Eosinophilic Esophagitis Associated With Anastomotic Strictures After Esophageal Atresia Repair

Abstract

Esophageal strictures are uncommon in the pediatric population. The etiologies of strictures in childhood include peptic, allergic, infectious, corrosive, congenital, and anastomotic. Peptic and anastomotic esophageal strictures are short and discrete, whereas strictures caused by the ingestion of corrosive agents are long. Allergic strictures are associated with eosinophilic infiltration of the esophagus and this type of stricture can be either short or long. Conventional therapies for stricture dilatation include string-guided Tucker's dilators, wire-guided Savary dilators, rubber bougie dilators, and pneumatic or balloon dilators (1). We present a series of three cases with a history of refractory anastomotic strictures after esophageal atresia repair that were associated with eosinophilic esophagitis (EE). CASE 1 Case 1 is an 8-year-old female born with esophageal atresia (EA) with tracheoesophageal fistula (TEF) repaired soon after birth. While in the hospital, she experienced vomiting with an intact protein formula and was discharged on a protein hydrolysate formula. She was started on a combination of ranitidine (4 mg/kg/day) and metoclopramide (0.4 mg/kg/day) for her gastroesophageal reflux (GER) symptoms. One year later this therapy was discontinued when her symptoms disappeared and she was able to tolerate a milk and soy protein containing diet. At 28 months she presented with history of vomiting and significant reactive airway disease and underwent an esophagogastroduodenoscopy (EGD). Esophageal biopsies showed moderate eosinophilic infiltration of the distal esophagus with more than 20 eosinophils per high-power field. Stomach biopsies revealed focal clusters of eosinophils in the lamina propria and duodenal biopsies were normal. She also had 15% peripheral eosinophilia on her leukocyte count. She was restarted on ranitidine (4 mg/kg/day) and metoclopramide (0.4 mg/kg/day). Skin testing was performed and she tested positive for milk, eggs, peanuts, pork, beef, animal dander, and dust mites. She was instructed to avoid these allergens and was placed on a soy-based formula. Her reactive airway disease was treated with inhaled albuterol and cromolyn sodium. At age 3½, she started with failure to thrive and dysphagia. A barium swallow documented an anastomotic stricture in the distal esophagus. The stricture was dilated using endoscopic balloon dilators to a maximum size of 12 mm in diameter, 8 cm in length and 8 atm of pressure. The esophageal biopsies demonstrated a dense eosinophilic infiltrate with more than 20 per high-power field. The biopsies of the stomach and duodenum were normal. Multiple balloon dilatations using dilators to a maximum size of 12 mm in diameter, 8 cm in length, and 8 atm of pressure, did not produce lasting improvement of the dysphagia. Her ranitidine was changed to omeprazole (1.2 mg/kg/day) with improvement of her symptoms. She became more compliant with her diet avoiding milk, eggs, peanuts, pork, and beef. Subsequent endoscopies showed improvement in her biopsies with less eosinophilic infiltration. She remains on omeprazole (1.2 mg/kg/day). Regarding her diet, certain foods including beef, pork, and eggs have been reintroduced without reappearance of vomiting and dysphagia. She has not required esophageal dilatation therapy in a 5-year follow-up. Her rate of growth has been normal. CASE 2 Case 2 is a 6½-year-old female born with EA with TEF repaired soon after birth. She was discharged on oral feedings with a soy-based formula and ranitidine (4 mg/kg/day). At approximately 1 year of age, while on ranitidine, she developed an anastomotic esophageal stricture that required several dilatations with Savary dilators to a maximum size of 32 French. Subsequently, she was able to tolerate a regular diet containing milk protein. At 2 years of age, she underwent an EGD with balloon dilatation using dilators to a maximum size of 12 mm in diameter, 8 cm in length, and 8 atm of pressure. Her esophageal biopsies revealed eosinophilic infiltration with distal esophagitis with more than 100 eosinophils per high-power field. The biopsies of the stomach and duodenum were normal. Her therapy for GER was maximized with omeprazole (1 mg/kg/day). After presenting with the stricture, she developed multiple episodes of reactive airway disease and eczema. She was diagnosed to be allergic to peanuts, peas, dust, and animal dander. The patient was instructed to avoid these allergens and her reactive airway disease was treated with inhaled corticosteroids, albuterol, and cromolyn sodium. Although her esophageal biopsies still have eosinophilia with more than 20 per high-power field, she has not had any further dilatations and her esophageal stricture has resolved. The esophageal stricture has not recurred in a 3½-year follow-up. She remains on omeprazole (1 mg/kg/day) and a diet avoiding the offending allergens. CASE 3 Case 3 is a 3-year-old male born with EA with TEF repaired soon after birth. He was on an intact protein formula during the first year of life and subsequently tolerated a regular diet. At approximately 1½ years of life he started with post-prandial vomiting, that did not improve with ranitidine (4 mg/kg/day). A barium swallow documented an anastomotic esophageal stricture. He underwent esophageal dilatation under fluoroscopy with Savary dilators to a maximum size of 38 French. A few months later, an EGD was performed using injections of 20% triamcinolone at a dose of 0.5 ml (40 mg/ml diluted 1:1 with normal saline) per injection in four quadrants at the level of the stricture. The injections were followed by esophageal balloon dilatation using dilators to a maximum size of 10 mm in diameter, 8 cm in length, and 3 atm of pressure. Initially, the stricture was restrictive to the neonatal endoscope which has a 5.3-mm external diameter. Mucosal biopsies were not taken during this procedure. One month later a repeat endoscopy was performed using 20% triamcinolone injections at a dose of 0.5 ml (40 mg/ml diluted 1:1 with normal saline) per injection in four quadrants at the level of the stricture. Esophageal biopsies revealed moderate inflammation with eosinophils in the distal esophagus with more than 20 eosinophils per high-power field. The stomach and duodenum were histologically normal. The patient was placed on omeprazole (1mg/kg/day) and metoclopramide (0.4 mg/kg/day) with moderate improvement of his symptoms. Skin testing was negative. Since then, he has had six sessions of endoscopic balloon dilations using dilators to a maximum size of 13.5 mm in diameter, 8 cm in length, and 4.5 atm of pressure. Esophageal biopsies have consistently revealed eosinophilic infiltration of the distal esophagus. At 2½ years of age, he was placed on a casein and soy-free diet with improvement on the vomiting episodes. Currently he continues on omeprazole (1.5 mg/kg/day) and metoclopramide (0.6 mg/kg/day). He has not required dilatations for 6 months and he remains on the milk and soy protein-free diet. DISCUSSION Esophageal strictures occur in 5 to 15% of cases of single-layer anastomosis for esophageal atresia (2,3). Anastomosis of esophageal atresia with wide gap requires extensive mobilization of the distal esophagus, which can predispose to GER. McKinnon and Kosloske (2), in a review of 64 infants, determined that the most important etiologic factor for the development of esophageal strictures was the tension placed on the anastomosis due to a wide gap between the esophageal segments. Severe GER disease is more common among infants with a wide gap. Conventional treatment for these strictures included dilatation therapy with acid suppression. Yanchar et al. reviewed 111 cases of EA; the authors concluded that the development of a post-anastomotic stricture requiring dilatation or resection was associated with a history of clinically evident GER. No correlation with anastomotic leaks or other postoperative complications were found (4). Sometimes bouginage or surgical therapy alone is not enough to effectively treat post-surgical strictures. Children with persistent GER despite appropriate medical therapy and/or anti-reflux surgical procedures may have a persistent or recurrent stricture. As has been shown by Kelly et al. (5), children with GER non-responsive to medical or surgical therapies may have eosinophilic esophagitis. Gastroesophageal reflux is a frequent complicating factor in patients with esophageal strictures and needs to be controlled adequately. Although eosinophilic infiltration of the esophagus is associated with GER disease when the infiltration persists despite adequate acid suppression therapy, a different entity called EE needs to be considered. The diagnosis of EE can be easily missed unless biopsies are obtained. Ruchelli et al. (6) concluded, in a review of 102 patients, that the presence of severe eosinophilic infiltration correlated inversely with the response to conventional GER treatment. Medical therapy, before considering a Nissen fundoplication, may be appropriate in the presence of severe esophageal eosinophilia. History of multiple allergies, particularly to food, is frequent in these patients (7). Most of our patients with EE had history of allergic disorders including reactive airway disease. EE can present in association with asthma, eczema, and chronic rhinitis in more than 50% of the patients (8). Skin testing can be negative in some patients. Eosinophilic gastroenteritis can be classified into three categories: serosal, muscular, and mucosal. EE is considered a subset of allergic or eosinophilic gastroenteritis Symptoms of EE include epigastric pain, nausea, vomiting, regurgitation, and dysphagia (8). Vasilopoulos et al. (9) have described the relationship of dysphagia with a small-caliber esophagus in EE. The authors preferred the term small-caliber esophagus over stricture due to the absence of cicatrization. The gold standard for the diagnosis of EE is the esophageal biopsy. Biopsies of other parts of the gastrointestinal tract can support the diagnosis of eosinophilic gastroenteritis (8). The eosinophilic infiltration in the esophagus is dense, usually more than 20 eosinophils per high-power field (10). Endoscopic findings in EE may include a granular mucosa that sometimes is seen as whitening, erythema, ulceration (7), or linear furrowing (11); however, many cases of EE have no specific visual findings. Bousvaros et al. (12) reported the association of ringed esophagus, visualized by endoscopy, with esophagitis. Years later, Siafakes et al. (13) demonstrated the relationship between EE and multiple esophageal rings. None of our patients had linear furrowing or ringed esophagus. Treatment modalities for eosinophilic gastroenteritis include avoidance of the dietary allergens, the use of elemental formulas (5), and oral (14) or inhaled (10) corticosteroids. Kelly et al. (5) described a series of 10 patients who presented with EE who were unresponsive to standard treatments for GER and improved with an amino acid-based formula. The authors reported reappearance of symptoms on these patients during open food challenges (5). Liacouras et al. demonstrated, in a prospective study of 20 patients with EE, histologic improvement with a four-week course of oral methylprednisolone. Faubion et al. reported, in a report of four patients with EE, success with inhaled corticosteroids when used without a spacer (a chamber that improves delivery to the lungs when attached to the inhaler) for 1 to 6 weeks. The use of cromolyn has been reported in eosinophilic gastroenteritis (15); however there are no controlled studies in children with EE. Intralesional corticosteroid injections have been shown to be effective for refractory esophageal strictures of various etiologies. The mechanism of action of this treatment is not known; it appears to be related to the disruption of the collagen linkages in the scar tissue (16). One of our cases (Case 3) was treated with endoscopic triamcinolone injections before the diagnosis of eosinophilic esophagitis was made. The use of intralesional corticosteroids in the management of EE has not been studied. In conclusion, refractory strictures may be secondary to an underlying process that requires identification and treatment. These three cases illustrate the association of esophageal atresia and eosinophilic esophagitis. Even though the cases benefited from esophageal dilatations, there was concomitant presence of eosinophilic and reflux esophagitis. Maximal acid suppression therapy, together with the avoidance of the dietary allergens, resulted in major improvement of the symptoms. It is important to emphasize that esophageal biopsies are imperative in making the diagnosis of EE. We advocate the use of dietary manipulation for the treatment of EE.