Screening and the incidence of newly diagnosed, metastatic prostate cancer over time

Another article in this week’s issue of the New England Medical Journal is going to re-fuel the fire among those who advocate for widespread, frequent screening of men for risk of prostate cancer.

Whether the new article by Welch et al. (which is available on line as a full text article) actually justifies the value of widespread screening based on the use of PSA testing is open to debate. What is probably not open to debate are the data that Welch and his colleagues present about the major decline in the incidence of initial diagnosis with metastatic prostate cancer since about 1988 (after the initiation of widespread use of the PSA test here in America) as compared to the complete lack of any significant decline in the incidence of initial diagnosis with metastatic breast cancer since the initiation of mammography to screen for breast cancer risk in the early 1980s (shown in Figure 1 of the article).

Welch et al. do not address the implications of whether mammography or PSA testing are valuable screening tests in this article. What they are focused on is the hypotheses that might explain the difference between the failure of widespread mammography to reduce the incidence of initial diagnosis with metastatic breast cancer as compared to the (apparent) success of PSA testing to reduce the incidence of initial diagnosis with metastatic breast cancer. (Note: When we say “the incidence of initial diagnosis with metastatic” breast or prostate cancer we are referring to the annual number of women or men who first present with metastatic cancer having never been diagnosed previously, i.e., excluding all patients who have progressive disease that has become metastatic after diagnosis earlier in the course of their disease.)

The article is relatively short, and it is well worth reading. Welch et al. discuss the three major hypotheses behind the development of metastatic cancers like breast and prostate cancer (the “solid” tumors) and how they can each, in different ways, be used to explain the differing findings about the value of mammography as compared to PSA testing and the major discrepancy between them. However, …

There is no escaping the very evident fact that widespread use of PSA testing seems to have cut the incidence of initial diagnosis with metastatic prostate cancer from about 65/100,000 in 1988 to about 25/100,000 in 2010. It also seems — for men of > 40 years of age — to have dropped the average (mean) age at initial diagnosis with metastatic prostate cancer from 71.8 to 69.8 years. By contrast, there has been no significant change at all in the incidence of initial diagnosis with metastatic breast cancer, which was about 18/100,000 in 1975 and was about 19/100,000 in 2010. Nor has there been any change in the mean age of initial diagnosis with metastatic breast cancer among women > 40 (which has been roughly 63.7 years of age for the past 37 years).

Even if annual or some other form of regular PSA screening has very little impact on prostate cancer-specific or overall mortality rates among men diagnosed with prostate cancer (a question that is still unresolved by the available data), if regular PSA testing can massively reduce the frequency of initial diagnosis with metastatic prostate cancer (by slightly more than 60 percent over the past 25 years), isn’t that a “good thing”? And can this fact help us to better plan out which patients really do need to be regularly and carefully tested to minimize the risk of initial diagnosis with metastatic disease on the grounds that many of them may be curable prior to progression? And does the USPSTF recommendation about the non-use of the PSA test not — as has been proposed many times by advocates for widespread PSA screening — not threaten the risk of a massive rise in the incidence of initial diagnosis of metastatic prostate cancer again, back up to something like 60/100,000 among men > 40 by about 2025?

8 Responses

  1. Results Are Even Better for Survival of Prostate Cancer Than Reflected in Statements by Authors of Paper

    This is an important paper, well covered by Sitemaster, and I very much appreciate the authors’ work here. However, there is an important point involving screening’s impact on survival of prostate cancer that should be in sharper perspective based on updated research, which apparently was not within the range of awareness of the authors.

    The paper observes that, while “PSA screening almost halved the risk of metastatic prostate cancer presentation, screening reduced the risk of death from prostate cancer only by about one fifth.3” (Not that even that premature improvement is bad! One fifth is pretty substantial!) Now reference 3 is the following paper, the 11-year update since enrollment, not since diagnosis of the European Randomized Study of Screening for Prostate Cancer, well known as the ERSPC: Hugosson J, et al. Prostate-cancer mortality at 11 years of follow-up. New Engl J Med. 2012;366:981-990 [Erratum, New Engl J Med. 2012;366:2137].

    However, a 13-year update of the ERSPC was initially published online in The Lancet in August 2014: Schröder FH, Hugosson J, et al. Screening and prostate cancer mortality: results of the European Randomised Study of Screening for Prostate Cancer (ERSPC) at 13 years of follow-up. Lancet. 2014:384:2027-35. Among salient points, the update documents a continuing plunge in the number of screenings needed to save a life with just 2 more years of follow-up. It also documents that at the 13-year point, such follow-up was still very short since diagnosis: only 6.2 years in the intervention (i.e., screening) group and 4.3 years in the control group! Therefore, as of the 11-year follow-up relied upon by the authors, follow-up since diagnosis was likely only about 4 years in the screening group and 2 years in the control group. The point is that, with the long but ultimately nasty survival fuse burning for prostate cancer, such follow-up was grossly inadequate to show what screening could do for survival.

