PSA Screening: “One of Medicine's Great Success Stories"?

.
Guest blog by Robert Ferrer, MD MPH

I spent 2 days at home recently with the H1N1 flu and caught up on some of my newspaper reading. In the Oct 11, 2009 New York Times Magazine, (p.38) was an advertising supplement, "Health and Wellness Outlook Special Report: Cancer Treatment Options," paid for by some of our finest cancer centers. [1] In the prostate cancer section, I found this interesting assertion, in big capital letters: "The PSA test for prostate cancer detection and management is one of medicine's great success stories." It goes on to say that 90% of prostate ca is now diagnosed when curable and that the death rate has declined by 40% since the PSA test began to be widely used in the 1980's. The source is the chair of urology and senior vice president for translational research at Roswell Park Cancer Institute in Buffalo, where the the PSA test was developed.

I found the assertion curious because this past March the New England Journal of Medicine published 2 long-awaited studies[2],[3] on whether PSA testing was effective: one found a modest benefit and the other virtually none. Both noted that a very large number of men had to be screened and treated for every one who benefited. These articles received extensive press coverage. And just as I was going to my laptop today to assemble this and other maybe-not-such-a-great-success-story evidence for PSA screening, this week's JAMA showed up with a terrific paper by Esserman, Shieh, and Thompson.[4] They have this to say about PSA screening:

"After 2 1/2 decades of screening for [breast and] prostate cancer, conclusions are troubling: Overall cancer rates are higher, many more patients are being treated, and the absolute incidence of aggressive or later-stage disease has not been significantly decreased. Screening has some effect, but it comes at significant cost, including over-diagnosis, overtreatment, and complications of therapy."

So how do we get from "great success story" to "troubling"? How can diagnosing cancer early not be a good thing? The answer lies in the kind of cancers we can detect with screening. Slow growing cancers, the kind unlikely to kill you, grow... slowly and so are around for a long time to be detected by screening. On the other hand, fast growing cancers can go from undetectable to lethal even in the year between cancer screenings. So the cancers we detect through screening are more likely to be the non-lethal kind. Well, isn't that still a good thing? Can't cancer harm without killing?

Yes, but the issue here is what we call "cancer." Our screening tests can detect collections of cells that are, by pathologists' standards, "cancer" when viewed under the microscope, yet not every collection of such cells is destined act like cancer; that is, to grow or spread (metastasize) to other parts of the body. Some are destined to remain dormant until the person eventually dies of something else. And therein lies the problem with PSA screening. It detects many of the ones destined to be dormant or slow growing for every one destined to be lethal. The exact number is uncertain, but the large European study in the NEJM this March estimated 1410 men needed to be screened and 48 cases of prostate cancer treated to prevent 1 death. [2] The American study released in parallel found the benefits to be even smaller. [3]

What this means, is that the consequence of PSA testing for many men is adding 6-12 years of life diagnosed -- and often treated -- as a cancer patient, without actually living any longer.

Just how much over-diagnosis can we attribute to PSA? In the August 2009 issue of the Journal of the National Cancer Institute, H. Gilbert Welch and Peter Alberson calculate than in the first 19 years of the PSA era, 1987 to 2005, about 1.3 million additional cases were diagnosed and 1 million more men treated.[5] They estimate that about half of these extra cases represent over-diagnosis, meaning that the diagnosed man was very unlikely to die from prostate cancer. So of the roughly 4 million men diagnosed from 1985 to 2005, half a million were over-diagnosed.

What about the fact that, as the Roswell Park urologists note, mortality rates have fallen since PSA testing began in the mid-80's? Doesn't that suggest that PSA is helping? The authors in this week's JAMA paper address this in their analysis. For that claim to be credible, we should be seeing a sharp fall in number of advanced stage prostate cancers, which is what would happen if screening was finding the "bad" cancers early, before they could reach an advanced stage. Although we have indeed seen a fall in advanced cancers it has been nowhere near as sharp as we would expect, given the many more cancers we are finding in the PSA era. We should thus probably look elsewhere to explain the fall in prostate cancer mortality, likely improvements in treatment.

So, given what we know about how well PSA testing performs as a screening test, how can it be advertised as one of medicine's great success stories? As potential explanations, I offer two themes that I believe also offer some larger lessons for why health care is less effective and more expensive than it should be.

Theme 1: Thinking about organs rather than people: If your focus is the prostate, then finding and removing cancerous prostates is the goal. This works well at the level of prostates, but not so well for whole men. With a test as imperfect as the PSA, a small or nonexistent reduction in the risk of dying from prostate cancer is sometimes traded for diminished quality of life, most commonly the incontinence and impotence that affect about 1/4 of men treated for prostate cancer.

Theme 2: Economic incentives favoring procedures. As the numbers above demonstrate, PSA has expanded the number of prostate cancer patients by about a third. The professional urology association has long recommended PSA screening even when the US Preventive Services Task Force, tasked with providing rigorous assessments for screening procedures, has consistently recommended against routine PSA screening.

Themes 1 and 2 intertwine. Greed is not what drives PSA testing. When a urologist can make a prostate cancer diagnosis and provide a "cure," doctor and patient alike perceive it as a valuable service. A life-saving intervention. That the service is well reimbursed appears justified when the stakes seem so high. It is only from the application of healthy skepticism and careful analysis -- of outcomes for people, not organs -- that we can reach better conclusions about the value of what we do.

The topic of PSA screening was previously addressed (if less well) in PSA Screening: What is the Value? March 21, 2009

[1] Anonymous. Health and Wellness Outlook Special Report: Cancer Treatment Options [advertising supplement]. New York Times Magazine, 11 October 2009. p. 33-46
[2] Schroder FH, Hugosson J, Roobol MJ, et al. Screening and prostate-cancer mortality in a randomized European study. New England Journal of Medicine. 2009a;360:1320.
[3] Andriole GL, Crawford ED, Grubb III RL, et al. Mortality results from a randomized prostate-cancer screening trial. N Engl J Med. 2009b;360:1310.
[4] Esserman L, Shieh Y, Thompson I. Rethinking screening for breast cancer and prostate cancer. JAMA. c;302:1686-1692.
[5] Welch HG, Albertsen PC. Prostate cancer diagnosis and treatment after the introduction of prostate-specific antigen screening: 1986-2005. J Natl Cancer Inst. 2009c;101:1325-1329

This topic was also addressed previously in PSA Screening: What is the Value? March 21, 2009
.