Eosinophilic Gastritis With Gastric Outlet Obstruction Mimicking Infantile Hypertrophic Pyloric Stenosis

PATIENT 1 A 1-month-old baby presented with persistent vomiting since birth. Because of his mother's misbelief that breast-feeding is associated with vomiting, breast-feeding was changed to cow's milk formula in just a few days after birth. Nevertheless, vomiting became worse and body weight did not increase satisfactorily (25th–50th percentile at birth; 10th–25th percentile 1 month later), which led the patient to be admitted to the hospital. Physical examination was unremarkable at admission. To exclude infantile hypertrophic pyloric stenosis (IHPS), ultrasonography (USG) was performed, demonstrating segmental concentric layered wall thickening in the gastric antrum and intact pylorus and duodenal bulb (Fig. 1A). Following the upper gastrointestinal (GI) series (UGI) showed severe gastric outlet obstruction and little passage of contrast to duodenum on 6 hours’ delayed image (Fig. 1B). Peripheral eosinophil count and serum IgE level were not elevated. Esophagogastroduodenoscopy (EGD) revealed prominent antral swelling with pyloric narrowing. Biopsy was performed and feeding tube (5 Fr, 500 mm, JMS-K, Hankuk Medical Supply, Republic of Korea) was inserted postpylorus. Histopathologic examination showed eosinophilic infiltration (>20 eosinophils/HPF) in the antrum and pylorus (Fig. 1C). Final diagnosis was eosinophilic gastritis.FIGURE 1: A 1-month-old baby with eosinophilic gastritis. A, ultrasonography shows prominent antral swelling (5-mm thick, 20-mm length) (dotted line) and intact pylorus (arrow). Distally, normal duodenal bulb is visible. B, upper gastrointestinal series shows gastric outlet obstruction. C, Before treatment, gastric mucosa with eosinophil (arrow) infiltration (22 eosinophils/HPF) is found in gastric antral mucosa. After steroid treatment, eosinophilic infiltration is improved (not shown).This patient was fed with amino acid–based formula by transpyloric feeding tube, treated with corticosteroid (Solumedrol [methyl prednisolone sodium succinate], through IV side), administered for a duration of 4 to 8 days, depending on the patient's symptom for five times and then followed by USG, UGI, and EGD. Oral ketotifen was added. Gastric wall thickening had hardly improved on follow-up USG, but barium passage had improved gradually. Transpyloric feeding tube was removed after 3 months. EGD performed 3 days later showed resolved antral swelling. On the last follow-up, the patient was about 1-year old with body weight in the 75th to 90th percentile, and GI symptoms did not relapse. PATIENT 2 A 3-week-old baby was presented with persistent projectile vomiting and loose stool for 1 week during breast-feeding. On admission, peripheral eosinophil count was in normal range and physical examination was unremarkable. USG revealed gastric wall thickening, especially in the middle hypoechoic layer, representative of muscularis propria. EGD revealed antral swelling, which caused stenosis and transpyloric tube insertion, and biopsy at the antrum was performed. The pathologic finding was chronic gastritis with eosinophilic infiltration (18 eosinophils/HPF) in antrum. So clinician and pathologist and radiologist made a diagnosis as eosinophilic gastritis. Treatment with solumedrol (methyl prednisolone sodium succinate, through IV side) for 3 days and transpyloric feeding by amino acid–based formula started. Parenteral nutritional support was also added. About 3 weeks later, USG showed improvement in gastric wall thickening. Four months later, the feeding tube was removed. The patient had no symptoms and grown well with ideal body weight by 7 months of age. PATIENT 3 A 7-month-old boy was presented with projectile vomiting for 1 week after the induction of soybean formula. Peripheral eosinophil count and serum IgE level was slightly elevated (709 cells per μL and 43 IU/mL, respectively); however, there was no evidence of parasite infestation or other reason of hypereosinophilia. In UGI, contrast agent did not pass through the pylorus. Endoscopy showed significant antral swelling with pylorus obstruction (Fig. 2). Histologically, there was eosinophilic infiltration (16 eosinophils/HPF).FIGURE 2: A 7-month-old male patient. A, Initial endoscopy revealed pyloric obstruction. B–D, Antral swelling had resolved progressively after solumedrol and hypoallergenic formula feeding. Follow-up endoscopic images obtained 3 weeks, 3 months, and 10 months after treatment, respectively.Total parenteral nutrition with IV prednisolone therapy started. Three days later, vomiting subsided and following EGD showed improved antral swelling and approximately 2 mm of pyloric canal opening. Water intake was permitted along with hypoallergenic formula feeding, and oral ketotifen was started. The amount of milk feeding increased and vomiting did not relapse. Prednisolne was tapered after approximately 1 month. On the last follow-up by 16 months of age, the patient showed normal growth. DISCUSSION Eosinophilic gastroenteritis (EGE), which is characterized by diffuse eosinophilic infiltration in the bowel, is a rare disorder; however, there are no standards for making histologic diagnosis of EGE (1). Talley et al (2) had proposed 3 main diagnostic criteria: the presence of GI symptoms, biopsies demonstrating eosinophilic infiltration of one or more areas of the GI tract, and no evidence of parasitic or extraintestinal disease. Because of the broad diagnostic criteria, a biopsy specimen with nearly any amount of eosinophilic infiltration from a patient with any GI symptom raises the possibility of EGE (3). There are 3 forms of gastroenteritis, according to the predominance of eosinophilic infiltration in different layers of intestinal wall: mucosal, muscular, and subseroal forms. Among them, muscular form leading to thickening of the bowel wall may result in GI obstructive symptoms (4,5). To our knowledge, although intestinal obstruction in adults was reported, there was no report regarding bowel stenosis in infants. In daily practice, when infants present nonbilous, projectile vomiting, EGE can be easily confused with IHPS. IHPS has similar clinical presentation, but treatments are different. Although an operation is treatment of choice for IHPS, goals of treatment for EGE are to relieve the symptom without operation and for patient to catch up on normal growth. Laboratory tests, imaging, and biopsy are needed to diagnose EGE. Peripheral eosinophilia can be commonly found, but 23% of patients had showed normal eosinophil count (3). Eosinophilia was present in one of our patients. USG played an important role for diagnosis in our cases. Anatomically, the antrum is approximately 2.5 cm in length and terminates at the pyloric orifice and pyloric ring. In infants with IHPS, USG findings show thickened pyloric muscle and normal antrum (1 mm) and no identifiable pyloric ring (6); however, in eosinophilic gastritis, USG shows identifiable normal pylorus, pyloric orifice, and thickened antral wall (>4 mm). EGD with biopsy is helpful to establish more definitive diagnosis. Mucosal biopsy can be nondiagnostic in case of a patchy process or mucosal sparing; therefore, multiple biopsies, full-thickness biopsy, or surgical resection can be required. Although surgical intervention may sometimes be required for patients with obstruction complication, treatment with steroid and sufficient nutritional support by transpyloric feeding tube for appropriate growth are the mainstay in the management of EGE, and our patients showed improvement in this way. The duration of transpyloric feeding tube depends on the infant's ability to tolerate sufficient oral feeding without vomiting. Breast milk, extensive protein hydrolysate, or amino acid–based formula could be used for transpyloric enteral feeding, which should be delivered by continuous infusion. Because the transpyloric feeding tube itself could partially block the pyloric canal in early infants with eosinophilic gastritis, the infant could be anticipated to tolerate more volume after tube removal. Its removal time could be considered when oral intake after tube removal would be satisfactory, so that growth continues to be age appropriate. CONCLUSIONS When infants present persistent vomiting, USG could differentiate eosinophlic gastritis from IHPS. In these cases, hasty surgery should be avoided. Appropriate growth can be achieved by tube feeding and nutritional support, and appropriate use of steroid therapy can also be a good method to treat and control the symptom.