End of an Era for PSA Screening:
A Newsmaker Interview With Thomas Stamey, MD
Laurie Barclay, MD
Sept. 17, 2004 — Editor's Note: The era of using prostate-specific antigen (PSA) levels as a screening test for prostate cancer is over, according to the results of a study published in the October issue of the Journal of Urology.
Using data from more than 1,300 prostate tissue samples collected at Stanford University during the last 20 years, divided into four 5-year periods between 1983 and 2004, the investigators compared the volume and the grade of cancer with clinical findings, including rectal examination and blood PSA levels. Over time, the correlation between PSA levels and the amount of prostate cancer weakened dramatically, from 43% predictive ability in the first five-year group to 2% in the past five years. However, PSA levels continued to be accurate as a direct measure of benign prostatic hyperplasia.
To learn more about the clinical implications of this study, Medscape's Laurie Barclay interviewed lead author Thomas Stamey, MD, a professor of urology at Stanford University in California.
Medscape: What were the main findings of your study?
Dr. Stamey: About 20 years ago in the New England Journal of Medicine, we reported that if you plotted the size of the largest cancer in the prostate that had been removed with the level of the blood PSA before it was removed, that there was a reasonable relationship between the blood PSA and the size of the largest cancer.
In fact, that relationship was approximately 50%, so it wasn't even perfect 20 years ago when we first started using PSA to diagnose prostate cancer.
This paper reports that relationship from 1983 through 2003. It showed that while that relationship in the first period, 1983 to 1988, was about 40% or 50%, in the last five years it's now only 2%, so that's negligible. That allows me to say that PSA today in the United States no longer has any relationship to the cancer except for 2% of men.
Now what is the PSA related to?
We show that PSA is now related to only benign enlargement of the prostate in the last five years.
In fact, it was also related to benign enlargement from 1983 to 1988, and that's why the relationship between PSA and prostate cancer was far from perfect 20 years ago. But now, instead of a 40% relationship — you could live with that as a test, although it was far from perfect — it's now only 2% related to the size of the largest cancer in the prostate.
Medscape: Why has the predictive value of PSA changed over time?
Dr. Stamey: That's a very good question.
One possible answer is that we have so overscreened men in the United States with PSA that the relationship that we had 20 years ago, 40% or so, is now only 2%. Maybe we have now removed all of the big cancers and we are left only with the small microscopic ones, so that's one possibility.
Medscape: How should PSA screening best be used in diagnosing or monitoring prostate conditions?
Dr. Stamey: I think that in this heavily screened country, there's hardly a man around that doesn't know what his PSA is, and that means most of them have been biopsied, too. I think that men need to recognize that a PSA between 2 and 10, and in a lot of cases between 2 and 20, is now unrelated to the cancer but is related to the size of the prostate. Almost all men over 50 years of age start to develop benign prostatic hyperplasia, so I think we need to recognize that PSA is related today to the benign enlargement of the prostate and not to the cancer. That says that we must find a new cancer marker.
I think that practicing urologists should recognize that a PSA between 2 and 10 is mainly caused by benign enlargement of the prostate. But if you biopsy men's prostates, you're going to find cancer, because we all have age-related prostate cancer. It begins in 8% of men in their 20s, based on a study of men dying accidentally on the streets of Detroit. Dr. Saker found that 8% of men in their twenties had prostate cancer, but every decade it increased, until 70% of men had prostate cancer in their 70s.
So I think that men need to realize that prostate cancer is something that we all get if we live long enough. But it's very important that when you say that, instead of scaring the hell out of the man, you follow it with the fact that the death rate as recorded by the National Cancer Institute from prostate cancer in the United States is 226 per 100,000 men over the age of 65. The bottom line is that prostate cancer is ubiquitous — we all get it, and it goes up with increasing age — but at the end of the road, the death rate of 226 per 100,000 men over the age of 65 is a pretty small death rate. Now a part of that, of course, is that when men get to be 60, the competing causes of death are huge — strokes, accidents, and other causes.
Medscape: In men who have already been diagnosed with and treated for prostate cancer, is the PSA of any use in monitoring recurrence?
Dr. Stamey: PSA will always be useful. PSA that reaches the blood serum only comes from the prostate, so PSA will always remain a very good marker when you go to treat the prostate, either treating it by surgery or irradiation. But because PSA is prostate-specific but not cancer-specific, it's a very good marker for any way that you treat.
Medscape: What do you believe currently is the best method of screening and diagnostic workup for prostate cancer?
Dr. Stamey: Patients must recognize two things:
Medscape: If a person happens to have an elevated PSA level because they had it done as part of a panel, what should they do if they are asymptomatic and not aware of any other problems?
- Prostate cancer is ubiquitous and age-related regardless of any markers.
- PSA is an excellent marker once you start to treat the prostate, whether it is for benign enlargement or cancer.
Dr. Stamey: An annual rectal exam is an important part of every examination. If you feel a hard nodule, it has a greater than 90% chance of being prostate cancer.
Medscape: Is there anything you would like to add in closing?
Dr. Stamey: To me, the biggest issue is that we need to recognize that we all get prostate cancer, but that the death rate is only 226 per 100,000 men over 65 years old.
Any excuse you use to biopsy the prostate — and we've been using PSA as an excuse — you're very likely to find cancer.
So the real need, and that's what I have PhDs and MDs in my laboratory working on all the time, is that we need to get a marker for prostate cancer that is proportional to the amount of cancer in the prostate. Then we might be able to make some intelligent decisions about who should be treated and who shouldn't.
Financial disclaimer: Donations to Dr. Stamey's Prostate Cancer Research Fund at Stanford funded this study.
J Urol. 2004;172(4, part 1 of 2):1297-1301
Reviewed by Michael W. Smith, MD