • They could not
    • Have been diagnosed with prostate cancer
    • Have used any amount of any 5α-reductase inhibitor (e.g., finasteride or dutasteride) within the previous 3 months
    • Have had a prior biopsy result that was positive or suspicious for prostate cancer (e.g., signs of high-grade PIN or atypia)
    • Have had a prior transurethral resection of the prostate (TURP)

In addition they couldn’t be included in the study if their physician gave them a phi test but didn’t wait for the result until he or she gave the patient a biopsy.

The data from these patients was then compared to data from a comparable group of 683 earlier patients at the same four large urology groups (and managed by exactly the same physicians) who met all the same criteria except that they hadn’t had a phi test — so-called “historical controls”.

And here is what the study found:

  • Among the men given the phi test
    • 184/506 (36.4 percent) were actually given a biopsy (after receipt of the result of the phi test)
    • 110/184 (59.8 percent) had a positive biopsy result
    • 50/184 (27.2 percent) had a positive biopsy result but only Gleason 3 + 3 = 6 disease on biopsy (and therefore had low-risk prostate cancer)
  • Among the historical controls
    • 412/683 (60.3 percent) were actually give a biopsy
    • 257/412 (62.4 percent) had a positive biopsy result
    • 126/412 (30.6 percent) had a positive biopsy result but only Gleason 3 + 3 = 6 disease on biopsy (and therefore had low-risk prostate cancer)

Also, among the physicians, here is how the results of the phi test made a difference to what actually happened:

  • Individual patient’s phi scores impacted management plans for that patient in 72.5 percent of cases.
  • In 28.7 percent of cases, the physician decided to monitor the patient instead of giving him a biopsy.
  • In 14.3 percent of cases, the physician decided to give the patient a biopsy instead of just monitoring him.
  • In 18.9 percent of cases, the physician modified the originally proposed monitoring strategy (type of monitoring and how often).
  • In 92.3 percent of cases, the physician stated that knowing the phi score helped to explain his or her recommendation to the patient.
  • In 28.3 percent of cases, the physician stated that knowing the phi score helped to alleviating patient anxiety.

It is very clear from these data that use of the phi test (a) was affecting physician decisions and behavior; (b) reduced the total number of of patients who were actually given a biopsy overall; (c) made no immediately apparent difference to the percentage of men being diagnosed with intermediate- or high-risk forms of prostate cancer.

The significant problems (“limitations”) of this study include the following:

  • It wasn’t randomized. In other words, it would have been a better study is 1,000 patients who met the study criteria were randomly being assigned to have or not have a phi test and the study was therefore prospectively comparing “apples to apples”.
  • The patients haven’t been followed for long enough to know (a) what percentage of men went on to be diagnosed with prostate cancer over time, and (b) what percentage of men (specifically) went on to be diagnosed with higher-risk forms of prostate cancer (especially if they had not been given a biopsy based on the results of their phi test).

In an ideal and patient-centric world, what would be wonderful would be a new trial in which a cohort of about 2,000 patients like those defined above were all given:

  • A phi test or
  • A 4KScore test and
  • A germline genomic test for risk of specific, pre-defined mutations

and then they were all followed for about a decade to see how these data affected decisions about their management and outcomes over time.

Your sitemaster is an optimist at heart, but he has his doubts about whether that particular trial is likely to be started soon!

In the meantime, it is clear from this trial that definable subsets of patients (and their physicians) do appear to clearly benefit from the use of phi testing prior to any decision about the need for a biopsy (which many men would certainly rather avoid if they could).


4 Responses

  1. Hi my opinion as a prostate cancer survivor, at least it seems that there is feelings that maybe another test could be developed that would ease the minds of men with the elevated PSA 4 – 10, so they would not have to endure the invasive “12 needle pierced prostate gland through the rectum” biopsy I had to have which came about before I could proceed with surgery to replace my left hip joint and was fairly close to not being able to walk at all.

    Of course, as many of us patients know, there is a possible side effect of creating problems involving the rectum tissues later. Some concerns also included of dragging cancer cells from the needle encasement tubes which could contaminate the rectum area when withdrawing the needles to be sent to pathology. I hope for other patients that research continues as mentioned in this info to seek a better way of testing for possible prostate cancer.

  2. In the real world, it is better to seek progress than to wait for perfection.

    My mentor used to remind me of Voltaire’s axiom that, “The perfect is the enemy of the good”.

  3. Dear William:

    There are already two such tests: The phi test and the 4KScore test. The real question is, why aren’t they being used as much as they could be by the urology community?

  4. Dear pfadtag:

    Please appreciate that I am in complete concurrence with Voltaire. I don’t think we should be “waiting for perfection” at all. But we should all want to know whether the 4KScore test is any better or worse than the phi test. Surely one would want to be given the test that is more accurate? If they are much the same, the phi test is a lot less costly!

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: