Treatment of eosinophilic esophagitis with inhaled corticosteroids.

Abstract

Eosinophilic esophagitis (EE) is a newly described inflammatory lesion characterized by dense eosinophilic infiltration of the esophagus that occurs in the absence of gastroesophageal reflux (GER), parasitic infection, or other known causes of eosinophilic inflammation (1,2). Clinically, patients have symptoms of dysphagia (most commonly), odynophagia, or chest pain (1,2). Whether EE is a distinct clinical entity or an unusual variant of eosinophilic gastroenteritis is unclear. Although a brisk response to systemic corticosteroid therapy is described (3), multiple courses are often required, with the potential for attendant side-effects. Because steroid inhalers, when used without a spacer (a chamber which, when attached to the inhaler, allows improved delivery of the product to the lungs), deposit a significant portion of the delivered dose to the esophagus (4), we theorized that these devices might provide a route for topical treatment of EE. We report the response of four children with EE treated topically with steroid inhalers. CASE REPORTS Patient 1 A 13-year-old boy with a 5-year history of dysphagia sought treatment. One month before, he had experienced a meat impaction requiring endoscopic removal. At the time of endoscopy, the esophagus was noted to be normal in appearance. No biopsies were performed. A subsequent barium esophagram was normal. He had no history of asthma or allergies (specifically, no food allergies). He was referred to Mayo Clinic for a second evaluation. Results of pertinent laboratory studies showed normal hemoglobin levels and no peripheral eosinophilia. A second endoscopic procedure was performed, and biopsies were performed in the proximal and distal esophagus, antrum, and duodenum. Histologic examination of the esophagus showed dense eosinophilic infiltration (>100 eosinophils per high-power field), consistent with EE. Histologic examination of biopsy specimens obtained from the duodenum and antrum produced normal findings. He was treated with fluticasone (220 µg/puff; four puffs, two times daily for 6 weeks). He was instructed to use the inhaler without a spacer, to swallow after inhalation, and to rinse out his mouth after administration. He experienced a rapid clinical response (within 1 week) with no apparent recurrence 3 months after treatment. Patient 2 A 12-year-old boy came to Mayo Clinic with an 18-month history of chest pain. He had undergone a previous esophagogastroduodenoscopy, which reportedly showed changes consistent with esophagitis, and he had been treated with metoclopramide and ranitidine without improvement. He had a history of asthma and a family history of seasonal allergies. He had no history of food allergies. Results of esophageal manometry and a 24-hour pH probe study were normal. Analysis of peripheral blood smears produced normal findings and no evidence of eosinophilia or anemia. After a month-long course of omeprazole, the patient underwent a second endoscopic examination that showed no mucosal lesions. Histologic specimens obtained from the proximal and distal esophagus showed a dense eosinophilic infiltrate (>50 eosinophils per high-power field) consistent with EE. Findings in analyses of biopsy specimens obtained in the antrum and duodenum were normal. This patient received beclomethasone (42 µg/puff; four puffs, twice daily), and similarly to patient 1, he was instructed on the use of the inhaler without a spacer. Rapid clinical response was observed within 1 week of initiation of therapy. He has used the inhaler three times since the initial diagnosis, but he uses the medication for only 1 week at a time. He has had no symptoms and required no medication for the last 1.5 years. Patient 3 A 13-year-old boy came to the clinic with a 2-year history of dysphagia. He had a history of hypertrophic obstructive cardiomyopathy. He had no history of asthma or allergies (including food allergies) and had a normal complete blood cell count (no anemia or peripheral eosinophilia). Results of a barium esophagram were normal and specifically showed no extrinsic compression. Endoscopic examination showed no mucosal lesions, and the histologic evaluation was consistent with EE. He underwent a 3-month course of oral prednisone and had a prompt symptomatic response. Ten months later, he experienced a meat impaction and again underwent emergency endoscopy. Analysis of biopsy specimens of otherwise normal-appearing mucosa showed recurrence of the dense eosinophilic infiltrate (>100 eosinophils per high-power field) in the proximal and distal esophagus (Fig. 1). Analyses of specimens from the antrum and duodenum produced normal findings.FIG. 1: Dense eosinophilic infiltrate (>100 eosinophils per high-power field) in the proximal and distal esophagus(original magnification 40×).Similar to patient 1, he used fluticasone (220 µg/puff; four puffs, two times daily for 6 weeks). He also experienced a rapid clinical response within 1 week, but approximately 3 months after his initial 6-week course, he experienced dysphagia at school, and his inhalational therapy was reinstated. He was again treated with fluticasone (220 µg/puff; four puffs, once daily for 4 months). He remains asymptomatic 8 months after institution of inhalational therapy. A second histologic examination after the initial 6-week course of fluticasone showed complete resolution of the eosinophilic infiltrate (Fig. 2).FIG. 2: Complete resolution of the eosinophilic infiltrate (magnification 40×).Patient 4 A 12-year-old boy with a 6-month history of dysphagia sought treatment. He had also been treated unsuccessfully for GER with famotidine. He had been healthy with no history of asthma or allergies (including food allergies). A complete blood cell count showed no anemia or peripheral eosinophilia. He also had negative findings in a barium esophagram. On upper endoscopy, diffuse linear erosions were described distally from the midesophagus to the gastroesophageal junction. Analysis of biopsy specimens taken from the proximal and distal esophagus showed a dense infiltration of eosinophils, consistent with EE. Results of analyses of small bowel specimens and specimens from the gastric antrum were normal. He was treated with fluticasone (220 µg/puff; four puffs, twice daily) for 6 weeks. He experienced rapid clinical improvement (within 1 week), and analysis of a specimen obtained in a follow-up biopsy performed at 6 weeks produced normal results. Tryptase Staining Biopsy specimens from all four patients were stained for tryptase, a marker denoting the presence of mast cells. The stain is a peroxidase stain that includes a primary mouse antitrypsin antibody and an antimouse secondary peroxidase staining antibody. At least a threefold increase in mast cells was observed in the biopsy specimens from all four patients. DISCUSSION Eosinophilic inflammation limited to the esophagus is an infrequently reported disorder. It is well accepted that eosinophilic infiltration is a hallmark of gastroesophageal reflux. The degree of involvement, however, is critical and can be used to differentiate GER from other disorders. In reflux, eosinophils usually number fewer than 10 per high-power field and typically are approximately 1 to 2 per high-power field(2). The presence of mast cells has been proposed to differentiate allergy-induced esophagitis from peptic esophagitis(5). Additionally, other signs of esophageal injury in the form of neutrophilic infiltration and squamous hyperplasia are present in GER (6,7). Basal layer thickness and papillary hyperplasia are less specific markers of inflammation. Three of the four patients had normal endoscopic appearance. Patient number 4 had the linear furrows occasionally seen in EE. The degree of eosinophilic infiltration and the presence of eosinophils, both proximally and distally, effectively ruled out GER and obviated the need for pH probe testing in three of our four patients. It is unclear whether allergic enteritis, eosinophilic gastroenteritis, and EE are different entities or they represent different manifestations of the same underlying pathophysiologic state. Case reports of allergic enteritis describe patients with high serum immunoglobulin (Ig) E levels and mucosal biopsy specimens that stain positive for IgE-producing lymphocytes(8,9). These patients typically have the clinical manifestations of allergy (asthma, eczema, allergic rhinitis) along with an elevated serum IgE and peripheral eosinophilia (8,9). It is this subgroup of patients that may respond to dietary modification (10), although the response is not universal (11). Determination of serum IgE levels, radioallergosorbent testing or skin testing, and special tissue stains (other than tryptase stains) were not performed in this series and indeed may identify those patients who are likely to respond to the strict dietary measures outlined in Kelly et al. (10) Defining a case as allergic esophagitis implies that there is an identifiable antigenic trigger. Our patients did not have allergic symptoms or eosinophilia; nor did they have an obvious allergic trigger. Likewise, the absence of eosinophils in gastric and small bowel specimens and the absence of other gastrointestinal complaints excludes eosinophilic gastroenteritis. Thus, we believe the diagnosis of idiopathic EE best fits the clinical and histologic features noted in our patients. The exact incidence of EE is unknown. There have been 34 reported cases of isolated EE in the adult and pediatric English literature (1,2,12-17). Eighteen of the cases involved pediatric patients, most of them boys. The burgeoning literature on this disorder probably reflects the fact that this diagnosis may be more common than was previously believed but that it is frequently missed. We have recently recorded a series of patients of EE with distinguishing histologic features (17). The eosinophilic infiltration is dense, usually more than 20 eosinophils per high-power field(2,17). The involvement of both the proximal and distal esophagus is an important diagnostic feature, a finding that may help to distinguish this disorder from GER. The clinical manifestation of EE depends on the area of involvement within the esophageal wall (12,14). Mucosal involvement tends to produce symptoms of dysphagia, whereas significant muscular involvement may cause dysmotility or a mass lesion (2,12,14). A medical history of hypersensitivity is reported to be common (1), although this was not seen in our patients. Peripheral eosinophilia may be the only laboratory finding(1), and a high serum level of IgE should raise the suspicion of an allergic stimulus and may identify a patient who will respond to dietary therapy. The most common radiographic finding is esophageal stricture, although in most cases, barium esophagrams are normal(15). Eosinophilic gastroenteritis and esophagitis reportedly respond readily to systemic corticosteroids and less well to cromolyn (18,19); but given the chronic nature of the disease, multiple courses of systemic steroids are frequently required. When used without a spacer (a chamber which, when attached to the inhaler, allows improved delivery of the product to the lungs), 80% of inhaled corticosteroid is deposited in the esophagus (4). These products may provide an effective route for the delivery of topical corticosteroid therapy in pure EE. The favorable responses of all four patients suggest that this is the case. Adverse effects caused by inhaled products include thrush and dysphonia(20). Other well-documented adverse effects of corticosteroids in general include adrenal suppression and growth retardation in children. The degree of growth retardation and adrenal suppression occurring with inhaled products is controversial (20), but clearly the potential for systemic effects is greatly diminished in comparison with the effects of oral prednisone. The use of the inhaler without a spacer alters the safety profile of these products, in that more steroid is deposited in the esophagus and may be absorbed to a greater degree through the gastrointestinal tract. However, it has been suggested that adverse systemic effects of inhaled steroids are dependent on pulmonary rather than gastrointestinal absorption, because steroid absorbed from the gastrointestinal tract undergoes first-pass metabolism in the liver(21). Although some systemic absorption may occur and the possibility of systemic effects exists, the risk should be significantly less than that of oral steroid therapy, the only consistently effective treatment of EE reported to date. In conclusion, we report a novel approach for the management of patients with EE. Inhaled corticosteroid appears to be a safe, effective alternative to oral therapy in patients with eosinophilic esophagitis. This approach may provide effective relief while attenuating the long-term side effects of conventional steroid therapy.