Low-dose aspirin and prostate cancer-specific mortality

A newly published article — and an associated editorial — in the Annals of Internal Medicine has addressed the possible benefits of taking low-dose aspirin as a method to prevent the progression of newly diagnosed prostate cancer.

The article by Skriver et al. reports on a retrospective analysis of data from a cohort of 29,136 men, all diagnosed with prostate cancer in Denmark between 2000 and 2011. About 25 percent of these men were treated with low-dose aspirin in the year after their diagnosis, and most of them had been taking low-dose aspirin prior to their diagnosis as well. The researchers knew this because, in Denmark, nearly all low-dose aspirin is administered by prescription — unlike the situation here in the US, where it is usually bought “over the counter”. In this study “low-dose” aspirin was defined as a daily dose of aspirin ranging from 75 to 150 mg per tablet (as opposed to the 300 mg in a standard aspirin tablet).

Here is what the authors report as their key findings:

  • The average (median) patient follow-up was 4.9 years.
  • The average (median) age of the patients was 70 years.
  • 7,633/29,136 men (26.2 percent) died of prostate cancer.
  • 5,575/29,136 men (19.1 percent) died of other causes.
  • At the 4.9-year median follow-up period,
    • Post-diagnosis use of low-dose aspirin was associated with multivariable-adjusted hazard ratios (aHRs) of
      • 0.95 for prostate cancer–specific mortality
      • 1.12 for other-cause mortality
    • Among men with less aggressive types of disease (with Gleason scores ≤ 6), compared to non-users, use of low-dose aspirin was associated with a significant reduction in prostate cancer-specific mortality (aHR = 0.82)
  • A secondary analyses showed that post-diagnosis use of low-dose aspirin was associated with aHRs of
    • 0.91 for prostate cancer-specific mortality at 5 years
    • 0.84 for prostate cancer-specific mortality at 7.5 years
  • The reduction in prostate cancer-specific mortality at 7.5 years was particularly evident (aHR = 0.77) among patients filling prescriptions for a large quantity of low-dose aspirin (> 1,096 tablets during the entire 7.5-year period).

Skriver and her colleagues conclude that:

… our study did not support an overall effect of postdiagnosis low-dose aspirin use on prostate cancer mortality. However, our results suggest that low-dose aspirin use might be inversely associated with prostate cancer mortality after 5 years from cancer diagnosis.

That conclusion is in line with prior data from Finland that also showed a lack of impact of a mini-aspirin on prostate cancer-specific survival.

In their editorial commentary on this paper, Murtola and Veitonmaki make a number of key observations, as follows:

  • Colorectal cancer is the type of cancer in which the evidence for aspirin’s preventive impact has been best established, but “the minimum required exposure time for benefit seems to be 10 years.”
  • Men with prostate cancer who take low-dose aspirin are likely to be different from those men who do not use aspirin — for a variety of reasons, but most specifically because “Aspirin is commonly used in low doses to prevent cardiovascular and cerebrovascular events”.
  • Aspirin does have biochemical properties that suggest it might have oncologic benefit in prostate cancer.
  • Aspirin’s anti-inflammatory effects might also be beneficial in the setting of diagnosed prostate cancer, as chronic inflammation within the prostate may increase the risk of men developing high-grade disease.
  • There is a logical argument supporting an association between improved prostate cancer-specific survival and long-term aspirin exposure, because the anti-inflammatory actions of aspirin probably need time to exert a meaningful effect on prostate cancer development.
  • There is also the possibility that tumor grade misclassification differs between aspirin users and non-users at time of diagnosis.

So what we have here is a set of intriguing data that suggest three possible hypotheses with respect to the long-term, daily use of low-dose aspirin among prostate cancer patients:

  • It might extend prostate cancer-specific survival at 5 or 10 years among men initially diagnosed with low-risk forms of the disease.
  • It might extend time on active surveillance by helping to prevent disease progression among men with low-risk forms of the disease.
  • It might extend prostate cancer-specific survival — albeit only slightly — at 10 years among all men initially diagnosed with the disease.

Of course the cost of conducting a prospective randomized trial to explore these possibilities would be high, and it might be better to limit such a trial to the men likely to obtain the greatest possible return on the investment — the men diagnosed with low-risk disease and most particularly those men who elect active surveillance as their first-line form of management.

Editorial note: We thank the publisher for provided us with the full texts of the article by Skriver et al. and the editorial by Murtola and Veitonmaki for our review in constructing this commentary.

7 Responses

  1. Seems like it’s such a low-cost preventive agent, considering both economics and health risk, that it is prudent to take it.

  2. 26.2% prostate cancer-specific mortality at 4.9 years? What’s up with that sample? I can’t account for that in a population-wide cohort. Can anyone?

  3. I think the issue is in the denominator, the 29,136 men selected as “diagnosed” over 12 years from a male population of 2.5 million. That’s approx. a sixth of a lifetime, suggesting a lifetime risk of 180,000/2,500,000 if it captures all diagnoses = lifetime risk of 7%. So it’s not an exhaustive sample, but under half of all cases. What gives? The good news is we are left with a recognizable but low lifetime mortality risk (26.2% of 7%). That all suggests this was a heavily high-risk sample.

  4. Dear SUM:

    I think you are over-analyzing these data. And of course this was a “heavily high-risk sample”. They had all already been diagnosed with prostate cancer.

    We have no idea what percentage of other men in Denmark over the same time period might have had low-risk forms of prostate cancer that never got diagnosed at all, and there is no way to know that statistic (unless we were to autopsy every man who died in Denmark over a 10-year period).

  5. No: sorry. The lifetime incidence is too low and we need to know nothing about undiagnosed cases to make that observation. All very odd.

  6. Dear SUM:

    These data may be relevant.

    The incidence of newly diagnosed prostate cancer in Denmark is about 4.500 patients a year since 2008.

    The age-adjusted incidence rate in Denmark increased dramatically from 1992 to about 2008 (just as it did in nearly every other “western” nation) whereas the prostate cancer-specific mortality rate remained pretty much stable (see Outzen et al., 2013)

  7. Thanks: @SM. I think I am starting to get it.

    I am forgetting how low the survival is of advanced cases which dominated the incidence before PSA testing became widespread. Those last links really helped; thank you. So it is, by current standards, a high-risk sample, lacking the PSA-detected cases that round out modern cohorts.

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 )

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: