Randomized, double-blind, placebo-controlled trial of prednisolone in post-infectious irritable bowel syndrome
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Aliment Pharmacol Ther 2003; 18: 77–84. doi: 10.1046/j.0269-2813.2003.01640.x
Randomized, double-blind, placebo-controlled trial of prednisolone
in post-infectious irritable bowel syndrome
S. P. D UNLOP*, D. JEN KINS*, K. R. NEAL*, J. NAESDAL , M. BORGA ONKERà, S. M. COLLINSà &
R. C. SPILLER*
*Divisions of Gastroenterology, Pathology and Public Health, University Hospital, Nottingham, UK; AstraZeneca, Hassle,
Sweden; àMcMaster University, Hamilton, Canada
Accepted for publication 16 April 2003
Results: Initial enterochromaffin cell counts were
SUMMARY
increased and correlated with initial lamina propria
Background: Post-infectious irritable bowel syndrome is T-lymphocyte counts (r ¼ 0.460, P ¼ 0.014). Entero-
associated with increased serotonin-containing entero- chromaffin cell counts did not change significantly after
chromaffin cells and lymphocytes in rectal biopsies. either prednisolone () 0.8% ± 9.2%) or placebo
Animal studies have suggested that steroids reduce the (7.9% ± 7.9%) (P ¼ 0.5). Although lamina propria
lymphocyte response and suppress some of the post- T-lymphocyte counts decreased significantly after
infectious changes in neuromuscular function. prednisolone (22.0% ± 5.6%, P ¼ 0.003), but not after
Aim: To evaluate whether steroids reduce the number placebo (11.5% ± 8.6%, P ¼ 0.1), this was not associ-
of enterochromaffin cells and improve the symptoms of ated with any significant treatment-related improve-
post-infectious irritable bowel syndrome. ment in abdominal pain, diarrhoea, frequency or
Methods: Twenty-nine patients with post-infectious urgency.
irritable bowel syndrome underwent a randomized, Conclusions: Prednisolone does not appear to reduce the
double-blind, placebo-controlled trial of 3 weeks of oral number of enterochromaffin cells or cause an improve-
prednisolone, 30 mg/day. Mucosal enterochromaffin ment in symptoms in post-infectious irritable bowel
cells, T lymphocytes and mast cells were assessed in syndrome. Other approaches to this persistent condition
rectal biopsies before and after treatment, and bowel are indicated.
symptoms were recorded in a daily diary.
anxiety,1 diarrhoea predominance3, 8 and increased
INTRODUCTION
numbers of immunocytes in rectal biopsies,9 which
Post-infectious irritable bowel syndrome, the onset of suggest a different pathogenesis from other forms of
new irritable bowel syndrome symptoms after gastro- irritable bowel syndrome. Quantitative studies of the
enteritis, is well documented,1, 2 and, depending on the rectal mucosa have demonstrated increased numbers of
setting, occurs in 4–32% of individuals after infection.3–7 chronic inflammatory cells4 and entero-endocrine cells10
Post-infectious irritable bowel syndrome (PI-IBS) has a in post-infectious irritable bowel syndrome. Furthermore,
number of distinctive features, including lower levels of enterochromaffin cell and lymphocyte numbers are
closely correlated,10 suggesting a possible causal link.
Enterochromaffin cells, which contain 85% of the body’s
Correspondence to: Professor R. C. Spiller, C Floor, South Block, University
Hospital, Nottingham, NG2 6DN, UK. total serotonin stores, play a key role in transducing
E-mail: robin.spiller@nottingham.ac.uk luminal stimuli, such as nutrients and pressure. These
2003 Blackwell Publishing Ltd 7778 S. P. DUNLOP et al.
stimuli release serotonin, which acts on visceral afferents, rhoea, vomiting, fever or positive stool culture.
both extrinsic and intrinsic, mediating both pain percep- The study was approved by the University Hospital
tion and the peristaltic reflex, in addition to regulating Nottingham Ethics Committees. All patients gave writ-
intestinal secretions.11 5-Hydroxytryptamine-3 (5HT3) ten informed consent.
