If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Background & Aims: Irritable bowel syndrome (IBS) is regarded as a functional bowel disorder. Few studies have looked for histopathologic changes in the gut and only then in biopsy specimens from intestinal mucosa. Because bowel function is governed mainly by nerve plexuses in the bowel wall, we have investigated full-thickness bowel biopsy specimens in patients with severe IBS. Methods: We used a laparoscopy-assisted technique to obtain full-thickness biopsy specimens from the proximal jejunum. Tissue specimens were investigated with light microscopy using routine stainings and immunohistochemical techniques. Horizontal sectioning was done to visualize large areas of the myenteric plexus. Fifteen autopsy specimens were used as controls regarding the myenteric plexus. Colorectal adenoma controls with terminal ileum biopsy specimens and full-thickness jejunal biopsy specimens from patients with degenerative enteric neuropathy were used as control groups for intraepithelial lymphocyte counts. Results: Ten patients (2 males, 8 females) were studied. In 9 patients, we found low-grade infiltration of lymphocytes in the myenteric plexus. Lymphocytes had peri- and intraganglionic location. The mean number of lymphocytes per ganglion ranged from 1.9 to 7.1 per patient, with an overall mean of 3.4. No intraganglionic lymphocytes were found in the control group and only a few periganglionic lymphocytes (mean, 0.2). Four patients had concomitant intraepithelial lymphocytosis. Neuron degeneration was evident in 6 of 9 patients with and 1 patient without ganglionic lymphocyte infiltration. Conclusions: Our findings indicate that inflammation and neuronal degeneration in the myenteric plexus are involved in the pathogenesis of IBS.
Irritable bowel syndrome (IBS) is currently viewed as a functional bowel disorder. This means that no identifiable disease process has been found to explain the patient's perception of ill health. Symptoms are believed to arise from disturbances of function rather than structure. There is no known etiology and no pathophysiologic mechanism unique to IBS that has been identified. Today, research needs to be performed on groups of patients with clusters of symptoms as the definition, the Rome-II criteria.
The use of symptom clusters to define a patient group leads to a heterogeneity within the group that may be one of the problems in our attempts to find pathogenetic mechanisms leading to IBS.
There are several observations that favor the notion that gastrointestinal function is disturbed in IBS. These include disturbances of anal pressure activity,
The reasons for these disturbances are unclear, but there is an increasing interest in inflammatory processes as a generator of IBS in at least a subset of patients (for review, see Collins
Prevalence of gastrointestinal symptoms six months after bacterial gastroenteritis and risk factors for development of the irritable bowel syndrome: postal survey of patients.
Increased rectal mucosal enteroendocrine cells, T lymphocytes, and increased gut permeability following acute Campylobacter enteritis and in post-dysenteric irritable bowel syndrome.
If these observations signify an underlying pathology in patients with IBS, the first place to look for an abnormality would be the enteric nervous system because this governs most of the functions studied. Very little has been done to find out whether this is a correct assumption or not.
With this as a background, we designed a project with the aim of finding out whether patients with severe IBS symptoms may have neuromuscular or inflammatory abnormalities in the small bowel. The group of patients recruited can be regarded as “the tip of the iceberg” in the IBS population, and the work-up leading to the diagnosis in these cases should not be misconstrued as a general strategy for diagnosing IBS.
Materials and methods
Patient selection
Consecutive patients who met the inclusion criteria and agreed to participate were included in the study between 1998 and 1999. All patients were seen at the Department of Gastroenterology and Hepatology, Huddinge University Hospital, Stockholm, Sweden. The inclusion criteria were the following: (1) fulfillment of the Rome-I Criteria for the Diagnosis of IBS, (2) need for referral to a specialist in gastroenterology, (3) sick listing from work because of IBS symptoms for more than 1 month during the last year, (4) normal or nonspecific findings in stationary or ambulatory small-bowel manometry, (5) age 18–60 years, and (6) no contraindications for anesthesia or laparoscopic surgery.
Criteria 2 and 3 were used in an attempt to define “severe IBS.” Small-bowel manometry was used to rule out severe motility disorders, primarily chronic intestinal pseudo-obstruction (CIPO). In 9 patients, stationary antro-duodeno-jejunal manometry was used and 1 had an ambulatory small-bowel manometry performed. We used the manometric abnormalities defined by Stanghellini et al.
All patients were thoroughly examined for other somatic diseases before entering the study. This included a minimum investigation as follows: laboratory tests (erythrocyte sedimentation rate, hemoglobin level, leukocyte count, electrolytes, liver function tests, pancreatic amylase, and thyroid function tests), gastroscopy with biopsy specimen from duodenum, colonoscopy or rigid or flexible sigmoidoscopy with biopsy specimens if diarrhea was a predominant symptom, x-ray investigations of colon and the small bowel, abdominal ultrasound or computed tomography, lactose tolerance test, and fecal culture including ova and parasites. If there was a clinical suspicion, additional tests regarding neuroendocrine tumors, porphyria, or spontaneous intestinal bacterial overgrowth were performed. A screening regarding relevant autoimmune disorders was done with the use of autoantibodies.
All patients were given oral and written information before acceptance into the study. They were all aware of the fact that the operative procedure most probably would not change their treatment options. The study was approved by the Ethics Committee at Huddinge University Hospital, Stockholm, Sweden.
Biopsy technique
A full-thickness biopsy specimen was taken from the proximal jejunum using a laparoscopy-assisted procedure. This technique is performed as follows: Two 2- to 4-cm incisions are made. One is situated below the umbilicus and 1 in the left fossa. A bowel loop of the proximal jejunum is identified laparoscopically and is then exteriorized through the incision below the umbilicus. Without interruption of the blood supply, the bowel loop is held with a surgical forceps while taking the full-thickness biopsy specimen with a surgical knife. A minimum size of 5 × 3-mm muscle area is required. Most often, the tissue specimen is at least 10 × 10 mm. The bowel loop is then closed with 2 layers of absorbable suture and repositioned into the abdomen. Visual inspection of the sutured bowel loop is performed laparoscopically before closing the incision in the abdominal wall. The patient is then observed for complications for 24 hours and allowed to eat and drink before leaving the hospital.
Biopsy specimen preparation
The full-thickness intestinal biopsy specimens were immediately brought to the Department of Pathology. There, they were divided into 2 parts, oriented on a cork plate with the mucosa downwards, and fixed in 4% buffered formalin for no longer than 72 hours. Following fixation, the specimens were embedded in paraffin: 1 for transversal and 1 for horizontal sectioning. Evaluation of all layers of the intestinal wall was possible from the 4-μm transverse sections, which were made at 3 different levels of the tissue block, containing 3 serial sections each. In the paraffin blocks used for horizontal sectioning, the approximate distance between the peritoneal surface and the intermuscular layer, as well as the thickness of this layer was determined on transversal sections. Afterwards, the whole intermuscular layer containing the myenteric plexus was cut in serial sections. The horizontal embedding and serial sectioning allowed us to visualize as much of the myenteric plexus as possible.
Biopsy specimen staining
The paraffin sections were stained with H&E for general histologic examination and morphometry, with Masson's trichrome for connective tissue, with Giemsa for mast cells, and with cresyl violet for neurons. Immunohistology was performed with the following antibodies: CD3 for T lymphocytes, neuron-specific enolase (NSE), neurofilament (NF), synaptophysin and bcl-2 for neurons, S100 for Schwann cells, C-kit (CD117) for interstitial cells of Cajal and general actin, and α-actin and desmin for the intermediate contractile filaments of smooth muscle cells.
Morphometry
The number of intraepithelial lymphocytes (IEL) was determined on CD3-stained sections by counting at least 300 epithelial cells (EC) of the small intestinal villi. Areas in the vicinity of lymphoid aggregates within the lamina propria were not counted. The IEL count was expressed as the number of IELs per 100 EC. The thickness of the circular and longitudinal muscle layers was determined by using an ocular micrometer. The periganglional/intraganglional lymphocyte count in the myenteric plexus was determined by choosing those ganglia that contained the most lymphocytes. At least 20 ganglia were counted.
Controls
The number of intraepithelial lymphocytes was determined on paraffin sections of ileum biopsy specimens stained with CD3 from 20 subjects who had undergone colonoscopy for adenoma control. IEL counts were also determined in full-thickness biopsy specimens of the jejunum from 10 patients who had visceral degenerative neuropathy. The upper normal limit for the intraepithelial lymphocyte count was defined as the mean + 3 SD.
For the presence of lymphocytes within the myenteric plexus, tissue slices of jejunum, ileum, and colon from 15 autopsies of deceased patients with no known bowel disorder were embedded and sectioned according to the same protocol as described previously for the patient group. The sections were stained with H&E and CD3. The same pathologist made all biopsy assessments nonblinded but unaware of the IBS diagnosis. The specimens were part of his routine assessments from patients with gastrointestinal dysmotility.
Results
Eight women and 2 men with severe IBS were studied. The average age was 40 years (range, 23–57 years). All patients underwent laparoscopy-assisted full-thickness biopsy of the jejunum from which the patient-related results below were obtained. The postoperative period was uneventful, and all patients were discharged on the first postoperative day. Brief patient characteristics were as follows:
Female 45 years of age, constipation-predominant IBS. Dyspeptic symptoms during many years. Developed prominent IBS symptoms in 1997, at age 43, without a distinct onset. Uses bulking agents. Cisapride, anticholinergics, paracetamol, dextropropoxyphene, and proton pump inhibitors without effect. Operated on twice because of in situ breast cancer, last time when she was 38 (1992). No chemo- or radiotherapy. Chronic low back pain.
Female 43 years of age, IBS with alternating constipation and diarrhea. Suddenly developed IBS symptoms at the age of 35 years (1981) following penicillin V treatment for respiratory tract infection. Uses bulking agents. Has had ketobemidon in periods for her abdominal pain. Chronic neck pain because of osteoarthritis.
Female 40 years of age, diarrhea-predominant IBS. IBS since she was a teenager without a distinct onset. Lactose intolerance. No change in IBS symptoms with a lactose-free diet. Uses bulking agents.
Female 47 years of age, IBS with alternating constipation and diarrhea. Sigmoid resection laparoscopically in 1996 because of repeated episodes of diverticulitis. After that, a sudden onset of IBS symptoms without any remaining signs of diverticular disease. Uses bulking agents.
Male 23 years of age, IBS with alternating constipation and diarrhea. Symptoms since he was 12 years without a distinct onset. Extensive investigations both in Sweden and United States because of bowel symptoms. No abnormalities found. Uses anticholinergics and papaverine on demand for his abdominal pain.
Female 24 years of age, diarrhea-predominant IBS. Gastroenteritis during a vacation in Turkey when she was 22 (1997). No pathogen found in fecal cultures. Developed severe IBS after that, problems to a lesser degree before the infectious episode. A huge number of medications has been used without effect: bulking agents, loperamide, papaverine, anticholinergics, tricyclic antidepressants, analgesics to the potency of opioids, a short trial of cortisone, octreotide, cholestyramine, pancreatic enzyme supplementation, and proton pump inhibitors. Has had a dramatic decrease in symptoms since the summer of 1999. No need for medications at the time of inclusion.
Male 55 years of age, constipation-predominant IBS. Dyspeptic symptoms since he was a teenager. Slowly developed more prominent IBS symptoms over the years. Became worse after laparoscopic fundoplication for gastroesofageal reflux disease 1995. H. pylori infected without ulcer disease. Eradication treatments have failed many times. Uses anticholinergics and tramadol on demand.
Female 47 years of age, diarrhea-predominant IBS. Developed IBS after an acute gastroenteritis when she was 42 years (1994). No fecal cultures were done. Before that, a long history of dyspeptic symptoms. Uses cholestyramine and anticholinergics on demand. Hypertension treated with carvedilol.
