SEARCH

SEARCH BY CITATION

Keywords:

  • robotic prostatectomy;
  • nerve sparing technique;
  • tri-zonal neural architecture;
  • anatomy

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

OBJECTIVE

To review the neural architecture around the prostate gland, as it is relevant for nerve-sparing robotic prostatectomy, including in particular the anatomy of the proximal neurovascular tissue, the neurovascular bundle (NVB), and accessory neural pathways (ANPs).

MATERIAL AND METHODS

The aims of this study were achieved in collaboration between the Cornell Institute of Robotic Surgery, New York, NY, USA and the Institute of Urology at the University of Innsbruck, Austria. The broad steps were: (i) anatomical studies of 10 fresh and two fixed male cadavers; and (ii) collection of videotape and still image data from 200 men undergoing radical prostatectomy by the athermal robotic technique at the Cornell Institute.

RESULTS

From a surgical standpoint there was a tri-zonal neural architecture including the proximal neurovascular plate (PNP), the predominant NVB (PNB) and ANPs. The PNP was a mean (range) of 5 (3–10) mm lateral to the seminal vesicles, was 3 (2–7) mm thick, 7 (5–25) mm wide and 9 (4–30) mm long. It was within 6 (4–15) mm of the bladder neck, 5 (2–7) mm of the endopelvic fascia and overlapped 5 (0–7) mm of the proximal prostate. The PNB varied in shape and size from the proximal to distal end, was thickest at the base and most variable near the apex. In eight of 12 cases, there was a medial extension behind the prostate, which converged medially at the apex in four cases. ANPs were noted within the layers of levator fascia and/or lateral pelvic fascia on the anterolateral aspect in five cases and in three on the posterior aspect of the prostate. In nine cadavers, the proximal third of the prostate was covered by the PNP where these ANPs were most prominent. The ANPs formed a plexus on the posterolateral aspect of the apex in four cases.

CONCLUSION

We have created an anatomical map of neurovascular tissue relevant to robotic prostatectomy. A tri-zonal neural architecture is described which has helped in standardizing the steps of robotic prostatectomy.


Abbreviations
PNP

proximal neurovascular plate

PNB

predominant neurovascular bundle

ANP

accessory neural pathway

RP

radical prostatectomy.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Although the anatomical map for nerve-sparing during retropubic radical prostatectomy (RP) was established by the pioneering contributions of Walsh et al.[1], the anatomical principles need to be re-emphasized in the context of robotic RP, because surgical steps are reversed, visual angles are different, and magnification and stereoscopy provide more detailed anatomical images than seen during open RP. We therefore felt the need to revisit the anatomical foundations, tailored to robotic RP. We integrated cadaveric studies to mimic robotic steps, and incorporated the critiques of a veteran open surgeon to comply with established anatomical principles. Anatomical studies are presented here, and oncological and functional outcomes will be addressed in future reports.

These anatomical studies were primarily aimed at depicting the location of zones of neurovascular tissue in relation to robotic RP. These zones (Fig. 1) include the proximal neurovascular plate (PNP), the predominant neurovascular bundle (PNB), and accessory neural pathways (ANPs). The first aim was to highlight the applied anatomical importance of the PNP (the pelvic plexus), which presents as a plate and includes ganglia and neuronal connections around the inferior vesicle pedicles, bladder neck and seminal vesicles [2]. Inadvertent transection, crush, traction or thermal injury to the PNP can significantly delay the return of sexual function, because it is a relay centre for distal neural impulses to the cavernosal nerves. The second aim was to delineate the applied importance, course and histological architecture of the PNB as it presents during antegrade robotic RP, by microscopic dissection and neural staining. The third aim was to depict the putative ANPs which might be additional pathways for transmitting neural impulses.

image

Figure 1. Tri-level neural zones concept. Lateral view from left side after removal of the pubic bone. Black arrowheads indicate the continuity of PNP and PNB (arrows), and white arrowheads indicate the continuity of PNP and ANP. Fresh cadaver dissection.

