Celiac disease (CD) and eosinophilic esophagitis (EE) are distinct disorders with specific clinicopathological characteristics. EE is defined as a primary clinico-pathological disorder of the esophagus characterized by esophageal or upper gastrointestinal symptoms in association with esophageal mucosal biopsy specimens containing more than 15 eosinophils per high-power field in 1 or more biopsy specimens and absence of pathological gastroesophageal reflux disease as evidenced by normal pH monitoring or lack of response to high-dose proton pump inhibitory therapy (1). EE is an emerging disorder that characteristically presents with acute or chronic dysphagia, food impaction, and gastroesophageal reflux symptoms unresponsive to acid suppression. Endoscopy may reveal longitudinal furrowing, circular rings, or white exudates in the esophagus. Although the true prevalence of EE is largely unknown, the incidence appears to be increasing in both the pediatric and adult populations (2–4). A recent study from our center demonstrated that the prevalence of EE in the pediatric population of Western Australia increased from 0.5/10,000 children in 1995 to 0.89/10,000 children in 2004 and suggested that this increase is real and not merely because of increased awareness and improved detection (4).
CD is an immune-mediated enteropathy induced by dietary gluten in genetically susceptible individuals. With the advent of highly sensitive serological markers, several epidemiological studies have shown that the prevalence is in the range of 1% of the populations tested (5). Probably less than 10% of those affected are being diagnosed (6). CD has an impressive list of associated disorders, which appears to be ever growing. Clinical presentation of CD has also changed over time, with many affected individuals being asymptomatic and identified by targeted screening (7).
In recent years, several children at our center who underwent upper gastrointestinal endoscopy for suspected CD, which was confirmed histologically, were also found to have endoscopic and histological evidence of EE. There are reports of coexistent CD and EE (8–12). Ooi et al (10) reviewed a cohort of 221 children with CD and found 3.2% of this cohort also had EE. We aimed to estimate the prevalence of EE in the cohort of Western Australian children diagnosed as having CD.
MATERIALS AND METHODS
Princess Margaret Hospital is the state pediatric referral center and the site of practice of all pediatric gastroenterologists in the state of Western Australia. The Department of Anatomic Pathology handles almost all of the pediatric gastrointestinal biopsy specimens from this state. All of the children with histologically confirmed CD, from January 2000 to November 2007 were identified by searching the laboratory database for systematized nomenclature of pathology codes for duodenal miscellaneous conditions (T6430M0002), duodenal inflammation (T6430M4000), and duodenal villus abnormality (T6430M7100). The histology reports were reviewed and all of the cases in which histology was compatible with CD were identified (numbers = 250). From these 250 cases, those in which a concurrent esophageal biopsy specimen had been obtained were identified (number = 121). Esophageal biopsies fulfilling the histological criteria for EE (1) were obtained and reviewed. The case records of children with both CD and EE were reviewed for demographic details, clinical symptoms, and endoscopic findings.
A total of 250 children were diagnosed with CD during the study period. Of these, 121 had concurrent esophageal biopsies available for examination. Of these esophageal biopsies, 101 (83%) were normal. Ten specimens showed mild nonspecific changes. Significant esophageal eosinophilia consistent with EE was present in 10 biopsies. None had unusually high eosinophils in their gastric or duodenal biopsies.
The prevalence of esophageal eosinophilia in children with CD who had concurrent esophageal biopsies was 8.2% (10 of 121) or 4% of all children diagnosed as having CD during study period (10 of total 250). Among the 10 children who had CD and esophageal eosinophilia, 6 were male and 4 were female. The male to female ratio for the cohort of 250 children with CD was 99:151. Clinical characteristics of the 10 children with CD and esophageal eosinophilia are given in Table 1.
Nine children underwent endoscopy for suspected CD, whereas 1 had endoscopy for food bolus obstruction. Endoscopic appearances suggestive of EE were seen in 7 children with the presence of linear furrowing or white exudates. Four children had undergone repeat endoscopic examinations to date. Duodenal mucosa had recovered on a gluten-free diet in all 4, but there was no resolution of esophageal eosinophilia.
One of these 4 children had treatment with proton pump inhibitor, which did not resolve esophageal eosinophilia. These 4 children subsequently had specific treatment of EE (2 swallowed topical fluticasone, 1 montelukast, and 1 dairy-free diet). Repeat esophageal biopsies showed resolution of esophageal eosinophilia in 3 children and improvement in the other.
Histological examination of the esophageal biopsies at diagnosis in children with esophageal eosinophilia demonstrated a median of 52 eosinophils per high-power field (range 23–80) with the presence of eosinophilic microabscesses in 5. All demonstrated some degree of basal hyperplasia. The number of esophageal eosinophils in 4 children who had repeat biopsies is shown in Table 1.