Female 24 years of age, IBS with alternating constipation and diarrhea. IBS symptoms since she was a teenager without a distinct onset. Previously treated for endometriosis. Uses rectal enemas on demand. Proton pump inhibitors in periods because of reflux-like symptoms.
Female 50 years of age, IBS with alternating constipation and diarrhea. Symptoms started at age 46 (1995) as intermittent epigastric pain. Has gradually developed IBS. Uses NSAIDs and anticholinergics. Proton pump inhibitors because of gastroesofageal reflux disease. Also has a whiplash injury to her neck.
Intraepithelial lymphocytes
The mean (SD) IEL count was 14.2 (4.2) in the adenoma controls and 12.6 (3.0) in the control patients with visceral degenerative neuropathy. There was no significant difference between the values of the 2 groups. The mean (SD) of both groups was 13.9 (4.0). Thus, the upper normal limit was defined as 26 IEL/100 epithelial cells in CD3-stained sections (Figure 1).
Fig. 1Control ileum with normal number of intraepithelial lymphocytes (arrows). (CD3 immunoperoxidase; original magnification ×380.)
Four patients had intraepithelial lymphocytosis, i.e., more than 26 lymphocytes per 100 epithelial cells (Figure 2).
Fig. 2Proximal jejunum with increased number of intraepithelial lymphocytes (arrows) (41 IELs/100 epithelial cells). (CD3 immunoperoxidase; original magnification ×380.)
The enterocytes were well matured in all of them without any sign of degeneration, and there was no villous atrophy or crypt hyperplasia. No inflammation or only a slight increase of lymphocytes and plasma cells was seen in the lamina propria.
Intra- and periganglionic lymphocytes
Altogether, 397 ganglia of the myenteric plexus and their interconnecting nerve trunks were identified in the horizontally embedded sections of the intestines from the 15 autopsied control subjects. Sixty-seven CD3+ lymphocytes were found in the sections. All of the lymphocytes were located outside the ganglia and nerve fibers within the connective tissue between the 2 muscle layers (Figure 3).
Fig. 3Normal myenteric ganglion. The neurons show normal cytomorphology with moderate or slight eosinophilic and finely granular cytoplasm. Note that there is no vaculization or deeply stained, dark cytoplasm. The nuclei are round with well-visible nucleoli. No signs of nuclear shrinkage or homogeneous dark basophilia of the chromatin as in pyknotic (shrunken) degenerating nuclei. The small, round nuclei belong to glia cells (thin arrows), whereas the elongated pale nucleus is that of a Schwann cell (thick arrow). (H&E; original magnification ×380.)
The mean (SD) of periganglional/perineural lymphocytes was 0.17 (0.63)/ganglion.
In 9 of 10 patients, we found infiltration of lymphocytes in the myenteric plexus. The lymphocytes had a peri- and intraganglionic location (Figures 4 and 5).
Fig. 4Lymphocyte infiltrate within myenteric ganglion. Thick arrow shows a neuron and a few lymphocytes to the right of the neuron. There are more lymphocytes within the area marked by thin arrows. (H&E; original magnification ×380.)
Fig. 5There is a focus of T lymphocytes within the myenteric ganglion marked by thin arrows. Thick arrow: neuron. (CD3 immunoperoxidase; original magnification ×380.)
The average number of lymphocytes per ganglia and/or nerve ranged from 1.9 to 7.1. Six of these patients also showed signs of neuron degeneration, i.e., cytoplasmic swelling and vaculization, shrunken (pyknotic) or ruptured nuclei (karyorrhexis) (Figure 6).
Fig. 6Degenerative neuropathy. Myenteric ganglion with 2 shrunken degenerated neurons marked by thin arrows. The cytoplasm is vacuolated and the nuclei are pyknotic. One normal neuron is shown by a thick arrow. (H&E; original magnification ×380.)
There were no signs of inflammation in the submucous plexus, but the number of ganglia examined was considerably lower than that of the myenteric plexus. In the 1 remaining patient (No. 4), the findings were those of a degenerative neuropathy, which mainly affected the myenteric plexus, but there were also discrete signs of neuron degeneration in the submucous plexus.
