SEARCH

SEARCH BY CITATION

Keywords:

  • capsule endoscopy;
  • gastrointestinal food allergy;
  • intestinal blood loss;
  • intestinal lymphoid hyperplasia;
  • small-bowel pathology

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

To cite this article: Hagel AF, de Rossi TM, Zopf Y, Lindner AS, Dauth W, Neurath MF, Raithel M. Small-bowel capsule endoscopy in patients with gastrointestinal food allergy. Allergy 2012; 67: 286–292.

Abstract

Background:  Food allergy may present with a plethora of gastrointestinal and extraintestinal symptoms such as abdominal pain, diarrhea, cardiocirculatory symptoms, cutaneous reactions, or rhinitis. Macropathological lesions like lymphofollicular hyperplasia and erosive or ulcerative lesions have seldom been described in gastroscopy and colonoscopy previously.

Methods:  Fifteen patients presenting with unspecific abdominal symptoms in which food allergy was detected in due course were included. During the examination process, those patients showed various indications for small-bowel capsule endoscopy, such as weight loss and anemia.

Results:  Fourteen (93.3%) of the 15 small-bowel capsule endoscopies could be assessed, showing nonerosive lesions such as erythema, swelling, and lymphoid hyperplasia in 8 patients (57.1%) and erosive lesions such as aphthoid lesions, erosions, and petechiae in 4 patients (28.6%) with food allergy.

Conclusion:  In 15 patients with confirmed food allergy and after exclusion of other diseases, 12 (85.7%) showed various unspecific nonerosive or erosive mucosal lesions within the small bowel, resulting, however, partially in grave consequences such as anemia. Lymphoid hyperplasia was the most prominent finding in 7 patients (50%), albeit infectious disease had been excluded. Anemia improved within 1 year after adequate antiallergic treatment.

Abbreviations
BPCFC

blinded placebo-controlled food challenge tests

CE

capsule endoscopy

GI

gastrointestinal

Incidence of food allergy has been reported to increase and currently affects 3–5% of the adult population (1–3). It is defined as an immune-mediated hypersensitivity to ingested allergens, whereby mostly IgE-mediated symptoms of type I allergy (Coombs and Gell) occur but also non-IgE-mediated reactions have been described (3–7). Food allergy may present with various clinical patterns, either as acute symptomatology like oral allergy syndrome, acute postprandial rhinitis caused by food allergens (e.g., pollen-associated food allergy), and food-dependent exercise-induced anaphylactic symptoms or as chronic disease like eosinophilic allergic enteropathy, irritable bowel syndrome, or microscopic and allergic colitis (5–11).

Gastrointestinal (GI) food allergy commonly presents with unspecific symptoms like nausea, colicky abdominal pain, reflux, vomiting, abdominal pain, diarrhea, or hematochezia, whereas extraintestinal signs of allergy such as skin reactions, tachycardia, flush, and urticaria may occur in a lower frequency (2, 8, 10–12). Thus, in patients showing the above-mentioned gastrointestinal symptoms, different diagnostic examinations are conducted, such as transabdominal ultrasound, small-bowel magnetic resonance imaging, endoscopy, and histologic processing, with often unspecific results at best and tissue eosinophilia in only a minority of patients. But until gastrointestinal allergy is diagnosed by demonstration of specific IgE sensitizations in skin, blood, or intestine directing the patients to confirmatory oral provocation tests (single- or double-blind placebo-controlled food challenge), a substantial amount of time is usually passed (6–8, 12, 13).

In literature, macropathological results during gastroscopy and colonoscopy have been mainly described as lymphofollicular hyperplasia, lymphonodular duodenitis or mucosal swelling, erythema, and sometimes erosive lesions like colitis and ulcerative lesions in some patients with gastrointestinal food allergy (4, 5, 10, 14–16). However, so far, systematic studies investigating the mid-gastrointestinal tract (jejunum, ileum) in food allergy are currently lacking. In this study, we evaluated a small cohort of patients with confirmed gastrointestinal food allergy using wireless capsule endoscopy (CE) for evaluation of the small intestine.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

In this study, we included 15 patients with proven gastrointestinally mediated food allergy. All patients presented initially abdominal symptoms, nausea, pain, vomiting, and/or diarrhea after certain meals. Postprandial extraintestinal signs of allergy such as pruritus, flush, bronchial asthma, and allergic rhinoconjunctivitis occurred in 6 of 15 patients (40%).

