The standard, widely used approach to screen men for prostate cancer--annual PSA tests after age 50--may be less efficient and cost-effective than one that tests men earlier and less frequently, according to a study in the Sept. 19 Journal of the American Medical Association.
The study used a computer model to simulate the outcomes of prostate cancer detection under a variety of scenarios. Researchers varied the age at which men would first start regular PSA testing, varied the interval between tests and varied the PSA cutoff level urologists use for recommending a biopsy.
The favored scenario showed that screening men twice before they're 50--once at age 40 and again at age 45--and then following that with every-other-year screening beginning at age 50 "would substantially reduce the cost we spend on prostate screening. It would also prevent more deaths from prostate cancer, compared with existing screening practices," says Hopkins urologist H. Ballentine Carter, a member of the research team.
Compared with the standard way of screening, he says, "the model showed you'd do 3,000 fewer PSA tests and 200 fewer biopsies for every 1,000 men screened over a lifetime."
"What we've found doesn't mean the world should change screening practices tomorrow. It is, after all, a computer simulation," Carter adds. "But because this simulation is trustworthy, it's very strong evidence indeed that what we're doing now is probably not the best way."
Researchers from Johns Hopkins, Merck Research labs and the University of North Carolina collaborated on the study.
Under current guidelines for prostate cancer screening, men get an annual PSA (prostate specific antigen) test after age 50, with a follow-up biopsy if the PSA level is higher than 4 nanograms per milliliter of blood. "There's no science to show that you should start at age 50 and test annually," Carter says. "That strategy was adopted because it seemed reasonable, especially in light of women getting an annual mammogram for breast cancer. But prostate cancer tends to grow and progress more slowly than breast cancer, so yearly screening is probably not necessary.
"In fact," Carter adds, "no good clinical research exists to confirm that PSA screening in general reduces deaths from prostate cancer. Such very expensive randomized studies, where large numbers of patients are followed, are now under way in Europe and the United States. But our study assumes PSA screening is useful and goes on to focus on the best way to do it."
The model simulated a population of 1 million men, followed beginning at age 40 for 40 years.
Under the "ground rules" used as a basis for their model--real data on the way prostate cancer progresses over the years, on results of treatment, on who gets the disease, on what PSA tests and biopsies can show--the researchers first produced a basic model of the natural history of unchecked prostate cancer. "That's what you'd get with no screening," Carter says.
Then they modeled different PSA screening strategies. The researchers compared each strategy, with outcomes translated into the number of deaths it would prevent and into the number of PSA tests and biopsies that strategy required.
The researchers began their work because one of Carter's earlier studies showed younger men tend to have a higher cure rate from prostate cancer, probably because their tumors are smaller and more confined. "Screening earlier and detecting the disease at a younger age may decrease the number of deaths from prostate cancer in this country," Carter says.
"We can't translate this yet into universal guidelines. The purpose of this study now is to raise awareness. What needs to take place are confirmatory studies with real patients. But based on this study, some physicians and their patients may choose to begin PSA testing before age 50."
The research was funded by grants from the National Cancer Institute. Other researchers were Kevin S. Ross, now a medical student at Hopkins and formerly with the University of North Carolina, Chapel Hill; Harry A. Guess, with the University of North Carolina and the Merck Research labs in Blue Bell, Pa.; and Jay Pearson, with Hopkins and Merck.
For more information on prostate research at Hopkins, logon to prostate.urol.jhu.edu.