SEARCH

SEARCH BY CITATION

Keywords:

  • cholinergic hyperfunction;
  • excitatory-inhibitory imbalance;
  • histologic morphology;
  • ineffective oesophageal motility;
  • myopathy;
  • nutcracker oesophagus

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

Abstract  The pathogenesis of nutcracker oesophagus (NE) and ineffective oesophageal motility (IEM) is unclear. Damage to the enteric nervous system or smooth muscle can cause oesophageal dysmotility. We tested the hypothesis that NE and IEM are associated with abnormal muscular or neural constituents of the oesophageal wall. Oesophageal manometry was performed in patients prior to total gastrectomy for gastric cancer. The oesophageal manometries were categorized as normal (n = 7), NE (n = 13), or IEM (n = 5). Histologic examination of oesophageal tissue obtained during surgery was performed after haematoxylin and eosin (H&E) and trichrome staining. Oesophageal innervation was examined after immunostaining for protein gene product-9.5 (PGP-9.5), choline acetyltransferase (ChAT) and neuronal nitric oxide synthase (nNOS). There were no significant differences in inner circular smooth muscle thickness or degree of fibrosis among the three groups. Severe muscle fibre loss was found in four of five patients with IEM. The density of PGP-9.5-reactive neural structures was not different among the three groups. The density of ChAT immunostaining in the myenteric plexus (MP) was significantly greater in patients with NE (P < 0.05) and the density of nNOS immunostaining in the circular muscle (CM) was significantly greater in IEM patients (P < 0.05). The ChAT/nNOS ratio in both MP and CM was significantly greater in NE patients. NE may result from an imbalance between the excitatory and inhibitory innervation of the oesophagus, because more than normal numbers of ChAT-positive myenteric neurones are seen in NE. Myopathy and/or increased number of nNOS neurones may contribute to the hypocontractile motor activity of IEM.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

Peristalsis in the smooth muscle oesophagus is coordinated centrally at the level of the dorsimotor nucleus of the vagus (DMV) and peripherally at the level of the myenteric plexus. Swallow-induced peristalsis in the smooth muscle oesophagus is initiated by the DMV, but its timing and force of peristaltic contraction are determined by myenteric neuromuscular interactions.1 Acetylcholine functions as an important excitatory neurotransmitter, and nitric oxide (NO) as the predominant inhibitory neurotransmitter that controls the latency gradient.2–4 The importance of peripheral neuromuscular mechanisms in the control of oesophageal motor function is illustrated by achalasia. Achalasia is a myenteric neuropathy characterized pathologically by a loss of inhibitory (NO synthase-containing) myenteric neurones.5 This manifests functionally as failure of both peristalsis in the smooth muscle oesophagus and lower oesophageal sphincter (LES) relaxation.

Smooth muscle hypertrophy, as assessed by high-frequency ultrasound, is reported in primary oesophageal motility disorders such as achalasia, diffuse oesophageal spasm and nutcracker oesophagus (NE).6 Whether muscle hypertrophy is a pathogenic mechanism causing these disorders or simply an epiphenomenon is not known. As the pathogenesis of most oesophageal motility disorders remains unclear, our aim was to determine if damage to the oesophageal neuromuscular apparatus might contribute to the pathogenesis of NE as representative of a hypercontractile oesophageal motility disorder, and ineffective oesophageal motility (IEM) as a typical hypocontractile disorder. For this, we observed histological morphology, thickness and degree of fibrosis of the circular muscle coat. We also quantitatively assessed the neuronal distributions. The densities of PGP-9.5 as pan-neuronal markers, choline acetyltransferase (ChAT) and neuronal nitric oxide synthase (nNOS) were measured.

Patients and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

Patients

We recruited patients who were diagnosed with gastric cancer of the mid- to upper body of the stomach and needed total gastrectomy. We excluded patients who had gastric cancer on fundus or cardia supposed to invade oesophagus. Oesophagogastroduodenoscopy identified clearly no oesophageal lesions or gastric cancer invasion up to the gastro-oesophageal junction. We performed conventional oesophageal manometry in patients who agreed to participate in our study. We first categorized the three groups as normal, NE and IEM, and assigned patients to each group. Twenty-five patients (male : female = 16 : 9; mean age, 56.8 ± 10.2 years) were included in this study. Oesophageal manometry identified seven patients with normal oesophageal motor function as a control, 13 with NE and five with IEM. This count is not representative of primary oesophageal motor disorder among general gastric cancer patients but the result of arbitrarily selected patients compatible with diagnostic criteria of each group. We took distal oesophageal tissue 3 cm above the gastro-oesophageal junction after total gastrectomy. Written informed consent was obtained from all patients before surgery. The Institutional Review Board of Yonsei University Medical Center approved this study protocol.

Oesophageal manometry

An eight-lumen polyvinyl manometry catheter (diameter, 4.5 mm; internal diameter of each lumen, 0.8 mm; EMC8-R, Synetics Medical Liberty System, Stockholm, Sweden) was used. A low-compliance pneumo-hydraulic capillary infusion system (Arndorfer Medical Specialities, Inc., Greendale, WI, USA) was used to continuously perfuse each catheter. Each lumen was connected to a pressure transducer, and pressure profiles were recorded (PC polygraph HR, Synetics Medical Liberty System) and analysed by a Polygram Upper GI Edition (Version 5.06C2; Gastrosoft Inc., Stockholm, Sweden). We measured the manometric parameters using a station pull-through technique. We characterized NE and IEM by criteria established by Spechler and Castell.7 NE was defined by a mean distal oesophageal peristaltic pressure wave of amplitude >180 mmHg with normal progression and normal or prolonged duration.7 IEM was defined as hypocontraction in the distal oesophagus with at least 30% of wet swallows exhibiting any combination of the following abnormalities: distal oesophageal peristaltic wave amplitude <30 mmHg, simultaneous contractions with amplitudes <30 mmHg, failed peristalsis in which the peristaltic wave does not traverse the entire length of the distal oesophagus, or absence of peristalsis.7

Histological preparation

Histology  Full-thickness oesophageal tissue specimens were taken 3 cm above the gastro-oesophageal junction. The segment measured 10 mm × 5 mm. The tissue was fixed on 4% formaldehyde and embedded in paraffin wax. Sections were cut at 4 μm thickness. The tissue was stained with haematoxylin and eosin (H&E) and trichrome. A single pathologist blinded to processes reviewed the tissue microscopically.

Immunohistochemistry  Whole-mount preparations were stained immunohistochemically for PGP-9.5 (pan-neuronal marker), ChAT (marker for cholinergic neurones) and nNOS (marker for nNOS neurones). Immunohistochemical staining was performed using a labelled streptavidin–biotin immunoperoxidase technique (LSAB2 kit; DAKO, Glostrup, Denmark). Sections for the same blocks used for routine histology were deparaffinized in xylene and immersed in 3% hydrogen peroxide for 10 min to block endogenous peroxidase activity. The tissue was incubated with the following primary antibodies: rabbit polyclonal antibody against PGP-9.5 (DAKO) (1 : 50; room temperature, 2 h), goat polyclonal antibody against ChAT (Chemicon, Temecula, CA, USA) (1 : 50; 4 °C, overnight), and rabbit polyclonal antibody against nNOS (Upstate, Lake Placid, NY, USA) (1 : 200; room temperature, 2 h).

The sections were washed with phosphate-buffered saline (PBS) and incubated with secondary antibodies with LSAB2 kit for PGP-9.5 and nNOS (DAKO) (room temperature, 30 min), biotinylated anti-goat IgG for ChAT (Vector, Burlingame, CA, USA) (1 : 100; room temperature, 2 h). The sections were washed with PBS and incubated for 30 min with streptavidin (1 : 200; room temperature, 30 min). After washing with PBS, the sections were processed with AEC chromogen solution for 10 min and counterstained with Mayer’s hematoxylin for 30 s.

Histologic analysis  A pathologist reviewed the tissue and described inflammation, vacuolization and other abnormal findings. Infiltration of inflammatory cells such as lymphocyte, plasma, polymorphonuclear, eosinophil, mast or macrophage was evaluated in mucosa, muscle layer and the myenteric plexus. The thickness of the inner circular muscle layer was measured and muscle thickness ratio was calculated as muscle layer thickness/(mucosa + submucosa + muscle layer thickness) × 100. The degree of fibrosis was determined semi-quantitatively (Fig. 1) as: no fibrosis – normal-looking muscularis propria; + fibrosis – interstitial fibrosis surrounding muscle fibres and large vessels replacing <10% of the muscularis propria; ++ fibrosis – fibrous tissue separating muscle fibres and replacing them by >10% but <50% replacement of the muscularis propria; and +++ fibrosis – fibrous replacement of >50% of the muscularis propria.

image

Figure 1.  Degree of endomysial fibrosis. (A) No fibrosis, normal-looking muscularis propria. (B) +, interstitial fibrosis surrounding muscle fibres and large vessels <10%. (C) ++, fibrous tissue separate muscle fibres and replace them 10% ≪ 50%, (D) +++, fibrosis replaced muscularis propria >50%. No significant differences were found among the three groups.

Download figure to PowerPoint

We then measured the stained densities of Masson’s trichrome, PGP-9.5, ChAT and nNOS under a microscope (ECLIPSE 80i; Nikon Corp., Kanagawa, Japan) with an image analysis program (i-Solution; IMT., Daejeon, Korea). The stained areas were measured under ×400 magnification. It represented as % of unit area (400 μm × 400 μm) : (stained area/160 000 μm2) × 100 for the myenteric plexus. In circular muscle, it is represented as 100% area : (stained area/160 000 μm2) × 10 000. Morphometric analyses were performed five times in each case. We also defined excitatory/inhibitory (EI) ratio, that is ChAT immunoreactive area divided by the nNOS immunoreactive area (Fig. 2).

image

Figure 2.  Immunohistochemical stains for PGP-9.5 (panel A,B), nNOS (panel C,D), and ChAT (panel E,F). These are under ×400 magnification. The panels on the left are in the circular muscle and panels on the right are in the myenteric plexus.

Download figure to PowerPoint

Statistical analysis

Two-tailed Kruskall–Wallis tests were carried out for statistical comparison among the three groups for sections from the myenteric plexus and circular muscle. Data are presented as median and range. A P-value of <0.05 was considered to be statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

Histology

Haematoxylin and eosin  Vacuolization and a ground-glass appearance of muscle fibres were observed in all groups. In NE patients, lymphocyte, plasma cell and eosinophil infiltration was seen around the myenteric plexus (3/13). The thickness of the inner circular muscle layer and muscle thickness ratio were not significantly different among normal, NE and IEM patients (Table 1). In patients with amplitude contraction greater than 220 mmHg (n = 2), muscle thickness ratio was not definitely increased either. Varying degrees of fibrosis were also seen in all groups but severe myolysis (3/5), wide intercellular space (3/5) and complete replacement of the muscularis propria with fibrous tissue after myolysis (1/5) were only seen in IEM patients.

Table 1.   The thickness of the circular muscle layer (mm) and muscle thickness ratio (%) were not found to be significantly different among the three groups
 ControlNEIEMP
  1. Muscle thickness ratio* was calculated as muscle layer thickness/(mucosa + submucosa + muscle layer thickness) × 100.

  2. NE, nutcracker oesophagus; IEM, ineffective oesophageal motility.

Circular muscle layer (mm)1.2 (1.0–2.0)1.5 (0.2–3.0)2.0 (1.7–5.0)ns
Muscle thickness ratio* (%)78.9 (62.9–87.5)64.7 (55.5–75.0)71.9 (63.6–76.0)ns

Masson’s trichrome  There were no significant differences in the subjective measure of fibrosis. The quantitative measure of % trichrome-stained area per unit area was also not different statistically among normal (5.2, 2.7–7.0), NE (5.2, 1.6–9.1), and IEM (6.8, 2.6–9.5).

Immunohistochemistry Table 2 shows the density of the PGP-9.5-, nNOS- and ChAT-positive areas. PGP-9.5-immunoreactive neuronal structures seem to be greater both in the myenteric plexus and circular muscle in NE, although this did not reach statistical significance. The nNOS expression of nerve fibres was greater in circular muscle in IEM compared with other groups but not in the myenteric plexus. Conversely, ChAT-immunoreactive nerve fibres were more prominent in the myenteric plexus in NE patients than in the other groups. EI ratio was significantly greater in both the myenteric plexus and circular muscle in NE patients (Table 2).

Table 2.   Immunohistochemistry
 PGP-9.5nNOSChATEI ratio
  1. *P < 0.05.

  2. We measured the stained densities of PGP-9.5, ChAT and nNOS under a microscope. The stained areas were measured under ×400 magnification. It is represented as % of unit area (400 μm × 400 μm) : (stained area/160 000 μm2) × 100 for the myenteric plexus. In the circular muscle, it represented as 100% area : (stained area/160 000 μm2) × 10 000. We also defined the excitatory/inhibitory (EI) ratio – that ChAT-immunoreactive area divides nNOS-immunoreactive area. ChAT-stained density in MP in NE and nNOS-stained density in CM in IEM were significantly increased. EI ratio was increased both in the myenteric plexus and circular muscle in NE.

  3. NE, nutcracker oesophagus; IEM, ineffective oesophageal motility; MP, myenteric plexus; CM, circular muscle; PGP-9.5, protein gene product-9.5; nNOS, neuronal nitric oxide synthase; ChAT, cholineacetyltransferase; EI ratio, excitatory/inhibitory ratio.

Control (n = 7)
 MP18.68 (13.78–20.53)7.99 (3.85–14.09)3.47 (0.64–6.56)0.38 (0.04–0.78)
 CM0.39 (0.22–0.95)0.18 (0.07–0.51)0.15 (0.08–0.48)0.85 (0.34–1.56)
NE (n = 13)
 MP20.02 (9.32–30.74)5.39 (1.04–17.04)7.64 (6.07–13.85)*5.81 (0.89–16.73)*
 CM0.86 (0.19–2.23)0.12 (0.02–0.23)0.33 (0.16–0.69)3.74 (0.71–5.03)*
IEM (n = 5)
 MP12.53 (11.87–23.27)10.95 (6.84–21.52)2.62 (2.50–5.77)0.30 (0.19–0.37)
 CM0.48 (0.41–0.62)0.61 (0.33–0.83)*0.17 (0.15–0.27)0.33 (0.20–0.39)

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

This is the first case–control study demonstrating histopathological abnormalities of oesophageal neuromuscular structures associated with primary oesophageal motility disorders other than achalasia. In this study, NE patients were more included than IEM patients because the general prevalence of NE in patients examined in motility laboratories in Korea is much higher than prevalence of IEM, when compared with rates in western countries.8 As best we could determine, the morphology and distribution of neural elements in the myenteric plexus and circular muscle layer were normal in NE patients; however, ChAT-immunoreactivity was greater in the myenteric plexus. This suggests the possibility of a greater cholinergic drive in NE patients. Oesophageal smooth muscle of patients with IEM frequently exhibited fibrosis, myolysis and widened intercellular spaces, suggesting the possibility of a myopathic process. In addition, we found that nNOS-immunoreactivity was greater in the circular muscle layer, suggesting the possibility of a greater inhibitory drive in IEM. We also described histological changes of the oesophageal tissues of patients with hypercontractile (NE) and hypocontractile (IEM) oesophageal disorders when compared with normal subjects. There was no thickening of the muscularis propria in NE patients, and it was not histologically different from normal.

Swallow-induced peristalsis in the smooth muscle oesophagus is a complex neuromuscular event. It initiates from the DMV but its timing and force of peristaltic contraction are determined by myenteric neuromuscular interactions.1 The major neurotransmitters that control the timing and amplitude of the peristaltic sequence are Ach, an excitatory neurotransmitter, and NO, an inhibitory neurotransmitter. It is likely that their integrated effects produce normal peristalsis. Many researchers of oesophageal motor function feel that most primary motor disorders of the oesophagus are caused by defects in either its inhibitory or excitatory innervation.1

Both the NO and Ach innervations affect the timing and amplitude of the peristalsis but for the most part, NO controls the timing of peristalsis in the distal oesophagus.9,10 There is a direct correlation between the degree of inhibition and the velocity of peristalsis: less inhibition is associated with more rapid peristalsis.11 An extreme example of this situation is achalasia, which is characterized functionally by simultaneous pressure waves in the smooth muscle oesophagus with failure of LES relaxation and histopathologically by an absence of NO neurones in the myenteric plexus.12 Overactivation of the NO-cGMP signalling system also alters oesophageal motor function. In humans, type-5 phosphodiesterase inhibitors, which increase cellular concentrations of cGMP, dramatically decrease the amplitude of peristalsis in the smooth muscle oesophagus without altering its propagation velocity.13 The amplitude of peristalsis decreases to the range that defines IEM. From these observations, we may be able to predict that overproduction of NO may diminish the amplitude of oesophageal peristalsis. In fact, more nNOS immunoreactivity was seen in the circular muscle layer of patients with IEM, thus excess NO production is one plausible mechanism for IEM.

Oesophageal tissues of patients with IEM also revealed histopathological changes of myopathy with a morphologically normal-appearing myenteric plexus. This suggests that the myopathic process may contribute to IEM. The manometric pattern of IEM is frequently observed in diseases such as scleroderma that affect oesophageal smooth muscle. Scleroderma is characterized histopathologically, as were our patients with IEM, by atrophy and fibrosis with a normal-appearing myenteric plexus.14

Inhibiting the cholinergic innervation of the oesophagus also alters swallow-induced peristalsis. In humans, cholinergic blockade either disrupts the peristaltic sequence or slows its onset in the smooth muscle oesophagus while altering its velocity little.15 It also markedly diminishes the force of peristaltic contractions, suggesting that cholinergic drive is an important determinant of the amplitude of peristalsis. Augmenting cholinergic tone increases the amplitude of swallow-induced oesophageal contractions.16 From these observations, we might be able to predict that increasing cholinergic tone may augment the force of peristaltic contractions. More ChAT-immunoreactivity was seen in patients with NE, and the EI ratio was commonly increased both in the myenteric plexus and circular muscle. Thus, either greater Ach production or an increase in the EI ratio is a plausible mechanism for NE.

Inflammatory cells were found infiltrating the myenteric plexus in a few patients with NE. Mediators of inflammation, particularly those agents released from degranulating eosinophils, can cause profound changes in the function of smooth muscle17 and/or enteric nerves.18 These observations suggest that in some cases, NE may result from inflammation.

Smooth muscle hypertrophy and hyperplasia are associated with oesophageal motility disorders or obstruction.6,19,20 Muscle thickness and cross-sectional area measured by high-frequency ultrasonography indicate that muscle mass varies as a function of the specific motor disorder: achalasia > diffuse oesophageal spasm (DES) > NE/high-amplitude oesophageal contraction (HAEC) > normal subjects. Although a thickened muscle layer is proposed as a distinctive finding associated with motility disorders characterized by high-amplitude oesophageal contraction, we did not observe wall thickening associated with NE. There was overlap in muscle thickness between the control and NE. In other studies, patients with contraction amplitudes greater than 250 mmHg showed definitely greater muscle thickness than normal subjects but they found that patients with contraction amplitudes close to 200 mmHg also showed overlap with control.21 Only one of the patients with NE in this study showed a contraction amplitude of over 230 mmHg. The relatively low-amplitude contraction could not obstruct bolus transit.

We assume that this is because true NE is mechanically quite different from spastic oesophageal motor disorders of the oesophagus, such as achalasia, DES or obstruction, each of which is characterized by some degree of outflow obstruction secondary to primary LES dysfunction or simultaneous body contraction. In their ultrasonographic study of the oesophageal wall, Meltzer et al. reported thickening of the oesophageal wall in only 30% of patients with NE.22 In a similar study exploring oesophageal wall thickness in patients with high-amplitude oesophageal contractions, Puckett et al. did find thickening of the oesophageal musculature.23 These patients, however, were not those with NE only. Many may have met the criteria for DES, almost half had a hypertensive LES, and more than half had incomplete LES relaxation. The authors conceded that the thickening of the muscular wall in patients might result from partial oesophageal outlet obstruction. Experimental partial oesophageal obstruction causes muscular hypertrophy, changes in the morphology of myenteric neurones and changes in muscle excitability.20,24,25 Manometric changes included high-amplitude, simultaneous, repetitive and multipeaked pressure waves, which are the many characteristics of spastic oesophageal motor disorders.26

The other possible reason that we did not see the muscular thickening noted on ultrasonography is purely technical: we made our measurements ex vivo on fixed tissues while ultrasound is a dynamic study performed in vivo. The resected specimen shrinks in ex vivo studies and can artificially increase in muscle thickness, not reflected in vivo size. So we compared the measurements relatively among the three groups, including normal subjects.

We should mention the possibility of oesophageal motility alterations secondary to a paraneoplastic syndrome of gastric cancer. Paraneoplastic syndrome can result from distant malignancy.27 Severe gastrointestinal motor dysfunction results from neurone and axon degeneration; lymphoplasmacytic infiltration within the myenteric plexus in seven patients with lung cancer had been reported.28 Neurologic forms of paraneoplastic syndrome are frequent in small cell lung cancer,28 but the association of paraneoplastic neurologic degeneration, especially visceral neuropathy and gastric cancer is rare. Several case reports suggested the possibility of association of autoantibodies and gastric cancers, such as anti-Yo antibody and anti-Ri antibody.29,30 Furthermore, the correlation between the pathology of hypercontractile and hypocontractile oesophageal motility disorders and gastric cancers associated with numerous plausible autoantibodies should be investigated.

We conclude that our studies support the hypothesis that NE is caused by cholinergic hyperfunction in the myenteric plexus and excitatory–inhibitory imbalance both in the myenteric plexus and the circular muscle layer. They also suggest that either a myopathic process or an increase in inhibitory tone contributes to IEM. Observations like these may help guide future therapies for these disorders.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References

This study was made possible by a grant from the Brain Korea 21 Project. We thank Dr Roberto De Giorgio for critiquing the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Conflict of interests
  9. References