Can the phi test accurately predict who should go on active surveillance?


In more data reported at ASCO yesterday afternoon, we got the first real look at what appears to be a helpful study designed to evaluate whether the Beckman Coulter Prostate Health Index (or phi) test is really able to identify men who are good candidates for active surveillance.

Eichholz et al. (abstract no. 5071) reported data from a study initiated in 2002 in which they looked at baseline phi values in a series of men who are on active surveillance with clinical stage T1/2 disease, a Gleason score of ≤ 3 + 4 = 7, and a PSA of < 15 ng/ml (very similar to the men in the long-term Sunnybrook patient cohort of Klotz et al.). Patients in the Eichholz cohort get PSA tests every 6 months and a repeat biopsy every 2 years. Treatment was recommended for any patient with a PSA velocity of > 1 ng/ml/yr or a Gleason score of 4 + 3 = 7 or higher on repeat biopsy.

Because serum was banked at baseline for all the patients in this study, the investigators were able to run phi tests on the baseline serum of all the patients in this study and compare those phi test results to the patients actual outcomes over time.

Here are the key study findings:

  • The study included 370 evaluable patients.
  • The average (median) follow-up was 5 years.
  • In men with a phi value of up to 31.4 (quartile 1, N = 94)
    • The median time to treatment could not be evaluated because too few men (only 6)  had gone on to have treatment
    • 95.1 percent of men were treatment-free at 5 years
  • In men with a phi value of between 31.5 and 42.9 (quartile 2, N = 91)
    • The median time to treatment was 11.4 years (but could still not be formally evaluated because only 17 men had gone on to have treatment)
    • 81.6 percent of men were treatment-free at 5 years
  • In men with a phi value of between 43.0 and 58.5 (quartile 3, N = 93)
    • The median time to treatment was 6.4 years
    • 58.9 percent of men were treatment-free at 5 years
  • In men with a phi value of ≥ 58.6 (quartile 4, N = 93)
    • The median time to treatment was 5.1 years
    • 51.4 percent of men were treatment-free at 5 years

Eichholz et al. conclude from their study, appropriately, that:

In men with favorable risk prostate cancer, phi at diagnosis was a significant predictor of the outcome of active surveillance. The data require validation, but suggest that active surveillance is particularly attractive to men with a low phi.

However, it has to be noted that a small percentage of the men in this study (6/94, 4.9 percent) with a low phi score (≤ 31.4) did go on to have treatment within 5 years (although it is not clear from the abstract whether all those men actually met the criteria for recommendation of treatment). It also has to be noted that a high percentage of the men in this study (41/93, 48.6 percent) with a high phi score (≥ 58.6) did not go on to have treatment within 5 years (and we can reasonably assume that most or all of these men did not meet the criteria for recommendation of treatment).

In other words, while the phi score was certainly useful in helping to predict groups of men who would do well on active surveillance over the next 5 years after diagnosis, it was clearly not able to predict this with anything approaching absolute accuracy for an individual patient — although the level of accuracy for men with a phi score of ≤ 31.4 was certainly high, at > 95 percent

5 Responses

  1. Mike:

    You should link to the poster on this as well. Pretty interesting. … PHI is one of quite a few biomarkers designed to cool down over-treatment. We need to give them very careful study analysis and see what comes out, but the industry is really focused now on finding curbs to treatment morbidity and that’s a good thing.

  2. Tony:

    Is the poster actually accessible to non-ASCO attendees? I thought it was only available as part of the “Virtual Meeting”.

  3. Where is the phi test available?

  4. Akai:

    I believe that any doctor can order this test now, but that specimens may have to be sent to Innovative Diagnostic Laboratory in Richmond, VA, for processing. This is the only laboratory currently listed by Beckman Coulter as an approved provider of the test.

  5. Thanks, I also called them today and found that there is only one lab that you listed above. Don’t know why there is only one lab, hope they licensed it to Quest soon.

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 )

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: