PSA Density a More Accurate Screen Than PSA

September 1, 1995

SEATTLE--A study of 1,695 cancer-free men found that PSA density provides a far more accurate screening assay for detection of prostate cancer than PSA serum concentration, Robert Kane, MD, of Harvard Medical School, said at the Pacific Northwest Cancer Foundation Meeting on Transperineal Brachytherapy for Early Stage Prostate Cancer.

SEATTLE--A study of 1,695 cancer-free men found that PSA densityprovides a far more accurate screening assay for detection ofprostate cancer than PSA serum concentration, Robert Kane, MD,of Harvard Medical School, said at the Pacific Northwest CancerFoundation Meeting on Transperineal Brachytherapy for Early StageProstate Cancer.

PSA density is calculated by relating the serum PSA concentration(ng/mL) to the volume of the prostate gland in a particular patient.

Serum PSA testing is very useful for following patients with knownprostate cancer, to assess the effectiveness of therapy and theprogression of the cancer, Dr. Kane said. But for prostate cancersurveillance, the considerable number of false-positives in menwho simply have enlarged glands results in emotional cost to thepatient and substantial economic cost to the health-care system.

Lacks Specificity in Both Directions

The serum PSA test also lacks specificity in the other direction.Dr. Kane reported that in several studies, 23% to 43% of patientswith organ-confined prostate cancer had normal PSA values.

The multicenter study, conducted by Dr. Kane and other investigatorsas part of the American Cancer Society National Prostate CancerDetection project (Cancer, March 1, 1992), analyzed the distributionsof PSA concentrations in cancer-free men overall and also accordingto differing prostate gland volumes, patient age, and the presenceor absence of symptoms of benign prostatic hypertrophy (BPH).

They found that neither age nor symptoms of BPH appeared to haveany effect on PSA levels when adjusted for gland volume. Theyconcluded that estimating PSA density (based on ng/mL/cc glandvolume) rather than relying on absolute serum levels may makeit possible to reduce the number of false-positive PSA resultsin the general population.

This approach would require a single initial transrectal ultrasonography(TRUS) examination to determine gland volume. Patients withinthe normal volume-adjusted PSA range could then be followed annuallywith serum PSA testing alone. Those with an abnormal volume-adjustedPSA level would be selected for further TRUS examination and ultrasound-guidedbiopsies as needed.

Dr. Kane noted that this approach might make the PSA test morecostly to apply on a population-wide basis, but the benefits oflowered false-positive results might outweigh those costs.

Another way to keep costs down, Dr. Kane said, would be to useultrasound only on those men who appear to have serum PSA levelsabove the normal cutoff level of 4 ng/mL for the monoclonal assay.