Predicting Prostate Cancer Death
Rapid Rise in PSA Levels Even Years Before Diagnosis May Foretell Outcome
WebMD News Archive
Oct. 31, 2006 -- Men with prostate cancer don't always need treatment, but there is no reliable way to tell which cancers are deadly and which are not.
Now, new research suggests a blood test widely used to screen for the disease can identify which patients are more likely to die from it -- and do so more than a decade before the cancercancer is even diagnosed.
In the study, Johns Hopkins School of Medicine researchers report that the rate at which prostate specific antigen (PSA) levels change over time is an accurate predictor of prostate cancer survival 25 years later.
They found that 92% of patients with a slower rise in PSA levels a decade or so prior to diagnosis were alive 25 years later. Meanwhile, only 54% of those with higher rises in PSA survived their disease 25 years later.
The PSA test measures blood levels of a protein made by the prostate. As prostate cancer cells increase in number, PSA levels can rise.
In the study, the rate of PSA rise, known as PSA velocity, was assessed among men who ended up dying of prostate cancer and those who had the disease but were still living.
The researchers found that PSA velocity 10 to 15 years before the cancer was diagnosed was a strong predictor of whether men lived or died decades later.
"It is absolutely amazing that a blood test can tell with relative accuracy who will die of prostate cancer and who will not years before diagnosis," researcher H. Ballentine Carter, MD, tells WebMD.
The PSA has become controversial as a screening tool, with critics charging it has led to the overdiagnosis and treatment of cancers that would have never affected the patient's life.
"Right now, 94% of men with a diagnosis of prostate cancerprostate cancer undergo active treatment, regardless of age," Carter says. "We are overdiagnosing and overtreating this disease, and part of my interest is to try and make an impact on that."
In the past, PSA levels of less than 4.0 ng/mL (nanograms of protein per milliliter of blood) have been considered normal, with any PSA number above that seen as suspicious.
But it is now known that men with PSA levels under 4.0 can have prostate cancercancer; and it is increasingly clear that a single PSA reading usually doesn't tell the whole story, Carter says.
"There is no single PSA level that can be used to determine if someone needs a biopsy," he says.
A better approach, Carter says, is to test PSA at regular intervals to determine how fast levels are rising.
He suggests men have a baseline PSA at age 40, with the timing of repeat tests determined by that PSA and other risk factors.
In the Johns Hopkins study, men with a PSA velocity above 0.35 per year – meaning their PSA level rose more than that per year -- were five times more likely to die of prostate cancer 25 years later than men with a slower rise in PSA.
The study is published in the Nov. 1 issue of the Journal of the National Cancer Institute.