Article Text


Eosinophilic oesophagitis and coeliac disease: is it just a casual association?
  1. F Verzegnassi1,
  2. J Bua1,
  3. P De Angelis2,
  4. L Dall’Oglio2,
  5. G Di Leo3,
  6. A Ventura3
  1. 1Department of Paediatrics, Institute of Child Health IRCSS Burlo Garofolo, University of Trieste, Trieste, Italy
  2. 2Surgical and Endoscopic Digestive Unit, Ospedale Pediatrico Bambino Gesù, Rome, Italy
  3. 3Department of Paediatrics, Institute of Child Health IRCSS Burlo Garofolo, University of Trieste, Trieste, Italy
  1. Correspondence to:
    Dr J Bua
    Department of Paediatrics, Institute of Child Health, IRCSS Burlo Garofolo, Via dell’Istria 65/1, 34100 Trieste, Italy; jennybua{at}

Statistics from

Eosinophilic oesophagitis (EO) is the leading cause of dysphagia in children.1 Long underestimated in the past, it is today being reported with increasing frequency. Acute and chronic dysphagia and food impaction are the typical symptoms in both children and young adults, whereas vomiting and refusal of food may be more frequent in infants.1,2 Diagnosis of EE is mainly based on the eosinophilic infiltration of the oesophageal mucosa (>20 eosinophils/high power field (HPF)) in association with some classic endoscopic features (adherent whitish plaques, oesophageal concentric rings, linear furrowing). The aetiology of EE is far from clear, but in about half of the cases reported it is associated with an allergy either to food or to aeroallergens.1,3

We describe three cases of EE, that we diagnosed by chance when performing endoscopy for the duodenal biopsy to confirm coeliac disease (CD). One case was diagnosed at the Institute of Child Health, IRCSS Burlo Garofolo, Trieste Italy, and, two at the Surgical and Endoscopic Digestive Unit of Ospedale Pediatrico Bambino Gesù, Rome, Italy.

Case 1

A 7-year-old boy was referred to the Gastroenterology Department, IRCSS Burlo Garofolo, Trieste, Italy with persistent sideropaenic anaemia since the age of 4-years. Given the unresponsiveness of his anaemia to iron supplementation, CD was suspected and confirmed by the positivity of antiendomysial serum antibodies (EMAs) and antitransglutaminase antibodies (tTG immunoglobulin (Ig) A 213 IU). The boy was allergic to dust mite, cat epithelium and betullaceae. His most common allergic manifestations included rhinitis and asthma. He was referred to our institute to confirm the diagnosis of CD. Endoscopy showed small white adherent plaques in the proximal tract of the oesophagus and haemorrhagic duodenitis with mucosal scalopping. Histopathological features of the duodenal biopsy were typical of CD (Marsch 3c atrophy). Histopathological examination of the oesophagus showed a large amount of eosinophilic cell infiltrate (>20/HPF) in the epithelium. CD was therefore confirmed and a gluten-free diet was started. No specific treatment was given for EE. After 6 months of gluten-free diet we reassessed the child. Laboratory tests showed unvaried haemoglobin and ferritin levels with higher minimum condensed volume, an eosinophilic count of 500/μl, lowered tTG IgA (44 IU) and high total IgE (271 IU/ml, normal range 0–52). Endoscopy was repeated, which showed no signs of duodenal inflammation and a marked reduction in the number and size of white plaques. Oesophageal biopsy specimens showed few eosinophils (<5/HPF). Figures 1 and 2 show the endoscopic and histological aspects of the oesophagus before and after the gluten-free diet, respectively.

Figure 1

 Endoscopic and histological aspects of the oesophagus of patient 1 (7-year old boy) before the gluten-free diet. The arrow indicates the eosinophilic infiltrate.

Figure 2

 Endoscopic and histological aspects of the oesophagus of patient 1 (7-year old boy) after the gluten-free diet.

Case 2

A 17-year-old asymptomatic sister of a girl with CD, underwent endoscopy to perform duodenal biopsy. Hence she tested positive for EMAs and tTG, duodenal biopsy confirmed the diagnosis of CD (Marsch 3a atrophy). However, endoscopy showed vascular congestion in the proximal and middle oesophageal mucosa, with bleeding at contact with the instrument, and white linear plaques. Histological examination showed, dense eosinophilic infiltration (>20 cells/mm3). She had never experienced dysphagia, vomiting or refusal of food. There was no allergy to inhaled or alimentary allergens. Serum IgE level was normal (54 IU/l) and skin tests for dust mite, graminaceae and common alimentary allergens (lactoproteins, ovalbumin, grain, rice and soja) were negative. A gluten-free diet was prescribed, but the girl admitted to not have followed it properly. After 1 year, EMAs were still weakly positive with persistent mucosal atrophy in the duodenum. At macroscopical and histological examination, EE remained almost unvaried, but the girl still had no EE-suggestive symptoms.

Case 3

A 19-year-old asymptomatic sister of a girl with CD, with positive DQ2 HLA and positive EMAs and tTG antibodies, underwent endoscopy to confirm CD. Endoscopy did not show macroscopic duodenal inflammation (but duodenal biopsy showed lymphocytic infiltration and initial mucosal atrophy), whereas on distal oesophageal mucosa white small plaques were reported, with bioptical examination revealing lymphocytic and granulocytic infiltration in the distal oesophagus, with eosinophilic count varying from 50 to 100/HPF. CD and EE were therefore diagnosed. Skin tests for common inhaled and alimentary allergens were negative, whereas only the patch test for nickel sulphate was positive. She was then prescribed a gluten-free diet and omeprazole for 3 months. A further clinical, endoscopic, bioptical and laboratory assessment was performed 7 months later. The girl was still asymptomatic, and she admitted to properly following the diet (except for possible external contaminations). Nonetheless, EMA and tTG were still weakly positive. Endoscopic examination showed a granular aspect of the distal oesophageal mucosa but no other macroscopic alterations. Biopsy specimens revealed lymphocytic epithelial infiltration with slightly dismorphic villi of the duodenal mucosa and an important eosinophilic count (90/HPF) in the distal oesophageal mucosa. Skin tests were repeated and tested negative, as were the radioallergosorbent tests for specific allergens. The histopathological diagnosis was of CD in infiltrative stadium and active EE. As the girl was asymptomatic, no specific EE treatment was started.

Conclusions and comments

The pathogenetic mechanisms that lead to EE are far from being understood. However, among paediatric patients, a large proportion of cases seem to be associated with food allergies and to respond to a specific elimination diet.4 High levels of circulating eosinophilic cells and IgE are usually found in patients affected by CD, thus supporting the hypothesis of an atopic allergic reaction as a possible pathophysiological trigger. Moreover, EE is often associated with conditions such as allergic rhinitis, asthma, atopic dermatitis and familiar history of atopy.1,3 Unlike most of better-known atopic syndromes, it is usually very difficult to identify the cause leading to EE (as non-food items are often involved); therefore immune modulating therapy (topical steroids, leucotrienes) is generally preferred to avoidance strategies (elementary or elimination diets), at least in the absence of indicative prick or radioallergosorbent test.

This is the first report of the association between EE and CD. Regardless of the intrinsic limitations of any isolated reports, our three cases suggest that EE may be a gluten-associated manifestation in subjects with CD. Between 2000 and 2006, 7440 endoscopies were performed at the Institute of Child Health, IRCCS Burlo Garofolo, Trieste, Italy, and at the Surgical and Endoscopic Digestive Unit, Ospedale Paediatrics, Bambino Gesù, Rome, Italy, leading to the diagnosis of 1919 cases of CD and 33 cases of EE. The prevalence of CD among patients with EE was 3 of 33 (9%), nine times higher than that expected in the general population (1:100).5 By contrast, the prevalence of EE among patients with CD was 3 of 1919 (1:640, 0.0015%), a 3.5 times higher prevalence than that calculated in a recent American study (1:2380, 0.00042%).3 In only one of the three cases that we have described, EE disappeared after the gluten-free diet, suggesting that EE may be a gluten-dependent—even if rare—manifestation of CD. In the two other cases, the gluten-free diet did not have any effect on the eosinophilic infiltrate but both girls were not compliant to the gluten-free diet. The association between EE and CD may be plausible, thus suggesting that one performs a duodenal biopsy in all patients with EE in association with anti-tTG and EMAs dosage.

These three reports further underline how EE may be completely asymptomatic even when endoscopic and histological diagnostic features are evident. In our clinical experience, some patients treated with either topical steroids and/or elimination diets became quickly asymptomatic while their endoscopic and histological alterations persisted (data not published). The possibility that EE may be completely asymptomatic raises the still open questions on its natural history, the goals and the best duration of treatment of this emerging clinical condition.


View Abstract


  • Competing interests: None.

    Informed consent has been obtained for publication of the patients’ details in this letter.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.