Percent Free PSA Test May Prevent Unnecessary Biopsies

June 1, 1997

NEW ORLEANS--For prostate cancer detection, the proportion of free PSA to total PSA, using a cut-off point of 25%, is more sensitive and specific than total PSA and avoids many unnecessary biopsies, according to a multicenter study presented at the American Urological Association (AUA) annual meeting.

NEW ORLEANS--For prostate cancer detection, the proportion of free PSAto total PSA, using a cut-off point of 25%, is more sensitive and specificthan total PSA and avoids many unnecessary biopsies, according to a multicenterstudy presented at the American Urological Association (AUA) annual meeting.

Free PSA is a "nicked version" of the prostate-specific antigenthat floats freely in the blood. The free-to-total PSA ratio is known aspercent free PSA.

William Catalona, MD, professor and chief of urology at Washington UniversitySchool of Medicine, said that the 25% cut-off for free PSA allows about95% of cancers to be detected and eliminates, on average, 20% of unnecessarybiopsies, largely in older men.

"This test provides useful information, particularly about menwho are in the 'gray zone,' with PSA levels between 4 and 10 ng/mL,"Dr. Catalona said. "The problem has been that when you biopsy menwith PSAs of 4 ng/mL, only one third will actually have cancer. But ifthe patient's free PSA is less than 25%, you know he has a much greaterchance of having cancer."

The study involved 773 men screened for prostate cancer who had a normalDRE but total PSA levels of 4 to 10 ng/mL. Among the men who were diagnosedwith cancer, 95% had free PSA levels below 25%; therefore, the 25% cut-pointwas determined to be reasonable for screening men with elevated total PSA.

The implications of a free PSA level actually depend somewhat on thepatient's age, prostate size, and total PSA. If two men, aged 65 and 75,have identical free PSA readings, the older man is slightly more likelyto have prostate cancer. If two men of the same age have identical freePSA readings, the man with the smaller prostate is more likely to haveprostate cancer, Dr. Catalona said.

"But we wanted to avoid using many different cutoffs and to developa test that was practical, so we settled on the single cutoff point of25%," he said.

The study was sponsored by Hybritech, which is seeking FDA approvalto use the free PSA test diagnostically. But most companies that markettotal PSA tests are already providing the free PSA component to physicians;therefore, it can be put to use now, Dr. Catalona said. He emphasized thatthe 25% cut-off is good for the Hybritech version, but may not be applicableto other assays.

Free PSA is also predictive of tumor aggressiveness, with very low percentagesof free PSA associated with high-grade tumors, Dr. Catalona said.

For example, in two men with total PSA levels of 6 ng/mL, the man whohas 10% free PSA has about a 50% chance of having cancer; the man with30% free PSA has a 10% chance, and the cancer is more likely to be indolent.

"So the same test that allows you to diagnose prostate cancer alsoprovides you with important information about its aggressiveness,"he said.

During a panel discussion at the AUA session, other speakers agreedthat free PSA will be very useful in difficult decision-making. Dr. AlanPartin, of Johns Hopkins, said that free PSA will be useful as a stagingtool and in deciding when to repeat a biopsy.

Dr. Thomas Stamey of Stanford said the test "adds additional informationwhen total PSA does not give enough," but cautioned "not to letany patient disappear" on the basis of the test.

Swedish Study

A large population-based randomized study from Sweden evaluated thediagnostic ability of PSA with a cut-off value of 3 ng/mL and the possiblebenefit of adding percent free PSA.

Dr. Jonas Hugosson, of Ostra Hospital, Goteborg, Sweden, reported theresults from the European Randomized Study of Screening of the Prostateat the AUA annual meeting.

The investigators concluded that the 3 ng/mL cut-off value combinedwith percent free PSA of less than 18% offers better detection and betterspecificity than PSA alone at a 4 ng/mL cut-off.

The study enrolled almost 10,000 men to an active screening group andevaluated a subgroup of 608 men who had PSA levels over 3 ng/mL and hadundergone biopsy.

A PSA over 3 ng/mL led to the identification of 137 cancers. A PSA over3 and percent free PSA of less than 18% found 120 cancers. A PSA over 4led to the diagnosis of 105 cancers. And PSA over 4 plus percent free PSAless than 18% identified 97 cancers. Over half of all the cancers werenot palpable, Dr. Hugosson reported.