O

PSRS

O

M

R

C

InwtvtpNhfaiM

SF

tp

UK

8

ncology

eriprostatic Distribution of Nerves inpecimens From Non–nerve-sparing Radicaletropubic Prostatectomy

eung Bae Lee, Sung Kyu Hong, Gheeyoung Choe, and Sang Eun Lee

BJECTIVES To investigate the distributive pattern of nerves surrounding the prostate by analyzing specimensfrom non–nerve-sparing radical retropubic prostatectomy (RRP).

ETHODS We histologically analyzed specimens from 95 cases of non–nerve-sparing RRPs performed at ourinstitution by examining 12 representative slides from each case. We reconstructed images fromscanning the slides and prepared them for analysis by placing them on a circle divided into 12designated zones. The number of total nerves and ganglionated nerves on the outside of theprostatic capsule was determined manually for each zone. We assessed the locations of gangli-onated nerves separately, and also assessed the localization of vessels and nerve trunks in theposterolateral region of the prostate along with the actual shape of the neurovascular bundle(NVB).

ESULTS Significant proportions (19.9%-22.8%) of total nerves counted were located on the anterior sideof the prostate. Ganglionated nerves, present in posterolateral sectors in all cases, were alsolocated on the anterior side of the prostate in some cases (21.1%). We observed NVB in arelatively round, bundle-like formation in 51.6%, whereas NVB was more spread anteriorly,rather than having a rounded structure, in 48.4%.

ONCLUSIONS Anatomic variation exists regarding the periprostatic distribution of nerves, including theformation of NVB. We observed a significant proportion of periprostatic nerves, particularlyganglionated nerves, on the anterior side of the prostate. Thus, further investigation is neededregarding the actual role of these observed nerves not included in NVB. UROLOGY 72:

878–881, 2008. © 2008 Elsevier Inc.

hn

dsabts

M

WunfawHpsbi

n the past few years, several researchers have chal-lenged the classical anatomic concept regarding neu-rovascular bundle (NVB). Some have indicated that

erves associated with erectile function may be moreidely spread around the capsule than was previously

hought.1-5 Others have even suggested that anatomicariations exist regarding the formation of NVB, and thathe definite localization of vessels and nerve trunks in theosterolateral region of the prostate forming a so-calledVB is actually present in only some cases.6 Such reportsave been mentioned as providing possible explanations

or unsatisfactory recovery of erectile function observedfter meticulously performed standard nerve-sparing rad-cal retropubic prostatectomies (RRPs) in some patients.

oreover, based on these recent reports, some surgeons

upported by a grant from the Seoul National University Bundang Hospital Researchund.From the Department of Urology, Eulji University School of Medicine, Daejeon; and

he Departments of Urology and Pathology, Seoul National University Bundang Hos-ital, Seongnam, Korea

Reprint requests: Sang Eun Lee, MD, PhD, Department of Urology, Seoul Nationalniversity Bundang Hospital, 300, Gumi-dong, Bundang-gu, Sungnam, Kyunggi-do,

Sorea 463-707. E-mail: selee@snubh.orgSubmitted: February 27, 2008, accepted (with revisions): May 4, 2008

78 © 2008 Elsevier Inc.All Rights Reserved

ave recommended modifying the techniques of standarderve-sparing RRP.7-11

A paucity of data exists in the literature on the actualistribution of nerves around the prostate. Althoughome have investigated the subject, the number of casesnalyzed has been limited, and the reported findings haveeen variable. Thus, we investigated the distributive pat-ern of nerves surrounding the prostate by analyzingpecimens from non–nerve-sparing RRPs.

ATERIAL AND METHODS

e histologically analyzed specimens from 95 patients whonderwent non–nerve-sparing RRP at our institution. In non–erve-sparing RRP, fascia on the lateral surface of the rectum

rom the level of the apex to the base was divided so that bundlend fascial tissue were included in the specimen. Specimensere processed by the 3- to 5-mm step section technique.istologic sections (whole-mount sections in some cases) were

laced onto slides that were blocked with 10% normal rabbiterum for 30 minutes. After we rinsed the slides in phosphateuffered saline containing the detergent Tween-20, we thenncubated them with an antibody against the nervous protein

100 for 1 hour at room temperature. A corresponding biotin-

0090-4295/08/$34.00doi:10.1016/j.urology.2008.05.038

ybdr(t

fiflS5otbsnosrdnzrop

RWpstsamwsw

7anlhTamoi

g5gioc

c(rcNrsm

CStscsaCcNcorn

Fs

Faa(

U

lated immunoglobulin G was applied for 30 minutes, followedy another 30 minutes of incubation with streptavidin peroxi-ase from an LSAB kit (Dako, Glostrup, Denmark). After moreinsing, we visualized the tissue using a diaminobenzideneDAB) solution and then counterstained it with routine hema-oxylin.

In each case, 20 to 30 slides were taken for each patient. Fornal examination by a single pathologist, 2 slides were takenrom the apex, the mid-body, and the base of prostate for theeft and right lobe, respectively, totaling 12 slides for each case.lides were analyzed by images scanned via a Nikon Coolscan000 ED (Nikon, Bangkok, Thailand). As done previously bythers, we reconstructed images from scanning slides represen-ative of the same section plane and prepared them for analysisy placing them on a circle divided into 12 zones designated ashown in Figure 1.5 Subsequently, we manually determined theumber of total nerves and ganglionated nerves on the outsidef the prostatic capsule for each zone. We also assessed dataeparately for the prostatic apex, mid-body, and base regions,espectively. The percentage of nerves in a certain zone wasefined as the number of total nerves in a zone divided by theumber of total nerves for the prostatic lobe encompassing thatone. We assessed the locations of ganglionated nerves sepa-ately. Also, in histologic analyses, we assessed the localizationf vessels and nerve trunks in the posterolateral region of therostate, along with the actual shape of NVB.

ESULTSe noted nerves in all designated zones surrounding the

rostate in all cases. Median numbers of observed nervesurrounding prostate were 81, 136, and 171 for the pros-atic apex, mid-body, and base regions, respectively. Ashown in Figure 2, the median number of nerves observedt the apex region was fewer than those seen at theid-body or base level. The highest number of nervesas seen in the zones encompassing the posterolateral

ides of the prostate (zones 4, 5, 8, and 9). In those zones,

igure 1. Zonal designations applied in the current analy-is and S-100-stained periprostatic nerves.

e found median percentages of 73.8%, 68.3%, and t

ROLOGY 72 (4), 2008

1.9% of counted nerves at the level of apex, mid-body,nd base regions, respectively. On the other hand, sig-ificant proportions of the total nerves counted were

ocated on the anterior side of the prostate above theorizontal line demarcating sectors 3 and 4 (or 9 and 10).he sum of median percentages of nerves in zones 1-3nd 10-12 were 19.9%, 22.8%, and 18.7% for apex,id-body, and base, respectively. As for the posterior side

f the prostate (zones 6 and 7), only 5.8%-9.9% ofdentified nerves were located.

Among the nerves surrounding prostate, we notedanglionated nerves at the posterolateral zone (in zone 4,, 8, or 9) in all cases (Fig. 3A). We also observedanglionated nerves on the anterior side of the prostaten some cases (Fig. 3B). Of the 95 total cases analyzed, webserved ganglia to be present in zone 1-3, or 10-12 in 20ases (21.1 %).

In all cases, nerves around the prostate were mostoncentrated in the posterolateral region of the prostatezones 4, 5, 8, or 9), forming NVB. NVB with a relativelyound, bundle-like formation was observed in only 49ases (51.6%) (Fig. 4A). In the other 46 cases (48.4%),VB was more widely spread rather than having a

ounded structure as it extended anteriorly (Fig. 4B). Inuch cases, larger vessels and S-100 staining were alsoost frequently observed in zones 4, 5, 8, or 9.

OMMENTeveral researchers have suggested, after performing his-ologic investigations with a limited number of adultubjects, that the nervous component supposed to beontained within NVB actually spreads on the lateralides of the prostate and extends more to the anteriorspect of the prostate than was traditionally believed.ostello et al. performed an anatomic analysis of 12 male

adavers and reported that most of the nerves included inVB originally descend posterior to the seminal vesicle,

onverging to the mid-prostatic level only to divergence again as they approach the prostatic apex.1 Theesearchers mentioned that the anterior and posteriorerves of NVB are extensively separated by about 3 cm at

igure 2. Zonal distribution of periprostatic nerves showns the percentage of median number of nerves per zonemong the median total number of nerves per prostatic lobeleft: 1-6; right: 7-12).

he level of the base of the prostate. In 2004, Takenaka

879

ertsr

tnrapsnscbtssollBvdtoAqscmFercn

asainsosoammsgiKebsoappf

Fsp

Fts

8

t al. performed fresh cadaver dissection of 7 males andeported that fibers of the pelvic splanchnic nerves joinhe hypogastric nerve fibers within the NVB in a con-iderably more spraylike manner and at a far more distant

igure 3. (A) Ganglionated nerve present in NVB (S-100tain). (B) Ganglionated nerve present in the anterior side ofrostate (S-100 stain).

igure 4. (A) NVB localized to the posterolateral aspect ofhe prostate. (B) NVB in the formation relatively morepread to the anterior side of the prostate. S-100 stain.

egion than previously believed.2 Lunacek et al. reported i

80

hat NVB has concave, curtainlike shape with cavernouserves running along the prostate displaced more ante-iorly and spread.3 They mentioned that, during theging process, cavernous nerve fibers along the surface ofrostate capsule increasingly disperse anteriorly with re-pect to the prostate. Therefore, they indicated that theerve-sparing technique should be modified to start dis-ection for NVB preparation more anteriorly than waslassically done. Our findings support these statements,ecause in this study we observed considerable propor-ions of nerves surrounding the prostate on the anterioride of the prostate. In addition, we observed NVB topread more anteriorly than others in some cases. More-ver, ganglionated nerve, which existed in the postero-ateral aspects of the prostate in all of our cases, was alsoocated on the anterior side of the prostate in some cases.ecause ganglionated nerves are more likely to be in-olved in autonomic signaling, observation of such aistribution may provide more support to the hypothesishat nerves associated with potency travel more anteri-rly than classically believed, at least in some men.lthough in our study we could not trace nerves ade-

uately to validate the findings of others, our resultsuggest support for those findings. A recent report indi-ated that cavernous nerve innervating erectile functionay be more posteriorly located than originally believed.

rom their electrophysiologic tracing of nerves, Tak-naka et al. reported that cavernous nerve existed on theectal wall, apart from the NVB.4 Thus, the controversyontinues with regard to the actual course of cavernouserve and NVB.Previously, some authors indicated that anatomic vari-

tion may exist regarding the formation of NVB. Kiyo-hima et al. analyzed non–nerve-sparing RRP specimensnd reported that NVB was situated at a localized spacen approximately half the cases; whereas in others cases,o definite formation of NVB was seen, with the nervesparsely spread to the lateral aspect of the prostate with-ut definite bundle formation.6 Our findings demon-trated that larger nerves or ganglions and vessels existedn the posterolateral side of the prostate in all specimensnalyzed in our series. In some cases, NVB were spreadore to the ventral side of the prostate in our study, asentioned by Kiyoshima et al.6 Although they did not

pecifically present relevant data, differences existed re-arding the number or size of nerves, vessels, or bothncluded in the NVB. Thus, similar to the findings fromiyoshima et al’s report, our study suggests that NVBxist demonstrating anatomic variations in the size, num-er, and location of nerves, vessels, or both. After clas-ically performed nerve-sparing RRPs, some patients arebserved to be potent without the aid of phosphodiester-se-5 inhibitors, whereas others are not. It can be hy-othesized that those with NVB well localized in theosterolateral aspects of the prostate may comprise theormer, whereas those with NVB spread more anteriorly

n the latter. Variations in the number of potency-asso-

UROLOGY 72 (4), 2008

cha

tcdctOeSEosnllastscrmnaspHi

spnltnirpwcsb

CActWnlNt

R

1

1

U

iated nerves spread outside the so-called NVB also mayave resulted in different outcomes regarding potencyfter nerve-sparing RRP performed in a standard way.

