Prostate Cancer: With PSA, It's Not How Much But How Fast
A man's risk of prostate cancer may have far less to do with how much of the protein called PSA he has in his blood than with how fast it is rising. A study released this week in the Journal of the National Cancer Institute suggests that if the PSA level is climbing faster than 0.35 billionths of a gram per milliliter of blood (ng/mL) a year, a biopsy is a good ideaeven if the total isn't within sight of the 4 ng/mL level often used as a threshold of concern. "You cannot assure someone with a low PSA that they will never get cancer," stresses lead author H. Ballentine Carter, director of adult urology at Johns Hopkins Hospital in Baltimore.
Nor, studies have found, is a high PSA reading a clear indicator of cancer. That's why PSA testing is regarded with some suspicion. The American Cancer Society and the American Urological Association endorse routine screening, but other physician groups do not, and a federal task force concluded that it cannot determine whether screening for prostate cancer by using PSA tests and digital rectal exams is worthwhile.
The study led by Carter looked at velocitythe rate at which PSA levels increased over timein 980 men whose medical status had been tracked since 1958 as part of another project. All told, 124 had been diagnosed with prostate cancer, and 20 had died of the disease. The key finding: In those who died, the average PSA level during the 10 to 15 years prior to their diagnosis was only 1.9 ng/mLbut was rising fast.
The notion that PSA velocity might be a warning sign has been around since the early 1990s, recognizing that the likeliest killer cancers are aggressive and fast-growing. But an annual rise of 0.35 ng/mL sets the bar far lower than has been proposed to date. "We are suggesting that velocity can be used to identify men with a life-threatening disease when they can still be cured," says Carter. That means much earlier testing, and moving away from reliance on absolute levels. To establish a baseline, Carter advocates PSA checks starting at age 40. (Existing guidelines recommend age 50 unless factors such as family history elevate the risk.) If a 40-year-old's PSA is below the average for that age of 0.6 ng/mL, Carter would next test him at age 45, then again at age 50, and after that every two years. If the initial PSA at age 40 is above average, the test should be done every two years after that, he says.
And even if the readings never reach 4 ng/mL, Carter would take action based on a rapid rise. "I'd biopsy him at 2 to 2.5that's my threshold," he says. Conversely, even with a PSA of 4, he would not do a biopsy if the rise were more gradual.
An ongoing trial administered by the National Cancer Institute to evaluate the merits of prostate cancer screening highlights the rough-hewn quality of PSA testing. Initial results published last year, based on more than 34,000 men, found that among the 1,112 who had PSA levels above 4 ng/mL and were biopsied, only 44 percent had cancer. And in 397 others, whose PSA levels were below 4 ng/mL but who were biopsied because a digital rectal exam raised a question, 17 percent had cancer.