Medical Device Daily

ATLANTA – The question of whether the prostate-specific antigen (PSA) test for prostate cancer still is relevant was reintroduced to current discussion with a study published in the October 2004 issue of the Journal of Urology, Joseph Presti Jr., MD, of Stanford University (Palo Alto, California), said at the American Urological Association (AUA; Linthicum, Maryland) annual meeting here Monday.

That article, authored by Thomas Stamey, MD, a professor of urology at Stanford, suggested that the era of the PSA test in the U.S. is over because tumor size no longer correlates to PSA values in American men.

According to Presti, the PSA test “wasn't really clinically utilized until 1986-87,” and even at that time, there was correlation between tumor size and PSA value only about 44% of the time.

“One could argue that this relationship was weak to begin with,” he said. However, it still finds more cancer than mammography, which is used routinely in screening for breast cancer in women, despite the belief that it misses tumors in some women.

Presti, who presented a study on the performance of PSA levels between 4.0 and 10 ng/ml in the era of extended biopsy schemes, maintained that the PSA test still is relevant and valuable.

“As the PSA goes up, the chance of finding prostate cancer goes up,” he said.

But in the current era, doctors also conduct extended biopsies, which means that 10 to 20 biopsies may be done of a single prostate once it is removed after being found to be questionable by a digital rectal exam.

Presti said that many cancers once were missed by doing fewer core biopsies of the prostate, but today, “we miss very few.”

His study states that “several variables can influence positive biopsy rates, including patient age and PSA levels.” The study “assessed the performance of PSA in a contemporary biopsy population stratified for patient age, digital rectal exam and PSA levels.”

There were 999 first-time biopsy patients between the ages of 50 and 79 with PSA levels between 4 ng/ml and 10 ng/ml, all of whom were referred to Stanford following an abnormal digital rectal exam. According to the study, the positive predictive value of PSA was determined for several outcomes: presence of any cancer; presence of a significant cancer, defined as a cumulative length of greater than or equal to 3mm of cancer; and presence of high-grade cancer, or a Gleason score of greater than or equal to 7.

The conclusion of the study – that “PSA is a good test for the detection of prostate cancer” – was based on the finding that in men with a normal DRE, the positive predictive value of a PSA between 4 ng/ml and 10 ng/ml ranges from about 25% to 68% to find any cancer and 12% to 53% to find any cancer.

In another study of 400 men presented by Atsushi Ochiai, of the M.D. Anderson Cancer Center (Houston), the relationship of PSA value to tumor volume and noncancerous prostatic tissue volume was evaluated using multivariate analysis in men undergoing prostatectomy during two different time periods.

Ochiai concluded that “the relationship of PSA to tumor volume as well as noncancerous prostate tissue volume has decreased in recent years, but PSA is still significantly associated with tumor volume and noncancerous prostate tissue volume.”

A third presenter, Danil Makarov, of Johns Hopkins University Medical Center (Baltimore), who addressed the question of whether PSA standard cut-off values should be lowered, evaluated “pathologic outcomes and biochemical progression” in men with normal rectal exams but undergoing radical prostatectomy and who had a PSA test value of 2.6 ng/ml -4.0 ng/ml vs. the “more traditional” value of greater than 4.0 ng/ml.

“We sought to compare outcomes between men with clinical stage T1c disease undergoing radical prostatectomy who had a 'low' versus 'slightly elevated' PSA,” the study said.

The study population included 2,896 men treated by radical prostatectomy between 1985 and 2004 who had normal rectal exams and a pre-operative PSA between 2.6 ng/ml and 6.0 ng/ml.

The study concluded that in “men with clinical-stage T1c treated by radical prostatectomy, a lower pre-operative PSA was associated with significantly more favorable pathological findings.” However, the study also concluded that “whether this degree of improved outcomes justifies the limitations associated with lower the PSA cut-point [e.g., increased biopsies performed and diagnosis of insignificant cancers] remains to be determined.”

“There is no PSA so low that we can guarantee there is no cancer,” Makarov said, noting later that a question to consider is whether physicians are getting better outcomes or are they diagnosing detectable disease earlier?

Presti added: “If PSA really meant nothing, then [the study at Johns Hopkins] shouldn't have seen a difference in PSA . . . I look at this paper telling us very strongly [that] PSA [does tell us something].”