    It is interesting that the authors of the study were not aware of the important 13 year update, despite the fact that it was published well over a year ago. We need to make sure that the USPSTF does become aware of that update as it reconsiders its previously horribly ignorant position against prostate cancer screening! (I’m trying to comment on the NEJM site for the paper but am having difficulty.)

  2. Dear Jim:

    (1) I am quite sure that Dr. Welch and his colleagues are well aware of the 13-year follow-up paper. It was simply irrelevant to the points they were making in their paper.

    (2) You can’t submit comments to articles in the NEJM in the way you may be used to. You would have to formally submit a Letter to the Editor.

    (3) The way to make sure that the USPSTF is aware of your opinions is to comment on their new research plan, as discussed here.

  3. Dear Sitemaster:

    One point that seems new and relevant in the 13-year follow-up is the first publication as far as I know of the very short follow-up since diagnosis, as noted in my earlier comment. Awareness of that type of follow-up is crucial to appreciating the mortality data, right?

  4. Dear Jim:

    We really don’t know yet how time from diagnosis to death (as opposed to time from initiation of screening to death) may affect the value of screening — for one very simple reason.

    Most men who die from prostate cancer are diagnosed at 60+ years of age because risk for prostate cancer is very certainly age related. Thus, if you are diagnosed with prostate cancer when you are 60, and you live with the consequences of that diagnosis for another 25 years, the issue of average length of life starts to affect whether you are at risk of dying from prostate cancer.

    For screening for prostate cancer to be effective, we therefore have to be able to prove a real and ongoing effect of the screening initiative on risk for prostate cancer-specific or (better still) overall mortality rates over something like 25+ years (not just 13). But during that time frame many men diagnosed with prostate cancer will also be dying from other causes. Until we have something like 25 years of follow-up data, it is not going to be possible to tell whether the time from initiation of screening to actual diagnosis in a study that only screened men for 4 years will make any difference to the potential benefits of screening at all. If there is a significant difference, it could well show up early but then vanish again over time.

  5. Mike,

    Why is this paper news? Names like those of Welch, Barry, Wilt, Woloshin, Brawley, et al. have been associated with publishing studies with a negative perception of the use of PSA to check for prostate cancer. For more than 15 years it has been known to them and many others (including us) that since the commercialization of the PSA test the number of men diagnosed with advanced forms of prostate cancer (including metastatic disease) has been reduced considerably. It has been their argument that over-diagnosis and over-treatment of prostate cancer needed to be addressed. And the powers to be did …

    Hence the recommendation D from the USPSTF. Instead of offering logical solutions to excessive treatment of disease that might be observed closely the recommendation was to not test with PSA. The consequence of that decision is potentially dangerous to men. Why has the current prostate cancer mortality reduction experienced here in the US been ignored? Why has it been misattributed to improved treatments when it is known that such reduction has not happened in countries in which the same treatments are available but the use of PSA has not been par to that used here? Early detection of disease that could kill and proper treatment is involved in the result.

    Vickers and Lilja have a nicer option to detect men at risk. Their proposal involves less testing, less biopsies and ultimately less prostate cancer deaths. Hopefully the new review by the USPSTF will be an initial step to change their D recommendation to a C. Smart testing with PSA will prevent unnecessary deaths.


  6. Ralph:

    The paper is interesting because it is the first time (as far as I am aware) that the effect on metastatic disease has been quantified, and it is also interesting because it coincides with the USPSTF’s decision to conduct a new research project which could lead to a re-assessment of the D recommendation.

    I would also note that one of the authors of the paper is Dr. Peter Albertsen, who has never been “anti” screening of appropriate men (although he has certainly been “anti” the over-treatment of low-risk prostate cancer).

  7. Mike,

    The SEER database (mentioned in the study) has reported such reduction in metastatic disease for years. The benefit of a longer 5-year relative survival associated to the US prostate cancer mortality reduction has been well known for years. The exaggerated reported computer model rate of over-treatment brought about the USPSTF D recommendation.

    It is amazing (to me) to see reports such as this one when the evidence has been there for a long time. Maybe the USPSTF will change their recommendation but unfortunately the spin about the value of PSA testing will continue. It is a sign of the times and not unique to prostate cancer. …

  8. Ralph:

    The fact that it is amazing to you doesn’t take away from the publicity this paper got in the New England Journal. I realize that this may be incomprehensible to you, but most of the US healthcare community doesn’t spend a lot of its time thinking about prostate cancer at all.

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: