Prostate biopsy: Lessons from the Prostate Cancer Prevention Trial

March 1, 2006

Prostate cancer is more prevalent than anticipated in patients with a normal DRE and a PSA <4.0 ng/mL.

The most significant event in the chemoprevention of prostate cancer occurred with publication of the results of the Prostate Cancer Prevention Trial (PCPT) (N Engl J Med 2003; 349:215-24). In this groundbreaking study, 18,852 men with a normal digital rectal exam and a PSA value <3.0 ng/mL were randomly assigned to finasteride (Proscar), 5 mg per day, versus placebo in a 7-year study that began in 1993. Patients were subject to close annual follow-up with DRE and PSA, and recommendations for biopsy included an abnormal DRE or a PSA >4.0 ng/mL.

The most important feature of the study was the recommended end-of-study sextant biopsy of the prostate that was performed in patients not previously diagnosed as having prostate cancer, regardless of the current PSA value. Therefore, the placebo arm of the PCPT provides insight into the natural history of prostate cancer, particularly with regard to the value of PSA as a screening tool.

Recently, the value of PSA for screening has been questioned, and its value as a marker for prostate cancer, as opposed to benign hyperplasia, has decreased in the last 20 years (J Urol 2004;172:1297-301). Many now question whether an across-the-board cutoff value for biopsy versus no biopsy is still valid.

PCPT: Key observations

The placebo arm of the PCPT included 9,459 men randomly assigned to receive placebo and to undergo annual DRE and PSA, as well as the 7-year end-of-study sextant biopsy. These patients all started the trial with a DRE that was not suspicious for cancer and with a PSA <3.0 ng/mL. All underwent biopsy either "for cause" (ie, elevated PSA or abnormal DRE) or at the end of the study. Sixty-three percent of the placebo group completed the study as outlined, so the cancer status of these patients thought to be at low risk of harboring cancer at the start of the trial as determined by contemporary screening criteria was ultimately determined. Several interesting observations were made in this group.

First, prostate cancer is more prevalent than anticipated in patients with a normal DRE and a PSA <4.0 ng/mL. Initial studies of PSA as a screening tool put the detection rate of biopsy with a PSA of 4.0 to 10.0 ng/mL at 22% to 26%, and much lower for those with PSA under 4.0 ng/mL (N Engl J Med 1991; 324:1156-61; J Urol 1992; 147:841-5).

In the PCPT, the overall incidence of prostate cancer was surprisingly high at 24.8%. In the cohort with a PSA <4.0 ng/mL, 15.2% were diagnosed with prostate cancer at the end-of-study biopsy, of which 14.9% were Gleason score 7 or higher, which is only slightly lower than in the population for which biopsy is classically recommended. Even 6.6% of patients with a PSA <0.5 ng/mL had prostate cancer, so there is no lower limit of PSA below which there is complete freedom from cancer risk.