Download figure to PowerPoint

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Data were acquired from anatomical studies of 10 fresh and two fixed male cadavers, and 200 nerve-sparing robotic RPs. The cadavers were from men aged >40 years, with no previous pelvic or urethral surgery, and with normal prostates and urinary bladders. Ten cadavers were frozen at <12–36 h after death and stored at − 20 °C until dissection. Two cadavers were fixed by arterial injection of 10% formalin solution. No macroscopic tumour was evident in the abdominal and pelvic regions of these cadavers. The emphasis on fresh frozen cadavers was to focus attention on the NVBs and to the pelvic plexus, which is a relay centre for erectile impulses, and the aim was to orientate surgeons to the location of these neurovascular structures during robotic RP. We undertook cadaveric dissections and traced the cavernosal neurovascular supply. The hypogastric nerve was dissected and labelled with tape. Before removing the prostate, the pelvic splanchnic nerve was dissected and labelled in the presacral area. An operating microscope (× 2.5, Surgical Acuity, Meddleton, WI, USA) was used to trace branches of the pelvic plexus from their origin to their exit from the pelvis around the urethra. The technical details were published previously [3].

The pelvic plexus, NVBs and periprostatic fascia were labelled with China ink (carbon particle suspension) to identify them during specimen trimming, embedding and sectioning. Specimens included all parts of the membranous urethra, the dorsal half of the prostate and small parts of the levator ani, bladder, seminal vesicle and rectum. After paraffin-wax embedding, serial horizontal sections of 10–20 µm were cut. Most sections were stained with haematoxylin and eosin, others were stained with S-100 neural stain to delineate the precise location of this neural tissue.

Between January 2005 and December 2005, 200 men had robotic RP using the da Vinci system (Intuitive Surgical, Moutainview, CA, USA) for clinically localized prostate cancer, using the athermal robotic technique described by Tewari et al.[4]. In terms of anatomy using the fresh cadavers, we re-evaluated every procedure of this technique in 200 videotape collections.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

During robotic RP, we encountered three broad zones of relevant neurovascular tissue: the PNP, the PNB and the ANPs (Fig. 1). The PNP processes and relays erectogenic neural signals. While the major components of the PNP include vesical and prostatic subdivisions of the pelvic plexus, it was composed of ganglia and interconnecting nerve fibres (Fig. 2). The PNP was located lateral to the bladder neck, the seminal vesicles and branches of the inferior vesical vessels. It was thick in the centre near the seminal vesicles. Depending on anatomy and prostate size, the PNP was located at a mean (range) of 5 (3–10) mm lateral to the seminal vesicles, was 3 (2–7) mm thick, 7 (5–25) mm wide and 9 (4–30) mm long. It was located within 6 (4–15) mm of the bladder neck, within 5 (2–7) mm of the endopelvic fascia and overlapped 5 (0–7) mm of the proximal prostate.

image

Figure 2. The ganglion cells and nerve fibres in the PNP. Haematoxylin and eosin staining.

Download figure to PowerPoint

The PNP was prone to injury during incision of the endopelvic fascia, incision of the posterolateral aspect of prostatovesical junction, lateral dissection of the seminal vesicles, application of a bulldog clamp [5] and/or electrocautery/cutting of the prostatic pedicles. The PNP extended lateral to the base of the prostate, and was often embedded in the fascial layers covering the prostatic capsule and intermingled with the pedicle of the prostate (Fig. 3). Distally, the PNP converged to continue as the classical PNB, while a few branches travelled through the fascial and capsular tissue of the prostate as accessory distal neural pathways (ANPs; see below).

image

Figure 3. Fixed cadaveric dissection showing the relationship between seminal vesicle and PNP according to robotic RP. The bladder neck is transected and the prostate is lifted up by the forceps. PNP is intermingled with vascular pedicle of the prostate. Arrow head, PNP; Black arrow, PNB; White arrow, intermingled structure of vascular (white star) and neural (black star) component.

Download figure to PowerPoint

The PNB is the classical bundle that carries neural impulses to the cavernosal tissue. The PNB is enclosed within the layers of levator fascia and/or lateral pelvic fascia, is posterolateral to the prostate and, as described by Walsh [6], tends to vary its course from the base to the prostatic apex (Fig. 4). As described previously [7,8], the PNB varied in shape and size from the proximal to the distal end. It occupied the groove between prostate and rectum, was thickest at the base, and had the most variable course and architecture near the apex. In eight of the 12 cases there was a medial extension of the PNB behind the prostate, which converged medially in four cases near the midline at the apex of the prostate (Fig. 5). These structures were identified as neural components by immunohistochemistry for S-100 (Fig. 5B). There were substantial variations in the architecture and thickness of the PNB.

image

Figure 4. Lateral view of PNP, PNB, and ANP. Fresh cadaver dissection showing that the neural pathway from the PNP is a spray-like distribution. The prostate and bladder are lifted up by the forceps.

