Randomized phase II trial evaluation of erectile function after attempted unilateral cavernous nerve-sparing retropubic radical prostatectomy with versus without unilateral sural nerve grafting for clinically localized prostate cancer

Abstract

Background: Nonrandomized studies of unilateral nerve-sparing (UNS) radical prostatectomy (RP) have reported improved recovery of erectile function if the sacrificed cavernous nerve is reconstructed with a sural nerve graft (SNG).

Objective: To determine whether UNS RP plus SNG results in a 50% relative increase in potency at 2 yr compared to UNS RP alone.

Design, setting, and participants: The study enrolled patients from October 2001-May 2006 from a single academic center and was randomized, open label. Participants were men with localized prostate cancer recommended for UNS RP, less than 66 yr old, normal baseline erectile function, and willing to participate in early erectile dysfunction (ED) therapy. Patients were followed up to 2 yr.

Intervention: Patients underwent UNS RP and ED therapy starting at 6 wk: oral prostaglandin type-5 (PDE5) inhibitor, vacuum erection device (VED), and intracavernosal injection therapy. In the SNG group, a plastic surgeon performed the procedure at the time of RP.

Measurements: The ability to have an erection suitable for intercourse with or without a PDE5 inhibitor at 2 yr. The hypothesis was that SNG would result in a 60% potency rate compared to 40% for controls (80% power, 5% two-way significance).

Results and limitations: The trial planned to enroll 200 patients, but an interim analysis at 107 patients met criteria for futility and the trial was closed. For patients completing the protocol to 2 yr, potency was recovered in 32 of 45 (71%) of SNG and 14 of 21 (67%) of controls (p=0.777). By intent-to-treat analysis, potency recovered in 32 of 66 (48.5%) of SNG and 14 of 41 (34%) of controls (p=0.271). No differences were seen in time to potency or quality of life scores for ED and urinary function. Limitations included slower-than-expected accrual and poor compliance with ED therapy: <65% for VED and <40% for injections.

Conclusions: The addition of SNG to a UNS RP did not improve potency at 2 yr following surgery.

Trial registration: ClinicalTrials.gov, Identifier: NCT00080808, http://www.clinicaltrials.gov/ct2/show/NCT00080808?term=NCT00080808&rank=1.

Figure 1

Cumulative probability of potency recovery (intent to treat) by treatment group: 32 of 66 in SNG, median recovery 26 mo (95% CI of 18, 50+), and 14 of 41 controls, median recovery 36 mo (95% CI of 34, 50+).

Figure 2

CONSORT flow diagram of enrollment, allocation, follow-up, and analyses of 107 patients randomized.

Figure 3

Cumulative probability of potency recovery by treatment group among patients with 2 yr of follow-up.

Figure 4

IIEF-6 scores for SNG versus controls with standard of error bars.

Figure 5

IIEF-6 scores for patients in either group reaching the endpoint of potency versus nonpotency.