Mast cells
No mast cells were seen within or in the vicinity of the myenteric plexus in any of the patients. The number of mast cells in the various layers of the bowel wall was judged to be normal by qualitative evaluation.
Interstitial cells of Cajal
Six patients were found to have abnormalities in the number or the size of the interstitial cells of Cajal (ICC). Four patients showed ICC hyperplasia combined with ICC hypertrophy within both the perimyenteric and the deep muscular ICC plexus. Two had focally decreased numbers of ICCs around the myenteric plexus.
Another feature that was seen in 9 of 10 patients was hypertrophy of the longitudinal muscle layer of the bowel wall (Figures 7 and 8).
Fig. 7Normal muscularis propria. The inner circular layer (C) is considerably thicker than the outer longitudinal muscle layer (L) (bar = 200 μm). (H&E)
Fig. 8The thickness of the 2 muscle layers are equal because of a thickened longitudinal muscle layer (L) secondary to hypertrophy of the myocytes (C, inner circular muscle layer). (Desmin immunoperoxidase; original magnification ×140.)
In one patient, this assessment was impossible because of a mechanical artefact. Histopathologic findings in the patients with IBS are summarized in Table 1.
Table 1Histopathologic findings in the IBS patients
Patient no.
L/G
EL
LMH
ICC
ND
IBS type
Duration (yr)
1 (F)
7.1
27
x
+
C
2
2 (F)
3.2
<26
x
−
x
A
18
3 (F)
3.2
41
x
+
x
D
~25
4 (F)
0.3
14
x
x
A
3
5 (M)
3.5
13
x
−
A
11
6 (F)
2.5
<26
x
+
x
D
2
7 (M)
1.9
23
x
x
C
~30
8 (F)
2.5
17
x
−
D
5
9 (F)
2.6
36
x
x
A
~10
10 (F)
4.3
37
l
+
x
A
4
M, male; F, female; L/G, number of lymphocytes per ganglion and/or nerve; EL, number of lymphocytes per 100 epithelial cells; LMH, longitudinal muscle hypertrophy; ICC, increase in number and/or size (+) or decrease in number and/or size (−) of the interstitial cells of Cajal; ND, neuron degeneration; C, constipation predominance; D, diarrhea predominance; A, alternating constipation and diarrhea; l, not assessed because of mechanical biopsy artefact.
The results of our study must be interpreted with caution. We found histopathologic abnormalities of the proximal jejunum in patients whose diagnosis was based on the absence of structural lesions and fulfilling the Rome I criteria for IBS. The patients were from different age groups and could be divided into all categories of IBS. There were coexisting diseases in the group but nothing that interferes with the IBS diagnosis and, most importantly, that can explain our findings. We observed 5 types of pathology: (1) All but 1 patient had an infiltration of lymphocytes in and around the nerves constituting the myenteric plexus. (2) Four patients had an increased number of intraepithelial lymphocytes. (3) Seven patients showed signs of neuron degeneration in the myenteric plexus, all but 1 in conjunction with lymphocyte infiltration. (4) Nine patients had longitudinal muscle hypertrophy. (5) Seven patients had abnormalities in ICC numbers and/or size.
Inappropriate section or staining techniques might explain why such pathologic features have never been reported. We found out that horizontal cuts are required to visualize large areas of the intermuscular layer containing multiple myenteric ganglia. Immunohistochemistry for lymphocytes is also a prerequisite. The need for a dedicated pathologist in this respect cannot be underestimated because the lymphocyte count in absolute numbers is comparatively low and both lymphocytosis and neuron degeneration are patchy. Hence, multiple ganglia and sections have to be investigated.