antagonists improve the symptoms of diarrhoea-predom-
inant irritable bowel syndrome,12 leading to the hypo-
Trial protocol
thesis that a reduction in the number of enterochromaffin
cells may benefit patients with post-infectious irritable The study was a randomized, double-blind, placebo-
bowel syndrome, whose symptoms are similar to those of controlled trial of prednisolone, 30 mg once daily, for
diarrhoea-predominant irritable bowel syndrome. 3 weeks. Prednisolone 30 mg or placebo was supplied
Animal models of post-infectious irritable bowel syn- within a single gelatine capsule. A designated phar-
drome suggest that T lymphocytes play an important macist generated random sequences in blocks of six,
role in mediating the long-term effects of infection on five times. The medication was allocated in numbered
neuromuscular function,13 and that steroids can pre- containers, which were dispensed by the pharmacist
vent the long-term dysfunction seen after Trichinella after patient enrolment by a clinician (S.P.D.).
spiralis infection.14 The aim of this study was to assess Patients were initially assessed at the hospital and
whether a short course of oral corticosteroids could then 6 weeks later. There was a 2-week run-in
reduce inflammation and enterochromaffin cell num- period, followed by 3 weeks of taking the medication
bers, and thereby improve symptoms, in post-infectious at breakfast time. There was a further week without
irritable bowel syndrome. medication prior to the second assessment. A symp-
tom diary was maintained over the full 6 weeks. As
this was a mechanistic study, the analysis was per
METHODS protocol rather than intention-to-treat, with dropouts
replaced. The inclusion criteria were as follows: age,
Patients
18–65 years; the presence of Rome I criteria irritable
Originally, we planned to recruit all patients from the bowel syndrome symptoms15 immediately following a
community who had developed new irritable bowel well-defined infective episode; and the presence of
syndrome symptoms following a positive Campylobacter these symptoms for more than 3 months. Exclusion
jejuni or C. coli stool culture. Between September 1999 criteria were as follows: abnormal screening investi-
and September 2001, questionnaires detailing the acute gations; previous diagnosis of irritable bowel
symptoms of gastroenteritis, and current and previous syndrome prior to gastroenteritis; known chronic
bowel habits, were sent to all subjects aged gastrointestinal disorder; use of systemic steroids
18–75 years, 3 months after a positive stool culture within the previous 4 months; clinically significant
for Campylobacter within the Nottingham Health hepatic, renal or cardiac dysfunction; diabetes melli-
Authority area. Those who responded and recorded tus; active infection; active peptic ulcer; immunosup-
new irritable bowel syndrome symptoms in accordance pression; alcoholism; gout; previous venous or arterial
with the modified Rome I criteria15 were invited to take thrombus; known thrombophilic disorder; current or
part. However, the number of patients with post- previous depression requiring ongoing drug therapy;
infectious irritable bowel syndrome who could be or pregnancy.
contacted and agreed to take part in the study was
very low. We therefore extended the inclusion criteria to
Clinical and laboratory assessments
include patients who had a clear history of new irritable
bowel syndrome symptoms following an acute infection At the screening visit and 6 weeks later, patients filled
and who had been referred to gastroenterology out- in questionnaires and underwent a general examina-
patients, often with a previously positive stool culture. tion, including rigid sigmoidoscopy and rectal biopsy.
We defined post-infectious irritable bowel syndrome as Blood was taken for full blood count, haematinics,
new bowel symptoms developing in a previously asymp- thyroid function, liver function, calcium, C-reactive
tomatic individual immediately after an acute illness protein and anti-endomysial antibody at the screening
characterized by two or more of the following: diar- visit. A repeat stool culture was also sent to exclude
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–84PREDNISOLONE IN POST-INFECTIOUS IRRITABLE BOWEL SYNDROME 79
recurrent or persistent infection. On both visits,
Immunohistochemistry
patients completed the following questionnaires: gas-
trointestinal symptom rating scale,16 irritable bowel Formalin-fixed, paraffin-embedded sections (4 lm) were
syndrome quality of life questionnaire,17 modified mounted on pre-coated slides (Dako, Glostrup,
Talley bowel symptom questionnaire,18 hospital anxi- Denmark). Following dewaxing, sections for subsequent
ety and depression scale19 and global score of well- staining with antibodies to CD3 were pre-treated with
being. The gastrointestinal symptom rating scale uses microwave antigen retrieval by heating to 101 C for
a Likert scale to produce separate summed scores for 12 min in 0.01 m citrate buffer at pH 6.0. Pre-
abdominal pain syndrome, reflux syndrome, indiges- treatment with a few drops of proteinase K (Dako) was
tion syndrome and constipation syndrome. The irrit- used for subsequent staining with antibodies to mast
able bowel syndrome quality of life questionnaire also cell tryptase. Sections for 5-HT staining did not undergo
uses a summed score for bowel symptoms, fatigue, pre-treatment. Slides were then washed for 15 min
limitations of activity and emotional function. Anxiety under running tap water and immersed in Tween 20
and depression scores indicate normal (0–7), mild (0.1% polyoxyethylene sorbitan monolaureate) for
(8–10), moderate (11–14) or severe (14–21) levels. 30 min. Slides were washed with tris-HCl buffer and
The global score of well-being is marked on a linear underwent automated staining using the Dako Tech-
scale, from the worst (0) to the best (100) health that mate 500 Plus and Techmate reagents (Dako, Glostrup,
the patient has ever known. Denmark). Enterochromaffin cells were stained with
Patients recorded in a diary the severity of abdominal primary rabbit polyclonal antibodies to serotonin (dilu-
pain or discomfort (0, none; 1, mild; 2, moderate; ted 1 : 100, Dako). Mast cells were stained with
3, intense; 4, severe), stool consistency (1, very hard; primary mouse monoclonal antibodies to mast cell
2, hard; 3, formed; 4, loose; 5, watery), urgency tryptase (diluted 1 : 250, Dako). Lamina propria and
(present or absent) and the number of stools per day. intra-epithelial T lymphocytes were stained with pri-
At the end of each week, there was a yes or no response mary rabbit polyclonal antibodies to CD3 (diluted
to adequate relief of irritable bowel syndrome pain or 1 : 100, Dako). All sections were stained with a
discomfort and adequate relief of diarrhoea within the secondary biotinylated goat anti-mouse and anti-rabbit
last 7 days. Adverse events were recorded and the antibody as part of the automated process.
questionnaires and rectal biopsy were repeated at week
6. A further modified Talley questionnaire was sent
Methods of cell counting
to patients at 3 months with a yes or no response to
whether their bowel symptoms were still troublesome. Sections were oriented so that the full mucosal height
A final questionnaire was sent 1 year later, asking was visible. Enterochromaffin cells, lamina propria
whether abdominal symptoms were still troublesome. lymphocytes and mast cells were counted within four
All data were analysed prior to the treatment code being non-overlapping, high-power fields (·200 magnifica-
broken. tion) on a single microscope by one person (S.P.D.).
Sections were assessed with the operator blind to the
clinical details. Entero-endocrine cells were counted if
Rectal biopsy
there was a visible nucleus surrounded by dark-stained
Rigid sigmoidoscopy was performed without bowel granules. CD3-positive lymphocytes and mast cells were
preparation. At each visit, two rectal biopsies were easily recognizable. Fields containing parts of lymphoid
obtained using endoscopic biopsy forceps (FB-13K-1, aggregates were avoided. All mast cells that were
Olympus, Japan), and these were gently flattened on filter quantified were within the lamina propria. The
paper, fixed in formalin and embedded in paraffin wax mean numbers of entero-endocrine cells, mast cells
with orientation optimized using a dissecting microscope and lamina propria T lymphocytes were expressed per
to ensure that sections were perpendicular to the mucosa. high-power field. Using four fields for cell counting at
A single expert pathologist (D.J.) performed conventional ·200 magnification gave coefficients of variability of
histological assessment on haematoxylin–eosin sections 7.2%, 6.8% and 6.9% for entero-endocrine cells, lamina
using standardized published criteria.20 All sections propria T lymphocytes and mast cells, respectively
appeared normal using these criteria. (n ¼ 10 biopsies). Intra-epithelial lymphocytes were
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–8480 S. P. DUNLOP et al.