This study showed that the prevalence of EE is at least 4% in our cohort of children with CD. This is likely to be an underestimation because only 121 of the total 250 children with CD had concurrent esophageal biopsies. This prevalence is significantly higher than the background population prevalence of EE in Western Australia, which is estimated to be 0.89/10,000 children, suggesting a true association between these 2 entities.
Several authors have reported coexistent CD and EE (8–12). Ooi et al (10) reported the prevalence of EE to be 3.2% among 221 children with CD, all of whom had undergone esophageal biopsies. Our result shows a slightly higher prevalence of EE among children with CD, which may reflect the fact that EE is rapidly increasing in Western Australian children (4).
In 3 of 10 children with CD and EE, esophagus appeared normal at endoscopy, but biopsies showed evidence of EE. This finding is similar to the previous study from our center, which showed that approximately 30% of children diagnosed as having EE had normal-appearing esophageal mucosa at endoscopy (4).
It is interesting to note that the effect of a gluten-free diet on EE in children with CD and EE is varied in the reported studies. In our study, children who had undergone repeat endoscopic examinations showed recovery of their duodenal mucosa but no resolution of esophageal eosinophilia on a gluten-free diet alone. Ooi et al (10) reported similar findings with 2 of 7 children with CD and EE who underwent repeat endoscopy showing improved duodenal histology but persistent EE on gluten-free diet. Quaglietta et al (9), however, reported different outcomes in their cohort of 6 children with CD and EE who presented primarily with symptoms of EE. All 6 children improved symptomatically on gluten-free diet, and 3 children in whom a second endoscopy was performed showed significant reduction in intraesophageal eosinophils. Verzegnassi et al (8), in their 3 cases with CD and EE, reported resolution of EE on gluten-free diet in 1 child. In the other 2 (adults) EE was unchanged after gluten-free diet.
These different responses of EE to gluten-free diet may suggest different mechanisms of esophageal eosinophilia in CD. EE pathogenesis is likely to be associated with allergen sensitization, either ingested or inhaled, in predisposed individuals (13). In some patients with CD, esophageal eosinophilia could be the expression of CD, the eosinophilic infiltration simply representing a less common esophageal manifestation of CD, as postulated by Quaglietta et al in their group of patients (9). Studies have also shown that eosinophils may play a role in gliadin-induced intestinal damage (14). However, not all children with EE and CD show resolution of their EE on gluten-free diet alone, suggesting involvement of antigens other than gluten. Increased intestinal permeability in CD (15) may facilitate the exposure of the intestinal immune system to various antigens and transport of these antigens to various body sites leading to hypersensitivity reactions in genetically predisposed individuals. CD is a TH1-mediated and EE is a TH2-mediated response. Although studies have suggested an inverse association between TH1- and TH2-related disorders (16), recent large-scale population studies failed to show this inverse association. On the contrary, they have shown that autoimmune and atopic diseases share risk factors that increase the propensity of the immune system to generate both TH1- and TH2-mediated inappropriate responses to nonpathological antigens (17,18). Recently, Nilsson et al (19) analyzed cytokine patterns in children with TH1 and TH2 diseases including CD, type 1 diabetes mellitus, and allergy. They found that children suffering from a combination of TH1 and TH2 diseases responded with a high TH1- and TH2-like responses to a mitogen compared with those suffering from only TH1 disease who showed hardly any response to food antigens and allergens. These findings suggest that occurrence of CD and EE may be more than just a casual association because of a more generalized immune dysregulation.
We acknowledge several limitations in the study. Esophageal eosinophilia may occur in a wide variety of conditions including gastroesophageal reflux disease, EE, eosinophilic gastroenteritis, Crohn disease, connective tissue diseases, infections, and drug hypersensitivity (1). Our study cohort consisted of children with histologically confirmed CD who did not have evidence of infection, Crohn disease, or eosinophilic gastroenteritis. However, none had undergone 24-hour esophageal pH monitoring or impedance study to rule out gastroesophageal reflux as the cause or contributing factor. Peptic esophagitis is shown to have endoscopic and histological findings similar to that of EE (20). Demonstration of absence of pathological GOR as evidenced by normal pH monitoring or lack of response to high-dose PPI medication is necessary before diagnosing eosinophilic esophagitis (1). In our study cohort, only 1 child had a trial of PPI, in whom the esophageal eosinophilia persisted. Two other children had resolution of esophageal eosinophilia on swallowed fluticasone, thus ruling out reflux as the cause.
The study is retrospective in nature with inherent variability in investigations and management of patients. Only 121 of 250 children with CD had concurrent esophageal biopsies at the time of diagnosis of CD. This is likely to underestimate the prevalence of EE in CD because it has been shown that esophagus may appear normal at endoscopy in some cases of EE. With increasing recognition of possible association of CD and EE there has been an increase in the number of esophageal biopsies being performed in children. Only 4 children in our cohort of 10 had undergone a repeat endoscopic assessment. Although some suggest improvement of EE on a gluten-free diet (9), this is not our experience. However, the numbers are too small to draw any conclusions.
Despite these limitations, our study suggests that the prevalence of EE in Western Australian children with CD is at least 4% and possibly higher. With the background prevalence of EE in Western Australia being 0.89/10,000 children, this suggests a true association between these 2 entities rather than a mere coincidence. Coexistence of EE needs to be considered in children with CD. This study highlights the importance of obtaining routine esophageal biopsies in children who undergo endoscopy for the diagnosis of CD irrespective of whether the esophagus appears normal or abnormal at endoscopy.
1. Furuta GT, Liacouras CA, Collins MH, Gupta SK, et al
. Eosinophilic Esophagitis
in Children and Adults: a systematic review and consensus recommendations for diagnosis and treatment. Gastroenterology 2007; 133:1342–1363.
2. Straumann A, Simon HU. Eosinophilic esophagitis
: escalating epidemiology? J Allergy Clin immunol 2005; 115:418–419.
3. Noel RJ, Putnam PE, Rothenburg ME. Eosinophilic esophagitis
. N Engl J Med 2004; 351:940.
4. Cherian S, Smith NM, Forbes DA. Rapidly increasing prevalence of eosinophilic oesophagitis in Western Australia. Arch Dis Child 2006; 91:1000–1004.
5. Rodrigues AF, Jenkins HR. Investigation and management of coeliac disease. Arch Dis Child 2008; 93:251–254.
6. Ravikumara M, Nootigattu VKT, Sandhu BK. Ninety percent of coeliac disease is being missed. J Pediatr Gastroenteorl Nutr 2007; 45:497–499.
7. Ravikumara M, Tuthill DP, Jenkins HR. The changing clinical presentation of coeliac disease. Arch Dis Child 2006; 91:969–971.
8. Verzegnassi F, Bua J, De Angelis P, et al
. Eosinophilic oesophagitis and coeliac disease: is it just a casual association
? Gut 2007; 56:1029–1030.
9. Quaglietta L, Coccorullo P, Miele E, et al
. Eosinophilic oesophagitis and coeliac disease: is there an association
? Aliment Pharmacol Ther 2007; 26:487–493.
10. Ooi CY, Day AS, Jackson R, et al
. Eosinophilic esophagits in children with coeliac disease. J Gastroenterol Hepatol 2008; 23:1144–1148.
11. Shah AA, McGreal N, Li B, et al. Coeliac disease in association
with eosinophilic oesophagitis: case series of 6 patients from 2 centres. J Pediatr Gastroenteorl Nutr
12. Kagalwalla AF, Shah A, Ritz S, et al
. Cow's milk protein-induced eosinophilic oesophagitis in a child with gluten sensitive enteropathy. J Pediatr Gastroenteorl Nutr 2007; 44:386–388.
13. Blanchard C, Rothenberg ME. Basic pathogenesis of eosinophilic esophagitis
. Gastrointest Endoscopy Clin N Am 2008; 18:133–143.
14. Colombel JF, Torpier G, Janin A, et al
. Activated eosinophils in adult celiac disease
: evidence for a local release of major basic protein. Gut 1992; 33:1190–1194.
15. Fasano A, Shea-Donohue T. Mechanisms of disease: the role of intestinal barrier function in the pathogenesis of gastrointestinal autoimmune diseases. Nat Clin Pract Gastroenterol Hepatol 2005; 2:416–422.
16. Rosenbauer J, Herzig P, Giani G. Atopic eczema in early childhood could be protective against type 1 diabetes. Diabetologia 2003; 46:784–788.
17. Simpson CR, Anderson WJ, Helms PJ, et al
. Coincidence of immune mediated diseases driven by Th1 and Th2 subsets suggests a common aetiology. A population based study using computerised general practice data. Clin Exp Allergy 2002; 32:37–42.
18. Sheikh A, Smeeth L, Hubbard R. There is no evidence of an inverse relationship between Th2 mediated atopy and Th1 mediated autoimmune disorders: lack of support for hygiene hypothesis. J Allergy Clin Immunol 2003; 111:131–135.
19. Nilsson L, Kivling A, Jalmelid M, et al
. Combinations of common chronic paediatric diseases deviate the immune response in diverging directions. Clin Exp Immunol 2006; 146:433–442.
20. Ngo P, Furuta GT, Antonioli DA, et al
. Eosinophils in the esophagus—peptic or allergic eosinophilic esophagitis
? Case series of three patients with eophageal eosinophilia. Am J Gastroenterol 2006; 101:1666–1670.