European Urology

1. Introduction

In nerve-sparing surgery in the male pelvis, major structures designated for preservation include nerve bundles such as the cavernous nerve and hypogastric nerve. In contrast, pelvic autonomic ganglion cells have received little consideration in nerve-sparing strategies. However, surgical damage to ganglion cells would seem likely to have far worse consequences than nerve bundle injury since, unlike nerve fibers, ganglion cells lack capacity for repair [1].

According to previous knowledge concerning distribution of pelvic autonomic ganglia, parasympathetic ganglia are located near or along the pelvic viscera, while sympathetic ganglia along the lumbar and sacral sympathetic nerve trunks [2], and [3]. Thus, one would expect that parasympathetic ganglion cells should remain intact after surgery if the visceral nerve target such as the bladder is successfully preserved. Moreover, the sympathetic trunk generally is located far dorsally, out of the surgical field. If these generalizations are true, little or no specific attention need be given for preservation of pelvic ganglion cells during surgery. We are not aware of a previous study specifically focused on qualitative and quantitative delineation of human pelvic ganglion distribution. The aims of this preliminary work were to describe distribution and number of ganglion cells at each site in the male pelvis using semiserial sections and to classify these cells as sympathetic or parasympathetic.

2. Materials and methods

Histologic specimens were obtained from 14 donated male cadavers (72 to 85 years old at death), that had been fixed by arterial injection of 10% formalin solution. The cadavers harbored no macroscopic tumor in thoracoabdominal regions. To facilitate sectioning and identification of structures, intrapelvic soft tissues usually were divided into three or four large blocks including the hypogastric nerve and distal ureter; the bladder, prostate, seminal vesicle, and rectum; the membranous urethra, penile hilum, and levator ani; and, for half of the cadavers, the pelvic splanchnic nerves. To isolate these nerves, we separated the rectum slightly from the presacral space to obtain a sheetlike specimen (fourth block) including nearly the entire course of the pelvic splanchnic nerves.

After routine processing for paraffin embedding, semiserial sections 6 to 10 μm in thickness were cut at 1-mm intervals. For each section stained with hematoxylin and eosin (H and E), two or three adjacent sections were processed for immunohistochemistry at sites chosen according to observations in H and E sections. The primary antibody used to identify sympathetic neurons was polyclonal rabbit anti-tyrosine hydroxylase (TH; Chemicon International, Temecula, CA; 1:400 dilution in phosphate-buffered saline, or PBS). TH is expressed by noradrenergic sympathetic neuron [4], and [5]. To identify parasympathetic neurons, we used rabbit anti-peptide histidine isoleucine (PHI) anti-serum (Yanaihara Institute, Fujinomiya, Japan; 1:200 dilution in PBS). PHI is considered to colocalize with vasoactive intestinal peptide (VIP), a major marker for parasympathetic nerve cells [6]. PHI- and VIP-immunoreactive nerves were similarly distributed in the human and experimental animal tissues [7], and [8]

3. Results

3.1. Routine histology

Pelvic ganglion cells displayed a homogeneous morphology, appearing large and eosinophilic, with a diameter of 25 to 30 μm. These cells were found along nerve fiber bundles, in enlargements of nerve bundles, and in round or oval, grossly identifiedble ganglia surrounded by a thick connective tissue capsule (Fig. 1, Fig. 2, and Fig. 3). However, because of numerous intermediate morphologies between a macroscopically typical ganglion and enlargements along a nerve, and because cell numbers found in a ganglion varied significantly between sites and individuals, strict definition of a “ganglion” proved to be difficult. We therefore adopted the term, “ganglion cell cluster” (GCC) for the present description. GCC were seen in or along nerve components such as the pelvic splanchnic, cavernous, and hypogastric nerves, and the pelvic plexus, as well as near or along pelvic visceral surfaces (Table 1, Fig. 4).

Create Platinum Slide Series

Fig. 1 Ganglion cell clusters at various sites in the male pelvis according to hematoxylin and eosin (H and E) ataining. (A) Dorsal aspect of the bladder-prostate junction. A ganglion cell cluster (arrows) is attached to the prostate. (B) Posterolateral surface of the prostate. A cell cluster (arrows) is attached to the prostate. (C) Lateral to the seminal vesicle. Portions of the pelvic plexus show focal enlargements (arrow) containing ganglion cells. (D) Levator ani muscle. This sometimes contains a small cluster of ganglion cells. Scale bars, 1 mm. P, prostate; B, bladder; SV, seminal vesicle.

Create Platinum Slide Series

Fig. 2 Immunohistochemistry of a ganglion cell cluster attached to the posterolateral surface of the prostate.
Tyrosine hydroxylase (TH; panel A) and peptide histidine isoleucine (PHI; panel B) are immunostained as sympathetic and parasympathetic markers, respectively. Sections are cut at 50-μm intervals. TH-positive cells and PHI-positive cells co-exist but tend to cluster with one another. Arrows, immunohistochemically negative cells. P, prostate. Scale bars, 0.3 mm.

Create Platinum Slide Series

Fig. 3 Ganglion cells along pelvic splanchnic nerve roots.
Panel A (H and E staining) is a transverse section of the pelvic splanchnic nerve roots 30 mm ventral to the S2 to 4 anterior sacral foramina. Some minimal dissection was performed to straighten the nerve courses. Arrows indicate transverse sections of nerves. Two nerve components found near the caudal and cranial margins of the section (open star and black star) contain ganglion cells. Panels B and C are higher-magnification views of the ganglia marked by the open star and black star, respectively. In panel B, TH-positive cells (61% of all ganlion cells in the panel) are intermingled with negative cells (arrows). In contrast, only one cell (3% of all ganglion cells in panel C; arrow; H and E staining) stained for TH (arrow; panel D). A nerve bundle (asterisk in panels C and D) is TH-positive. Scale bars, 5 mm (panel A) and 0.3 mm (panels B to D).

Table 1 Numbers and distribution of extramural ganglion cells in the male hemipelvis: An evaluation using semiserial sections at 1-mm intervals

Specimen	M1	M2	M3	M4	M5	M6	M7	Average No.
NERVE
Hypogastric nervea	604	945	276	–	248	–	825	579.6 ± 314.8
Pelvic splanchnic nervesb	1262	396	223	84	853	285	765	552.6 ± 420.9
Pelvic plexus	1113	332	250	–	534	411	575	535.8 ± 307.7
Sacral sym. ganglionc	1566	687	–	249	1092	849	1884	1054.5 ± 596.2
VISCERA
Bladder, dorsal aspect	172	48	44	0	10	162	567	143.3 ± 199.1
B/P junctiond	78	280	135	53	50	232	211	148.4 ± 93.2
SVe, dorsal aspectf	163	25	78	–	212	45	273	132.7 ± 99.2
Prostate
Dorsal aspect	155	101	0	65	230	15	535	157.3 ± 184.7
Ventral aspect	0	0	10	–	3	0	0	2.2 ± 4.0
Near the apexg	15	0	10	109	177	104	387	114.6 ± 136.8
NVBh	698	230	448	66	908	96	500	420.9 ± 313.3
In levator ani	0	0	–	16	0	3	0	3.2 ± 6.4
Total numbers of cells	5826	3044	1474	642	4317	2202	6522
% in VISCERAi	22	22	–	–	37	–	38

^a Along the hypogastric nerve and distal ureter.

^b From 0 to 30 mm from the anterior sacral foramina.

^c Macroscopically Sacral sympathetic ganglion.

^d Bladder-prostate junction.

^e Seminal vesicle.

^f Histologically, the seminal vesicle completely overlapped the upper half of the prostate. Thus, the “upper half of the prostate” in typical surgical views was included in these sections.

^g Within 10 mm from the apex, not including the rhabdosphincter area.

^h Neurovascular bundle, posterolateral aspect of the prostate.

ⁱ (Number of cells in VISCERA/total cell number) x 100.

Create Platinum Slide Series

Fig. 4 Localizationof the ganglion cells in the male pelvis. Panel A displays the ganglion localization descrived in Table 1. The 4 sites along the macroscopically identified nerves, those are shown in the upper part of Table 1 and contain more than half of the ganglion cells, correspond to the hypogastric nerve, pelvic splanchnic nerve, pelvic plexus and sacral sympathetic ganglia. In addition, as shown in the lower major part of Table 1, some parts of the ganglion cell population are present immediately around the the dorsal aspect of bladder, bladder/prostate junction, apex of the prostate, dorsal aspect of the prostate, seminal vesicle, neurovascular bundle. Panel B demonstrates the distribution of the ganglion cells (stars) immediately around the seminal vesicle and prostate (posterior view). These stars show the average numbers of ganglion cells at each sites. Even in the nerve-sparing retropubic radical prostatectomy, the ganglion cells at the dorsal aspect of seminal vesicle, bladder/prostate junction, and along the prostate must be excised, due to the location close to the visceral capsules. These visceral ganglion cells are attached to the prostatic capsule or vesical smooth muscle, or even embedded within the capsule or smooth muscle. By contrast with these sites, the ganglion cells at the pelvic plexus might be preserved because these are separated from the seminal vesicle by at least 1.0 mm. AP; apex of the prostate, VA; ventral aspect of the prostate, DA; dorsal aspect of the prostate, BL; dorsal aspect of the bladder, B/P; bladder-prostate junction, SV; dorsal aspect of the seminal vesicle, UR; ureter, HGN; hypogastric nerve, SSG; sacral sympathetic ganglion, PSN; pelvic splanchnic nerve, PP; pelvic plexus, LA; levator ani, NVB; neurovascular bundle, DEN; Denonvilliers’ fascia, REC; rectum. A large and small star show a hundred and ten ganglion cells, respectively.

In and along macroscopically identified nerve components, many ganglion cells could be found (Table 1, upper portion). Along pelvic splanchnic nerves, ganglion cells usually were distributed distally, 15 to 30 mm ventral from the anterior sacral foramina (Fig. 3). However, in some individuals, these cells were found along nearly the entire nerve course, including a site immediately ventral to the anterior sacral foramen. Great interindividual differences in cell numbers were evident along splanchnic nerves (84 to 1262 cells). Because the cavernous nerve runs near or along the levator ani, GCC sometimes were located upon the fascia covering this muscle. The hypogastric nerve, previously believed to contain mainly sympathetic nerve fibers, also contained GCC, even a large ganglion-like mass, along its distal course near the distal ureter at levels caudal to S2 (not illustrated). The pelvic plexus along the lateral aspect of the seminal vesicle also contained many ganglion cells (250 to 1113). GCC were not attached to the seminal vesicle, but were sepsrated from it by at least 1.0 mm (Fig. 1C). Macroscopically, grayish ganglia found along the sacral sympathetic trunk numbered 2 to 8 per pelvis. However, the size of these ganglia (1 to 6 mm in maximum length) and cell numbers (249 to 1566) varied widely between individuals. In addition, we sometimes found ganglion cells along nerves passing through the levator ani [9] (Fig. 1D).

Along the pelvic visceral surface GCC were not distributed evenly, but rather at limited sites located on the dorsal aspect of the bladder, the bladder/prostate junction (Fig. 1A), the dorsal aspect of the seminal vesicle, and along the prostate (Table 1, lower portion). At these sites, one to five cell clusters were found per section, each composed of 5 to 65 ganglion cells. These visceral GCC were attached to the prostatic capsule or vesical smooth muscle, or even embedded within the capsule or smooth muscle. In particular, the posterolateral aspect of the prostate, which is identified as the neurovascular bundle in the surgical field, contained many ganglion cells (66 to 908; Fig. 1, and Fig. 2). Comparatively large numbers of ganglion cells were present even near the prostatic apex, the most peripheral portion of the pelvis. In relation to the ventral aspect of the prostate, we found a few GCC in only two specimens; these clusters were located ventrally but close to the apex. No ganglion cells were found along the extrapelvic pudendal nerve, in or around cavernous tissues including the penile hilum, in the area of the rhabdosphincter, in retropubic fat, or in well-developed recto-urethral muscle.

Those results, including interindividual differences, are summarized in Table 1 for seven cadavers in which essentially all sites in the pelvis were investigated. Total number of cells and cell numbers at each site displayed significant interindividual variation (total, 3044 to 6522 cells). The ratio of viscerally located ganglion cells to all ganglion cells was 22% to 38%. Seven other specimens in which ganglion cell distribution was investigated surrounding the prostate showed similar results (data not shown).

4. Discussion

Urologists who perform nerve-sparing radical prostatectomy pay careful attention to the course of the cavernous nerve, but seldom if ever consider the distribution of ganglion cells. Using a rat model, Kato et al. [1] recently reported difficulty in regeneration of injured ganglion cells, even following simple damage to nerve fibers. Accordingly, we need to know the distribution, number, and character of ganglion cells at each site in the male pelvis. In this study we demonstrated three features of human pelvic autonomic ganglia: relative scattering of ganglion cells, significant interindividual differences in cell numbers, and co-existence of sympathetic and parasympathetic cells within a GCC.

The present study demonstrated a site-specific distribution of pelvic autonomic ganglion cells. Even when the large numbers of cells found along the sacral sympathetic trunk were excluded, most remaining cells still were located in and along macroscopically identified nerve components, especially in parts of the pelvic plexus near the seminal vesicle and along the proximal course of the cavernous nerve. The latter corresponds to the most caudal portion of the pelvic splanchnic nerve [10]. GCC attached to the pelvic viscera were more limited in number (22% to 38% of all ganglion cells). Visceral GCC included those on the dorsal aspect of the bladder, at the bladder/prostate junction, and at the dorsal aspect of the seminal vesicle and along the prostate. No GCC were found along the extrapelvic pudendal nerve, in or around the cavernous tissues including the penile hilum, in the area of the rhabdosphincter, in retropubic fatty tissue, or in the recto-urethral muscle. Nevertheless, human pelvic ganglion cells were distributed over a large area at various sites, in contrast to pelvic ganglion cells in experimental animals, where ganglion cells appear largely limited to major pelvic ganglia [11]. This is consistent in part with interindividual variation in cavernous nerve course [12]. In nerve-sparing radical prostatectomy, urologists are likely to believe that parasympathetic ganglion cells are essentially preserved if the neurovascular bundle also preserved. However, visceral GCC often are attached to the prostatic capsule or even embedded in the capsule (Fig. 1B), and may be removed with the prostate. Additionally, when we separate the seminal vesicle from the dorsal aspect of the bladder, or cut between the bladder and prostate with ordinary procedures using an electric scalpel, we may inadvertently injure many ganglion cells (Fig. 4B). Taguchi et al. [13] carried out whole-mount staining of fresh human pelvic viscera using acetylcholineesterase enzyme histochemistry, dividing parasympathetic ganglia into primary and secondary. The former were located along the inputs to the pelvic plexus, while the latter connected to postganglionic fibers, apparently acting as interneurons. Because secondary ganglia tended to be located along and near pelvic viscera, they seemed more vulnerable to injury during radical prostatectomy than primary ganglia. Thus, we can not preserve all extramural GCC in the male pelvis even in the nerve-sparing surgery.

The present study suggested great interindividual differences in ganglion cell numbers, more than twofold in the whole pelvis and more than fifteenfold in the pelvic splanchnic nerve, although we examined only semiserial sections. For comparison, counts of the total number of autonomic ganglion cells in experimental animals were 540 to 1080 in the rat pulmonary nerve plexus [14] and 5000 to 7000 in the rat cardiac nerve plexus [15]. This interindividual variation in cell numbers and GCC ratio attached to the pelvic viscera appears likely to influence functional outcome after surgery. Although much further study is required, we suspect that greater numbers of cells, especially an abundance of secondary ganglion cells, may confer greater resistance to dysfunction from surgical injuries.

If TH-positive cells are sympathetic and PHI-positive cells are parasympathetic, the present study indicated that both kinds of cells intermingled and co-existed in individual ganglion cell clusters in the male pelvis. Notably the hypogastric nerve, believed to convey sympathetic input to pelvic viscera [3], included cells that stained as parasympathetic ganglion cells. These ganglion cells appear to give rise to ascending nerve fibers to the ureter and descending colon [16]. In fact some studies showed the colocalization of TH and VIP in the nerve fibers of human ureter [17], and [18]. Thus, damage to parasympathetic rather than sympathetic elements in the hypogastric nerve would appear to result in dysfunction of these organs. Likewise, the pelvic splanchnic nerve appeared to contain sympathetic fibers as well as parasympathetic fibers. Co-existence or intermingling of the two types of ganglion cells has been described in experimental animals [19] and also in human cadavers [4], [20], and [21]. Our result confirmed the previous studies. Consequently, the usual simple classification of pelvic nerve components as “sympathetic hypogastric nerve” or “parasympathetic pelvic splanchnic nerve,” appears to distort understanding of autonomic physiology, and could impede evaluation of nerve damage during surgery.

5. Conclusions

Human pelvic autonomic ganglia existed not only in nerve components but to a lesser extent along visceral surfaces. The usual simple designation of grossly identified pelvic nerve components as either sympathetic or parasympathetic is likely to cause misunderstanding of pelvis physiology. Distribution, numbers, proportions of sympathetic and parasympathetic pelvic ganglion cells are likely to influence outcome after nerve-sparing surgery.