A study in the Journal of the National Cancer Institute examined the use of a blood test to find prostate cancer in a group of patients 60-84 over a 10-year period. It concluded that 29 percent to 44 percent of the men were "over-diagnosed."
Those patients may have received surgery or radiation treatment for prostate cancer that would never have progressed so far that it threatened their health, said Ruth Etzioni, a biostatistician at the Fred Hutchinson Cancer Research Center in Seattle.
"Over those 10 years, of the men diagnosed as a result of the PSA test, up to 30 percent may have been treated unnecessarily," said Etzioni.
The findings, sure to be controversial in a society where men are often reluctant to be screened for cancer anyway, suggest doctors should be careful when discussing a prostate cancer diagnosis with patients.
The prostate specific antigen or PSA test was approved in 1986 by the U.S. Food and Drug Administration. It measures levels of a molecule produced only by prostate cells, and overproduced by prostate cancer cells.
The higher the levels, the more likely a man is to have prostate cancer. Prostate cancer is the second leading cancer killer of men, after lung cancer, affecting 189,000 men this year in the United States and killing 30,000.
The test doubled the number of men diagnosed with the disease.
Etzioni said prostate cancer is a slow-moving disease. Many older men may have prostate tumors that will never become life-threatening because they will die of other causes before the cancer has a chance to spread.
Over-diagnosis for prostate cancer is a concern because active treatment - surgery or radiation - can have debilitating side effects. Etzioni said that about 80 percent of men who have prostate cancer surgery experience impotence, often permanently. For those treated with exterior beam radiation, the impotence rate three years later is about 43 percent. About 10 percent are incontinent after surgery, she said.
Etzioni and colleagues used a computer simulation of 2 million hypothetical men aged between 60 and 84.
They compared what their model predicted would have been the prostate cancer rate between 1988 and 1998 if there was no PSA test, with what was actually registered in the National Cancer Institute's Surveillance, Epidemiology and End Results database.
They found about 29 percent of white men diagnosed with prostate cancer, and 44 percent of black men, would not have ever been diagnosed in the days before PSA testing. This suggests they had such a slow-growing form of cancer that it would never have caused significant symptoms.
The same researchers found in 1998 that only 9 percent of all American men were ever diagnosed with prostate cancer before 1986. Autopsy evidence has shown that 36 percent of white men and 28 percent of black men turn out to have had prostate cancer when they get an autopsy after dying of something else.
Based on this, the researchers said, only 4 percent of white men and 7 percent of all black men in the population will be diagnosed with cancer using a PSA test and thus over-diagnosed -- an argument that suggests most men do not risk getting an "over-diagnosis" of prostate cancer.
Dr. Darracott Vaughan, a New York urologist and a recent president of the American Urological Association, said that the Etzioni study fails to address the effect of PSA tests on the rate of survival among prostate cancer patients. Such studies are not yet complete, he said, and until they are "it's anybody's guess" about whether prostate cancer is being over-diagnosed.
He said some prostate cancer patients - those who have low PSA tests or who may be expected to die within 10 years - are now treated with "watchful waiting" instead of aggressive surgery or radiation. But he said the level of treatment should be a patient decision.
Vaughan said that PSA tests help find patients when they have a 75 percent chance of being cured of the disease. Without the test, the cancer is often not detected until symptoms develop; by then, it is 75 to 80 percent incurable because it has spread to the bones or other organs, he said.
"Our position on PSA is that we find out and then inform the patient," said Vaughan. "We give them the data about being at low risk, intermediate risk or high risk" and let them make the treatment decision.