Sept. 10, 2004 -- PSA screening for prostate cancer is dead, one of the test's earliest advocates now says.
PSA stands for prostate-specific antigen. It's a protein thrown off by a growing prostate gland. Growing prostate cancers give off a lot of PSA. So blood tests for PSA are widely used as an early warning system for prostate cancer.
Experts disagree on how often, or even if, men should have their PSA regularly checked. Some experts recommend annual PSAs starting at age 50 -- or age 40 for men at high risk of prostate cancer, such as black men or men with a family history of it.
But there's no proof that PSA screening saves lives. Thomas A. Stamey, MD, founding chairman of the Stanford University urology department, led many of the studies that led to the wide acceptance of PSA screening. Now, Stamey says, the PSA era is over.
"You can biopsy according to whether a man has blue eyes or green eyes and get pretty much the same results as biopsying according to PSA," Stamey tells WebMD. "We can't understand this disease except by first knowing we men all get it, and, second, by knowing that we are very unlikely to die from it."
Stamey says that to put PSA screening into context, two things must be understood:
- Nearly all men eventually get prostate cancer. It's found in some 8% of men in their 20s and in 80% of men in their late 70s, Stamey says.
- Relatively few men die of prostate cancer. In the U.S., Stamey says, there are 226 prostate-cancer deaths for every 100,000 men. So a man has only a 0.2% chance of dying from the disease.
20 Years and 1,317 Prostates
Stamey's latest findings -- reported in the October issue of the Journal of Urology -- come from a study of 1,317 prostate glands removed by the same surgeon between 1983 and 2003. All of the patients had PSA tests. All had biopsies that showed signs of cancer.
In prostates removed during the earliest five years of the study, PSA levels predicted cancer severity. That is, the higher the patient's PSA, the larger and more aggressive the patient's prostate cancer.
But in the last five years of the study, PSA levels had little if anything to do with cancer severity. All PSA levels predicted, Stamey and colleagues found, was an enlarged prostate -- and all men's prostates enlarge as they age.
"In the first five years, we had a correlation of 60% between high PSA levels and severe cancer -- so the PSA test wasn't perfect even in the early years," Stamey say. "But my paper shows it is a straight curve down to a 2% correlation in the last five years. We show PSA is driven by benign enlargement of the prostate."
What's going on? Urologist H. Ballentine Carter, MD, a researcher at the Johns Hopkins Brady Urological Institute in Baltimore, explains.
"At the beginning of the PSA era, we were pulling out of the population patients with larger tumors, where there was better correlation between PSA and cancer volume," Carter tells WebMD. "Dr. Stamey's contention is that as the PSA era has gone on, we have culled those large tumors. Now we are detecting very, very small tumors where there is less correlation between PSA and cancer."
Lots of Prostate Cancers, Small Chances of Death
Prostate cancer, Stamey says, is very common. He points to autopsy data showing that while many men have prostate cancer, few die of it.
"Prostate cancer is a paper tiger in terms of the number of men who die from it," Stamey says.
PSA tests send many men to their urologists for biopsies. And these biopsies often find cancer. The question is whether these really are dangerous cancers, says Howard Parnes, MD, chief of the prostate and urological cancer research group at the National Cancer Institute.
"Sixteen percent of men in the U.S. will be diagnosed with prostate cancer," Parnes tells WebMD. "About 86% of men diagnosed with prostate cancer end up with definitive treatment, either surgery or radiation. But only 3% of these men are destined to die. ... What is scary is that if we look hard for it, if we test men every year, and if we biopsy men with a PSA over 4, we can find prostate cancer in about 25% of men."
Stamey agrees that if you look hard for prostate cancer, you'll often find it. But that alone doesn't tell doctors or patients who is at risk of dying.
"When a man submits to biopsy, he might as well be resigned to a positive result," Stamey says. "But if we don't counteract this with knowing how low the death rate is, we overestimate the significance of a positive biopsy."
PSA Still Useful
Even Stamey agrees that PSA tests are useful for some purposes, such as monitoring the success of prostate cancer surgery.
Carter argues that PSA tests aren't used properly.
"We could improve the test by focusing our efforts on younger men, where the discovery of cancer is more relevant and PSA is a much better test," he says. "But we are screening the wrong age group. We are treating the wrong age group."
Carter also argues that single PSA measurements give much less information than multiple tests. He'd like to see the test used differently.
"I think the way to use this test correctly, which gets rid of a lot of the noise in the test -- the false positives -- is to start PSA screening at age 40. Then you'd test again at age 45, and every year after 50, and put together a long PSA history. Then you could use that PSA trajectory to tell us who is gong to get into trouble."
By the time a man is 70, Carter says, PSA testing no longer yields valuable information.
Parnes says there are three things men should know about PSA screening:
- It's not yet known whether prostate cancer screening saves lives. A major government study is looking at the question, but the results won't be known for years.
- Screening poses risks as well as benefits. Yes, early detection can save the life of a person with an aggressive cancer. But biopsies cost money, and they are no picnic. And prostate cancer treatments run high risks of leaving men with urinary incontinence and erectile dysfunction.
- In Parnes' opinion, PSA tests are hard to evaluate. "There is no magic PSA threshold below which a man can be assured there is no risk, or above which a biopsy should be automatically performed," he says.
Stamey, Parnes, and Carter all advise men to discuss their risk of prostate cancer -- and the risks of prostate cancer screening -- with their doctors.