Download figure to PowerPoint

image

Figure 5. Medial extension at the apex of the prostate. A, in a fresh cadaver, several nerve fibres and blood vessels behind the prostate. B, immunohistological study of this area, using S-100 neural stain.

Download figure to PowerPoint

The PNB also contained ganglion cells like those in the PNP (Fig. 6); these were attached to the prostate capsule or even embedded within the capsule. Consequently, the PNB should be separated carefully and athermally from the prostate capsule to avoid the ganglion cells.

image

Figure 6. Transverse section of the posterolateral aspect through the middle of the prostate. There are many ganglion cells in the PNB attached to the prostate capsule or embedded within the capsule (arrows). There also are few ganglion cells outside of the lateral pelvic fascia (dotted arrows). Arrowhead, lateral pelvic fascia (haematoxylin and eosin stain).

Download figure to PowerPoint

ANPs are putative accessory (i.e. other than the PNB) neural pathways around the prostate that might be additional conduits for neural impulses. These are usually described as within the layers of the levator fascia and/or lateral pelvic fascia, on the anterolateral and sometimes the posterior aspect of the prostate. As mentioned, many cadavers (eight of 12) had the proximal third of prostate covered by the PNP where these nerve fibres were most prominent (Fig. 1).

ANPs were seen in five of the 12 dissections, predominantly as a triangular extension of the PNP, which converged near the apex (Fig. 4). In three specimens, ANPs arose posteriorly from the medial aspect of the PNP near the base of the seminal vesicles. Importantly there were two sets of neural tissue on the surface of prostate: one was superficial and lay between the prostatic capsule and fascia, while deeper groups of nerves travelled within the capsule that are possibly involved with prostatic innervation. The superficial group was most prominent on the posterolateral aspect.

The ANPs occasionally formed an apical plexus on the posterolateral aspect of the prostatic apex and urethra incorporating fibres from the PNBs (of both sides) and ANPs. This distal plexus was seen in four cases and might act as a neural pathway for not only cavernosal tissue but also the urethral sphincter. It might also provide supporting neural ‘crosstalk’ between the sides. In one case, the fibres were circumferential at the apex. When the dorsal vein complex and bladder neck are sutured, it should be assumed that it is almost circumferential except for the dorsal vein complex, and knowing the position of the ANP is anterior if the suture is too deep.

The neural zones around the prostate have important implications for athermal robotic RP. Table 1 summarizes the critical steps during robotic RP which might result in injury to these zones, and suggests preventive measures to minimize this injury and improve the probability of early recovery of sexual function.

