Study: Rapidly rising PSA signals more aggressive prostate cancer

WASHINGTON (AP) Prostate cancer is more likely to be life-threatening if the man's PSA level rose rapidly during the years before he was diagnosed, says a new study that may help change how PSA tests

Wednesday, November 1st 2006, 3:20 am

By: News On 6

WASHINGTON (AP) Prostate cancer is more likely to be life-threatening if the man's PSA level rose rapidly during the years before he was diagnosed, says a new study that may help change how PSA tests are used.

The finding could help doctors diagnose aggressive cancers earlier, when they might be easier to fight.

Perhaps more important, it suggests a more in-depth evaluation of the common blood tests could better predict who needs aggressive treatment and who has a slower-growing tumor that may be OK to monitor instead.

"This is a test that doesn't just diagnose prostate cancer. It diagnoses prostate cancer that's going to actually cause harm,'' said Dr. H. Ballentine Carter, urology chief at Johns Hopkins University, who led the research published Tuesday in the Journal of the National Cancer Institute.

The study is far from proof that making health decisions based on so-called PSA velocity can really save lives.

But Carter contends the findings suggest that men should consider getting a baseline PSA test around age 40, instead of the more usual 50, to use as a comparison for future changes.

PSA tests are used to screen men for prostate cancer, but they're imprecise. Too much PSA, or prostate-specific antigen, in a man's blood can indicate that he has either a benign enlarged prostate or cancer. Only a biopsy can tell the difference.

It's not even clear when is the best time to do a biopsy. Some men have cancer despite a "normal'' PSA count of 4 or below. Yet routinely biopsying men with low PSA would worsen another problem, overdiagnosis. Many specialists say too many men today are undergoing side effect-prone treatment for tumors too small and slow-growing to ever threaten their lives.

Two years ago, Boston researchers reported that men whose PSA levels jumped more than 2 points the year before diagnosis were more likely to relapse and die despite prostate surgery. But those were men whose PSA levels were already fairly high.

Hopkins' Carter wondered if doctors could catch such men far sooner, when the cancer might be more treatable.

He turned to a study of aging that has been collecting and freezing blood samples from participants since 1958. The Hopkins team tracked PSA changes in that blood from 980 men, 20 of whom eventually died of prostate cancer and 104 of whom survived it.

How fast a man's PSA was rising a decade before his cancer was diagnosed, even before it reached that biopsy-triggering level of 4, predicted his survival 25 years later, regardless of his ultimate cancer treatment, Carter concluded.

Those with a higher PSA velocity (the level rose more than a count of 0.35 a year) had a 54 % survival rate, while those whose PSA rose more slowly had a 92 % survival rate.

What does that mean for men today? That it's a good idea to order a biopsy for a man with a low but fast-rising PSA, Carter said. And men diagnosed with prostate cancer whose PSA is rising slowly may be ideal candidates for monitoring instead of surgery or other treatment, he added.

A study with just 20 deaths is far too small to prove the value of PSA velocity, cautioned Dr. Durado Brooks, a prostate specialist with the American Cancer Society.

Still, growing numbers of doctors are using the method already to help decide when to order a biopsy, and "I think the study does raise the question as to whether PSA velocity may at some point be a helpful factor in determining prognosis,'' he said.

The work is "another step on the road to more sophisticated'' prostate cancer screening and treatment, Dr. Timothy Church of the University of Minnesota wrote in an editorial accompanying the work.

Some 234,000 U.S. men will be diagnosed with prostate cancer this year, and just over 27,000 of them will die, the cancer society estimates.