Objectives. To evaluate prostate cancer detection and prostate-specific antigen (PSA) among men with benign prostatic hyperplasia treated with finasteride.

Methods. Three thousand forty men 45 to 78 years of age with PSA less than 10 ng/mL and no history of prostate cancer were randomized in a double-blind, placebo-controlled trial to finasteride (n = 1524) or placebo (n = 1516) for up to 4 years. A prerandomization biopsy negative for prostate cancer was obtained in 98% of patients with a screening PSA of 4.0 ng/mL or more, and an end-of-study biopsy was requested of all such patients without a recent second negative biopsy or a prostate cancer diagnosis.

Results. Overall, 644 patients (21%) underwent biopsy and 201 (6.6%) underwent transurethral resection of the prostate. Prostate cancer was diagnosed in 4.7% of men on finasteride and 5.1% on placebo (P = 0.7). Elevated PSA prompted diagnosis in 35% of cases on finasteride and 34% on placebo. The area under the receiver operating characteristic curve for last PSA was 0.84 on finasteride and 0.79 on placebo (P = 0.07). Use of an upper limit of normal for last PSA of 2.0 ng/mL for finasteride and 4.0 ng/mL for placebo yielded similar sensitivity (66% versus 70%, P = 0.6), higher specificity (82% versus 74%, P < 0.0001), and a higher likelihood ratio (3.6 versus 2.7, P < 0.05) for finasteride than for placebo.

Conclusions. In men treated with finasteride, multiplying PSA by 2 and using normal ranges for untreated men preserves the usefulness of PSA for prostate cancer detection.