To compare the preoperative characteristics, postoperative prostate-specific antigen (PSA) doubling time (DT), and prostate cancer-specific mortality (PCSM) estimates after PSA failure in men diagnosed during a screening study versus a community referral population. A PSA-DT of less than 3 months is a surrogate endpoint for PCSM.

From 1988 to 2002, 1492 of 9637 patients with clinically localized prostate cancer underwent radical prostatectomy and experienced PSA failure. They were either participating in a screening study (n = 841) or attended 1 of 44 community-based practices (n = 611). The distributions of PSA, Gleason score, tumor stage, and PSA-DT were compared using chi-square metric. The estimates of PCSM after PSA failure were compared using Gray’s P value.

Compared with the community population, the annually screened men experiencing PSA failure had a lower PSA level at diagnosis (5.1 versus 9.5 ng/mL, P <0.0001), were less likely to have Gleason score 7 to 10 cancer (25.1% versus 42.1%, P <0.0001), and were more likely to have low-risk disease (64.5% versus 23.8%, P <0.0001). Furthermore, the screened cohort had a reduction (P <0.0001) in the proportion with a PSA-DT of less than 3, 3 to 5.99, and 6 to 11.99 months and a significant increase in the proportion with a PSA-DT of 12 months or longer. After a median follow-up of 4.5 and 4.1 years after PSA failure in the screened and community cohorts, respectively, the PCSM estimates were lower (P = 0.0002) in the screened cohort (10-year estimate 3.6% [95% confidence interval 1.3 to 5.8] versus 11.3% [95% confidence interval 5.9 to 17.4]).

Patients diagnosed by annual prostate cancer screening appeared more likely to experience an indolent PSA recurrence and less likely to die of prostate cancer after PSA recurrence compared with patients referred from the community.