Prostate cancer mortality in Scandinavia: 1965 to 2006

A new epidemiological analysis of data from Denmark, Finland, Norway, and Sweden suggests that despite a significant increase in the incidence of prostate cancer (the number of men diagnosed each year), prostate cancer-specific mortality rates have been minimally affected over the past 40 years in these four countries.

We know that since the introduction of the PSA test in the late 1980s, the incidence of prostate cancer has significantly increased not only in Scandinavia, but in most of the Western world. However, the impact of early detection of prostate cancer on overall and prostate cancer-specific mortality has been a great deal more controversial.

Meyer et al. have used data from the online NORDCAN (“Nordic Cancer”) database and associated software to analyze the incidence and mortality data from the four Nordic nations between 1965 and 2006, basing their analysis on the cumulative incidence, the cumulative mortality data, and the age-standardized mortality rates for selected calendar years over the 40-year period.

Their analysis shows the following results:

  • From 1965 to 2006, nearly 173,000 prostate cancer-specific deaths were reported in the four countries.
  • As one would expect, there has been a substantial rise in the cumulative incidence of prostate cancer (from 9 percent to > 20 percent), with some degree of geographic variation.
  • However, neither the prostate cancer-specific mortality rate nor the cumulative risk of prostate cancer-specific mortality have shown consistent trends over this 40-year period.
  • Cumulative prostate cancer-specific mortality rates have ranged between 3.5 and 7.5 percent in Scandinavia over this timeframe.
  • Prostate cancer-specific mortality has remained fairly constant among the four Nordic nations, with a minimally lower risk in Finland.

These data are likely to add to the continuing controversy over whether early detection of prostate cancer has actually had any meaningful impact on the risk of death from this malignancy.

Much as we would like to believe that early detection has been valuable in affecting overall prostate cancer-specific mortality, the evidence for this effect is still not compelling. There is certainly little doubt that we are diagnosing and then over-treating many men with low-risk disease who will never have  clinically significant disease. There is also no doubt whatsoever that early detection has meant that far fewer men are initially diagnosed with metastatic disease. And, the PSA test has most certainly allowed us to detect prostate cancer early on in younger men with clinically aggressive prostate cancer who quite certainly needed early intervention.

The question we are still trying to come to grips with, on a population basis, as well as on an individual patient basis, is whether we are implementing curative treatment for those men who really need such curative treatment or whether we are only giving curative treatment — for the most part — to men who had little need for that treatment? It seems to this author that, sadly, and despite protestations to the contrary from some quarters, Willet Whitmore’s original conundrum is still alive and well.

6 Responses

  1. I decided to look at disease-specific mortality in those four countries. It seems more important to look at the rates in the immediate past. That is since the inception of the PSA test. The data is from 1987 to 2005 from the WHO database (except Denmark were data is from 1994 to 2001).

    In all cases the mortality rate in the four countries increased in those four years. That seems to imply that the use of PSA testing is not that prevalent and when detected by PSA testing the cancer is treated conservatively in many cases.

    During those years, in Sweden the mortality rate increased by 24.3%. Finland experienced a 35.7% increase and the other two countries had increases of 7% to 8%.

    The study implied that disease-specific mortality during those 40 years was constant in spite of an increased use of PSA. The use of data back to 1965 seems to affect results. These countries have one of the highest prostate cancer mortality rates in the world and the results of the study were surprising to read. I do not see how over-treatment of cancer that would never impact mortality can affect these results.

  2. I wonder if the mortality experience in these Scandanavian countries mirrors that in the US and UK where there was a rise in mortality rates that tracked the rise in incidence, followed by a fall in mortality rates as the incidence rates started to dip.

    Here is a link to a good piece demonstrating some of these issues

    This quote from that site sums up the position pretty clearly in my opinion:

    “As yet it is not possible to say what proportion of the fall in prostate cancer mortality is the result of improvements in treatment, changes in cancer registration coding, the attribution of death to prostate cancer, and the effects of PSA testing. Only the ongoing randomised controlled trials can provide definitive answers about the efficacy of screening.”

  3. Ralph: I don’t think that the WHO data are as accurate as the NORDCAN data, and I also don’t think that the use of the PSA test in the Nordic countries is anything like as limited as you suggest.

    Much as I would like to believe that early detection is significantly impacting mortality, I am becoming very suspicious about the validity of this claim. Part of the problem, for course, is that with the increased awareness about prostate cancer, I have no doubt that a higher percentage of men who would previously have been classified as dying “with” prostate cancer are now being classified as dying “of” their prostate cancer. The assignment of cause of death has always been a variable that is difficult to quantify whern it comes to mortality statistics.

  4. Mike,

    Things are relative. It is difficult to establish a real number for the use of PSA in different countries. From searching the medical literature, the closest figure for the Scandinavian countries as compared to the US is 23%:54%. Also look at the ratio of diagnoses to deaths. In the Scandinavian countries the ratio is from 27% to 33% while in the US it is 15%

    If early detection is not impacting prostate cancer mortality how do you explain the reduction in mortality in the past 15 years in the US? This is the opposite of what you say about death attribution (less men are actually dying “of” their prostate cancer). In the case of the latest European screening results, at least there is a 20% reduction in PCa-specific mortality and that was done with screening every 4 years and at less than 10 years of follow-up.

    Then there is the issue of over-diagnosis and over-treatment. This is a non-issue as far as the mortality rate is concerned. These over-treated men are not the ones impacting mortality. The quoted numbers of over-diagnosis are highly questionable. For mortality reductions, the natural course of the disease needs to be altered or delayed. It is clear that men destined to die of prostate cancer are now living longer and dying of something else.

    As far as the mis-attribution issue, Albertsen looked at is hard and could not find it to be the cause of reduction in mortality. Until something better than PSA comes along its use should be recommended and the information obtained judiciously used for the patient’s benefit.

  5. Ralph:

    I am not for one moment suggesting that PSA shouldn’t be used to identify the men who need treatment. We have no other really useful test. What I am saying is that the apparent decline in prostate cancer-specific mortality rates in the USA does not correlate with the apparent non-decline of mortality rates in the Scandinavian countries. There are a thousand and one possible reasons for this. (For starters, Scandinavian males live slightly longer on average than American males, so presumably their risk for death from prostate cancer is inevitably slightly higher.)

    The point that I am trying to make is only that the supposed connection (that we would all like to be able to prove) between early detection and reduction in prostate cancer-specific mortality has yet to be demonstrated in a compelling manner. In the meantime (as I know you agree), in the USA, we continue to over-treat many men who would be better off without treatment because of ignorance, fear, and the socioeconomics of medicine.

  6. The problem I see, Mike, is that in trying to “prove” the connection, the message delivered is that PSA has little value and they even go to show (as is the case here) that there is no benefit in reducing mortality. My point is that the use of PSA in the Scandinavian countries is not very widespread and as a consequence men there are diagnosed late and die at a higher rate. Why use data from 1965? Life expectancy then was much lower and by necessity affects results.

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 )

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: