The effect of preservation of neurovascular bundles (NVBs) during radical prostatectomy (RP) on continence remains controversial.
To analyze if the differing surgical techniques of nerve-sparing (NS) versus non–nerve-sparing (NNS) RP and not the preservation of the NVB itself may be responsible for differences in continence rates.
Design, setting, and participants
A total of 18 427 men who underwent RP from 2002 to 2014 in a single high-volume center were analyzed retrospectively. Patients with bilateral NS RP, with primary NNS RP, and with bilateral secondary resection of the NVBs for positive frozen-section results after an initial bilateral nerve sparing (secNNS) RP were studied.
NS, NNS, or secNNS RP.
Outcome measurements and statistical analysis
Multivariable and propensity score matched analyses adjusting for age, prostate volume, and year of surgery were performed to assess differences in continence rates after RP. Continence was defined as the use of no or one safety pad per day.
Results and limitations
Post-RP urinary continence rates at 1 wk, 3 mo, and 12 mo were 59.8%, 76.2%, 85.4% in the NS group, 39.5%, 59.5%, and 87.0% in the secNNS group, and 29.1%, 52.8%, and 70.5% in the NNS group. Continence rates at 12 mo after surgery did not differ significantly between patients who had bilateral NS and patients who had resection of both NVBs after an initial nerve-sparing technique (secNNS). In contrast, when comparing the NNS study groups with initial NNS versus secNNS, the latter group had significantly higher continence rates after 12 mo.
Our results indicate that the meticulous apical dissection associated with the NS RP technique rather than the preservation of the NVBs itself may have a positive impact on long-term urinary continence rates.
We looked at continence rates after nerve-sparing (NS) versus non–NS radical prostatectomy (RP). NS surgery technique but not the preservation of the neurovascular bundles led to improved long-term continence rates after RP.
Keywords: Prostate cancer, Radical prostatectomy, Nerve sparing, Continence.
Radical prostatectomy (RP) is a common treatment option for prostate cancer but carries a significant risk of urinary incontinence and erectile dysfunction . Although the effect of the preservation of the neurovascular bundles (NVBs) during RP on erectile function is evident, its influence on continence remains controversial. In fact, to date it is unknown if NVBs contribute to the neural supply of the urethral sphincter .
Reeves and colleagues recently summarized the existing studies on the influence of the preservation of the NVBs on continence after RP and conducted a meta-analysis using data from 13 749 patients in 27 studies . Their meta-analysis demonstrated improved early urinary continence rates (up to 6 mo postoperatively) for patients undergoing nerve-sparing (NS) RP compared with patients undergoing non–nerve-sparing (NNS) RP. Long-term continence did not differ in the two patient groups. Steineck et al reported a strong association between the degree of NVB preservation and urinary incontinence 1 yr after surgery .
Some authors have argued that the difference in continence rates may be attributable to the differing surgical techniques of NS versus NNS RP . The meticulous apical dissection associated with the NS RP technique may be responsible for the improved continence outcome in this patient group  and .
To answer this important question, it is necessary to compare continence rates in patients with and without preservation of the NVBs who otherwise underwent the same surgical approach. Since the introduction of a neurovascular structure-adjacent frozen-section examination (NeuroSAFE) in our institution, most RPs were performed using a NS approach with secondary resection of the NVBs in case of positive frozen section .
The objective of the present study was to compare continence rates of patients from a large European high-volume center undergoing primary NNS RP, NS RP with preservation of the NVBs, and initially NS RP with secondary resection of the NVBs.
2. Patients and methods
2.1. Patient population
A total of 18 427 consecutive patients who underwent RP from January 2002 to October 2014 in the Martini-Clinic were analyzed retrospectively. Patients with unilateral NS (n = 4485), salvage prostatectomy (n = 70), or missing follow-up (n = 1339) were excluded from the study. Data were collected prospectively into a database approved by the institutional review board. Functional data were assessed using self-administrated questionnaires. RP was performed using an open retropubic approach or robot-assisted laparoscopic approach, as previously described  and . The following three patient groups were studied: patients with bilateral NS RP, patients with primary bilateral NNS RP, and patients with bilateral secondary resection of the NVBs for positive frozen-section results after an initial bilateral NS (secNNS) RP. Continence was defined as the use of zero or one safety pad per day.
Baseline characteristics between the three patient groups were compared using the likelihood ratio chi-square test. We first compared the functional outcome in the three investigated patient groups by comparing the use of pads per day at 1 wk, 3 mo, and 1 yr after RP using the chi-square test. In logistic regression analysis including age at surgery, prostate volume, and year of surgery as confounders we compared the continence status defined as zero or one safety pad in a multivariable setting. We validated our results using propensity score matching analysis. Propensity score matched analysis was performed in a regression model using age, prostate volume, and year of surgery as covariates. Based on estimated propensity scores, one patient from the NS or NNS group was matched (best match) to one patient of the secNNS group by caliper matching  and  and using nonrandom R package (R Project for Statistical Computing; www.R-project.org) . The selection process was conducted without replacement so that each patient was selected only once. After matching, differences in continence rates 1 wk, 3 mo, and 12 mo after RP between the study groups were assessed. All tests were two tailed, and p values <0.05 were considered statistically significant. Statistical analyses were performed with JMP software, v.9.0.2 (SAS, USA) and RStudio, v.0.98.945 (http://http://www.rstudio.com), an integrated environment development for R (v.3.1.0, http://www.R-project.org).
3.1. Clinical and pathologic characteristics
Table 1 compares the clinical and pathologic features of the patients. Median age in the entire group was 64 yr (interquartile range: 59–68). Bilateral NS RP was performed in 11 204 patients (89.4%), NNS RP in 1128 patients (9.0%), and secNNS in 201 patients (1.6%).
|Parameter||Overall||NS (%)||secNNS (%)||NNS (%)||p value|
|Patients, n||12 533||11 204 (89.4)||201 (1.6)||1128 (9)|
|RARP||1731||1582 (14.1)||41 (20.4)||108 (9.6)||<0.001|
|Open RP||10 802||9622 (85.9)||160 (79.6)||1020 (90.4)|
|Age at surgery, yr|
|Prostate volume, ml|
|Year of surgery|
|2002||387||338 (3)||1 (0.5)||48 (4.3)||<0.001|
|2003||447||420 (3.7)||0 (0)||27 (2.4)|
|2004||546||525 (4.7)||2 (1)||19 (1.7)|
|2005||573||527 (4.7)||3 (1.5)||43 (3.8)|
|2006||756||676 (6)||3 (1.5)||77 (6.8)|
|2007||924||821 (7.3)||14 (7)||89 (7.9)|
|2008||1128||990 (8.8)||13 (6.5)||125 (11.1)|
|2009||1215||1052 (9.4)||28 (13.9)||135 (12)|
|2010||1416||1246 (11.1)||27 (13.4)||143 (12.7)|
|2011||1419||1288 (11.5)||19 (9.5)||112 (9.9)|
|2012||1366||1230 (11)||27 (13.4)||109 (9.7)|
|2013||1326||1174 (10.5)||33 (16.4)||119 (10.5)|
|2014||1030||917 (8.2)||31 (15.4)||82 (7.3)|
|<4||1711||1587 (14.7)||11 (5.8)||113 (10.3)||<0.001|
|4–10||7558||7115 (65.8)||108 (56.5)||335 (30.7)|
|10–20||2075||1759 (16.3)||40 (20.9)||276 (25.3)|
|>20||753||353 (3.3)||32 (16.8)||368 (33.7)|
|pT2||9664||9387 (83.9)||38 (18.9)||239 (21.2)||<0.001|
|pT3a||1645||1262 (11.3)||102 (50.7)||281 (25)|
|pT3b||1161||538 (4.8)||55 (27.4)||568 (50.5)|
|pT4||45||2 (0)||6 (3)||37 (3.3)|
|Nx||5414||5326 (47.7)||12 (6)||76 (6.8)||<0.001|
|N0||6361||5613 (50.3)||155 (77.1)||593 (52.7)|
|N+||719||229 (2.1)||34 (16.9)||456 (40.5)|
|3 + 3||3438||3398 (30.4)||5 (2.5)||35 (3.1)||<0.001|
|3 + 4||6947||6518 (58.3)||121 (60.2)||308 (27.5)|
|4 + 3||1471||1052 (9.4)||45 (22.4)||374 (33.3)|
|4 + 4||641||206 (1.8)||30 (14.9)||405 (36.1)|
|R0||10 411||9742 (87.1)||125 (62.2)||544 (48.3)||<0.001|
|R1||2106||1448 (12.9)||76 (37.8)||582 (51.7)|
IQR = interquartile range; NNS = non–nerve sparing; NS = nerve sparing; secNNS = secondary non–nerve-sparing; PSA = prostate-specific antigen; RARP = robot-assisted radical prostatectomy; RP = radical prostatectomy.
3.2. Continence rates
In the unadjusted analysis, urinary continence rates (defined as zero or one safety pad) at 1 wk, 3 mo, and 12 mo after surgery were 59.8%, 76.2%, and 85.4%, respectively, in the NS group; 39.5%, 59.5%, and 87.0%, respectively, in the secNNS group; and 29.1%, 52.8%, and 70.5%, respectively, in the NNS group (Table 2).
|Time point||No. of pads in 24 h||NS (%)||secNNS (%)||NNS (%)||p value*||p value #|
|1 wk||0 or 1 safety||3315 (59.8)||49 (39.5)||125 (29.1)||<0.001||0.004|
|2||1033 (18.6)||31 (25.0)||73 (17.0)|
|3||601 (10.8)||18 (14.5)||88 (20.5)|
|≥4||594 (10.7)||26 (21.0)||144 (33.5)|
|3 mo||0 or 1 safety||3172 (76.2)||44 (59.5)||152 (52.8)||0.02||0.23|
|1–2||780 (18.7)||23 (31.1)||82 (28.5)|
|3–5||184 (4.4)||6 (8.1)||42 (14.6)|
|>5||26 (0.6)||1 (1.4)||12 (4.2)|
|1 yr||5269 (70.8)||74 (68.5)||365 (50.6)||0.45||0.001|
|1 safety||1087 (14.6)||20 (18.5)||144 (19.9)|
|1–2||884 (11.9)||10 (9.3)||152 (21.1)|
|3–5||163 (2.2)||4 (3.7)||46 (6.4)|
|>5||34 (0.5)||0 (0)||15 (2.1)|
* NS vs secNNS.
# secNNS vs NNS.
NNS = non–nerve sparing; NS = nerve sparing; secNNS = secondary non–nerve-sparing.
When comparing the NS and the secNNS groups, who were subject to the same surgical technique, 1 wk and 3 mo continence rates were higher in the NS group (59.8% vs 39.5%; p < 0.0001 and 76.2% vs 59.5%; p = 0.018). Urinary continence rates in both groups at 12 mo after surgery were comparable (85.4% vs 87.0%; p = 0.5). In contrast, when comparing the NNS study groups with initial NNS (NNS) versus secondary NNS (secNNS), the secNNS group had significantly higher continence rates after 12 mo (87.0% vs 70.5%; p = 0.001). The 1 wk (39.5% and 29.1%; p = 0.004) continence rate was also superior in the secNNS patients compared with the NNS group; the 3 mo (59.5% and 52.8%; p = 0.23) continence rate only tended to be better without reaching statistical significance.
Multivariable logistic regression analysis showed a higher probability of being continent in the NS group compared with the secNNS group at 1 wk and 3 mo (Table 3). No statistical difference between both groups was found at 1 yr after surgery. NNS and secNNS showed no statistical significant differences in continence rate at 1 wk and 3 mo. However, at 1 yr after surgery there was a significantly lower probability of being continent in the NNS group when compared with the secNNS group (Table 3).
|Parameter||1 wk||3 mo||1 yr|
|Odds ratio||95% CI||p value||Odds ratio||95% CI||p value||Odds ratio||95% CI||p|
|NS vs secNNS||2.29||1.58–3.35||<0.001||2.28||1.41–3.67||0.001||1.04||0.56–1.78||0.9|
|NNS vs secNNS||0.71||0.46–1.09||0.11||0.87||0.51–1.46||0.6||0.43||0.23–0.75||0.005|
|Age at surgery||0.96||0.96–0.97||<0.001||0.97||0.96–0.98||<0.001||0.97||0.96–0.98||<0.001|
|Year of surgery||0.9||0.87–0.93||<0.001||1.08||1.04–1.12||<0.001||1.15||1.13–1.18||<0.001|
CI = confidence interval; NNS = non–nerve sparing; NS = nerve sparing; secNNS = secondary non–nerve-sparing.
Propensity-matched analyses were performed for the NS versus secNNS groups and the secNNS versus the NNS groups for each of the three time points (1 wk, 3 mo, and 12 mo) after RP (Supplementary Table 1–3). These analyses confirmed the results from the multivariable analyses just described (Supplementary Table 4).
Despite a relatively large body of literature including a recent meta-analysis on the influence of NS on post-RP continence, the true effect of the preserved NVBs on continence remains unclear because published studies have not only compared the preservation of NVBs versus nonpreservation of them, but moreover one surgical approach to the other. Existing studies have been clearly imbalanced with regard to other parameters influencing post-RP continence rates such as age .
To overcome these limitations, in the present study we compared continence rates in patients with and without preservation of the NVBs who otherwise underwent the same surgical approach. To further exclude inherent biases, propensity score matching methodology was used. During the last decade most RPs in our institution have been performed utilizing the NeuroSAFE technique that includes an identical surgery technique for both the NS and the secNNS groups apart from the secondary resection of both NVBs in the latter group for positive frozen-section results .
Reeves and colleagues recently summarized the literature on the subject and performed a meta-analysis reporting 6 wk, 3–4 mo, 6 mo, and 12 mo continence rates of 42.2%, 64.8%, 88.9%, and 83.9% in the NS patients compared with 26.8%, 47.6%, 69.8%, and 75.2% in the NNS patients, respectively . The difference in continence rates of both patient groups was significant up to 6 mo after RP in their meta-analysis but not beyond that.
In contrast, our results showed a significant difference for long-term continence rates (12 mo) between the NS and NNS groups (85.4% vs 70.5%). This divergent finding may be due to important bias in the studies used for the meta-analysis that may have influenced the results . Reeves et al documented significant variations in baseline characteristics of the compared groups. In addition, the authors could not control for surgical factors such as surgeon experience and differences in surgical technique.
Interestingly, we even found a significant difference in long-term continence of the secNNS and NNS groups (87.0% vs 70.5%), but not of the secNNS and NS groups. When comparing the latter groups, who were subject to the exact same surgical technique, 1 wk and 3 mo continence rates were higher in the NS group. Urinary continence rates at 12 mo after surgery were similar in both groups (87.0% vs 85.4%). In contrast, patients with an initial NNS approach did not reach early and long-term continence rates comparable with patients with a NS surgical approach and possible secondary resection of NVBs (ie, NS and secNNS). These results indicate that the meticulous apical dissection associated with the NS RP technique may spare important nerve supply along its course other than the NVBs themselves to the urethral sphincter and is important for early and long-term urinary continence.
The urethral sphincter is composed of an inner smooth muscle sphincter and an outer rhabdosphincter of skeletal muscle  and . Autonomic branches innervate the smooth muscle of the urethra . Bladder neck closure by smooth muscle contraction is achieved by sympathetic stimulation and leads to urinary continence and prevention of reflux of the ejaculate into the bladder . The rhabdosphincter is supplied by the pudendal nerve  and .
Narayan and colleagues conducted anatomic studies on 18 adult male cadavers and showed that the rhabdosphincter may additionally be innervated by branches in close proximity to the prostatic apex originating from the dorsal nerve of the penis after it splits from the pudendal nerve . Similarly, Song et al studied 15 male cadavers and found intrapelvic branches from the pudendal nerve innervating the membranous urethra and located 5.3 ± 1.8 mm from the prostatic apex . Zvara et al reported that the nerve supply to the striated urethral sphincter varies greatly and is composed of branches from the pudendal nerve as well as “extrapudendal nerves” that run inside the pelvis . Recently it was described that cavernous nerves are not limited to the periprostatic region, but also cover the rhabdosphincter and perirectum caudal to the level of the apex of the prostate .
Taken together, these studies have supported the notion that the meticulous apical dissection associated with the NS RP technique, but not the NVBs preservation itself, may be responsible for the improved continence observed in these patients due to protection of the described pudendal branches. Besides that, meticulous apical dissection also may allow the preservation of a longer functional urethral length due to better visibility, which has been shown to be associated with improved continence rates , , and . Schlomm et al described a modified surgical technique for full functional-length urethra preservation during RP resulting in significantly increased early urinary continence results .
To our knowledge, we are the first group utilizing the data of patients with and without NS who otherwise underwent an identical surgical approach to assess the role of NVBs for post-RP continence. To minimize the influence of other risk factors of post-RP continence, we controlled for age, prostate volume, and year of surgery that were known risk factors for incontinence in our study population .
Our study had several limitations. First are the limitations inherent to retrospective analyses and a single-center location. It is possible that despite propensity score matching other unknown factors influencing continence may have affected continence rates. Also, we did not assess resection status of NVB electrophysiologically during surgery. Finally, all RPs were performed by experienced high-volume surgeons, and thus results may differ in other centers.
Despite these limitations, this study is to our knowledge the first to control for the NS surgical technique. The presented results are based on a large number of patients and provide strong evidence that neural innervation of the urethral sphincter does not depend on the NVBs.
Our results indicate that the meticulous apical dissection associated with the NS RP technique rather than the preservation of the NVBs itself may have a positive impact on long-term urinary continence rates. We confirm that the preservation of NVBs is important for early continence, and thus a NS approach should be attempted whenever it is safe oncologically.
Author contributions: Derya Tilki had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Tilki, Michl.
Acquisition of data: Michl, Tennstedt, Feldmeier, Mandel, Oh, Ahyai, Budäus, Chun, Haese, Heinzer, Salomon, Schlomm, Steuber, Huland, Graefen, Tilki.
Analysis and interpretation of data: Michl, Tennstedt, Feldmeier, Mandel, Oh, Ahyai, Budäus, Chun, Haese, Heinzer, Salomon, Schlomm, Steuber, Huland, Graefen, Tilki.
Drafting of the manuscript: Tilki.
Critical revision of the manuscript for important intellectual content: Michl, Tennstedt, Feldmeier, Mandel, Oh, Ahyai, Budäus, Chun, Haese, Heinzer, Salomon, Schlomm, Steuber, Huland, Graefen, Tilki.
Statistical analysis: Tennstedt, Tilki.
Obtaining funding: None.
Administrative, technical, or material support: None.
Supervision: Tilki, Michl.
Other (specify): None.
Financial disclosures: Derya Tilki certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None.
Funding/Support and role of the sponsor: None.
-  P. Mandel, M. Graefen, U. Michl, H. Huland, D. Tilki. The effect of age on functional outcomes after radical prostatectomy. Urol Oncol. 2015;33 203.e11–8
-  D.G. Murphy, A.J. Costello. How can the autonomic nervous system contribute to urinary continence following radical prostatectomy? A “boson-like” conundrum. Eur Urol. 2013;63:445-447
-  F. Reeves, P. Preece, J. Kapoor, et al. Preservation of the neurovascular bundles is associated with improved time to continence after radical prostatectomy but not long-term continence rates: results of a systematic review and meta-analysis. Eur Urol. 2015;68:692-704
-  G. Steineck, A. Bjartell, J. Hugosson, et al. Degree of preservation of the neurovascular bundles during radical prostatectomy and urinary continence 1 year after surgery. Eur Urol. 2015;67:559-568
-  P. Narayan, B. Konety, K. Aslam, S. Aboseif, W. Blumenfeld, E. Tanagho. Neuroanatomy of the external urethral sphincter: implications for urinary continence preservation during radical prostate surgery. J Urol. 1995;153:337-341
-  L.J. Song, H.K. Lu, J.P. Wang, Y.M. Xu. Cadaveric study of nerves supplying the membranous urethra. Neurourol Urodyn. 2010;29:592-595
-  T. Schlomm, P. Tennstedt, C. Huxhold, et al. Neurovascular structure-adjacent frozen-section examination (NeuroSAFE) increases nerve-sparing frequency and reduces positive surgical margins in open and robot-assisted laparoscopic radical prostatectomy: experience after 11,069 consecutive patients. Eur Urol. 2012;62:333-340
-  L. Budaus, H. Isbarn, T. Schlomm, et al. Current technique of open intrafascial nerve-sparing retropubic prostatectomy. Eur Urol. 2009;56:317-324
-  T. Schlomm, H. Heinzer, T. Steuber, et al. Full functional-length urethral sphincter preservation during radical prostatectomy. Eur Urol. 2011;60:320-329
-  P.C. Austin. Optimal caliper widths for propensity-score matching when estimating differences in means and differences in proportions in observational studies. Pharm Stat. 2011;10:150-161
-  P.C. Austin. Some methods of propensity-score matching had superior performance to others: results of an empirical investigation and Monte Carlo simulations. Biom J. 2009;51:171-184
-  Stampf S. nonrandom: Stratification and matching by the propensity score [R package v.11]. http://cran.r-project.org/web/packages/nonrandom/index.html.
-  M.M. Koraitim. The male urethral sphincter complex revisited: an anatomical concept and its physiological correlate. J Urol. 2008;179:1683-1689
-  J. Walz, A.L. Burnett, A.J. Costello, et al. A critical analysis of the current knowledge of surgical anatomy related to optimization of cancer control and preservation of continence and erection in candidates for radical prostatectomy. Eur Urol. 2010;57:179-192
-  K. Akita, H. Sakamoto, T. Sato. Origins and courses of the nervous branches to the male urethral sphincter. Surg Radiol Anat. 2003;25:387-392
-  T.M. Kessler, F.C. Burkhard, U.E. Studer. Nerve-sparing open radical retropubic prostatectomy. Eur Urol. 2007;51:90-97
-  P. Zvara, S. Carrier, N.W. Kour, E.A. Tanagho. The detailed neuroanatomy of the human striated urethral sphincter. Br J Urol. 1994;74:182-187
-  N. Hinata, G. Murakami, H. Miyake, S. Abe, M. Fujisawa. Histological study of the cavernous nerve mesh outside the periprostatic region: anatomical basis for erectile function after nonnerve sparing radical prostatectomy. J Urol. 2015;193:1052-1059
-  P.D. O’Donnell, B.F. Finan. Continence following nerve-sparing radical prostatectomy. J Urol. 1989;142:1227-1228 discussion 1229
-  P. Paparel, O. Akin, J.S. Sandhu, et al. Recovery of urinary continence after radical prostatectomy: association with urethral length and urethral fibrosis measured by preoperative and postoperative endorectal magnetic resonance imaging. Eur Urol. 2009;55:629-637
a Martini-Clinic Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Hamburg, Germany
b Department of Urology, University Hospital Hamburg-Eppendorf, Hamburg, Germany
c Department of Anatomy and Experimental Morphology, University Hospital Hamburg-Eppendorf, Hamburg, Germany
Corresponding author. Martini-Clinic Prostate Cancer Center, University Hospital Hamburg-Eppendorf, Martinistrasse 52, 20246 Hamburg, Germany.
Both authors contributed equally.
© 2015 European Association of Urology, Published by Elsevier B.V.