Introduction
Gastrointestinal symptoms assumed to be caused by food intolerance are reported frequently
in the general population. An estimated one-fifth of the population believe that they
have adverse reactions to food [1]. This proportion is even higher in patients with disorders of the gut-brain axis
(formerly functional gastrointestinal disorders) and may reach up to 80 % [2]. These symptoms include bloating, abdominal discomfort or pain reported soon after
food ingestion. They may occur in different clinical conditions, such as disorders
of the gut-brain axis, adverse reaction to food and gluten-related syndromes, which
frequently are interrelated [3]. In clinical practice, diagnostic tools to identify food components that trigger
gastrointestinal symptoms are limited. They comprise nutrition diary, identification
of lactose, fructose, sorbit and histamine intolerance, food allergies, mast cell
disturbances, and elimination diets, respectively. However, there is a significant
difference between self-reported and objective proven food intolerance, as shown by
placebo-controlled, double-blind randomized trials [4]
[5]. This discrepancy may be overcome by endoscopic confocal laser endomicroscopy (eCLE).
It has been shown recently that eCLE can provide an objective measure to test immune-mediated
reaction to food [6]
[7]. During eCLE, duodenal application of specified food via the endoscope channel may
induce immediate fluid extravasation through epithelial leaks.
Patients and methods
In an observational study we evaluated patients with functional abdominal pain who
presented at the Department of Internal Medicine and Gastroenterology of the Helios
Clinic Krefeld from January 2021 to June 2022. The study was approved by the local
ethic commission (IRB Fr-21–01). Symptoms were evaluated via a standardized DSFQ symptom
questionnaire [8]. All patients received a standardized diagnostic investigation, which was unremarkable.
This included upper and lower gastrointestinal endoscopy with biopsies, magnetic resonance
enteroclysis, Doppler sonography of the visceral vessels, and breath tests for sugar
intolerances and small intestinal bacterial overgrowth (SIBO), respectively. Blood
analysis included markers for inflammation/infection (c-reactive protein [CRP], blood
sedimentation rate [BSG], white blood cell count), mast cell dysfunction (tryptase,
N-methyl-histamine urine excretion), histamine intolerance syndrome (diamine oxidase),
immunoglobulin E (IgE), autoimmune (antinuclear antibodies, [ANA] and celiac disease
(immunoglobulin A [IgA] anti-transglutaminase antibodies), respectively. Stool analysis
consisted of calprotectin as a marker for gastrointestinal inflammation that was below
50 µg/g stool and microbiological analysis without evidence for pathogenic viruses
and bacteria. Before eCLE, patients received an allergen-reduced diet consisting of
boiled rice (954 kcal, 18.9 g protein, 1.8 g fat, 211.5 g carbohydrate, 1.8 g fiber)
and/or boiled potatoes (648 kcal, 16.2 g protein, 0.9 g fat, 153 g carbohydrate, 1.8 g
fiber) and water or coffee without milk and sugar ad libidum for 3 days. eCLE was
performed after an 8-hour fasting period. Standardized eCLE (Cellvizio System, Mauna
Kea Technologies, Paris, France) was applied as previously reported [6]
[7]. eCLE was performed after an 8-hour fasting period. The investigations were performed
during conventional upper gastrointestinal endoscopy under propofol sedation. Spontaneous
transfer of fluorescein into duodenal lumen was detected 10 minutes after intravenous
application of fluorescein and 10 minutes after duodenal food challenge (DFC). Local
food challenge was performed always by the same sequential application of five different
main food allergens diluted in 30 mL water (280–310 mosm/L) to the duodenal mucosa.
These were 1.5 g dry bio-yeast, 31.5 g dry egg, 1.5 g bio-milk, and 3 g soy flour
and 3 g wheat flour, respectively. Local application of sodium chloride solution 10 %
to the duodenal mucosa before food allergen exposure served as a control. The applications
into the duodenal lumen were always above the major papilla. The diluted food allergens
were applied directly to the duodenal mucosa through the biopsy channel of the endoscope.
At the end of the exposure time to the diluted food allergens, the remaining fluid
was withdrawn by suction through the endoscope. There was no indication of aspiration.