counted per 500 epithelial cells as described previ- ylobacter. There were 635 replies, 75 of which were
ously.21 Surface intra-epithelial lymphocytes were excluded (incomplete, n ¼ 29; pre-existant irritable
counted between crypts and crypt intra-epithelial bowel syndrome, n ¼ 37; other significant disease,
lymphocytes were counted within longitudinal crypts n ¼ 9), leaving 560 questionnaires suitable for evalu-
spanning the full mucosal thickness. ation (Figure 1). Of these, 66 patients had new
symptoms of irritable bowel syndrome following infec-
tion (12% incidence). A further 47 patients (8%)
Outcome measures
developed two or more significant bowel symptoms.
The objective of the study was to investigate whether Fifty-one patients were invited to take part in the
prednisolone reduced the number of serotonin- study. Twenty of the 36 patients with post-infectious
containing enterochromaffin cells. The primary out- irritable bowel syndrome from the community study
come measure was therefore a reduction in the number were randomized and 17 completed the study. Fourteen
of enterochromaffin cells per high-power view. Secon- of the 15 patients with post-infectious irritable bowel
dary outcome measures included a reduction in the syndrome from the clinic were randomized and 12 com-
number of lamina propria T lymphocytes, a reduction in pleted the study (Figure 1). Of the 29 patients who
the modified gastrointestinal symptom rating scale score completed the study, 22 had diarrhoea-predominant
and the presence of troublesome symptoms at irritable bowel syndrome, two had constipation-
3 months. A modified gastrointestinal symptom rating predominant irritable bowel syndrome and five had
scale was used employing the four variables of abdomi- alternating symptoms. There were no differences in
nal pain, looseness, urgency and frequency, because symptoms between those recruited from the community
diarrhoea predominance is common in post-infectious or the clinic, or between those randomized to receive
irritable bowel syndrome.8 Each variable was scored in prednisolone or placebo (Tables 1 and 2). Overall, the
the range 1–7 (no symptoms at all to very severe mean number of days per week with abdominal pain
symptoms),22 so that a maximum score of 28 was or discomfort was 2.4 ± 0.4 (graded between mild and
possible. moderate in severity) and the mean bowel frequency
was 2.6 ± 0.3 per day. The relative frequencies of
symptoms occurring more than 25% of the time were as
Power calculation
follows: loose motions or diarrhoea, 86%; urgency,
We estimated from our previous study10 that a total of 79%; sense of incomplete evacuation, 79%; bloating,
30 patients would give 90% power to show a reduction 69%; more than three stools per day, 41%; straining,
of 33% in enterochromaffin cell counts at the 5% 34%; hard motions, 32%; passing mucus, 24%; less
significance level. than three stools per week, 10%.
Overall, the treatment was well tolerated. No side-
effects were reported by seven of 14 patients (50%)
Statistical analysis
taking prednisolone and 10 of 15 patients (67%) taking
Data were analysed per protocol. Parametric data are placebo. The reported side-effects in the prednisolone
shown as the mean (standard error of the mean), whereas group were heartburn (n ¼ 3), mood disturbance
non-parametric data are reported as the median (inter- (n ¼ 4), acne (n ¼ 2) and single reports of insomnia,
quartile range). Comparisons between parametric data weight gain, palpitations, bruising, arthralgia, increased
were made using paired t-tests for individual responses appetite and moon facies. In the placebo group, patients
and unpaired t-tests for initial values of the groups using reported heartburn (n ¼ 3), mood disturbance (n ¼ 2)
SPSS for Windows Version 9 (SPSS Inc., Chicago, IL, and single reports of insomnia, weight gain and
USA). Correlations were assessed using the Pearson headache. Five patients were screened and randomized
correlation coefficient. but did not complete the study. Three patients were
given the medication but did not take any of it for fear of
possible steroid side-effects. Two patients, randomized to
RESULTS
receive prednisolone, failed to complete the treatment
Over the 2-year period, questionnaires were sent to course. One was withdrawn due to severe heartburn
1407 patients with positive stool culture for Camp- and acid regurgitation symptoms after 1 week, and the
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–84PREDNISOLONE IN POST-INFECTIOUS IRRITABLE BOWEL SYNDROME 81
Community cases Table 1. Demographic details at entry of patients with post-
infectious irritable bowel syndrome. Data as mean ± S.E.M. or
1407 Campylobacter-positive stool cultures median (interquartile range). Duration (months)
Prednisolone Placebo
(n ¼ 14) (n ¼ 15) P value
635 replies to questionnaire
Age (years) 40.1 ± 3.0 38.3 ± 2.8 0.7
Male/female 8/6 6/9 0.6
560 eligible subjects Duration of symptoms 4 (3.75–60) 5 (3–36) 0.9
Community/clinic 8/6 9/6 0.9
A score 8.5 ± 0.9 8.2 ± 0.9 0.8
D score 4.6 ± 1.0 4.9 ± 0.8 0.8
n=66 post-infectious IBS Global well-being (%) 64 ± 4.5 64 ± 6.0 1.0
unable to contact n=30 Positive stool cultures 10 12 0.4
A, anxiety; D, depression scores.
36 contactable
other declined to continue after noticing moderate ankle
13 excluded swelling after 3 days. Both patients were replaced
4 better without the randomization code being broken.
9 declined
Rectal histology
23 screened
The number of enterochromaffin cells was elevated
3 excluded
alcoholic n=1 (mean, 37.3 ± 1.3 per high-power field; range, 22–50)
pre-existing IBS n=1 and was correlated with the number of lamina propria
on MST for chronic back pain n=1 T lymphocytes (r ¼ 0.460, P ¼ 0.014), confirming
previous results. However, there was no significant
reduction in the number of enterochromaffin cells
20 randomized after prednisolone compared with placebo (Table 3).
didn't take treatment n=1 Although there was a significant reduction in the
withdrawn due to side-effects n=2 number of lamina propria T lymphocytes after predn-
isolone (22.0% ± 5.6%, P ¼ 0.003), but not after
placebo (11.5% ± 8.6%, P ¼ 0.1), these differences
were not statistically significant. There were no signifi-
17 completed study cant reductions in mast cells, crypt intra-epithelial
lymphocytes or surface intra-epithelial lymphocytes in
Clinic either treatment group (Table 3).
15 screened
Symptoms
declined n=1
Neither prednisolone nor placebo treatment was
14 randomized associated with any significant reduction in the daily
scores of abdominal pain, consistency, urgency or stool
didn't take tablets n=2 frequency throughout the 6-week study period.
Although there was a decrease in the modified gastro-
12 completed study intestinal symptom rating scale score in the placebo
group (P < 0.02), this decrease was not statistically
Figure 1. The recruitment process of patients from the commu- different from that observed in the prednisolone group
nity and out-patient clinic. (P ¼ 0.1) (Table 2). Using summed scores from the
irritable bowel syndrome quality of life questionnaire
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–8482 S. P. DUNLOP et al.
Table 2. Symptoms before and after inter-
Prednisolone Prednisolone Placebo Placebo
vention. Data as mean ± S.E.M.
before after before after
Days of pain per week 2.0 ± 0.5 2.4 ± 0.5 2.7 ± 0.6 2.6 ± 0.5
Bowel frequency per day 2.3 ± 0.4 2.1 ± 0.3 2.9 ± 0.4 2.1 ± 0.3*
Modified GSRS score 11.9 ± 1.3 10.6 ± 1.0 13.4 ± 1.6 9.5 ± 1.0*
GSRS, gastrointestinal symptom rating scale.
* Difference before compared to after, P < 0.02, paired t-test.
Table 3. Percentage reductions ± S.E.M. in
Prednisolone Placebo P value
quantitative rectal histology after
Reduction in EC cells ) 0.8 ± 9.2 7.9 ± 7.9 0.5 treatment
Reduction in CD3 LP lymphocytes 22.0 ± 5.6 11.5 ± 8.6 0.3
Reduction in mast cells per hpf ) 9.6 ± 14.3 ) 5.0 ± 11.3 0.8
Reduction in crypt IEL ) 0.8 ± 16.6 ) 4.0 ± 11.4 0.9
Reduction in surface IEL 2.0 ± 11.4 ) 6.0 ± 16.7 0.7
EC, enterochromaffin; hpf, high-power field; IEL, intra-epithelial lymphocyte; LP, lamina
propria.