Table 1.  Importance of neural zones around the prostate; implications for athermal robotic RP
Step of robotic RPPotential and mechanism of damage to neurovascular zones
PNPPNBsANP
Dissection of the endopelvic fasciaYes. If dissection is too far proximal it can damage the PNP by cautery, traction or blunt injuryYes. If dissection is too deep it can injure the PNB by electrocautery, traction or blunt injuryYes. Anterior and lateral ANP can be damaged by electrocautery or blunt injury
Preventive measuresUse of athermal technique using sharp dissection and referring to location of three neural zones at all times.
Dorsal venous sutureNoNoMaybe, if the suture is placed into a deeper plane and injures ANP near urethra
Preventive measuresUse of athermal technique using sharp dissection and referring to location of three neural zones at all times The suture can be placed later before disconnecting the apex when the prostate is relatively free and the venous complex can be better visualized.
Back bleeder sutureMaybe. If the suture is in the proximal two-thirds of prostateUnlikelyMaybe, if the suture is too deep and goes through the anterior ANP
Preventive measuresPlace suture distally and more in the midline strip.
Bladder neck transectionQuite possible. If incision is extended laterally or excessive electrocautery during posterior dissection.UnlikelyUnlikely
Preventive measuresAvoid lateral transection and limit the dissection to the anterior third of the prostato-vesical junction. Use bimanual pinching technique to define the junction.
Seminal vesicle and vas dissectionYesNoNo
Preventive measuresAvoid (or minimize) electrocautery during this phase. Small clips preferred.
Posterior dissectionYesYesPosterior ANP
Preventive measuresAvoid (or minimize) electrocautery during this phase. Small clips preferred.
Control of pedicleYesYesMaybe (proximal ANP)
Preventive measuresAvoid electrocautery during this phase. Small clips and individual pedicle controls are preferred.
Release of NVBsYesYesYes
Preventive measuresAvoid electrocautery during this phase. Development of the correct plane is of utmost importance. Small clips and individual pedicle controls preferred. No bulldog clamps because they can cause crush injury to the nerves. We avoid traction injury to the bundles by using sharp dissection and avoiding excessive pull and blunt dissection.
Apical transectionNoYesYes
Preventive measuresAvoid electrocautery around the nerves. Small clips and individual pedicle controls preferred. Keep PNB and ANP in mind while disconnecting the urethra. Complete haemostasis is essential.
Urethral anastomosisNoYesYes
Preventive measuresPrecise suturing to avoid strangulation of neural tissue in the ‘bites’.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The present study summarizes the neuroanatomy of the pelvic erectile nerves as relevant to robotic RP. We grouped important neural structures (Fig. 1) into the PNP, the PNB and ANPs. Clear dissection images are provided and correlation made with surgical steps. Critical measures for nerve preservation are outlined in Table 1. When surgeons attempt to preserve neural tissue during robotic RP they must cope with anatomical variations, obscure tissue planes [9], bleeding, narrow pelves, large prostates [10], high pelvic fat contents, differences in the shape of the prostatic apex, and, most importantly, cancer-related variables [11]. A clear anatomical map of the neurovascular tissue and a technique standardized to spare the periprostatic neural tissue is immensely helpful during robotic RP.

In a milestone study using male fetuses and newborn cadavers, Walsh and Donker [1] first reported the course of the cavernosal nerve. They later presented a gross description of the PNB, which enclosed nerve branches responsible for erection, i.e. the cavernosal nerves, located at the lateral aspect of the prostate [12]. Since then, nerve-sparing has become a standard technical step during RP for prostate cancer. Although several surgeons have mastered the technique of preserving the PNB, sexual outcomes are still variable [13–17]. The reasons could be: (a) differences in the architecture, location and size of the PNP; (b) variability in the composition and course of the cavernosal nerves [18]; and (c) the relative roles of the PNB and ANPs in an individual patient [19,20]. While most surgical techniques focus on release of the PNB during robotic RP, little mention is made of preserving the PNP. As seen in our studies [8], the PNP is a spray-like structure located close to the bladder neck and seminal vesicles. As robotic RP involves the preliminary transection of the bladder neck and dissection of seminal vesicles, it is very important to be aware of the PNP and to avoid injuring it by cautery, cutting, traction, or crushing by bulldog clamps (Figs 3 and 7). Using the images in the present study, surgeons should be able to appreciate the critical location of the PNP and thus to preserve it.

image

Figure 7. Surgical procedure, cutting the vessel pedicle of the prostate athermally using the small clips. During this procedure, care should be taken not to injure the PNP. A, transection of the left PNP (closed star) and inferior vesical vessels (open star). Viewing these structures laterally, we estimate the approximate border between PNP and vessel component, although they are intermingled. We have already cut a part of the vessels using a clip. B, We thread the left hand instrument though the border, ligate the residual vessels using a clip, and cut it sharply. UR, urethra; LA, levator ani.

Download figure to PowerPoint

The PNB is classically described as a tubular structure running along the dorsolateral aspect of the prostate gland. However, there has been little histological or physiological investigation of the gross and microscopic anatomy of putative PNB. Several papers suggested that parasympathetic neural components join the PNB in a spray-like distribution, and the classical cord-like architecture is seen in only a few patients [7,8]. The fresh cadaver dissections and histological findings in the present study favour an origin of the cavernosal nerve from the most caudal components of the parasympathetic neural branches showing a spray-like distribution (Figs. 4 and 8). Most importantly, the PNB varies in its shape and size from its proximal to distal end, is thickest at the base and was most variable in course and architecture near the apex. Often there was a medial extension of the PNB behind the prostate. Thus, the classical nerve-sparing approach will sacrifice most of the proximal and posterior extensions of the neurovascular tissue. The present maps and images will be helpful in avoiding injury to proximal, posterior and apical components of NVBs during RP (Fig. 9).

image

Figure 8. A, Transverse section of the posterolateral aspect through the middle of the prostate. The branches from the pelvic splanchnic nerve containing the ganglion cells go to the PNB. B, Magnification of the square in A. Nerve component, red area in A and yellow area in B. Arrow, ganglion cells, haematoxylin and eosin stain.

