*Division of Pediatric Gastroenterology and Nutrition and †Department of Pathology, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada; ‡Division of Pediatric Gastroenterology, Sydney Children's Hospital, Sydney, Australia; §Division of Pediatric Gastroenterology and Nutrition, IWK Health Centre, Department of Pediatrics, Dalhousie University, Halifax, Nova Scotia, Canada
Received June 19, 2003; accepted September 22, 2004.
Address correspondence and reprint requests to Dr. Ahmed, Division of GI/Nutrition, IWK Health Centre, 5850/5980 University Ave, PO Box 3070, Halifax, NS B3J 3G9. (e-mail: email@example.com).
Eosinophilic gastroenteritis (EG) is a chronic inflammatory disorder of the gastrointestinal tract characterized by the infiltration of eosinophils into the mucosa (1). Characteristic symptoms and a peripheral eosinophilia suggest the diagnosis. Confirmation of the diagnosis hinges on the finding of eosinophilic infiltration into one or more areas of the gastrointestinal tract in the absence of parasitic or extraintestinal disease (2).
The spectrum of disease varies according to the severity and extent of eosinophilic infiltration (3). The disease can be localized or generalized. In 1970, Klein et al (4), proposed a classification of EG based on histologic evidence of eosinophilic infiltration. The disease can be classified as mucosal, muscular or serosal based on the depth of eosinophilic infiltration. There is some correlation between the histologic findings and clinical presentation but this is not consistent. However, mucosal disease tends to present with abdominal pain, diarrhea or hypoalbuminemia. Failure to thrive and anemia may also be present. Colonic involvement may result in hematochezia. Muscular disease seems to differ from mucosal disease by causing obstructive symptoms, and serosal disease may cause eosinophilic ascites (5). There are limited descriptions of the endoscopic findings in this condition. In this report we describe two patients with the dramatic endoscopic finding of extensive gastric pseudopolyps.
Patient 1 is a 14-year-old boy who presented with an 18-month history of intermittent periumbilical abdominal pain. He had episodes of pain lasting several hours 2-3 times a week. There was no associated diarrhea, nausea or vomiting. There was no history of mucus or blood in the stools, and there were no extraintestinal symptoms. His growth and weight gain had not been impaired despite the elimination of numerous foods in an effort to control symptoms. There was no family history of dyspepsia, celiac disease, inflammatory bowel disease or atopy.
His initial investigations included a complete blood count that showed a hemoglobin of 140 g/L, a white blood cell count of 10 × 109/L with an eosinophil count of 3 × 109/L (normal range, 0.02 to 0.5 × 109/L), platelet count of 238 × 109/L, serum albumin of 27 g/L (normal range, 33-55 g/L) and normal transaminase levels. An anti-gliadin antibody test was positive at 1100 for immunoglobulin G, and an immunoglobulin A tissue-transglutaminase was negative with normal immune globulin levels. An abdominal ultrasound and upper gastrointestinal series done before referral were normal.
Based on the presence of the positive anti-gliadin antibody test, the patient was put on a trial of a gluten-free diet by the referring physician with partial response. Subsequent episodes of severe abdominal pain occurred and he was referred to our center for further investigations.
At the time of assessment, his weight was 57.2 kg (50%) and his height was 166.9 cm (50%). He appeared clinically well. He had no digital clubbing. His chest and cardiac examinations were normal. His abdomen was soft and non-tender with no hepatosplenomegaly. We intended to repeat his celiac antibody panel on a gluten containing diet. However, before reintroducing gluten, an anti-endomysial antibody was weakly positive with a titer of 1:40. An upper endoscopy was performed. At endoscopy, subtle white plaques were seen in his esophagus. Numerous polypoid lesions were seen in the antrum arranged in a radial pattern (Fig. 1). There was no gastric ulceration. The duodenum appeared mildly erythematous. A repeat anti-endomysial antibody on the day of the endoscopy was negative.
Histopathology of the antral mucosa showed diffuse eosinophilic infiltrates forming aggregates in the lamina propria (Fig. 2). Biopsies of the antral polypoid lesions showed hyperplastic foveolae, microcyst formation, inflammatory cells in the stroma and an increase in stroma disproportionate to the amount of epithelium (Fig. 3). Staining for Helicobacter pylori and culture of the antral biopsies were negative. Duodenal biopsies showed mild eosinophilic infiltration with normal villous architecture.
A diagnosis of EG was made and ketotifen (6) was given without improvement. Corticosteroid therapy was then initiated at a dose of 1 mg · kg−1 · day−1. Within 1 month of starting oral prednisone, his symptoms improved dramatically. He was able to eat a variety of foods including those that he previously could not tolerate. Two months after starting prednisone, a repeat upper endoscopy was performed. A limited colonoscopy was also performed to assess any colonic involvement. The upper endoscopy showed a decrease in the number of antral lesions with decreased nodularity. Histologic examination showed focal mild to moderate eosinophilic gastritis and normal villous architecture with mild eosinophilic duodenitis. Antral biopsies suggested the presence of a retention polyp with small eosinophilic infiltrates in the lamina propria. The colon appeared normal grossly and no polypoid changes were seen. Laboratory tests showed an improvement in the serum albumin from 27 to 44 g/L. The white blood cell count was 16.7 × 109/L, with a decrease in the peripheral eosinophil count to 0.84 × 109/L. The patient initially successfully tolerated weaning of his corticosteroid therapy but has since had an increase in symptoms requiring reinitiation of this therapeutic modality. He subsequently became steroid dependent despite attempts to introduce other therapies such as montelukast. He was ultimately successfully weaned off prednisone with the introduction of azathioprine as a steroid-sparing agent.
Patient 2 is a 16-year-old boy referred to our clinic with a 6-year history of poor weight gain and growth failure. His main symptom was abdominal discomfort associated with the ingestion of dairy products. He denied diarrhea or constipation. More recently, he had noted reduced energy and fatigue. Medical history included hemoglobin E trait and asthma in infancy. There was no other history of atopy. His medications included calcium supplements and he was on a lactose-free diet.
On examination, his weight was 31 kg (<3%) and his height 143.5-cm (<3%). At the age of 16 years, his pubertal development was Tanner stage 1. His physical examination revealed no pathologic findings other than growth retardation. A complete blood count showed a white cell count of 13.5 × 109/L with an elevated eosinophil count of 4.86 × 109/L, the hemoglobin was 125 g/L, platelet count was 499 × 109/L and the ferritin level was 9.4 μg/L (normal range, 22 to 400 μg/L). Antiendomysial antibody was negative (titer of 1:10) with normal immune globulins. Inflammatory markers were negative. His bone age was delayed more than 2 years.
Upper endoscopy showed erythema in the distal esophagus. Multiple polypoid lesions were noted in the stomach with the most striking findings in the antrum. There was no mucosal ulceration (Fig. 4). The duodenal mucosa appeared erythematous with no ulceration. A sigmoidoscopy showed mucosa with a normal vascular pattern and a nodular appearance. Histopathologic examination of the biopsy specimens showed superficial eosinophilic infiltration in the esophageal mucosa consistent with eosinophilic esophagitis. The antral mucosa showed large aggregates of eosinophils with regenerative changes. Eosinophils were not seen in the duodenum, and the villous architecture appeared normal. Stains for Helicobacter pylori and culture of the antral biopsy were negative.
Based on these findings, a diagnosis of EG was made and the patient was started on corticosteroid therapy at 1 mg · kg−1 · day−1. After 2 months, the patient's clinical condition significantly improved with a weight gain of 4 kg. Laboratory tests showed a decrease in the peripheral eosinophilia (eosinophil count 0.25 × 109/L) and an improvement in the serum albumin to 37 g/L. Attempts to wean the prednisone resulted in exacerbation of symptoms. However, the patient was subsequently successfully weaned off prednisone and is currently maintained on ketotifen and omeprazole.
EG was first described by Kaisjer in 1937 (6). Since then, multiple case series have been published (3,7,8). Clinical presentations include abdominal pain, vomiting and dysphagia (7). Others (10) have reported dyspepsia as the primary presenting symptom. In children, EG often presents with peripheral eosinophilia, as in our two patients (8). Rarely, surgical complications such as obstruction, perforation, bleeding and fistulae occur (9). These complications are most often seen in patients with muscular disease.
Histologic reports of eosinophilic gastroenteritis have mentioned the presence of scattered mucosal blebs and nodularity but never the extensive pseudopolyp formation seen in our patients. Steffen et al. (8) reported pseudopolyp formation in the antrum in one child from a series of six children with EG. In our patients, it appeared that despite clinical and biochemical improvement in symptoms, the endoscopic appearance did not change dramatically after treatment.
The pathogenesis of eosinophilic gastroenteritis is not fully understood. Recent data (10) suggest that eosinophil recruitment and activation is induced by three main cytokines: interleukin 3, granulocyte-macrophage colony-stimulating factor and interleukin 5. This appears to be a local effect because serum interleukin levels are reportedly normal. These cytokines have the ability to prolong the survival of mature eosinophils and therefore perpetuate disease activity (11). The polypoid appearance seen at endoscopy is related to the extensive eosinophilic infiltration and inflammation present in this affected tissue.
Therapeutic options for EG include inhaled (12) and systemic (13) corticosteroids and antihistamines such as ketotifen, which stabilize mast cells (14). Ketotifen is believed to act on both histamine release and leukotriene production. More recently, leukotriene receptor antagonists such as montelukast (15) have been tried in isolated cases. Patient 1 was tried on all of these medications but remained steroid dependent. Similarly, patient 2 had a recurrence of symptoms and hypoalbuminemia when the prednisone dose was initially decreased.
The spectrum of disease in eosinophilic gastroenteritis is becoming more evident as our understanding of this condition develops. Our cases illustrate a dramatic endoscopic appearance with extensive gastric pseudopolyp formation. The findings at endoscopy were a result of the severity of eosinophilic infiltration in the antral mucosa. Chronic eosinophil accumulation and the accompanying inflammatory response led to the gross changes that were visualized as pseudopolyps. Interestingly, it appears that the endoscopic findings do not resolve as quickly as the clinical symptoms.
In conclusion, EG can present with a wide variety of clinical manifestations and endoscopic findings. EG should be included in the differential diagnosis of gastric polyps.
1. Kelly KJ. Eosinophilic gastroenteritis. J Pediatr Gastroenterol Nutr 2000;30 Suppl:S28-35.
2. Cello JP. Eosinophilic gastroenteritis: a complex disease entity. Am J Med 1979;67:1097-104.
3. Guajardo JR, Plotnick LM, Fende JM, Collins MH, Putnam PE, Rothenberg ME. Eosinophil-associated gastrointestinal disorders: a world-wide-web based registry. J Pediatr 2002;141:576-81.
4. Klein NC, Hargrove RL, Sleisenger MH, Jeffries GH. Eosinophilic gastroenteritis. Medicine (Baltimore) 1970;49:299-319.
5. Kravis LP, South MA, Rosenlund ML. Eosinophilic gastroenteritis in the pediatric patient. Clin Pediatr (Phila) 1982;21:713-7.
6. Kaijser R. Kenntnis del Allergischen Affektionen des verdauungskanals vom Standpunkt des Chirurgan aus. Arch Klin Chir 1937;188:36-64.
7. 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.
8. Steffen RM, Wyllie R, Petras RE, Caulfield ME, Michener WM, Firor HV, et al. The spectrum of eosinophilic gastroenteritis. Report of six pediatric cases and review of the literature. Clin Pediatr (Phila) 1991;30:404-11.
9. Hoefer RA, Ziegler MM, Koop CE, Schnaufer L. Surgical manifestations of eosinophilic gastroenteritis in the pediatric patient. J Pediatr Surg 1977;12:955-62.
10. Desreumaux P, Bloget F, Seguy D, et al. Interleukin 3, granulocyte-macrophage colony-stimulating factor, and interleukin 5 in eosinophilic gastroenteritis. Gastroenterology 1996;110:768-74.
11. Levy AM, Kita K. The eosinophil in gut inflammation: effector or director? Gastroenterology 1996;110:952-4.
12. Faubion WA Jr., Perrault J, Burgart LJ, Zein NN, Clawson M, Freese DK. Treatment of eosinophilic esophagitis with inhaled corticosteroids. J Pediatr Gastroenterol Nutr 1998;27:90-3.
13. Khan S, Orenstein SR. Eosinophilic gastroenteritis: epidemiology, diagnosis and management. Paediatr Drugs 2002;4:563-70.
14. Melamed I, Feanny SJ, Sherman PM, Roifman CM. Benefit of ketotifen in patients with eosinophilic gastroenteritis. Am J Med 1991;90:310-4.
15. Schwartz DA, Pardi DS, Murray JA. Use of montelukast as steroidsparing agent for recurrent eosinophilic gastroenteritis. Dig Dis Sci 2001;46:1787-90.
© 2005 Lippincott Williams & Wilkins, Inc.