Sundaram, Shikha*; Sunku, Bhanu*; Nelson, Suzanne P.*; Sentongo, Timothy*; Melin-Aldana, Hector†; Kumar, Raj‡; Li, B U.K.*
*Division of Gastroenterology, Hepatology and Nutrition, †Division of Pathology, and ‡Division of Allergy and Immunology, Children's Memorial Hospital, Chicago, Illinois, U.S.A.
Received January 15, 2003;
revised September 23, 2003; accepted October 3, 2003.
Address correspondence and reprint requests to Dr. Shikha Sundaram, Division of Gastroenterology, Hepatology and Nutrition, Children's Memorial Hospital, 2300 Children's Plaza, Box 57, Chicago, IL 60614 (e-mail: firstname.lastname@example.org).
Eosinophilic esophagitis (EE) is increasingly recognized as a clinical entity by pediatric gastroenterologists. Although originally confused with severe gastroesophageal reflux disease (1), EE may be a condition truly distinct from gastroesophageal reflux disease based on its poor response to antireflux therapy, normal 24-hour pH studies, higher density of mucosal eosinophils (≥20 per high-powered field [hpf]), and histologic improvement on an elemental diet or topical steroids (2,3). Patients present with many symptoms, including abdominal pain, dysphagia, vomiting, and weight loss (4). The endoscopic appearance of EE is also varied. The gross appearance of the esophagus in EE has been described as shiny, granular and with vertical furrowing and concentric rings (4–6). We report 13 children presenting with a variety of gastrointestinal complaints and who at esophagogastroduodenoscopy (EGD) had an under-recognized gross finding of adherent white plaques on the esophageal wall suggestive of a candidal infection. None of these children had fungal esophagitis, and all had the histologic findings of EE.
The following two cases highlight typical presentations of EE seen in our patient population. Eleven additional cases are summarized in Table 1.
A.O., a 2½-year-old boy, experienced nonbilious, occasionally blood-tinged, postprandial emesis for 2 months. He had a poor appetite and poor weight gain, despite supplemental feedings of PediaSure (Ross Laboratories, Columbus, OH). He had no abdominal pain or diarrhea. He had experienced eczema, asthma, and a skin rash associated with the ingestion of milk and orange juice. He was an active, playful child with height between the 25th and 50th percentile and weight between the 50th and 75th percentile. He had diffuse, end expiratory wheezing. Results of abdominal examination were unremarkable. On EGD, linear (longitudinal axis) furrowing and adherent white plaques were seen throughout the esophagus. The histologic findings were consistent with EE.
B.Z., a 5½-year-old boy had had poor weight gain for several years. He was anorectic and calorie counts confirmed inadequate nutritional intake. He had never had nausea, vomiting, abdominal pain, or diarrhea. He had atopic eczema, asthma, and multiple food allergies. Radioallergosorbent test (RAST) results were positive for antibodies to nuts, nutmeg, dairy, wheat, and sesame seeds. His height and weight were both less than the 5th percentile. He had digital clubbing. Results of abdominal examination were normal. EGD revealed edema, linear (longitudinal axis) fissuring, and adherent white plaques in the esophagus. The histologic findings were consistent with EE.
Each patient had obvious adherent white plaques (0.5–1.0 mm) in the esophagus at the time of EGD (Fig. 1 A, B). In addition, each patient had at least some of the other visual findings previously described in EE, including linear fissuring/furrowing, nodularity, erythema, and edema (Table 2). Each patient met the histologic criteria for EE (i.e., ≥20 eosinophils/hpf) (Table 3, Fig. 2 A, B). Eight of 13 patients had eosinophilic microabscesses, defined histologically as four or more contiguous mucosal eosinophils (7).
No patient was evaluated for gastroesophageal reflux using pH probe testing. Rather, patients were treated presumptively for reflux with a proton pump inhibitor. A lack of response to aggressive therapy was used as the clinical indication to perform endoscopy. Most of our patients were treated with a proton pump inhibitor at a dose of 1 to 2 mg/kg/day for 4 to 6 weeks before endoscopy. We thought this approach would improve histologic changes that may have resulted from mild to moderate reflux, which could have confounded the clinicopathologic diagnosis.
At the time of writing, 6 of 13 children had been treated with an elimination diet. RAST was performed for the most common antigens implicated in food allergies (cow's milk, soy, egg white, peanut, and wheat), along with any other foods identified by the parents as related to symptoms. Four of the children's elimination diets were directed by RAST, and the remaining two were based on the five most common antigens. Three of these children had both symptomatic and histologic improvement of their EE. New foods were methodically reintroduced, starting with the least allergenic. Initially rice; yellow, green, and orange vegetables; and selected fruits (apples, pears, bananas) were tried. Next, meats and fish (excluding shellfish) and grain, and finally the foods with high allergic potential, such as eggs, tomatoes, and dairy were introduced. New foods were tried for 5 days before introduction of another food. Foods identified on RAST as allergic were eliminated completely. In three patients, the elimination diet failed, and the patients were switched to an elemental diet. Four children were initially started on an elemental diet. Three had complete resolution of clinical and histologic symptoms, and the other is awaiting follow-up. Three children are currently being considered for elimination diets. Table 4 provides the details of each patient's treatment plan and clinical course.
We report 13 pediatric patients with adherent white plaques in the esophagus associated with the diagnosis of EE. Eight of these patients also had eosinophilic microabscesses on histopathologic analysis. All children seen in follow-up have had clinical improvement on either elimination or elemental diets. In addition, patients who have had follow-up biopsies after treatment have shown disappearance of the adherent white plaques and histologic improvement.
The histologic criteria that define EE have not been established. Kelly et al. (2) described a group of patients with 15 to 100 eosinophils/hpf (median, 41) in the distal esophagus who improved on elemental diets. Liacouras et al. (3) found patients with gastroesophageal reflux have 2.3 ± 1.2 eosinophils/hpf compared with 34.2 ± 9.6 eosinophils/hpf in patients with EE. Orenstein et al. (4) reported a group of children with at least 20 eosinophils/hpf and indicated that those with 5 to 20 eosinophils/hpf should be considered in the EE category. Finally, Walsh et al. (7) described two groups of patients with 31 ± 19.5 eosinophils/hpf and 28 ± 23.7 eosinophils/hpf, as having EE if they also had incomplete response to antireflux therapy and clinical improvement with antiallergy therapy. All of our patients meet the most stringent criteria for EE proposed by these authors.
A variety of visual findings may be observed in patients with EE. Some children meeting the histopathologic criteria of EE have a grossly normal appearing esophagus. Some have nonspecific erythema, friability, or ulcerations. More characteristically, the mucosa may appear granular, occasionally with circumferential creases, rather than its normal smooth, shiny appearance (4). Fox et al. (5) describe longitudinal furrowed lines in the distal esophageal mucosa in patients with moderate to severe EE. Rings, webs, and a corrugated appearance attributable to repetitive rings or webs, also have been reported (6). Several previous reports of patients with EE, without specifically describing white plaques, do note a negative or inadequate evaluation for candidal infections, thus implying that white plaques may have been present (8–10). Ahmed (11) and Lim et al. (12) described adherent white plaques in the esophageal mucosa of patients with EE. The adherent white plaques seen in our patients may be present in a few discrete areas or more diffusely. Most of our patients had eosinophilic micro-abscesses of the esophageal mucosa on histologic analysis (Fig. 2B). We speculate that the adherent white plaques may represent the visual correlate of eosinophilic abscesses.
It is crucial that the adherent white plaques of EE not be confused with candidal esophagitis. The absence of yeast must be demonstrated, either by a brushed potassium hydroxide preparation or by specific histologic staining. Brito and Barbosa (13) described a patient with intermittent dysphagia and partial stenosis of the gastroesophageal junction who had esophageal candidiasis and an associated heavy infiltration of the esophagus with eosinophils. This patient's symptoms resolved after antifungal therapy, but the eosinophilic infiltration remained. (13). This example underscores the importance of excluding a superimposed or primary fungal infection. In each of our patients, fungal infection was systematically excluded by negative fungal histologic findings and often supported by the addition of a negative potassium hydroxide preparation.
EE has been classified as a mixed IgE and non-IgE mediated disorder. Total serum IgE levels may be normal or slightly elevated, but peripheral eosinophilia is uncommon (14). All of our patients in whom serum IgE was measured had elevated levels, but only a few had peripheral eosinophilia. Similar to other reports, we found an association between the eosinophilic gastrointestinal disorder and other allergic conditions, such as eczema and asthma (15). The presumption is that eosinophilic infiltration of the esophageal mucosa represents an adverse immunologic response. Eosinophils may serve as links between the immune system and inflammatory effector cells. IL-5, a cytokine secreted by lymphocytes, attracts and activates eosinophils (15). T cells from patients with eosinophilic gastroenteritis secrete more IL-5 than those of control subjects (16), although the exact immunopathogenesis is poorly understood.
Many patients with EE are initially thought to have gastroesophageal reflux disease. Because of the variability in clinical presentation, the diagnosis of EE often is not suspected before EGD. Although white exudate has been described in association with severe reflux, this report did not consider the possibility of EE. Thus, an evaluation of allergic gastrointestinal disorders, including eosinophil count, was not performed (17). Our patients were treated presumptively with a proton pump inhibitor before endoscopy. Lack of response to therapy was used as a clinical indication to evaluate endoscopically for refractory gastroesophageal reflux or allergic EE. After the diagnosis of EE was made, if precipitating foods could be identified, patients began an elimination diet. If precipitants could not be identified or an elimination diet failed, an elemental diet, such as EO-28 or Neocate One + (Scientific Hospital Supplies, Inc.) (2), was begun. These amino acid-based formulas were used to remove all potential food antigens and induced prompt remission of disease. Once symptoms improved, typically after 6 weeks, another EGD was done to assess visual and histologic changes. In our series, RAST testing proved useful in several patients in designing an elimination diet that excluded identified allergens. Many patients were allergic to common antigens (milk, soy, eggs, and wheat), whereas a few manifested unusual food allergies. None of our patients were treated with topical or systemic steroid therapy.
In summary, we have described an under-recognized endoscopic finding in EE, adherent white plaques in the esophagus, that resembles candidal esophagitis. This visual finding should prompt clinicians and pathologists to seriously consider EE and evaluate for fungal infections. Eosinophilic abscesses or collections may account for this visual finding. The prognostic implications of this finding require further investigation.
1. Winter HS, Madera JL, Stafford RJ, et al. Intraepithelial eosinophils: a new diagnostic criterion for reflux esophagitis. Gastroenterology 1982; 83:818–23.
2. Kelly KJ, Lazenby AJ, Rowe PC, et al. Eosinophilic esophagitis attributed to gastroesophageal reflux: improvement with an amino acid-based formula. Gastroenterology 1995; 109:1503–12.
3. Liacouras CA, Wenner WJ, Brown K, et al. Primary eosinophilic esophagitis in children: successful treatment with oral corticosteroids. J Pediatr Gastroenterol Nutr 1998; 26:380–5.
4. Orenstein SR, Shalaby TM, Di Lorenzo C, et al. The spectrum of pediatric eosinophilic esophagitis beyond infancy: a clinical series of 30 children. Am J Gastroenterol 2000; 95:1422–30.
5. Fox VL, Nurko S, et al. Eosinophilic esophagitis: it's not just kid's stuff. Gastrointest Endosc 2002; 56( 2):260–70.
6. Siafakas CG, Ryan CK, et al. Multiple esophageal rings: an association with eosinophilic esophagitis. Am J Gastroenterol 2000; 95( 6):1572–5.
7. Walsh SV, Antonioli DA, Goldman H, et al. Allergic esophagitis in children: a clinicopathologic entity. Am J Surg Pathol 1999; 23:390–6.
8. Straumann A, Bauer M, et al. Idiopathic eosinophilic esophagitis is associated with a TH-2 type allergic inflammatory response. J Allergy Clin Immunol 2001; 108:954–61.
9. De Agustin JC, Sanz N, et al. Successful medical treatment of two patients with eosinophilic oesophagits. J Pediatr Surgery 2002; 37:207–13.
10. Picus D, Frank PH. Eosinophilic esophagitis. Am J Roentegenol 1981: 136:1001.
11. Ahmed A. A novel endoscopic appearance of idiopathic eosinophilic esophagitis. Endoscopy 2000: 32( 6):S33.
12. Lim JR, Gupta SK, et al. White specks in esophageal mucosa (WSEM): a true endoscopic manifestation of severe eosinophilic esophagitis (EE) in children. J Pediatr Gastroenterol Nutr 2001; 33:411–2.
13. Brito EM, Barbosa AJA. Symptomatic eosinophilic esophagitis and esophageal candidiasis. Am J Gastroenterol 1998; 93:2003–4.
14. Sampson HA, Anderson JA. Summary and recommendations: classification of gastrointestinal manifestations due to immunologic reactions to foods in infants and young children. J Pediatr Gastroenterol Nutr 2000; 30:5287–94.
15. Kelly KJ. Eosinophilic gastroenteritis. J Pediatr Gastroenterol Nutr 2000; 30:S28–35.
16. Jaffe J, James S, et al. Evidence for an abnormal profile of interleukin-4 (IL-4), IL-5 and gamma interferon in peripheral blood T cells from patients with allergic eosinophilic gastroenteritis. J Clin Immunol 1994; 14:299–309.
17. Hatleback JG, Berstad A. Endoscopic grading of reflux oesophagitis: what observations correlate with gastro-oesophageal reflux? Scand J Gastroenterol 1997: 32:760–5.
© 2004 Lippincott Williams & Wilkins, Inc.