and the gastrointestinal symptom rating scale, there syndrome is not a transient condition, with more than
were no significant differences in score reductions for one-half of patients remaining symptomatic 6 years
any of the syndromes of bowel symptoms, fatigue, after infection.8 Our earlier studies suggested that
activity limitation or emotional function, apart from increased serotonin-containing entero-endocrine cells
constipation. The constipation syndrome score was were a frequent finding in such patients,9 and that
reduced more in the placebo group (1.0 ± 0.33) than enterochromaffin cell numbers correlated with lympho-
in the prednisolone group () 0.4 ± 0.55) (P ¼ 0.035). cyte counts,10 suggesting a possible causal link. This
Scores for anxiety, depression and global well-being trial aimed to test whether suppression of the lympho-
showed no significant change in either treatment group. cyte count with prednisolone would reduce enteroch-
The questionnaire at 3 months was returned by 21 of romaffin cell numbers and possibly improve symptoms.
the 29 patients. Of the 19 who answered the question, During the study, we encountered a number of obstacles
‘Are your symptoms still troublesome?’, three of nine in which need to be taken into account when assessing the
the prednisolone group and four of 10 in the placebo significance of our negative result.
group replied ‘yes’ (P ¼ 0.8). Patients were asked at Firstly, it proved to be extremely difficult to recruit
1 year if, over the last 12 months, troublesome bowel patients with post-infectious irritable bowel syndrome, as
symptoms were still present. Replies were received from they felt that their symptoms did not warrant the
20 of the 29 patients. Bowel symptoms remained in perceived risk of steroid therapy. This is a common
eight of 11 given steroids and seven of nine given feature in irritable bowel syndrome, indicating that even
placebo (P ¼ 0.8). an effective medication would not be acceptable unless
it had a better safety profile than prednisolone, even at
doses similar to those in frequent use in the community
DISCUSSION
for asthma. Despite this perception, prednisolone was in
There is a growing body of evidence indicating that low- fact well tolerated overall, with only one patient being
grade inflammation may underlie some cases of withdrawn because of a significant adverse reaction:
diarrhoea-predominant irritable bowel syndrome,23, 24 severe gastro-oesophageal reflux. Therefore, we had to
particularly those developing after infection.10 Our extend our study criteria to include those referred to
study is therefore of particular importance, as it is the gastroenterology out-patients, who proved to be slightly
first report of an attempt to modify the disease process in easier to recruit. However, there were no differences in
irritable bowel syndrome rather than to simply suppress terms of rectal histology, severity of symptoms or anxiety
symptoms. This attempt is justified by our recent study, or depression scores between these groups, suggesting
which showed that post-infectious irritable bowel that our results can be generalized to patients with post-
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–84PREDNISOLONE IN POST-INFECTIOUS IRRITABLE BOWEL SYNDROME 83
infectious irritable bowel syndrome in both primary and The inflammatory response to Campylobacter infection
secondary settings. Secondly, our previous experience is most obvious in the terminal ileum and on the right
indicated that the need for rectoscopy and biopsy was a side of the colon. We used enterochromaffin cells in
major deterrent to recruitment, and so this was kept to rectal biopsies as markers of inflammatory damage
a minimum. However, this meant that we could only elsewhere, as we did not feel that we could justify
obtain histology at one time point. Our decision to biopsy repeated colonoscopy. We are currently attempting to
at 4 weeks after the start of prednisolone was a compro- develop non-invasive measures of disturbed function in
mise between the desire to avoid too long an interval, post-infectious irritable bowel syndrome which assess
with the risk of other intervening factors diluting the drug other involved areas, including gut permeability and the
effect, and too short an interval, which might miss a more release of serotonin after a meal. These may be more
slowly developing response. We cannot exclude the strongly associated with symptom severity and may
possibility that we may not have allowed enough time provide better ways of assessing new treatments in the
for enterochromaffin cell numbers to change. Animal future.
studies have suggested that there are two populations of At present, in spite of the small size of our study, we
enterochromaffin cells: most have a half-life around four can be confident that, for reasons of patient acceptabil-
times that of surrounding enterocytes, whereas a smaller ity and efficacy, prednisolone is not likely to be an
longer lived population have a half-life around 30 times effective treatment for post-infectious irritable bowel
that of surrounding enterocytes.25 Extrapolation from syndrome. Whether a more prolonged course of topic-
this would indicate that most human enterochromaffin ally active steroid or a cyclo-oxygenase-2 inhibitor
cells have half-lives of 1–2 weeks, although this has not would be more effective remains to be determined.
been measured directly. If this were true, then changes
should have been apparent at 4 weeks. Mucosal lym-
ACKNOWLEDGEMENT
phocyte counts decreased with prednisolone, but the
effect was small in comparison with placebo, probably The authors are grateful to AstraZeneca, Hassle,
because the baseline continued to fall. Larger numbers of Sweden, for providing financial support.
patients would have helped here, but the difficulty in
recruitment precluded this.
REFERENCES
One explanation for our findings is that the previously
noted correlation between lymphocyte and enteroch- 1 Chaudhary N, Truelove S. The irritable colon. Q J Med 1962;
romaffin cell numbers is not due to the stimulation of 31: 307–22.
2 McKendrick MW, Read NW. Irritable bowel syndrome — post
the production of enterochromaffin cells from stem cells
salmonella infection. J Infect 1994; 29: 1–3.
by lymphocyte-derived factors, but rather because an 3 Neal KR, Hebden J, Spiller R. Prevalence of gastrointestinal
enterochromaffin cell product, namely serotonin, symptoms six months after bacterial gastroenteritis and risk
encourages the migration of circulating lymphocytes factors for development of the irritable bowel syndrome: postal
into the mucosa. This pro-inflammatory effect of survey of patients. Br Med J 1997; 314: 779–82.
4 Gwee KA, Leong Y-L, Graham C, et al. The role of psycholo-
serotonin is thought to be due to the stimulation of
gical and biological factors in postinfective gut dysfunction.
5HT2 receptors on vascular smooth muscle cells causing Gut 1999; 44: 400–6.
increased permeability.26 Previous studies using the 5 Bergin AJ, Carmell Donelly T, McKendrick MW, Read NW.
same animal model have indicated that, although post- Changes in anorectal function in persistent bowel disturbance
infective neuromuscular dysfunction appears to be following salmonella gastroenteritis. Eur J Gastroenterol Hep-
lymphocyte dependent in the first few days, later, atol 1993; 5: 617–20.
6 Rodriguez LA, Ruigomez A. Increased risk of irritable bowel
cyclo-oxygenase-2 induction may be more important,27
syndrome after bacterial gastroenteritis: cohort study. Br Med J
but this has never been assessed in humans. 1999; 318: 565–6.
Finally, we should consider the possibility that 7 Munzer D, Eyad C. Spastic colitis and irritable bowel syn-
increased mucosal lymphocytes and enterochromaffin drome: which expression is prevalent? (A review of 120
cells do not cause symptoms, but are merely markers of cases). Trop Gastroenterol 1992; 13: 27–35.
8 Neal KR, Barker L, Spiller RC. Prognosis in post infective
previous inflammation, which causes symptoms by
irritable bowel syndrome: a 6 year follow up study. Gut 2002;
other mechanisms, such as damage to mucosal nerves, 51: 410–3.
as reported by several authors.28–30
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–8484 S. P. DUNLOP et al.
9 Dunlop SP, Jenkins D, Spiller RC. Distinctive histological pat- 20 Jenkins D, Balsitis M, Gallivan S, et al. Guidelines for the initial
terns of chronic inflammatory cells in rectal biopsies of biopsy diagnosis of suspected chronic idiopathic inflammatory
patients with different clinical subtypes of IBS. Gastroenter- bowel disease. The British Society of Gastroenterology
ology 2002; 122: A316(Abstract). Initiative. J Clin Pathol 1997; 50: 93–105.
10 Spiller RC, Jenkins D, Thornley JP, et al. Increased rectal 21 Austin LL, Dobbins WO. Studies of the rectal mucosa in coe-
mucosal enteroendocrine cells, T lymphocytes, and increased liac sprue: the intraepithelial lymphocyte. Gut 1988; 29:
gut permeability following acute Campylobacter enteritis and 200–5.
in post-dysenteric irritable bowel syndrome. Gut 2000; 47: 22 Likert R. A technique for measurement of attitudes. Arch
804–11. Psychol 1932; 140: 55.
11 Gershon MD. Review article: roles played by 5-hydroxytryp- 23 Chadwick VS, Chen W, Shu D, et al. Activation of the mucosal
tamine in the physiology of the bowel. Aliment Pharmacol immune system in irritable bowel syndrome. Gastroenterology
Ther 1999; 13(Suppl. 2): 15–30. 2002; 122: 1778–83.
12 Camilleri M, Mayer EA, Drossman DA, et al. Improvement in 24 Salzmann JL, Peltier-Koch F, Bloch F, Petite JP, Camilleri JP.
pain and bowel function in female irritable bowel patients Morphometric study of colonic biopsies: a new method of
with alosetron, a 5-HT3 receptor antagonist. Aliment Phar- estimating inflammatory diseases. Lab Invest 1989; 60:
macol Ther 1999; 13: 1149–59. 847–51.
13 Barbara G, De Giorgio R, Deng Y, Vallance B, Blennerhassett 25 de Bruine AP, Dinjens WN, Zijlema JH, Lenders MH, Bosman
P, Collins SM. Role of immunologic factors and cyclooxyge- FT. Renewal of enterochromaffin cells in the rat caecum. Anat
nase 2 in persistent postinfective enteric muscle dysfunction in Rec 1992; 233: 75–82.
mice. Gastroenterology 2001; 120: 1729–36. 26 Abbott NJ. Inflammatory mediators and modulation of blood–
14 Barbara G, Vallance BA, Collins SM. Persistent intestinal brain barrier permeability. Cell Mol Neurobiol 2000; 20:
neuromuscular dysfunction after acute nematode infection in 131–47.
mice. Gastroenterology 1997; 113: 1224–32. 27 Collins SM, Piche T, Rampal P. The putative role of inflam-
15 Thompson D, Creed FH, Drossman DA, Heaton KW, Mazzaca mation in the irritable bowel syndrome. Gut 2001; 49:
G. Functional bowel disease and functional abdominal pain. 743–5.
Gastroenterol Int 1992; 5: 75–91. 28 Barbara G, Stanghellini V, De Giorgio R. Neuro-immune
16 Svedlund J, Sjodin I, Dotevall G. GSRS — a clinical rating relationships in the colonic mucosa of irritable bowel
scale for gastrointestinal symptoms in patients with irritable syndrome patients. Neurogastroenterol Motil 2000; 12:
bowel syndrome and peptic ulcer disease. Dig Dis Sci 1988; A272(Abstract).
33: 129–34. 29 Vallance BA, Galeazzi F, Collins SM, Snider DP. CD4 T cells
17 Wong E, Guyatt GH, Cook DJ, Griffith LE, Irvine EJ. Development and major histocompatibility complex class II expression
of a questionnaire to measure quality of life in patients with influence worm expulsion and increased intestinal muscle
irritable bowel syndrome. Eur J Surg Suppl. 1998; 583: 50–6. contraction during Trichinella spiralis infection. Infect Immun
18 Talley NJ, Phillips SF, Wiltgen CM, Zinsmeister AR, Melton LJ 1999; 67: 6090–7.
3rd. Assessment of functional gastrointestinal disease: the 30 Galeazzi F, Haapala EM, van Rooijen N, Vallance BA, Collins
bowel disease questionnaire. Mayo Clin Proc 1990; 65: SM. Inflammation-induced impairment of enteric nerve
1456–79. function in nematode-infected mice is macrophage
19 Zigmond AS, Snaith RP. The hospital anxiety and depression dependent. Am J Physiol Gastrointest Liver Physiol 2000;
scale. Acta Psychiatr Scand 1983; 67: 361–70. 278: G259–65.
2003 Blackwell Publishing Ltd, Aliment Pharmacol Ther 18, 77–84You can also read
