May 30, 1999
Is Prostate Cancer Test Reliable?
by Isadore Rosenfeld
A RECENT PIECE I WROTE ABOUT prostate cancer has generated an interesting and
The essence of the dissent is that the Prostate Specific Antigen test is
unreliable at detecting a malignancy and that there is no proof that treating
most prostate cancers is effective.
My recommendation- that PSA testing should be done for all men aged 50 to 70
reflects the opinion of most "establishment" urologists, including those at the
hospitals at which I work -- Memorial Sloan-Kettering Cancer Center and the New
York Presbyterian Hospital. Here are the critics' specifics:
These are compelling arguments. However, the American Cancer Society recommends
testing and treatment for prostate cancer, because 179,000 men in the U.S. will
learn they have the disease this year, and 37,000 will die from it. Although
many tumors remain innocuous, others grow and spread, and some do so early on.
Right now, there is no way to tell which tumor will follow which course. But
there is growing evidence that PSA testing has in fact lowered the death rate.
- Widespread use of the PSA test has not resulted in a drop in the death
rate from prostate cancer.
- Prostate cancer is almost universal. It is estimated that 70% of all-:men
age 70 and over have prostate cancer but that it rarely causes their
- There is no conclusive evidence that beating prostate cancer lengthens
life. The malignancy is slow-growing, and most older men will die from
other causes (which is why I don't recommend the test for those over 70).
- The PSA test is unreliable. Only 30% of men with an "abnormnl" reading
(greater than 4) have cancer. By the same token, the results are falsely
negative in about 30% of cases, failing to detect the malignancy when
- A positive or suspicious test result may lead to an in- vasive surgical
procedure - - such as a biopsy - that in itself may expose a patient to
the usual roster of post-operative problems, as well as unnecessary
- Even though the American Urological Association recommends the PSA test,
the equally prestigious U.S. Preventive Services Task Force does not.
I worry that while I monitor the progress of a cancer, it will suddenly spread.
So I inform my patients about the limitations of PSA testing; I tell them
prostate cancer is often not life-threatening. Nevertheless, most of them ask
for the PSA test because, if they turn out to have cancer, they want it found
early enough to be cured.
When I tell them that a third of all men over the age of 50 probably have some
form of prostate cancer but that it will cause death in less than 10%, they ask
whether I can guarantee that they won't be among that 10%. I can't. So, since
neither they nor I want to play Russian roulette with cancer, I recommend PSA
testing for all men between 50 and.70.