To determine the value of intraoperative biopsy during radical retropubic prostatectomy.

Between October 2000 and August 2002, 500 men with clinically localized prostate cancer underwent radical retropubic prostatectomy by a single surgeon. A 2 to 3-mm circumferential biopsy was routinely obtained from the apical and bladder neck soft-tissue margin and submitted for frozen section examination. In selective cases suspicious for capsular incision, a biopsy was sent from what was believed to be the contiguous neurovascular bundle/lateral pedicle.

Prostate cancer was observed in 4.5%, 0.8%, and 1.6% of the intraoperative biopsies sent from the apical, bladder neck, and neurovascular bundle/lateral pedicle soft-tissue margins, respectively. Patient age, Gleason score, perineural invasion on diagnostic prostate biopsy, and clinical stage were not associated with prostate cancer at the apical soft-tissue margin. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of the surgical specimen to predict cancer in the apical soft-tissue margin was 57.7%, 98.2%, 62%, 97.7%, and 96%, respectively. Intraoperative biopsy of the apical soft-tissue margin reduced the positive margin rate by 3.8%.

The yield of intraoperative biopsy of the bladder neck and neurovascular bundle/lateral pedicle is too low to justify it in routine practice. Biopsy of the apical soft tissue should be routinely performed to reduce the positive surgical margin rate.