Before suspicion of gastrointestinal food allergy, an extensive exclusion of various other differential diagnoses (e.g., infections, celiac disease, lymphoma, and neuroendocrine tumor mastocytosis) had already been conducted. Every patient was assessed on grounds of their history and detailed skin prick tests of environmental allergens (molds, fibers, pollen, dust) and food allergens (fish, fruit, vegetables, grains, and different types of wholemeal, flour, bran, egg, milk, meat, soy, spices, and cheese). Carbohydrate malabsorption was known in four patients, but they had persistent symptoms despite avoidance of lactose and fructose. One patient had associated salicylate intolerance, and in two patients histamine intolerance was suspected. In all patients, stool cultures and Helicobacter pylori testing were performed with negative findings. Total and antigen-specific serum IgE detections of the putative allergens were performed according to the patients’ history or skin tests. Case history, skin test reactions, and serum antigen-specific IgE testing were supplemented in 13 of 15 patients (86.6%) with determination of intestinal antigen-specific IgE during endoscopically guided segmental gut lavage as described previously (11) before performing blinded placebo-controlled food challenge tests (BPCFC). In the case of uncertainties about nontolerated foods, food challenge tests were conducted for basic foodstuff. Confirmatory diagnosis of gastrointestinal food allergy was achieved in nine patients by single-blind and in four patients with double-blind PCFC using a naso-gastric tube (11, 12), while two patients refused to undergo oral provocation because of a history of anaphylactic symptoms.

During the initial clinically indicated work-up, small-bowel CE was performed in these patients because of several symptoms and to exclude other significant pathology of the small bowel. This examination was conducted using Pillcam SB2 (Given Imaging Ltd, Yoqneam, Israel) after overnight fasting and oral ingestion of 2 l PEG solution. The recorded data were searched, documented, and evaluated by two experienced investigators who were blinded to each other’s findings. Timing of CE was under an unrestricted diet when patients described episodes of symptoms related to food intake. During the week of CE, no medication was given to any patient. Clinical symptoms, indications for CE, signs of atopy, type of allergic reactions involved, and the main causative food allergens detected during single- or double-blind PCFC are summarized in Table 1. The study was conducted in accordance with the Helsinki declaration, and all patients gave their informed consent.

Table 1.  Clinical symptoms, atopy status, indications for capsule endoscopy, and causative allergens identified by blinded placebo-controlled food challenge (BPCFC)
Allergy typeClinical reactions during food challengeEliciting allergensIndication for capsule endoscopy
  1. n, Patient number; OAS, oral allergy syndrome.

  2. Atopy status was defined as positive, when history or clinical manifestation of the patient gave evidence for milk crust, atopic dermatitis or eczema, bronchial asthma, and/or allergic rhinoconjunctivitis.

  3. For definition of the allergy type, the most dominant immunologic signs were chosen to classify the ongoing allergic mechanisms.

  4. Type I allergy (systemic IgE sensitization) was recognized when positive skin and/or antigen-specific IgE levels were present in serum (>0.35 U/ml); type I allergy (local IgE sensitization) was diagnosed when intestinal lavage fluid contained elevated food antigen–specific levels of IgE (>0.35 U/mg protein) (11, 12).

  5. Type IV allergy was suspected in four patients who showed markedly increased production of serum TNF levels during or after food challenge, while prechallenge TNF levels were normal during potato–rice diet.

Type I allergy (systemic food antigen–specific IgE) n = 9 8 atopics (88.8%) Median serum IgE 168 (84.5–398)8 diarrhea 4 pruritus, vomiting, abdominal pain, tachycardia 4 epigastric or abdominal pain and colics, bloating 2 nausea, hypotension, GI bleeding, OAS 1 urticaria6 nuts 3 egg, wheat 3 pollen-associated fruits 1 milk 1 soy flour3 weight loss, malabsorption 3 macroscopic or occult GI bleeding (anemia) 2 abdominal pain and colics
Type I allergy (local food antigen–specific IgE) n = 2 1 atopics (50%) Median serum IgE 39 (23–77.5)2 diarrhea 2 abdominal pain 2 iron deficiency anemia1 milk 1 nuts, pork, egg, wheat1 iron deficiency anemia 1 exclusion of celiac disease or lymphoma
Suspected type IV allergy (cellular hypersensitivity, no systemic or local food–specific IgE) n = 4 3 atopics (75%) Median serum IgE 68 (12–570.5)2 diarrhea 2 bloating, vomiting, flatulence, hypotension 1 tachycardia, flush, GI bleeding2 wheat, beef 2 milk, pork, soy flour, eggDiarrhea work-up, suspicion of neuroendocrine tumor, GI bleeding

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

Clinical symptoms and indications for capsule endoscopy

Ten of the patients examined were women, and five were men. The mean age was 39.6 years (range, 18–53 years). As illustrated in Table 1, in 12 of 15 patients (80%), clinical signs of atopy were present. Three different pathophysiological allergy types could be identified in this population with nuts, egg, wheat, and milk as the most frequent allergens. Clinical symptoms during unrestricted diet or displayed after specific food challenge during BPCFC included diarrhea (80%), abdominal pain or colics (66.6%), tachycardia (60%), skin reactions (40%), and hypotension (26.6%, Table 1). Endoscopy of the upper and lower gastrointestinal tract was usually performed before CE, and their macroscopic and histologic findings are listed in Table 2. In general, no similarities between CE and endoscopy were found.

Table 2.  Endoscopic and histologic findings in 15 patients with GI food allergy at the upper, mid and lower GI tract
Patient no.Esophago-gastro-duodenoscopyHistology – upper GI tractCapsule endoscopyIleo-colonoscopyHistology – lower GI tract
  1. GI, gastrointestinal; IBD, inflammatory bowel disease; neg., negative; LH, lymphofollicular hyperplasia; NERD, nonerosive gastroesophageal reflux disease, NSAID, nonsteroidal anti-inflammatory drugs.

  2. Endoscopic and histologic results from upper and lower gastrointestinal tract were compared with capsule endoscopy findings. Some patients had more than one qualitative pathological finding during endoscopies.

 1NormalNormalNonerosive lesionsNormalNormal, exclusion IBD
 2Chronic erosionsHelicobacter pylori -neg. chronic gastritisNonerosive lesions, delayed capsule excretionHyperplastic polyp, patchy erythema, hemorrhoidsHyperplastic polyp, normal
 3Fundic glands atrophy antrumNormalNonerosive lesionsHypertrophic anal papillaeLH ileum, edema, and mild tissue eosinophilia
 4Nonerosive reflux lesions at cardia, erythema corpusH. pylori-neg. chronic gastritisErosive lesionsLH, patchy erythematous colonic lesions, rectal tubular adenomaLH, tubular adenoma
 5NormalNormalNormalNormalNormal
 6Axial hiatal hernia and nonerosive esophageal reflux lesionsNERD, chronic type C gastritisErosive lesionsCecum erythema hyperplastic rectal polyps (<5 mm)Normal ileum, mild eosinophilia right hemicolon, hyperplastic polyps
 7NormalNormalNonerosive lesions, lipid islets jejunumNormalNormal
 8Atrophy antrum and duodenumNormalNormalNormalPatchy edematous colonic mucosa
 9Aphthous lesions antrum, chronic erosionsAcute and chronic pangastritis with lymphocyte proliferationErosive lesions, lipid islets duodenum4 colorectal adenoma (2–14 mm)2 tubular and 2 tubulovillous adenoma
10Ectopic pancreas antrum 2 acute erosions duodenal bulbH. pylori-neg. chronic corpus gastritisNonerosive lesions, lipid islets jejunum, delayed capsule retentionNormalLipid islets ileum, edema ileocecal valve
11Erosive esophageal refluxGastroesophageal reflux disease with suspected short-segment Barrett mucosaNonerosive lesionsNormalNormal
12Atrophy antrum, bile reflux at pylorusAtrophy and intestinal metaplasiaNonerosive lesions, lipid islets duodenumHyperplastic polyps rectumHyperplastic polyps rectum
13Chronic erosions, erythema, stomach and duodenum, suspected unspecific duodenitisAcute gastritis, increased lymphocyte numbers in duodenum, exclusion IBD celiac disease and Behcet’s diseaseErosive lesions, delayed capsule excretionNormalNormal, exclusion IBD
14Acute and chronic stomach erosions, ectopic gastric mucosa duodenumChronic type C gastritisCapsule retention stomachHemorrhoidsNormal
15Nonerosive reflux at cardiaEdema at cardia and mild eosinophiliaNonerosive lesionsIsolated colonic ulcer at ileocecal valve (6 mm)Benign ulcer with surrounding mild tissue eosinophilia, suspected NSAID lesion

Indications for CE were macroscopic or occult gastrointestinal bleeding resulting partly in anemia in 5 of 15 patients (33.3%), weight loss and signs of malabsorption in 6 (40%) patients, suspicion of small-bowel (neuroendocrine) malignancy in 2 (13.3%) patients, or diffuse abdominal pain in 2 patients.

Small-bowel findings in capsule endoscopy

Out of the 15 conducted CEs, 14 could be assessed (93.3%), with one capsule being retained in the stomach.

Among the remaining 14 patients, various pathological findings could be detected in 12 of 14 patients (85.7%). Table 3 lists all pathological findings detected during small-bowel CE. These findings were divided into erosive lesions, nonerosive lesions, findings of unknown functional significance, and clinically irrelevant lesions. Most patients had more than one lesion. Four patients (28.6%) had erosive mucosal lesions, eight patients (57.1%) showed non-erosive lesions, and four patients each (28.6%) had findings of unknown functional significance and/or clinically irrelevant lipid islands, respectively.

Table 3.  Findings during capsule endoscopy (CE) in 15 patients with gastrointestinal food allergy
 Frequency of findingsnLocations of findings
DuodenumJejunumIleum
  1. The patient numbers of nonerosive or erosive lesions, findings of unknown functional significance, or clinically irrelevant findings during CE are given in Table 3 for the whole group and according to their locations within the small bowel. n represents the number of findings, with some patients showing more than one qualitative pathological finding.

  2. In 8 patients, a total of 21 nonerosive lesions were found (3 per patient), with predominant location at jejunum and ileum.

  3. In 4 patients with erosive lesions, a total of 10 erosive lesions (2.5 per patient) were found with predominant location at proximal parts of the small bowel.

Nonerosive lesionsLymphoid follicular hyperplasia7 25
 Edema, swelling4 22
 Erythema32 1
 Submucosal lesions3 12
 Focal lymphangiectasia211 
 Relative stenosis2 11
8 patientsSum of all types of nonerosive lesions213711
ErosiveAphthoid lesions3111
LesionsErosions321 
 Petechiae2 11
 Inflammatory polyps211 
4 patientsSum of all types of erosive lesions10442
Not relevant findingsLipid islands4 22
4 patients     
Findings of unknown functional significanceCapsule retention in stomach (swelling?)1   
4 patientsDelayed capsule excretion3   

Interestingly, 8 of 12 patients (66.6%) showed one or more nonerosive lesions like edema, erythema, swelling, lymphoid follicular hyperplasia, or submucosal lesions in mid-jejunum and ileum (Fig. 1). Lymphoid follicular hyperplasia was the most common pathological finding detected in seven patients (58.3%).

image

Figure 1.  (A and B) Lymphofollicular hyperplasia in capsule endoscopy and endoscopically. These are common in patients with gastrointestinally mediated food allergy. Aphthoid lesions (C) and acute erosion (D) are also regularly found in patients with other diseases such as gastrointestinal bleeding and Crohn’s disease. These lesions can lead to chronic anemia caused by intermittent blood loss in gastrointestinal food allergy.

Download figure to PowerPoint

The remaining four patients (33.3%) showed one or more erosive mucosal lesions such as aphthoid lesions, erosions, petechiae, or two inflammatory polyps (Fig. 1). Erosive mucosal lesions were distributed throughout the whole small bowel from duodenum to ileum. Three of those patients with erosive and inflammatory lesions suffered of chronic iron deficiency anemia, despite negative findings during upper or lower endoscopy including negative hemoccult tests before CE. Although after introduction of antigen-specific diet no CE have been conducted, two of three patients with anemia showed an increase in hemoglobin levels within normal range within 1 year of follow-up.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

Gastrointestinal food allergy has been reported to induce a plethora of clinical symptoms after consuming antigen-containing foodstuffs, but objectification and visualization of corresponding or associated mucosal lesions are difficult (5–7, 9–11, 16–18). Small-bowel CE was not performed directly after antigen-specific oral food challenge procedures, but during an unrestricted diet when patients complained their typical symptoms. Small-bowel CE revealed in 12 of 14 patients (85.7%) various types of mucosal lesions, which were classified as erosive lesions, nonerosive lesions, findings of unknown functional significance, or clinically irrelevant findings. A priori, all findings appear to be of unspecific nature, but when considering that all patients had excluded a huge number of differential diagnoses (e.g., infections, celiac disease, inflammatory bowel disease, lymphoma, NSAID intolerance, and malignancy), the high frequency of mucosal lesions in this group of patients with proven gastrointestinal allergy (2.6 lesions per patient) at a mean age of 39.6 years is surprising. This is even more surprising when compared to existing studies, in which patients of a comparable age (34.5 and 43 years, respectively) presenting with chronic abdominal pain and negative endoscopic and radiological examinations but without common underlying disease (19, 20). In these groups, pathological findings were detected in up to 43%.

Nonerosive and other unspecific inflammatory changes like edema, swelling, and erythema were found as most frequent findings in gastrointestinally mediated allergy, but such lesions usually require above-mentioned differential diagnosis. As such diseases have been excluded in all patients, these unspecific nonerosive mucosal lesions demonstrate that intestinal mucosal physiology may become altered in food allergy and they reflect probably (subtile) intestinal allergic manifestations as has been reported from other investigators at other anatomic sites (7, 16–18). Although of unspecific nature, some of these mucosal features may be responsible for certain clinical and functional symptoms of patients with gastrointestinal allergy as dysmotility, bloating, heat sensations, or increased intestinal secretion. Delayed capsule excretion occurred in two of three patients with typical nonerosive lesions like mucosal swelling, edema, or lymphofollicular hyperplasia. Thus, it might be speculated whether these nonerosive lesions contribute to motility complaints, altered small-bowel function with reduced emptying of intestinal contents to the colon, or abdominal heat sensation in food-allergic individuals. Interestingly, in one patient with the above-listed symptomatology, nonerosive swelling was accompanied by extreme edema of the ileocecal valve (12 mm), which could also be observed during ileo-colonoscopy as an extremely rare event. But among the eight patients with nonerosive small-bowel lesions, ileo-colonoscopy was normal or inconclusive in at least five of eight patients. Thus, CE should be performed after inconclusive ileo-colonoscopy, and its clinical value is to supplement usual gastrointestinal diagnostics regarding the presence of small-bowel pathology. Unfortunately, at present, CE is not able to take biopsies for histology, but in patients with nonerosive mucosal lesions of the small bowel, performance of modern enteroscopy (e.g., double-balloon enteroscopy) should be discussed for further histologic assessment, measurement of food-specific intestinal IgE, or local food challenge within the jejunum. Topographically, nonerosive lesions were more often located in distal small-bowel segments (ileum), while erosive lesions were found in duodenum and proximal jejunum (Table 3).

Interestingly, lymphofollicular hyperplasia, which has also been described by several other investigators at the upper (duodenum) or lower gastrointestinal tract (ileum, cecum, colon ascendens), was the most prominent finding during small-bowel CE in our allergic patient cohort, albeit enteric infections or Helicobacter gastritis had been carefully excluded (4, 5, 15). So far, only one radiological study using enteroclysis was conducted to inspect larger parts of the small intestine in gastrointestinal allergy, but because of the more luminal orientation and their (sub)-mucosal location, conventional radiology failed to detect the increased occurrence of lymphoid follicles in this condition (3–5).

Erosive mucosal lesions were detected in a minor frequency, but may explain positive hemoccult tests and iron deficiency anemia in gastrointestinal allergy as has been reported in children and adults as a rare complication of food allergy (15–17) and may resolve after adequate antiallergic treatment (16). As has been shown in our cohort, 3 of 15 patients (20%) with normal or inconspicuous findings at upper or lower gastrointestinal endoscopy suffering from anemia caused by occult gastrointestinal blood loss had unexpected erosive or inflammatory lesions of the small bowel at the time of CE, and they recovered their hemoglobin levels within 1 year after allergen-specific avoidance. Thus, in contrast to conventional radiology, wireless small-bowel CE helps to visualize minute mucosal lesions in patients with gastrointestinal allergy and identifies the causes of otherwise unexplained anemia in certain allergic individuals as erosive, inflammatory, or ulcerative mucosal lesions. It demonstrates further that mucosal integrity in gastrointestinal allergy may become altered, manifesting with erosions or, in more severe cases, with ulcers (16, 17).

The reason for erosive and nonerosive lesions in patients with food allergy is partially the ingestion of food antigens triggering the development of an acute local anaphylactic inflammatory reaction (17), with production of short- or long-living mediators. While histamine and eicosanoids, for instance, may be responsible for the induction of observed nonerosive lesions, longer-acting mediators like cytokines, mast cell tryptase, or cationic proteins are suspected to be involved in erosive lesions, resulting in further alterations of the mucosal barrier and the lymphatic cell population (11, 15–18).

Thus, in summary, wireless CE detected a high frequency of various nonspecific mucosal lesions in the small bowel of patients with known gastrointestinal allergy. It is concluded that nonerosive, erosive, and lymphonodular lesions within the jejunum and ileum are an important morphologic feature of intestinal allergic inflammation and mucosal immune system activation, which may explain, at least in part, some gastrointestinal, functional, neurovegetative, and subjective symptoms in patients with food allergy.

Author contributions

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

All the authors contributed to the work in substantial matters, such as conception and design of study, acquisition of data, and interpretation of data, significantly involved in drafting the article or revising it critically for important intellectual content, and approved the final version to be published.

Conflict of interest

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References

The authors declare that they have no competing interests.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Author contributions
  7. Conflict of interest
  8. References