No general agreement has yet been reached regardinghe proper level of dissection for the nerve-sparing pro-edure, or the exact anatomy of NVB. Still, it can beeduced from previous reports, along with our data, thatavernous nerves or nerves innervating erectile func-ion may be more spread than was previously believed.ur results show that considerable amounts of nerves

xist on the anterior and lateral sides of the prostate.uch observations are in line with a recent report fromichelberg et al.6 Also, variations in the shape of NVB inur study suggest that NVB itself may be more anteriorlypread in some men. In addition to simple-lookingerves, ganglionated nerves were observed on the antero-

ateral side of the prostate, not included in the bundle-ike formation in some of our cases. Montorsi et al.dvocated performing a high incision to maximize nervespared during the nerve-sparing procedure. They statedhat incision of the levator fascia during RRPs shouldtart high up to preserve the largest possible fraction ofavernous nerves.7 Improved potency rates have beeneported with application of this approach. Such resultsay only reflect the fact that the classically understoodeurovascular bundle was less damaged. Considering thatsignificant amount of nerves exist on the anterolateral

urface of the prostate, Menon et al. advocated preservingrostatic fascia for nerve-sparing during robotic RRPs.8,9

owever, such an approach obviously warrants further val-dation regarding cancer control via long-term follow-up.

Our study may have been limited by the relativelymall number of subjects compared with other contem-orary series on RRP. However, to our knowledge, theumber of subjects in our study is currently one of the

argest among those histologically evaluating nerve dis-ribution around the prostate in specimens from non–erve-sparing RRPs. The actual role of nerves observed

n our study can only be theorized. Debate over the actualole of nerves spreading on the anterolateral side of therostate continues, because they may not be associatedith erectile function at all. On the other hand, be-ause ganglionated nerves may be more than simpleensory nerves, our findings on nerve distribution may

e significant.

ROLOGY 72 (4), 2008

ONCLUSIONSlthough periprostatic nerves were most commonly lo-

ated on the posterolateral aspect of the prostate, ana-omic variations exist regarding the formation of NVB.

e observed a significant proportion of periprostaticerves, particularly ganglionated nerves, on the antero-

ateral side of the prostate, not included in the so-calledVB. Further investigation regarding the actual role of

hese observed nerves not included in NVB is needed.

eferences1. Costello AJ, Brooks M, Cole OJ. Anatomical studies of the neu-

rovascular bundle and cavernosal nerves. BJU Int. 2004;94:1071-1076.

2. Takenaka A, Murakami G, Soga H, et al. Anatomical analysis ofthe neurovascular bundle supplying penile cavernous tissue to en-sure a reliable nerve graft after radical prostatectomy. J Urol.2004;172:1032-1035.

3. Lunacek A, Schwentner C, Fritsch H, et al. Anatomical radicalretropubic prostatectomy: “curtain dissection” of the neurovascularbundle. BJU Int. 2005;95:1226-1231.

4. Takenaka A, Tewari A, Hara R, et al. Pelvic autonomic nervemapping around the prostate by intraoperative electrical stimula-tion with simultaneous measurement of intracavernous and intra-urethral pressure. J Urol. 2007;177:225-229.

5. Kiyoshima k, Yokomizo A, Yoshida T, et al. Anatomical features ofperiprostatic tissue and its surroundings: a histological analysis of 79radical retropubic prostatectomy specimens. Jpn J Clin Oncol. 2004;34:463-468.

6. Eichelberg C, Erbersdobler A, Michl U, et al. Nerve distributionalong the prostatic capsule. Eur Urol. 2007;51:105-110.

7. Montorsi F, Salonia A, Suardi N, et al. Improving the preservationof the urethral sphincter and neurovascular bundles during openradical retropubic prostatectomy. Eur Urol. 2005;48:938-945.

8. Menon M, Kaul S, Bhandari A. Potency following robotic radicalprostatectomy: a questionnaire-based analysis of outcomes afterconventional nerve sparing and prostatic fascia sparing techniques.J Urol. 2005;174:2291-2296.

9. Kaul S, Savera A, Bhandari A, et al. Functional outcomes andoncological efficacy of Vattikuti Institute prostatectomy with veilof Aphrodite nerve sparing: an analysis of 154 consecutive patients.BJU Int. 2006;97:467-472.

0. Sato Y, Rehman J, Santizo C, et al. Significant physiological rolesof ancillary penile nerves on increase in intracavernous pressure inrats: experiments using electrical stimulation of the medial preopticarea. Int J Impot Res. 2001;13:82-88.

1. Stolzenburg JU, Rabenalt R, Tannapfel A, et al. Intrafascial nerve-sparing endoscopic extraperitoneal radical prostatectomy. Urology.

2006;67:17-21.

881