Positive mucosal reaction following food antigen exposure consisted of evoked leakage
of intravenously applied fluorescein into the duodenal lumen as previously described
[6]
[7]. The mucosal reaction was always visible clearly and scanned at three different
sites of the duodenal mucosa by two investigators. There was no intra-observer or
interobserver variability. If a reaction to any of the food components took place,
further food applications were discontinued. In this case, a second eCLE was performed
after several weeks to complete the sequence of the food challenge. Images of eCLE
findings before and after food challenge were documented and interpreted by two independent
observers. After food challenge, six duodenal biopsies were taken to analyze for mucosal
inflammation, intraepithelial lymphocytes (IELs) as well as number, distribution,
and morphology of mast cells by standard immunohistochemistry (CD117 and CD 25) and
counted per mm2 tissue in each patient. Patients received food exclusion dietary advice focused on
the results of eCLE. Clinical response to the dietary therapy was controlled 4 weeks
after eCLE by repeating the symptom questionnaire. Statistical analysis was performed
by Chi Quadrat and Mann Whitney U test and data were expressed as mean + SD.
Results
We evaluated 34 patients, 27 female, 46.4 + 15.0 years old. In all patients, diagnostic
evaluation as described in detail previously revealed no evidence of organic diseases
and no organic correlate that could explain their abdominal pain. Nine patients showed
elevation in IgE, one patient together with elevation of IgG4 and three patients had
a positive marker for Hashimoto thyroiditis. All patients reported their complaints
to be independent of their bowel habits. Therefore, the patients fulfilled the diagnostic
criterion for unspecified functional bowel disorder according to the Rom IV classification
[9]
[10] or for irritable bowel syndrome (IBS) according to German guidelines [11]. Of the patients, 73.5 % (n = 25, 20 female, 48.4 + 15.9 years) reported that their
abdominal pain was triggered by food (FI + ), whereas nine patients (7 female, 41.0 + 11.3
years) did not notice food intolerance (FI-).
Overall, eCLE showed spontaneous leakage of fluorescein in eight patients (23.5 %,
50 % female) that was not different from the subgroups with or without reported food
intolerance (IF + : 14.7 %/75 %, IF-: 33 %/66 %). Three patients (eCLE-, 8,8 %, 3
female) who reported food intolerance had neither spontaneous nor food-induced fluorescein
leakage. Twenty-three patients (eCLE + , 67,6 %) responded to the duodenal food challenge
([Table 1]). Frequency rank order of food antigens that induced a response were soy (50 %),
wheat (46.1 %), milk (20 %), egg (12%) and yeast (11.5 %), respectively. In 10 patients
with a positive eCLE, a second eCLE was performed after several weeks to complete
exposure to the remaining food allergens. Two patients responded to soy and wheat
and one patient to milk and yeast. Duodenal biopsies collected after food challenge
showed normal histology and no evidence of inflammation, mucosal atrophy or increase
in IELs. Mucosal mast cells appeared to be normal in morphology and distribution.
Average mast cell number in duodenal mucosa was 99.86 + 55.24 /mm2, 14–270 /mm2 and there was no significant difference between FI + and FI–, patients with and without
spontaneous leakage of fluorescein and patients responding and not responding to food
challenge, respectively ([Table 2]). Similarly, laboratory analysis of mast cell function measured by serum tryptase
and N-methylhistamine urine excretion as well as histamine intolerance as measured
by diamine oxidase was normal in all patients and subgroups ([Table 2]).
Table 1
Endoscopic confocal laser endomicroscopy with Fluorescein leakage into duodenal lumen
following food challenge (eCLE + ). FI + : patients reporting food intolerance, FI–:
patients reporting no food intolerance.
|
eCLE +
|
Yeast
|
Egg
|
Soy
|
Milk
|
Wheat
|
|
overall
|
11.5 %
|
12.0 %
|
50.0 %
|
20.0 %
|
46.1 %
|
|
FI +
|
15.0 %
|
10.5 %
|
41.%
|
12.5 %
|
46.0 %
|
|
FI–
|
0 %
|
16,7 %
|
100 %
|
50.0 %
|
0 %
|
Table 2
Number of mast cells in duodenal mucosal biopsies, serum tryptase, N-Methhylhistamine
excretion in urine and serum diamine oxidase overall, in patients with (FI + ) and
without (FI–) food intolerance, spontaneous (SL + ) and no (SL–) i. v. fluorescein
leakage into duodenal lumen and positive (FC + ) or negative (FC–) food challenge
of duodenal mucosa. Mean ± SD, (range).