The problem of finding a control group for this study is of major concern. Our solution for intraepithelial lymphocytes was to use ileum biopsy specimens taken in connection with colonoscopy for adenoma control and full-thickness jejunum biopsy specimens from patients with degenerative neuropathy. We examined the myenteric plexus in deceased patients, who did not have any known bowel disorder, as controls for lymphocytes in the myenteric plexus. We found only a few lymphocytes (less than 1 lymphocyte per 5 ganglia) near the enteric nervous system and never inside the nerve ganglia in those controls. This finding supports our view that more than 1 lymphocyte per nerve ganglion is a true abnormality and not a normal phenomenon. The localization of lymphocytes also seems to be of importance, i.e., within the ganglia in patients, outside the ganglia in controls.
Only a few researchers have previously looked for inflammatory changes in IBS. Most often, they have used mucosal biopsy specimens, which cannot be compared with full-thickness intestinal biopsy specimens because only a minimal part of the bowel wall is examined. In one previous study,
histology has been performed on full-thickness surgical specimens containing all layers of the bowel wall. The authors looked for the presence and number of mast cells in colectomy specimens from 4 cases of spastic colitis with severe symptoms. Elevated numbers were seen in the muscularis in all of them. We found no increase of mast cells in the mucosa or in the bowel wall of the jejunum in our patients. The findings in the previous study could, therefore, not be confirmed by us, but we did not perform quantitative assessment of mast cells.
The finding that 4 of our patients also had an intraepithelial lymphocytosis is interesting, and similar data were recently reported by Wahnschaffe et al.
in patients with diarrhea-predominant IBS. Whether intraepithelial lymphocytosis is a marker for a luminal factor that triggers a low-grade inflammatory response in the deeper layers of the bowel or an “immunological memory” that persists after earlier antigenic provocations or the hallmark of a “hyperreactive gut” in the same way that respiratory tract infections can lead to hyperreactive airways remains to be elucidated. In our series of patients, this finding was not restricted to patients with diarrhea-predominant IBS.
The hypertrophic thickening of the longitudinal muscle layer in the bowel wall of all our patients except 1, in whom a mechanical artefact made assessment impossible, is intriguing. In a rat model in which the myenteric plexus had been chemically destroyed,
the denervation was associated with an increase in muscle thickness both in denervated and distant bowel segments. Acute inflammation induced by Trichinella spiralis infection has also been shown to induce thickening of the muscle layers in the gut in yet another rat model.
These phenomena may be involved in creating structural abnormalities in the human gut wall as well. The finding in our study that the patient without a lymphocyte infiltration still showed longitudinal muscle thickening may favor neuropathy as the pathogenetic mechanism.
Nothing conclusive can be said about the relationship between observed pathology and the generation of IBS symptoms. There is no animal model used in studying what a low-grade inflammation of this kind does to the structure and function of the neuromuscular tissues of the gut. The results of acute inflammation in rats and the clinical observations in humans makes it tempting to believe that inflammation plays a central role in at least a subset of IBS patients, especially in creating visceral hypersensitivity. The findings in our small series of patients points in the direction that inflammation or neural pathology may be a common denominator in the IBS group and that the clinical phenotype resulting from it may be of great diversity.
Our present study indicates that future studies of IBS should focus more on histopathology than on symptom clusters. The findings need to be challenged by other research groups. If the results can be confirmed, the nature and significance of the inflammatory process must be characterized to develop new therapeutic options for patients with IBS.
References
Thompson WG
Longstreth GF
Drossman DA
Heaton KW
Irvine EJ
Müller-Lissner SA.
Functional bowel disorders and functional abdominal pain.
Prevalence of gastrointestinal symptoms six months after bacterial gastroenteritis and risk factors for development of the irritable bowel syndrome: postal survey of patients.
Increased rectal mucosal enteroendocrine cells, T lymphocytes, and increased gut permeability following acute Campylobacter enteritis and in post-dysenteric irritable bowel syndrome.
☆Address requests for reprints to: Hans Törnblom, M.D., Karolinska Institutet Department of Medicine at Huddinge University Hospital, Division of Gastroenterology and Hepatology, K63 SE-141 86 Huddinge, Sweden. e-mail: [email protected] ; fax: (46) 8 585 823 35.
☆☆Supported by a research grant from the Swedish Society of Medicine (Bengt Ihres fond).