Who actually dies from prostate cancer?


Some readers are probably going to find this very hard to believe, but … according to a newly published study, men initially diagnosed with metastatic prostate cancer are actually less likely to die from their cancer than men who are initially diagnosed with non-metastatic disease but who progress to having metastatic disease over time.

The new paper by Borno et al. is based on data from the CaPSURE registry and includes > 14,000 men diagnosed with prostate cancer at 43 different centers (academic, Veterans Administration, and community-based) between 1990 and 2016. And it also means that all of the authors are working at or formerly worked at the University of California, San Francisco, which is a highly reputable research center.

Although it is well understood that the majority of men who get diagnosed with prostate cancer do not die of this disease, this set of findings comes as something of a surprise to your sitemaster!

Here is what Borno et al. actually found, based on their data from 14,753 patients:

  • Just 669/14,753 men (5 percent) had metastatic disease — either at diagnosis or at some time post-diagnosis, and of those men
    • 303/669 (45 percent) had metastatic disease at initial diagnosis
    • 366/669 (55 percent) progressed to having metastatic disease over time
    • 461/669 (69 percent) were ≥ 65 years old at the time of their diagnosis
    • 342/669 (51 percent) had a PSA level of >20 ng/ml at diagnosis
    • 386/669 (58 percent) had a biopsy-based Gleason score of ≥ 4 + 3 = 7 at diagnosis
  • Of the men who had metastatic disease at diagnosis,
    • 31/303 (10 percent) were given initial local therapy
    • 272/303 (90 percent) were given initial androgen deprivation therapy (ADT) of some type
  • Of the men who progressed to having metastatic disease over time,
    • 239/366 (65% percent were given initial local therapy
    • 127/366 (35%) were given initial ADT of some type
  • After making relevant adjustments for sociodemographics, marital status, diagnosis year, and comorbidities,
    • Risk for prostate cancer-specific mortality (PCSM) was significantly lower for men initially diagnosed with metastatic disease than it was for men who progressed to have metastatic disease over time (hazard ratio [HR] = 0.66)
    • There was no evident difference in PCSM for patients treated with initial hormonal vs. combined local and hormonal therapy

The authors conclude that:

In this analysis of a nationwide cohort of men treated for [prostate cancer] with all types of therapy over 25 years, we observed that among men with metastatic [prostate cancer], the risk of PCSM was lower for de novo vs. recurrent metastatic disease. Additionally, no difference was observed based on initial treatment with combined local and hormonal therapy vs. hormonal therapy alone.

Now it is worth noting that, when the CaPSURE study was initially started — back in 1990 — the expectation was that men with an initial diagnosis of metastatic prostate cancer would die from their disease within about 3 years. What this study does not tell us (at least in the abstract) includes

  • The average (median) follow-up for all the 669 men with metastatic disease in this study
  • The median follow-up of the 303 men with an initial diagnosis of metastatic prostate cancer
  • The median follow-up from time of evidence of metastatic disease of the 366 men who progressed to having metastatic disease over time
  • What percentage of the men who progressed to having metastatic disease did so only after they had been started on ADT

Your sitemaster if going to see if he can get hold of a copy of the full text of this paper to read more about the details. Maybe some of these questions will be answered in the full paper.

4 Responses

  1. It strikes me that this finding is somewhat meaningless. It is dividing metastatic cancer patients into two cohorts and then comparing mortality. Any metastatic cancer is dangerous, and I doubt every variable of health history, heredity, and lifestyle is controlled for. Perhaps those later developing it have a more aggressive cancer or a genetic propensity.

    Obviously I could be wrong, so it will be interesting to see the take of others.

  2. Sitemaster,

    I was not surprised at all by the first conclusion of the study (newly diagnosed have less of a death risk than recurrent metastatic patients). I believe what has been shown is that newly diagnosed, hormone- and treatment-naive metastatic prostate cancer is less life threatening than recurrent, metastatic castration-resistant disease. The newly diagnosed patient has all of the treatment modalities lying ahead of him, whereas the mCRPC patient, by definition, has already failed at least one mode of ADT and probably second-generation ADT too.

    The results of this study lend support to my personal prostate cancer treatment philosophy that castration-resistant disease is, in part, driven by treatment-mediated selection pressure. Now, if the study measured the risk of death from the point of diagnosis of the non-metastatic patient (who becomes recurrent then mCRPC many years later) compared to the newly diagnosed patient with metastatic disease, I would be very surprised if the risk of death were lower for the newly diagnosed metastatic patient.

    Best regards,

    Richard

  3. This result is so counterintuitive that I would wonder about sample bias.

    For example, if the sample of 669 consisted only of men not lost to followup — a very practical restriction, but not stated — then the result might be considerably less surprising.

    The takeaway result might be that men who sign up for a cancer registry are far more likely to keep up with it if they started out healthy than if they started out very sick.

    It would be interesting to learn if and how this study compensated for possible biases, most of them less obvious than this.

  4. Paul:

    I tend to agree with you. Did you read my follow-up commentary after I had managed to get my hands on the full text of the paper?

    I heard from UroToday that Alicia Morgans is going to be discussing this paper with two of the authors relatively soon.

    Mike

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: