SEARCH

SEARCH BY CITATION

Keywords:

  • diverticular disease;
  • enteric nervous sys-tem;
  • enteric neuropathy;
  • morphometry;
  • pathogenesis

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

Background  The pathogenesis of diverticular disease (DD) is attributed to several aetiological factors (e.g. age, diet, connective tissue disorders) but also includes distinct intestinal motor abnormalities. Although the enteric nervous system (ENS) is the key-regulator of intestinal motility, data on neuropathological alterations are limited. The study aimed to investigate the ENS by a systematic morphometric analysis.

Methods  Full-thickness sigmoid specimens obtained from patients with symptomatic DD (n = 27) and controls (n = 27) were processed for conventional histology and immunohistochemistry using anti-HuC/D as pan-neuronal marker. Enteric ganglia, nerve and glial cells were quantified separately in the myenteric, external and internal submucosal plexus compartments.

Key Results  Compared to controls, patients with DD showed significantly (P < 0.05) (i) reduced neuronal density in all enteric nerve plexus, (ii) decrease of ganglionic nerve cell content in the myenteric plexus, (iii) decreased ganglionic density in the internal submucosal plexus, (iv) reduced glial cell density in the myenteric plexus, (v) decrease of ganglionic glial cell content in the myenteric plexus and increase in submucosal plexus compartments, (vi) increased glia index in all enteric nerve plexus. About 44.4% of patients with DD exhibited myenteric ganglia displaying enteric gliosis.

Conclusions & Inferences  Patients with DD show substantial structural alterations of the ENS mainly characterized by myenteric and submucosal oligo-neuronal hypoganglionosis which may account for intestinal motor abnormalities reported in DD. The morphometric data give evidence that DD is associated with structural alterations of the ENS which may complement established pathogenetic concepts.


Abbreviations:
DD

diverticular disease

ENS

enteric nervous system

GINMP

gastrointestinal neuro-muscular pathologies

Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

Diverticular disease (DD) is the most common morphological abnormality of the large intestine characterized by multiple mucosal herniations throughout the colonic wall. Due to its high prevalence especially in the elderly and the associated complications, DD represents the 5th most important gastrointestinal disease in western countries in terms of direct and indirect costs with a mortality rate of 2.5 per 100 000 population per year.1 Inspite of the worldwide distribution, the pathogenesis of DD is still poorly understood and considered to be multifactorial. Classical aetiological factors which have been shown to favour the formation of colonic diverticula include increasing age, low fibre diet, alterations of intramural connective tissue and genetic factors.2

However, the final pathophysiological pathway resulting in mucosal herniation appears to be an increased intraluminal pressure. It is thought that chronic excessive segmental contractions induce a concertina-like colonic wall (bladder colon) causing both symptoms of functional obstruction and painful sensations. Evidence for disturbed motility patterns underlying DD is mainly derived from manometric and myoelectric in-vivo studies showing that patients with DD are characterized by higher intracolonic pressure, enhanced high-amplitude propagated contractions and increased motility indices after administration of meals or provocative pharmacological agents.3–6

Although there is substantial evidence that DD is related to colonic motor disturbances, few studies have evaluated the enteric nervous system (ENS) as the pivotal regulator of intestinal motility in patients with DD. However, the hypothesis of an enteric neuropathy underlying DD seems to be justified, since a variety of colonic motor dysfunctions (e.g. irritable bowel syndrome, chronic intestinal pseudoobstruction, slow-transit constipation, idiopathic megacolon) are associated with distinct abnormalities of the ENS.7,8

More than four decades ago, Macbeth and Hawthorn9 were the first to described enlarged and ectopically located myenteric ganglia in colonic specimens of DD and postulated that DD is primarily defined by an ‘acquired neuromuscular derangement’. Since then, histopathological data related to alterations of the ENS in DD have been scarcely reported and were based on different methodical approaches with fairly controversial findings: whereas two studies yielded unchanged enteric nerve cell numbers,10,11 other groups reported a decrease of myenteric nerve cells,12 but an increase of submucosal nerve cells.13

Due to the inconclusive data the present study aimed at re-evaluating the ENS in patients with DD by means of a comprehensive morphometric analysis carried out for the entire ganglionated enteric nerve plexus. The systematic quantitative assessment involved all functionally relevant key-structures of the ENS (ganglia, nerve cells, glial cells) and is the first examination strictly adhering to international guidelines on histological reporting for gastrointestinal neuromuscular pathologies (GINMP).14 In particular, the study addresses the question whether DD is associated with an enteric neuropathy that might have a relationship to reported intestinal motor disturbances in patients with DD.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

Patients

Control group  Segments of sigmoid colon were obtained from patients (n = 27, mean age 64.4 years, 13 females, 14 males) who underwent anterior recto-sigmoid resection for non-obstructive colorectal carcinoma. All patients reported normal bowel habits and showed no evidence of diverticula nor anorectal out-let obstruction.

Patients with DD  Segments of sigmoid colon were obtained from patients (= 27, mean age 61.7 years, 18 females, nine males) who underwent sigmoid resection or left hemicolectomy (depending on the extent of diverticula) for symptomatic DD. All patients had reported recurrent abdominal pain and additionally change in bowel habits (25.9%) and episodes of peranal bleeding (11.1%). Symptoms had lasted between 1 year (40.7%), 2–3 years (33.3%) and 4–5 years (25.9%). The number of reported acute attacks of diverticulitis was 2 (63%) and 3 (37%). Preoperative colonoscopy and radiological studies (colonic enema, CT, MRI) showed left sided multiple diverticula which were confined to the sigmoid colon in 70.3% or additionally extended to the descending colon in 22.2% and up to the transverse colon in 7.4% of the cases. In 29.6% of the patients a partial luminal stenosis due to strictures was found. Operations were performed as elective (77.8%) or early elective (22.2%) surgery. Two patients showed colonic perforation with abscess formation covered by the greater omentum. Six patients (22.2%) presented intraoperatively with an encapsulated peridiverticular mesocolic abscess corresponding to Hinchey grade I. The study of human tissue received approval from the Local Ethics Committee of the Faculty of Medicine, University of Luebeck.

Tissue processing and conventional histology

Specimens from the control group were harvested at safe distance (>5 cm) from the tumour. Specimens from patients with DD were harvested from those sites without evidence of macroscopic shortening of the bowel wall (e.g. due to fibrotic contraction). After surgical removal all specimens were transferred into phosphate-buffered saline (PBS) at 37 °C to allow adaption and further dissection. Full-thickness rectangular tissue blocks (30 mm × 10 mm) were pinned out flat on a cork plate by fine needles without artificial stretching nor shortening thereby preserving the original size. The longer border of the tissue block was orientated perpendicular to the gut axis and corresponded to the cutting surface for histologic sections, so that myocytes of the circular muscle layer were cut along their longitudinal axis. After fixation (4% paraformaldehyde PBS) for 24 h and dehydration tissue blocks were transferred into paraffin wax and cut in sections (6 μm) for conventional histology (haematoxylin/eosin, Azan) and immunohistochemistry. Sections displaying transmural diverticula and/or fibrosis with a deranged architecture of the gut wall were excluded from the morphometric analysis of the ENS to prevent biased data sampling.

Immunohistochemistry

Among the available neuronal markers tested in preliminary experiments (protein gene product 9.5, neurofilaments, neuron-specific enolase, peripherin, HuC/D) anti-HuC/D (monoclonal murine antibody; Molecular Probes, Invitrogen, CA, USA) was used for the morphometric analysis of the ENS based on the following properties: (i) pan-neuronal marker labelling virtually all enteric nerve cells, (ii) visualization of neuronal somata but not neuronal processes to allow convenient counting of nerve cells, (iii) robust immunoreactive signal in routine paraffin sections.

Immunohistochemistry was carried out using the avidin-biotin-complex system (Vectastain ABC Standard; Vector Laboratories, Burlingame, CA, USA) combined with heat-induced epitope retrieval. Briefly, sections were incubated with 3% hydrogen peroxide to block endogenous peroxidase activity, rinsed in 10 mmol L−1 TRIS in 0.15 M sodium chloride, pH 7.4 (TRIS-buffered saline, TBS) and pretreated with citrate buffer (pH 6.0, 95 °C waterbath) for 45 min. Sections were incubated overnight with anti-HuC/D (1 : 500) diluted in antibody diluent (Zymed, Invitrogen, CA, USA), biotinylated goat anti-mouse IgG (1 : 200; Jackson Immuno Research, PA, USA) for 30 min and ABC conjugated with horseradish peroxidase for 1 h. 3-amino-9-ethylcarbazol (AEC ready-to-use, DakoCytomation, Hamburg, Germany) was used as substrate chromogen. Sections were counterstained with Meyer′s haematoxylin. Omission of the primary or secondary antibody served as negative controls.

Morphometry of the ENS

Technical devices, software  Morphometric analysis was carried out by using a light microscope (Axiophot, Zeiss, Jena, Germany) coupled to a digital camera (Axiocam, Zeiss). A specially edited software program (KS 100; Zeiss) was used to record the relevant structures. The data were transferred into Excel software (version 9.0) and further processed for statistical analysis (spss, statistical package for social sciences, version 12.0; SPSS Inc., Chicago, IL, USA).

Principle of morphometric analysis  Quantification of enteric ganglia and its components was performed for each of the ganglionated enteric nerve plexus (Fig. 1A): (i) myenteric plexus (Auerbach plexus) located between the circular and longitudinal muscle layer, (ii) external submucosal plexus (Schabadasch plexus) located within the submucosa adjacent to the circular muscle layer, (iii) internal submucosal plexus (Meissner plexus) located adjacent to the mucosa. All data concerning ganglionic, neuronal and glial cell density were referred to the intestinal length measured in each specimen (ca. 30 mm) and extrapolated to 100 mm intestinal length to allow comparison of the data. The following structures were recorded based on anti-HuC/D labelled, haematoxylin counterstained full-thickness sections (Fig. 1B):

image

Figure 1.  (A) Enteric nerve plexus compartments included in the morphometric analysis. Nerve cells (black circles with grey dots) and glial cells (black dots) were recorded separately within ganglia of the myenteric plexus (MP), external submucosal plexus (ESP) and internal submucosal plexus (ISP). Data for ganglionic, neuronal and glial cell density were referred to the intestinal length (IL). Muc, mucosa; SM, submucosa; CM, circular muscle layer; LM, longitudinal muscle layer. (B) Identification of intraganglionic cell types based on anti-HuC/D immunohistochemistry and haematoxylin counterstaining. HuC/D-immunoreactive nerve cells were characterized by dark-red brownish nuclei [NC (n)] and light-red granular pericarya [NC (pc)]. Glial cells were non-HuC/D-immunoreactive and identified by blue stained small nuclei [GC (n)]. G, myenteric ganglion; dotted line, ganglionic border. Original magnification ×63.

Download figure to PowerPoint

Ganglia  Enteric ganglia were identified by groups of HuC/D-immunoreactive neuronal somata surrounded by glial cells. Both the area and number of ganglia were recorded. The following variables were calculated for statistical comparison: (i) ganglionic area, (ii) ganglionic number per intestinal length reflecting ganglionic density.

Nerve cells  Enteric nerve cells were identified by HuC/D-immunoreactivity characterized by a dark-red brownish nucleus and a light-red granular pericaryon. HuC/D-immunoreactive cellular fragments without a discernible nucleus were excluded from the recordings for stereologic reasons to avoid overestimation of the neuronal number within a given section. The neuronal number was recorded and the following variables were calculated for statistical comparison: (i) neuronal number per ganglion, (ii) neuronal number per intestinal length reflecting neuronal density.

Glial cells  Enteric glial cells corresponded to those intraganglionic cellular elements not immunoreactive for HuC/D and were identified by their haematoxylin-stained blue small nuclei. The glial cell number was recorded and the following variables were calculated for statistical comparison: (i) glial cell number per ganglion, (ii) glial cell number per intestinal length reflecting glial cell density.

Glia index  The glia index corresponded to the ratio of glial cells to neurons within a given specimen reflecting the numerical relationship between glial cells and neurons.

Statistical analysis

Statistical comparison of the data between the control group and patients with DD was carried out by using non-parametric Mann–Whitney U-tests for two independent samples with P ≤ 0.05 considered as an indicator of significance. The data were compared separately for each enteric nerve plexus and are shown as median ± interquartiles and graphically presented by box-whisker plots indicating the median (horizontal line), 50% of values (box) and 99% of values (whiskers). Professional statistical consulting was provided by the Institute of Medical Biometry and Statistics (University of Luebeck).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

Morphometric data

Ganglia  Whereas the internal submucosal plexus showed significantly (P = 0.001) less ganglia per 100 mm intestinal length in DD (81.8 ± 44.2) than in controls (123 ± 98.5), the ganglionic density did not differ significantly between both groups in the myenteric plexus (controls: 174.5 ± 93.1, DD: 161.6 ± 50.8) and external submucosal plexus (controls: 97.8 ± 63.8, DD: 71.3 ± 47.5) (Fig. 2A). The average ganglionic area (μm2) was similar in the myenteric plexus (controls: 8840.9 ± 5464.5, DD: 8154.6 ± 4585.8), but differed significantly in both external submucosal plexus (controls: 1439.1 ± 1086.3, DD: 1819.0 ± 1158.0, P = 0.013) and internal submucosal plexus (controls: 774.8 ± 501.4, DD: 1023.9 ± 637.7, P = 0.002) (Fig. 2B).

image

Figure 2.  Statistical comparison of ganglionic density (A) and mean ganglionic area (B) between control group (white boxes) and patients with DD (grey boxes). MP, myenteric plexus; ESP, external submucosal plexus; ISP, internal submucosal plexus. *P ≤ 0.05; **P ≤ 0.01.

Download figure to PowerPoint

Neurons  The neuronal number per 100 mm intestinal length was significantly decreased in patients with DD compared to the control group. The numerical decline of enteric nerve cells was evident in the myenteric plexus (controls: 1607.6 ± 543.9, DD: 886.7 ± 564.4, P < 0.001), external submucosal plexus (controls: 228.3 ± 191.7, DD: 162.6 ± 126.4, P = 0.039), and internal submucosal plexus (controls: 259.6 ± 253.9, DD: 141.0 ± 114.3, P = 0.002) (Fig. 3A). The average neuronal number per ganglion was also decreased in patients with DD compared to controls. However, the reduced ganglionic nerve cell content was statistically significant only in myenteric ganglia (controls: 7.9 ± 3.5, DD: 5.6 ± 2.1, P < 0.001), but not in ganglia of the external submucosal plexus (controls: 2.3 ± 0.9, DD: 2.1 ± 0.7) and internal submucosal plexus (controls: 1.84 ± 0.8, DD: 1.80 ± 0.5) (Fig. 3B).

image

Figure 3.  Statistical comparison of neuronal density (A) and mean neuronal number per ganglion (B) between control group (white boxes) and patients with DD (grey boxes). MP, myenteric plexus; ESP, external submucosal plexus; ISP, internal submucosal plexus. *P ≤ 0.05; **P ≤ 0.01.

Download figure to PowerPoint

Glial cells  Whereas the myenteric plexus showed a significantly (P = 0.008) reduced glial cell number per 100 mm intestinal length in DD (6341.3 ± 3893.5) compared to controls (9869.2 ± 3566.2), the glial cell density did not differ significantly between both groups in the external submucosal plexus (control: 616.4 ±779.7, DD: 683.8 ± 561.0) and internal submucosal plexus (control: 510.0 ± 494.4, DD: 526.0 ± 407.2) (Fig. 4A). The average glial cell number per ganglion was significantly decreased in patients with DD in the myenteric plexus (controls: 51.9 ± 34.3, DD: 38.6 ± 23.3, P = 0.031). In contrast, ganglia of both the external submucosal plexus (controls: 7.5 ± 5.8, DD: 8.6 ± 3.5, P = 0.048) and internal submucosal plexus (controls: 4.4 ± 2.1, DD: 6.2 ± 2.3, P < 0.001) showed a significantly increased glial cell content in patients with DD (Fig. 4B).

image

Figure 4.  Statistical comparison of glial cell density (A) and mean glial cell number per ganglion (B) between control group (white boxes) and patients with DD (grey boxes). MP, myenteric plexus; ESP, external submucosal plexus; ISP, internal submucosal plexus. *P ≤ 0.05; **P ≤ 0.01.

Download figure to PowerPoint

Glia index  In all enteric nerve plexus compartments the number of glial cells outranged the number of nerve cells. In both groups the glia index increased continuously in the following order: internal submucosal plexus, external submucosal plexus and myenteric plexus. Comparison of controls with patients with DD showed that in each of the enteric nerve plexus the glia index was significantly higher in DD: myenteric plexus (controls: 6.5, DD: 8.0, P = 0.012), external submucosal plexus (controls: 2.7, DD: 4.5, P < 0.001), internal submucosal plexus (controls: 2.1, DD: 3.4, P < 0.001).

Histopathological pecularities of the ENS in DD

Although myenteric ganglia of patients with DD generally showed a distribution pattern and sizes comparable to those of the control group, they contained less neurons and glial cells (Fig. 5) as confirmed by morphometric analysis. Compared to the apparent nerve cell decline in myenteric ganglia, the decreased neuronal density in submucosal ganglia evidenced by morphometric analysis was less obvious at histological examination (Fig. 5).

image

Figure 5.  Ganglia from the control group (A–C) and DD (D–F). Whereas the boundaries (dotted line) of myenteric ganglia showed similar dimensions, the number of HuC/D-immunoreactive myenteric nerve cells (NC) was markedly reduced in DD (D) compared to controls (A). In both groups ganglia of the external (ESP) and internal (ISP) submucosal plexus were readily discernible extending adjacent to the circular muscle layer (CM) and close to the mucosa (Muc). Whereas submucosal ganglia of DD (E,F) were characterized morphometrically by an increased size and glial cell content compared to controls (B,C), the overall nerve cell density was decreased, however, not as pronounced as in the myenteric plexus. CM, circular muscle layer; LM, longitudinal muscle layer. anti-HuC/D immunohistochemistry with haematoxylin counterstaining, original magnification ×20 (A,D), ×40 (B,C,E,F).

Download figure to PowerPoint

Apart from regularly shaped myenteric ganglia, a subgroup of specimens obtained from DD exhibited considerably enlarged myenteric ganglia with hypertrophied neuropil (Fig. 6). The bulbous protrusions extended into adjacent muscle layers and were mainly composed of glial cells but bare of nerve cells. These gliosis-like formations were observed in 44.4% (12/27) of patients with DD and appeared at irregular intervals along the myenteric plexus plane.

image

Figure 6.  Histopathological pecularities of myenteric ganglia from DD (A,B) in comparison to control group (C). The enlarged ganglia (A,B) showed hypertrophied neuropil with bulbous protrusions extending into adjacent muscle layers (dotted lines). The ganglionic bulges (asterisks) were mainly composed of glial cells but bare of nerve cells (NC) as illustrated by both immunohistochemical (A) and Azan (B) stainings. CM, circular muscle layer; LM, longitudinal muscle layer. anti-HuC/D immunohistochemistry with haematoxylin counterstaining (A), Azan staining (B,C), original magnification ×10.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

Morphometric methodologies for quantitative analysis of the ENS

Precise quantification of enteric ganglia, neurons and glial cells is an essential tool for the reliable histopathological diagnosis of intestinal innervation disorders, but remains a challenging task reflected by the disparity of previously published quantitative data. The diverging ranges for normal values and the lack of reliable reference data on the human ENS can be attributed to multiple reasons.8 (i) various methods to visualize components of the ENS, (ii) ongoing search for optimal enteric pan-neuronal and pan-glial cell markers, (iii) different morphometric approaches, (iv) limitation of quantitative analysis to selected enteric nerve plexus. Being aware of these pending issues, an International Working Group has currently established guidelines on histological reporting for gastrointestinal neuromuscular pathologies (GINMP) on behalf of the upcoming GASTRO congress 2009.14

According to these consensus-based recommendations, the present study has implemented a systematic morphometric approach which has (i) focused on all ganglionated enteric nerve plexus compartments, (ii) used anti-HuC/D as an appropriate pan-neuronal immunohistochemical marker to perform enteric nerve cell counts,15–18 (iii) recorded all key-structures of the ENS (ganglia, neurons, glial cells) thereby, (iv) allowing the determination of overall ganglionic, neuronal and glial cell density and the average nerve and glial cell content per ganglion. By using these variables, a reliable quantification and statistical comparison of data retrieved from a sufficiently sized control group and patients with DD have been attempted. Biased sampling of quantitative data due to bowel wall shortening frequently seen in DD has been prevented by standardized orientation of tissue blocks and exclusion of colonic segments from the morphometric analysis characterized by the presence of transmural diverticula and/or fibrosis.

Alterations of enteric nerve cells in patients with DD

The morphometric analysis of the ENS revealed a statistically significant decrease of neuronal density in patients with DD evident within all enteric nerve plexus compartments. Compared to controls enteric nerve cell numbers decreased by 44.9% in the myenteric plexus, by 28.8% in the external submucosal plexus, and by 45.7% in the internal submucosal plexus. The discrepancy between the highly significant decline of overall neuronal density and a less evident decrease of the average ganglionic nerve cell content is most likely due to a concomitant decrease of ganglionic density in DD illustrating at the same time that morphometric assessments limited to neuronal numbers per ganglion may mask a general shortfall of nerve cells related to intestinal length.

According to a consensus conference on colonic innervation disorders,19 the present findings resemble typical histopathological features of oligo-neuronal hypoganglionosis. In contrast to intestinal aganglionosis characterized by a complete lack of intramural neurons, oligo-neuronal hypoganglionosis is defined by a significant numerical decrease of enteric neurons and has been described previously in several intestinal motility disorders such as in slow-transit constipation,20,21 adult megacolon,22,23 chronic intestinal pseudoobstruction and chronic constipation in children.24,25 These studies have suggested that a decline of enteric nerve cells beneath a physiologically required critical amount may compromise proper enteric neurotransmission resulting in disturbed intestinal motor functions.

The reduced myenteric neuronal density observed in the present study confirms a previous quantitative analysis in patients with DD also reporting decreased myenteric ganglion cell numbers per intestinal length.12 Similar findings have been described in cases of jejunal diverticulosis characterized by a degenerative loss of enteric neurons consistent with a visceral neuropathy.26 Accordingly, a PGP 9.5 based immunohistochemical study showed reduced and less densely distributed nerve tissue within muscle layers of patients with DD.27 In contrast, other studies either based on neuronal cell counts per microscopic field11 or using silver impregnation techniques10 did not find quantitative alterations of myenteric neurons in DD. These inconsistent findings may partly be explained by the different morphometric approaches and staining methods applied. According to the currently released guidelines on histological reporting for GINMP14 immunohistochemical methods are superior to silver impregnation techniques and determination of the neuronal number per intestinal length appears to be a more appropriate parameter than estimation of neuronal numbers in preselected microscopic fields. Intestinal neuronal dysplasia defined by enlarged submucosal ganglia with an increased nerve cell content previously described in patients with DD13 could only partially be confirmed in the present study: although the ganglionic area of submucosal plexus layers was increased in DD, the ganglionic nerve cell content was not elevated and overall neuronal density was even decreased in submucosal ganglia.

Alterations of enteric glial cells in patients with DD

Patients with DD showed a decreased glial cell density in myenteric ganglia. This finding is in accordance with a previous study yielding a significantly reduced number of protein S-100 immunoreactive myenteric glial cells in patients with DD.11 As the decline of nerve cells was more pronounced than that of glial cells, the glia index (glia:neuron ratio) turned out to be increased in patients with DD compared to controls. Similar as in patients with DD we have observed an increased glia index also in patients with slow-transit constipation likewise characterized by oligo-neuronal hypoganglionosis.20

Generally, data on the glia index in the human colon are scarce. The glia index yielded for the control group (6.5) fall within the range of glia indices described in previous studies using Sox8/9/10 immunohistochemistry (5.9–7.0)28 or PGP 9.5 immunohistochemistry combined with haematoxylin counterstain (8.7 ± 1.9).20 The relatively low glia index (3.7 ± 0.3) based on double-labelling HuC/D-S100β immunohistochemistry reported by Ippolito et al.18 was most likely due to cell counts also including those enteric nerve cells without a visible nucleus.

Although the overall glial cell density was reduced in DD, a subgroup of myenteric ganglia displayed bulbous protrusions almost exclusively composed of glial cells. These gliosis-like formations were clearly distinct from excessive enteric glial cell proliferation reported in peripheral neurofibromatosis (von Recklinghausen disease) or ganglioneuromatosis in multiple endocrine neoplasia type 2B,29 but rather resembled findings of earlier studies which have described the presence of hypertrophic myenteric ganglia in patients with DD.9,12 In other studies on neuropathological lesions in DD10,11,13 such an enteric gliosis has not been consistently reported – which might be attributable to the fact that the peculiar glial cell hyperplasia was confined to 44.4% of the specimens examined and appeared only at irregular intervalls along the myenteric plexus plane.

The functional significance of quantitative glial cell alterations is still a matter of debate. There is growing evidence that apart from their supportive role glial cells are also actively involved in the pathogenesis of intestinal motility disorders.30 Thus, depletion of enteric glial cells have been reported in patients with slow-transit constipation, idiopathic megacolon and obstructive defecatory syndrome suggesting that intestinal motor functions also depend on intact glial cells. Indeed, animal studies have shown that glial cells produce neurotrophic factors for the survival and proliferation of enteric neurons, modulate neuronal gene expression, influence the neurochemical coding and, thus, may alter both the neuronal phenotype and enteric neurotransmission including nociception.31

Impact of enteric neuropathy in DD

The impact of an enteric neuropathy associated with DD may be two-fold: first, an intestinal innervation disorder would result in abnormal intestinal motor functions. This assumption is supported by several in-vitro motility studies of muscular specimens from DD which have revealed abnormal contractility responses to nitric oxide,32,33 tachykinins,34,35 substance P,36 endocannabinoids,37 and acetylcholine.27,38 Obviously, structural lesions of the ENS such as demonstrated in the present study are reflected by functional deficiencies of the ENS to properly mediate intestinal motility.

Second, an impaired ENS may contribute to the generation of symptoms in DD. This concept has been particularly advocated by Spiller, Simpson and co-workers who could show a selective upregulation of neuropeptides (e.g. tachykinins and galanin with a nearly 10-fold increase) in the mucosa of patients with symptomatic diverticulosis.39 In the same study an increased density and pronounced angulation of nerve fibres have been demonstrated in chronic DD suggesting an active sprouting of nerve cell processes. It was hypothesized that inflammation may lead to nerve tissue remodelling and altered enteric neurochemical coding underlying the persistent sensory and motor dysfunctions in DD. Thus, neural abnormalities may be responsable for the visceral hypersensitivity reported in patients with DD.40 It can not be excluded that inflammation has also triggered those structural changes in the ENS documented in the present study. Although intraganglionic inflammatory infiltrates could not be observed in specimens with DD, it is conceivable that a degenerative loss of nerve cells may have taken place after an acute episode of inflammation has been settled.

Implications for pathogenetic concepts of DD

Although disturbed intestinal motor functions are consistently associated with DD favouring the development of diverticula, their emergence has been largely neglected. The evidence of morphometrically confirmed alterations of the ENS may partly close this gap. Taken together the present findings and previous data on both functional and structural alterations of the ENS, a paradigm change may be postulated for the pathogenesis of DD: mucosal herniations represent an epi-phenomenon evolved from an underlying intestinal motility disorder due to distinct abnormalities of the ENS.

This revised pathogenetical concept may be conceivable. However, the hitherto existing data are not yet sufficient to determine whether the enteric neuropathy is a primary event or secondary to an impairment of other tissues of the colonic wall (e.g. muscular and connective tissue pathologies), to mechanical stress (e.g. increased intraluminal pressure)11 or to inflammatory processes as outlined above. To better encircle the role of an enteric neuropathy possibly underlying the formation of diverticula further longitudinal studies are required in which alterations of the ENS are compared between patients with clinically apparent DD and patients with asymptomatic diverticulosis characterized by the presence of non-inflamed ‘cold’ diverticula. Thus, in spite of the increasingly acknowledged impact of enteric neuropathies on the generation of intestinal motor dysfunctions and painful sensations, classical pathogenetical concepts of DD (e.g. low fibre intake, connective tissue weakness, ageing) still remain valid suggesting that DD is caused by the coincidence of more than one single aetiological factor.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

We are indebted to the staff members at the Department of Surgery (University Hospital of Schleswig-Holstein, Campus Luebeck) for their collaboration in collecting the specimens and to Kathy Budler, Gudrun Knebel and Uschi Almert (Department of Anatomy, University of Luebeck) for skilful technical assistance.

Grant support

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References

This work was supported by research grants from the German Research Society (Deutsche Forschungsgemeinschaft, DFG WE 2366/4-1) and the Faculty of Medicine, University of Kiel.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosures
  9. Grant support
  10. References