|
Overall
|
FI +
|
FI–
|
SL +
|
SL–
|
FC +
|
FC–
|
|
Mast cells
(n/mm2)
|
99.86 ± 57.24
(14–270)
|
104.95 ± 60.54
(14–270)
|
83.85 ± 45.46
(20–147)
|
133.42 ± 54.10
(70–220)
|
89.18 ± 53.83
(14–270)
|
82.19 ± 36.54
(14–150)
|
168 ± 144.25
(66–270)
|
|
Tryptase
(ug/l)
|
3.96 ± 1.59
(1.1–7.2)
|
4.13 ± 1.52
(1.1–7.2)
|
3.55 ± 1.79
(1.8–7.0)
|
3.28 ± 1.23
(1.8–4.9)
|
4.19 ± 1.66
(1.1–7.2)
|
3.94 ± 1.47
(1.8–7.2)
|
4.16 ± 3.00
(1.1–7.1)
|
|
N-Methylhistamine
(ug/l)
|
102.7 ± 70.47
(14–250)
|
94.38 ± 59.88
(14–226)
|
119.37 ± 92.38
(23–250)
|
126 ± 85.32
(14–233)
|
94.9 ± 66.21
(23–250)
|
104.20 ± 72.21
(14–250)
|
76.5 ± 53.0
(l39–114)
|
|
Diamine oxidase
(U/l)
|
19.43 ± 18.16
(4.7–88)
|
21.11 ± 20.18
(4.7–88)
|
14.06 ± 8.69
(8.1–29)
|
13.13 ± 7.33 U/l,
(5.4–20)
|
20.48 ± 19.33
(4.7–88)
|
19.33 ± 18.42
(5.4–88)
|
20.35 ± 22.13
(4.7–36)
|
In the 23 patients who responded to the food challenge, the effect of dietary therapy
was evaluated with a second questionnaire 4 weeks after eCLE ([Table 3a] and [Table 3b]). Overall, 69.5% (n = 16) of the patients reported improvement in pain intensity
and reduction in pain frequency. Three patients (13.0 %) had no symptoms and five
patients (21.7 %) reported a reduction in pain frequency < 1×/week. Seven Patients
(30.4 %) reported receiving no benefit from the dietary therapy.
Table 3a
Efficacy of dietary therapy on pain intensity as illustrated by the results of the
first and second symptom questionnaire [8], n = 23 patients.
|
1. Questionnaire
|
2. Questionnaire
|
|
None
|
Slight
|
Moderate
|
Severe
|
Very severe
|
|
None
|
0
|
0
|
0
|
0
|
0
|
|
Slight
|
0
|
0
|
0
|
0
|
0
|
|
Moderate
|
1
|
1
|
3
|
0
|
0
|
|
Severe
|
2
|
4
|
3
|
3
|
0
|
|
Very severe
|
0
|
3
|
2
|
0
|
1
|
Table 3b
Efficacy of the dietary therapy on pain frequency as illustrated by the results of
the first and second symptom questionnaire [8], n = 23 patients.
|
1. Questionnaire
|
2. Questionnaire
|
|
< 1×/week
|
1×/week
|
2–3×/week
|
4–6×/week
|
Every day
|
|
< 1×/week
|
0
|
0
|
0
|
0
|
0
|
|
1×/week
|
0
|
0
|
0
|
0
|
0
|
|
2–3×/week
|
2
|
0
|
2
|
0
|
0
|
|
4–6×/week
|
1
|
0
|
0
|
2
|
0
|
|
Every day
|
2
|
4
|
5
|
2
|
3
|
Discussion
To our knowledge, this is the first report describing the application of eCLE in patients
with functional abdominal pain. The results of our study show that eCLE is a useful
tool for evaluating functional abdominal pain associated with food intolerance in
patients classified as having a nonspecific functional bowel disorder or IBS. The
response to food challenge was robust and indicated by clearly visible leakage of
IV fluoresceine into the duodenal lumen. This occurred always at different locations
of the duodenal mucosa. More than two-thirds of the patients reported that their abdominal
pain was triggered by food. This high rate of self-reported adverse reaction to food
in patients with functional bowel disease also has been reported in the literature
[12]. In contrast, the rate of objective proven food intolerance as shown by placebo-controlled,
double-blind randomized trials is very low [4]
[5]. The findings of our study suggest that eCLE could reduce the gap between subjective
feeling and objective measurable adverse reaction to food. In our study, eCLE could
detect immune-mediated mucosal reaction with leakage of fluorescein into the duodenal
lumen following mucosal food exposure in almost 70 % of patients. A comparable high
rate of positive eCLE in IBS according to Rom III also has been described in other
studies [6]
[7]. However, to our knowledge, our findings are the first in patients with functional
abdominal pain. Interestingly, soy and wheat were the food allergens that most frequently
evoked a mucosal response. With regard to wheat, milk, egg and yeast, this is in line
with other studies [6]
[7]. However, the high response rate to soy (50 %) in our study has not been reported
before. The reason for this is unclear but it may be caused by patient selection.
Twenty-three percent of patients showed spontaneous leakage of fluorescein before
duodenal food challenge, suggesting leaky gut syndrome. This also has been reported
in patients with functional dyspepsia, suggesting loss of mucosal integrity, which
could be triggered by stress-induced activation of mast cells [13]
[14]
[15]. A tight junction barrier defect that enhances disease progression also has been
suggested in post-infectious IBS and IBD in which the barrier loss induced by infection
may be the trigger that drives pathogenesis [16]. A tight junction barrier defect could also explain adverse reaction to different
food components. However, further studies are needed to clarify this potential relationship.
In our study, histologic evaluation of duodenal mucosal biopsies after the duodenal
food exposure revealed no pathologic findings, such as mucosal inflammation, increased
intraepithelial lymphocytes or mucosal mast cells. In addition, morphology and distribution
of mucosal mast cells appeared to be unremarkable and serologic markers for mast cell
dysfunction, histamine intolerance syndrome or autoimmune diseases were within normal
range. Subtle activation of mucosal immune cells following duodenal exposure to food
allergens has been reported in IBS [6]
[7] and functional dyspepsia [13]
[14]. In IBS [6]
[7], a significant increase in IELs in eCLE images has been reported in eCLE-positive
patients following food exposure, although IELs in histology were not different between
CLE-positive and eCLE-negative patients [6]. In another study [7], eosinophilic counts were not different between eCLE-positive and eCLE-negative
patients and IELs did not differ before and after exposure in the same patients. However,
post-exposure IELs were significantly higher in eCLE-positive patients compared to
eCLE-negative patients. In functional dyspepsia [13]
[14], significantly higher epithelial gap density compared to controls has been described.
This corresponded to impaired mucosal integrity, as shown by reduced transepithelial
electrical resistance, increased number of epithelial cells undergoing pyroptosis,
and altered duodenal expression of claudin-1 and interleukin-6. The trigger for leaky
gut is unknown but could be mediated by the central nervous system (e. g. stress)
as well as luminal factors such as food, acid, bile acids, and microbiota.
In our study, we did not evaluate the number of mucosal immune cells before and after
food challenge. Therefore, a potential mucosal immune reaction evoked by food allergens
remains to be proven. However, neither did we find an increase in IELs above normal
range nor abnormalities in mucosal mast cell morphology and distribution after exposure.
In addition, the number of mast cells was not different between the subgroups or between
eCLE-positive and eCLE-negative patients.
Interestingly, almost 70 % of patients reported a clinical benefit with reduction
of abdominal pain and 13 % were free of symptoms following use of a food exclusion
diary. This is a significant finding because it suggests the potential benefit of
selective dietary treatment guided by results of eCLE with food challenge in these
patients. The results of our study suggest an opportunity for use of individual and
tolerable dietary therapy in patients with adverse reactions to food and abdominal
pain. However, the magnitude of the placebo effects remains to be proven.
We did not apply the food allergens to the duodenal mucosa in a randomized order and
the applications were always above the major papilla. This may have caused a bias,
but that appears unlikely to us. We have no information about the timing or local
variance in response to the food challenge and cannot rule out this possibility. As
far as we know, there are no data in the literature available to prove this assumption.
In addition, we cannot rule out that part of the mucosal response to the food challenge
was mediated by a non-immune mechanism such as local release of nitric oxide-mediated
vasodilation by soy, wheat or the other food allergens [17]
[18]. However, we did not see a response to local application of sodium chloride solution
10 % to the duodenal mucosa, which makes osmotic effects unlikely. In addition, if
this effect of soy or wheat would have been the main mechanism for the mucosal reaction,
we would expect such a response in a greater number of patients. However, this was
not the case because only 50 % and 46 % of patients reacted to duodenal challenge
with soy and wheat, respectively.
Conclusions
In summary, the results of our study indicate that eCLE is a clinically useful tool
for evaluating patients with functional abdominal pain/IBS and adverse reaction to
food and to create individualized dietary therapy that may be clinically beneficial
for patients.