Download figure to PowerPoint

image

Figure 9. As indicated by the cadaveric and histological study, there are many nerve fibres behind the apex of the prostate during robotic RP. It might be very important to dissect them athermally for nerve-sparing. In this case, bilateral PNB overlapped behind the apex, and formed posterior plexus. Arrows, PNB; arrowhead, posterior plexus; REC, rectum; LA, levator ani.

Download figure to PowerPoint

As seen in Fig. 5, there was fatty tissue and several nerve fibres and blood vessels in the NVBs. Importantly, the nerve bundles are composed of many neural fibres. Surgeons should not approach nerve-sparing as an all-or-none phenomenon, but as an incremental act in which all, most, or none of the nerves can be spared. Robotic magnification and the extended degree of freedom of the instrument tip allows the surgeon to develop dissection planes within the fatty tissue of the NVBs (incremental nerve sparing) and therefore stay away from the capsule in cases where there is a high risk of extracapsular extension. This is a novel concept, and allows the surgeon to do partial nerve-sparing rather than completely excising the NVBs in cases at high risk for T3 cancer (Fig. 10). The long-term functional results of this approach are being collected.

image

Figure 10. These show the release of the left PNB. We can make the cutting line as we choose. In A, the PNB is completely spared. In B, the Gleason score is 4 + 3 at the apex and middle of the prostate; we choose a cutting line that is a little lateral (partial sparing). We cut sharply and athermally in both cases, although there is some oozing.

Download figure to PowerPoint

ANPs were recently proposed to carry neural impulses to penile tissue outside the classical PNB [21]. These pathways might help to explain the unpredictability of the return of sexual function and the lack of correlation between the surgeon’s perception of the quality of nerve-sparing and functional results after RP. Some novel approaches were developed to preserve ANPs, including the Veil technique by Kaul et al.[19], and the curtain dissection by Lunacek et al.[20]. We use the Veil technique athermally in appropriate cases where it is oncologically safe to come close to the capsule. However, in every case (Table 1) we also preserve the PNP and the PNB by using sharp dissection, miniature clips, and by avoiding traction and vascular clamps.

We previously focused on the qualitative and quantitative delineation of human pelvic ganglion distribution [2]. In particular, ganglion cells were distributed over a large area at various sites: the dorsal aspect of the bladder, the bladder-prostate junction, the dorsal aspect of the seminal vesicle, the apex of the prostate, and the PNB. By contrast, in experimental animals, ganglion cells appear largely limited to major pelvic ganglia. Using a rat model, Kato et al.[22] reported difficulty in regenerating injured ganglion cells, even after simple damage to nerve fibres. To avoid ganglion injury, an athermal approach not only to the PNP but also to the PNB and ANP might be very useful.

Limitations of the present study are: (i) the lack of physiological correlation of the PNP with erectile function; (ii) incompleteness of data on ANPs, leaving uncertain the relevance of neural tissue on and within the prostatic capsule; (iii) imprecise mapping of parasympathetic and sympathetic pathways around the prostate; and (iv) lack of long-term functional outcome data. However, we confirmed our findings in several fresh cadavers, performed specialized histological staining of samples, and verified these maps during 200 nerve-sparing RPs using the athermal robotic technique. In addition, using the original intraoperative stimulation device, we made preliminary maps of the nerves responsible for erectile function and urinary continence [23]. Future studies involving functional mapping and long-term outcomes using validated instruments will clarify the relevance of these findings.

In conclusion, we show three-dimensional computer graphic images of the neurovascular anatomy around the prostate (Fig. 11). The information is not new, but the description and technical details are uniquely relevant to robotic RP, which uses the antegrade approach rather than the retrograde approach commonly used for open RP. We illustrated the existing neurovascular anatomy, and introduced the concept of the PNP, three zones of neural tissue, ANPs, and incremental nerve-sparing. These might benefit new surgeons undertaking nerve-sparing robotic RP.

image

Figure 11. Three-dimensional computer graphic images of the neurovascular anatomy around the prostate.

Download figure to PowerPoint

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES