PSA Screening Imperfect, but It Works

Despite Its Failings, Prostate Cancer Test Catches Early Tumors

From the WebMD Archives

May 22, 2006 – The PSA test may not be perfect, but it's a good way to screen for prostate cancerprostate cancer, a leading urologist says.

Perhaps no subject in medicine is more controversial than the PSA test. This simple blood test measures the level of PSA -- prostate specific antigen. Since only prostate cells give off PSA, a rise in PSA could mean a prostate tumor is growing.

The problem is that not everyone with a high PSA level has prostate cancercancer. And not everyone with prostate cancer has a high PSA level. This uncertainty leads to many unnecessary prostate biopsies -- and to many unnecessary prostate surgeries or radiation treatments.

Is it time to toss the test? No, argues Stanford researcher Joseph C. Presti Jr., MD. Presti spoke at a news conference at this week's annual meeting of the American Urological Association in Atlanta.

"We now recognize there is no normal PSA level," Presti says. "But as PSA levels go up, the percentage of patients with prostate cancer goes up. So PSA is telling us something about cancer."

How Well Does PSA Predict Cancer?

Presti and colleagues studied 999 first-time prostate biopsy patients to see how well an elevated prostate level -- a score of 4 to 10 on the PSA test -- predicted cancer. Digital rectal exams failed to find anything abnormal in all of these men.

Presti's team found that PSA scores of 4-10 predicted cancer less often in younger (50-ish) men and more often in older (70-ish) men. In men with a normal rectal exam, a PSA of 4 to 10:

  • Predicted any prostate cancer 25% to 68% of the time.
  • Predicted high-grade cancer 12% to 53% of the time.

"PSA is still very useful in the early detection of prostate cancer, and does relate to the grade of the cancer and the chance of a cure," Presti said. "We don't have definitive answers. We do overdiagnose this disease. But most of us still contend it is better to find the cancer than not to find it."

Lower Pre-Op PSA Linked to Better Outcome

A meeting report by Johns Hopkins researcher Danil V. Makarov, MD, highlights the uncertainty surrounding PSA tests.


Makarov and colleagues analyzed prostates removed from nearly 3,000 patients. All these patients had relatively low PSA scores -- from 2.6 to 4.0.

The researchers found that patients with the lowest preoperative PSA scores had the least dangerous cancers.

"Why do men undergoing radical prostatectomy with lower PSA scores have better outcomes?" Makarov asked. "Are we treating men that never needed to be treated in the first place? Or are we recognizing truly significant disease earlier? Unfortunately, our data cannot distinguish between these outcomes."

Presti said the Makarov results support PSA testing.

"If PSA really meant nothing regarding cancer, Markov should not have seen differences in pathologic outcomes," he said. "What is impressive is that he saw that small incremental increases in PSA levels make a difference in the cancer volume we find when the prostate is removed. PSA is telling us something really strongly about that cancer."

WebMD Health News


SOURCES: Annual Meeting of the American Urological Association, Atlanta, May 20-25, 2006. Joseph C. Presti Jr., MD, Stanford University; abstract 468. Danil V. Makarov, MD, Johns Hopkins University, Baltimore, Md; abstract 474.
© 2006 WebMD, Inc. All rights reserved.